This is not the Shlav A study plan. It is a consolidation block for the months before you start at Sheba — September to the end of January — built to harden what you already know rather than to get you through an exam that is still nine or ten months away and whose date is not yet fixed.

Daily shape

Six days a week, one fixed day off — pick it now and keep it. The day off is scheduled, not earned; the plans that fail are the ones that run seven days until a bad week ends them. Each working day: 75 min on the chapter, 35 min on 50 questions from the whole bank, 10 min reviewing every miss. That is 7.5 hours of reading a week, which is what the schedule below is built on.

Read the chapter in your own copy. The sections in this file are a scaffold and a drill set — they are not a substitute for the text the questions are written from.

January: no new chapters

The reading ends on 3 January — 16 weeks, about 112 hours. The rest of January is for going back over the nine core chapters, clearing the missed cards in the drill, and picking up anything from the deferred list below that you have appetite for. Arriving at Sheba mid-consolidation is better than arriving three chapters into something new.

These appeared in two of the four sittings rather than three or four, which is weaker evidence. They are worth the time only once the core is solid.

Cross-checked against the הר״י Scientific Council syllabus (July 2024, v3.0) נושאי הבחינה list. These are on the official exam topic list but have no entry in the frequency-derived 46 — the bank under-represents them, which tells you about the bank, not the exam.

Syllabus coverage check

One caveat carried forward: the chapter evidence comes from four sittings, so a chapter at 4 pooled questions could be noise. The nine core chapters are not — each appeared in every single sitting, which is a different kind of evidence from a high count in one year. Verified against the primary document: the 42 exam headings in version 4.0 are identical to version 3.0, item for item and in the same order. Version 4.0 took effect 07.06.2026 for residents below the one-third mark of their residency, which governs rotation structure rather than the examination.

• Harrison's — 3–5 questions per sitting in your track, and the 50-question Harrison block belongs to the base track you are not taking
• The AGS GRS — 30% of the June 2022 paper and zero in every paper before or since. The syllabus still mandates it; four consecutive papers say otherwise
• The Beers Criteria as a document — 2 questions in 500. Its content still arrives through Hazzard chapters, so the polypharmacy material here is not wasted, but it does not earn a reading slot of its own
• Hazzard chapters below four pooled questions — the tail is long and flat, and the Q-bank is a better way to meet it than reading

An explicit, US, consensus list of potentially inappropriate medications for adults 65 and over. Learn the architecture, because questions test which table a drug sits in as often as whether it is on the list.

The criteria tables — 2 to 10

What changed in 2023

New criteria: warfarin · dextromethorphan-quinidine in heart failure · opioids as delirium precipitants · anticholinergics in falls and fractures · ticagrelor (caution) · SGLT2 inhibitors (caution) · skeletal muscle relaxants folded into the ≥3 CNS-drug rule · lithium + ARB/ARNI · warfarin + SSRI · baclofen in renal impairment.

The avoid list, by mechanism

A prescription can be wrong in two directions. Beers only sees the drug that is present and shouldn’t be. STOPP/START adds the drug that should be there and isn’t.

Version 3 (2023), by O’Mahony and colleagues; v1 2008, v2 2015 (80 + 34). 190 criteria in total. Organised by physiological system, not by drug — you work down the problem list. Each criterion carries its reason in-line, so it can be defended at the bedside rather than cited as authority. Both halves are applied as one review. 5 minutes for someone fluent, 20–30 otherwise.

STOPP — section by section

START — the omissions

Limitations, stated honestly

190 criteria is a lot for a 15-minute clinic slot, and adoption suffers for it. The criteria are consensus-based, not individually evidence-graded, so they carry less transparency than Beers on how strong each recommendation is. And the evidence that applying them improves hard outcomes — mortality, hospitalisation — remains contested, even though they reliably reduce PIM counts. A question asking whether STOPP/START improves clinical outcomes has a nuanced answer, not a yes.

The 2025 Alternatives paper is organised as “if not this drug for this condition, then what” — which is the exact shape of a single-best-answer stem. Thirty of the 36 Table 2 criteria, covering 21 conditions, plus two drug–drug interactions (opioid + benzodiazepine, opioid + gabapentinoid).

Two limits the panel states: it is guidance for ambulatory care primarily, built from a limited set of guidelines rather than a primary-literature review; and some listed alternatives are recommended for purposes or doses not in the manufacturer's label.

Six places the two tools give different answers. This is where questions live, because a stem can only have one right answer once it names the tool.

Structural differences

Corrections to an earlier version of these notes

Three of these came from israeli_drug_database.js or geriatrics_research_library.js, which means those files encode one side of a disagreement as settled fact. Treat both as unverified until audited end to end.

Further disagreements worth holding

The pattern across all five: Beers states a general prohibition with exceptions, STOPP states a conditional one tied to a comorbidity or a number. When a stem gives you a comorbidity or a renal figure, it is usually asking the STOPP question.

Eight past papers, 2020 – 06.2026, and their מראה מקום tables

What the papers say about where the questions come from, reduced to the three lists worth acting on: the chapters that appear in every sitting, the sources no textbook contains, and the specific Hazzard tables the committee keeps citing. The full chapter index sits underneath for resolving a מראה מקום reference without opening the book.

The nine chapters that appear in every sitting

Across the four sittings whose tables state chapter numbers — 400 questions — 299 came from Hazzard, spread over 69 chapters, very unevenly. The nine above carry 96 of them. Chapters seen in three sittings of four are the “next tier” in the rail and the chapter index.

Non-textbook sources, question by question

19 of 100 questions come from material no textbook contains. That is more than dementia's share under the old weighting, and the plan had none of it scheduled.

Which tables the committee actually cites

The מראה מקום entries often name a table, not just a page. Pooling the table references across the four usable-edition sittings gives a ranked list of the specific tables that have produced questions — a much sharper reading instruction than a chapter number. The ones below were each cited two or more times.

Figures are cited far less often, and only one recurs: Figure 15-2, from the emergency-department chapter, three times. The rest are one-offs. If you are short of time in a chapter, read its cited tables and the text immediately around them.

The whole book, chapter by chapter

Transcribed from the Hazzard 8th-edition contents so that a chapter number in a מראה מקום table can be resolved without opening the book. The last column is the pooled question count across the four chapter-stating sittings; bold means the chapter appeared in all four. Where this file has notes on a chapter, the link opens them.

Does the schedule still match the evidence?

Every chapter in the reading schedule carries its own question count, so the plan can be checked against the thing it was built from rather than taken on trust. Each point below is one chapter: how often it has been examined, against how many hours it is given.

Seven of 100 questions in the 2026 geriatrics block. The plan had it at 2.1% and scheduled for week 18. It is also the cheapest material on the paper: closed, factual, and asked as direct recall rather than through a vignette.

Homeostenosis — the organising concept

Asked by name in 2026. Ageing narrows the reserve between resting function and maximal function in every system. Baseline performance is often preserved, so the older person looks well at rest; what is lost is the capacity to respond to a challenge. The consequence is the whole of geriatric medicine: a trivial insult produces a disproportionate failure, decompensation appears in the weakest system rather than the affected one, and recovery is slower and often incomplete.

Cognitive ageing — what is preserved and what declines

The examinable line: in normal ageing retrieval is slow but cueing helps; in Alzheimer's, encoding fails and cueing does not rescue recall. That single distinction separates benign forgetfulness from disease in most vignettes.

System by system

The mechanism layer the criteria papers do not give you. This is from Hazzard's and the general pharmacology literature, not from the three documents in Sources — flagged so you know which claims have a paper behind them in this file.

Definitions to have cold

What changes with age, and the clinical consequence of each

Five prescribing cascades

Classic single-best-answer material, and two of them are named outright in STOPP.

Anticholinergic burden

The single mechanism that generates the most questions, because it links polypharmacy to delirium, falls, cognition, retention and constipation in one arc. Peripheral: dry mouth, blurred vision, constipation, urinary retention, tachycardia, anhidrosis. Central: confusion, delirium, sedation, impaired memory. Older adults are more susceptible because of reduced central cholinergic reserve and a more permeable blood–brain barrier.

Scoring tools — ACB (Anticholinergic Cognitive Burden), ADS, ARS — score each drug 0–3, with a total of 3 or more commonly taken as clinically significant. The tools disagree with each other, which is itself a valid exam point. None of these thresholds is in the three source documents; Beers gives a list (Table 7) rather than a score, and STOPP sets its threshold at two or more anticholinergic drugs concurrently.

Beers Table 7 — strong anticholinergics, by class

The panel's footnote: evidence on whether some bladder antimuscarinics are worse than others lacks consistent quality. Oxybutynin has the best evidence for adverse cognitive effects, but caution is warranted for all of them.

Hyponatraemia, because half this file causes it

SSRIs, SNRIs, tramadol, carbamazepine, thiazides, mirtazapine and antipsychotics all appear above with a hyponatraemia warning attached, and none of them tells you how to work it out. Two numbers and the volume status do almost all of it.

Check thyroid and cortisol before settling on SIADH; stop the drug that is causing it; and correct slowly — the ceiling in common use is 8 mmol/L in 24 hours, less if the hyponatraemia is chronic, malnourished, alcoholic or hypokalaemic, because osmotic demyelination is the iatrogenic harm here. Acute symptomatic hyponatraemia with seizures or coma is the exception and is treated with hypertonic saline.

The deprescribing method

Have a named, ordered process you can write out.

1. Ascertain what is actually taken, and why — including OTC, supplements, eye drops, inhalers, and the gap between the list and what gets swallowed
2. Estimate the overall risk of drug-induced harm — number of drugs, high-risk agents, frailty, cognition, renal function — to set the intensity of intervention
3. Assess each drug for eligibility to stop — no indication, indication expired, PIM, a preventive drug whose time-to-benefit exceeds life expectancy, harm outweighing benefit
4. Prioritise — lowest benefit-to-harm ratio first, and the easiest withdrawals first, because early wins build trust
5. Withdraw one at a time, with a monitoring plan and an agreed plan for what to do if symptoms return

Tapering — what must not stop abruptly, and what happens if it does

Can simply be stopped: most vitamins, laxatives, NSAIDs, most antihistamines, a short course of PPI, and most statins at end of life.

One worked mechanism: SSRIs and bleeding

Worth carrying because it explains a 2023 Beers addition (warfarin + SSRI, Table 5) and three separate STOPP criteria, and because the reasoning generalises.

Platelets have no ribosomes and cannot synthesise serotonin. They acquire all of it through SERT — the serotonin transporter, gene SLC6A4, the same protein the drug was designed to block in neurons — and store it in dense granules. Blocking SERT starves them over 1–2 weeks, the time the platelet pool takes to turn over. Serotonin is an amplifier in primary haemostasis: released on activation, it recruits passing platelets through 5-HT2A and constricts the injured vessel. A serotonin-poor platelet still adheres and aggregates, but recruitment is slower and the plug is looser.

So the coagulation cascade is untouched — INR and factor levels are normal — and the phenotype is mucosal ooze and bruising, never haemarthrosis. It is additive with aspirin rather than redundant: aspirin removes the thromboxane arm, the SSRI removes the serotonin arm, leaving ADP alone to recruit. Potency tracks SERT affinity, so paroxetine, sertraline and fluoxetine carry the most, and mirtazapine or bupropion the least. With warfarin there is a second mechanism: fluoxetine and fluvoxamine inhibit CYP2C9 and 2C19 and raise INR independently.

The Israeli layer

Questions on גריאטריה בישראל frequently attach to polypharmacy. סל הבריאות coverage often decides what is realistically prescribable — DOACs for AF, SGLT2 inhibitors for defined cardiovascular and renal indications — so coverage rather than evidence is often the rate-limiting step. Kupot Holim run their own medication review and מרשם חוזר processes, and the geriatrician's letter is frequently what triggers a change. Influenza and pneumococcal vaccines are covered annually from 65. Opioids require a special prescription. And the long-term-care settings (סיעודי, סיעודי מורכב, תשושי נפש) carry their own antipsychotic-oversight themes.

Note also that Beers itself says the criteria are not equally applicable across countries because of differences in drug availability — which is the formal basis for preferring STOPP/START where the two conflict on a drug that behaves differently in the Israeli formulary.

Hazzard ch 88, p1381 — 4 q, two of four sittings · deferred tier

Rank 2, 5.2% of past questions. The paper does not test oncology — it tests whether you can decide who should be treated, how hard, and when to stop looking.

The decision framework

1. Estimate life expectancy independent of the cancer — from comorbidity, function and frailty. Tools exist (ePrognosis aggregates the Lee and Schonberg indices), but the examinable point is that the estimate comes first
2. Compare it with the time to benefit of the intervention — screening, adjuvant therapy and preventive drugs all pay out in years
3. Establish the goal — cure, life prolongation, or symptom control — and say it out loud with the patient before choosing a regimen
4. Assess for treatment tolerance, not just performance status. ECOG alone is insufficient in older adults; it misses cognition, nutrition, falls and social support

Screening: when to stop

The examinable generalisation: screening harms are immediate, benefits are delayed, so a shortening life expectancy inverts the balance. This is the same time-to-benefit reasoning that governs statins and bisphosphonates.

Geriatric assessment in oncology

The recommended sequence is screen, then assess. The G8 is the commonest screening instrument (a score of 14 or below is abnormal and prompts full assessment); VES-13 is an alternative. A positive screen leads to a full CGA, which changes the treatment plan in a substantial minority of cases.

Domains that independently predict chemotherapy toxicity and mortality: function (IADL before ADL), comorbidity (CIRS-G, Charlson), nutrition (MNA, unintentional weight loss), cognition, falls, polypharmacy, depression, and social support. A falls history and IADL dependence are among the strongest and the most often overlooked.

Toxicity prediction: the CARG score (Hurria) and the CRASH score both estimate severe chemotherapy toxicity in older adults using a mix of geriatric and laboratory variables, and both outperform performance status alone.

Treatment modification

• Dose reduction is not the same as de-escalation. Starting lower with planned escalation preserves intent; abandoning a curative regimen does not
• Primary G-CSF prophylaxis is generally recommended from 65 with regimens of intermediate febrile-neutropenia risk, where a younger patient would not receive it
• Renal dosing matters more than in younger patients — carboplatin is dosed by target AUC from an estimated clearance, so an overestimated GFR overdoses directly
• Agent-specific toxicities that fall hardest on older patients: anthracycline cardiotoxicity (cumulative, worse with existing cardiac disease); cisplatin nephrotoxicity, ototoxicity and peripheral neuropathy; taxane neuropathy compounding an existing falls risk; fluoropyrimidine mucositis and diarrhoea driving dehydration and AKI
• Immune checkpoint inhibitors are relatively well tolerated with age, but immune-related adverse events — colitis, hepatitis, thyroiditis, hypophysitis, pneumonitis — present atypically and can be mistaken for infection or delirium
• Radiotherapy is often the better-tolerated alternative where surgery is high risk, and hypofractionation reduces the transport burden, which is itself a geriatric consideration

Supportive care, and the crossover with the rest of the syllabus

• Pain: the WHO ladder, and the Beers and STOPP rules still apply — no meperidine, no chronic NSAIDs, a laxative with every regular opioid
• Delirium is a common presentation of hypercalcaemia, brain metastases, opioid initiation, infection and dehydration in this group
• Anaemia, cachexia and sarcopenia overlap with frailty; megestrol is a Beers avoid
• Oncological emergencies that present atypically: hypercalcaemia (confusion, constipation, dehydration), cord compression (back pain before weakness), febrile neutropenia (often afebrile), SVC obstruction, tumour lysis
• Goals of care and advance directives — rank 12 on this syllabus, and almost always examined through an oncology vignette

The Israeli layer

Oncology drug access runs through סל הבריאות and the annual ועדת הסל, so an agent with good evidence may still not be available for a given indication — a real constraint on the honest answer to “what would you give?”. הוספיס בית and community palliative services are covered and are the usual route for symptom control at home. Cancer is a qualifying condition for מחלה קשה arrangements, which matters for the family conversation as much as for the chart.

Hazzard ch 61, p935 — 18 questions across four sittings. The single highest-yield chapter in the book for this exam, and the one the old topic plan had at rank 17.

Parkinsonism, defined the way the chapter defines it

Parkinsonism is the unifying term, not a diagnosis: a variable combination of six independent motor features. Learn them as six, because a stem listing three of them is testing whether you know the set.

The lesion behind the bradykinesia is loss of dopaminergic neurons in the substantia nigra pars compacta, which innervates the basal ganglia. Idiopathic Parkinson disease is the commonest cause, affecting roughly 1–2% of people over 60 — and nearly every other cause of parkinsonism also becomes commoner with age, which is why the differential below matters more in your clinic than in a neurologist's.

Parkinson disease versus parkinsonism

Asked directly in 2026: which clinical pattern supports PD rather than parkinsonism.

The atypical syndromes, by their giveaway

What levodopa actually fixes

Treatment in the older patient

• Levodopa/carbidopa is first-line in older adults — not dopamine agonists. Agonists (ropinirole, pramipexole, rotigotine) carry markedly more hallucinations, somnolence, impulse-control disorders and oedema with age, and the historical reason to delay levodopa does not survive contact with a frail 78-year-old
• MAO-B inhibitors for mild disease or once-daily preference — rasagiline is the best tolerated. Watch the serotonergic interactions
• Amantadine helps levodopa-induced dyskinesia and “off” time, but causes confusion, hallucinations, livedo reticularis and ankle oedema, and needs renal dose reduction
• COMT inhibitors (entacapone) extend each levodopa dose for wearing-off
• Anticholinergics are effectively contraindicated here — Beers Table 2 lists benztropine and trihexyphenidyl, and the cognitive cost outweighs any tremor benefit

Motor fluctuations

• Wearing-off — benefit fading before the next dose, asked in 2026 at 30 minutes before the next dose. Shorten the dosing interval, or add a COMT or MAO-B inhibitor. Raising the individual dose worsens peak dyskinesia
• Peak-dose dyskinesia — reduce the individual dose and give it more often, or add amantadine
• Delayed on / dose failure — protein competition for absorption; take levodopa 30–60 min before meals, and redistribute protein to the evening
• Freezing — poorly dopaminergic. Visual and auditory cueing, physiotherapy

Non-motor disease, which is where the questions live

• Psychosis — withdraw in this order: anticholinergics, then amantadine, then MAO-B, then the dopamine agonist, and reduce levodopa last. Treat an intercurrent cause first. If a drug is needed: pimavanserin (FDA-approved for PD psychosis), or quetiapine or clozapine. Never haloperidol, risperidone or olanzapine
• Dementia — rivastigmine, named in START D4 for PDD and DLB
• Orthostatic hypotension — withdraw the offending drug, then non-pharmacological measures, then midodrine or fludrocortisone. Common enough in PD to be part of the diagnosis
• Constipation often precedes motor onset by years; RBD by a decade or more
• Depression and apathy are intrinsic, not reactive, and are under-treated

The hospital rules, which are examined and also kill people

• Contraindicated antiemetics: metoclopramide, prochlorperazine, and any typical antipsychotic. Use domperidone or ondansetron
• Contraindicated for agitation: haloperidol and risperidone. If something is unavoidable, quetiapine or clozapine, and pimavanserin where available
• Watch for the complications that end the admission: aspiration from dysphagia, constipation and ileus, orthostatic hypotension made worse by bed rest, urinary retention, and delirium

Device therapy

Deep brain stimulation and levodopa–carbidopa intestinal gel are for motor fluctuation in a patient who still responds to levodopa and whose cognition is intact. They do nothing for the non-dopaminergic features, and stimulation can worsen cognition and mood, which is why age and cognition are selection criteria rather than afterthoughts. A patient whose disability is falls, freezing and dementia is not a device candidate however severe the disease.

What motor fluctuation actually is

Early on, a levodopa dose produces benefit that outlasts the drug in the blood — the long-duration response — so the patient is smoothly “on” all day and dose timing barely matters. As dopaminergic terminals are lost, that buffering disappears: the striatal dopamine level starts to track the plasma level, and the therapeutic window between too little and too much narrows.

Key Clinical Points — ch 61

Hazzard ch 46, p679 — 16 questions across four sittings. The second highest-yield chapter, and one the original topic plan put in the cheap-points lane.

Staging — NPIAP

How common, and what they cost

• About 70% of pressure injuries occur in people over 65, in 9–22% of nursing home residents and 5–32% of hospital patients
• Healing is slow and stage-dependent: 75% of stage 2 wounds heal within 60 days, against only 17% of stage 3 or 4. A full-thickness injury should at least show improvement every 2 to 4 weeks, and if it does not, the plan is wrong
• Osteomyelitis is reported in 32% of stage 4 injuries — and the strongest clinical indicator is a probe reaching bone, not the swab
• 60% of chronic wounds contain a biofilm, which is why debridement matters more than antiseptics

Pain is the part that gets missed: most patients with a pressure injury report pain at rest, a substantial minority at the highest level — and only a small fraction receive any analgesia for it. Give analgesia 30 minutes before a dressing change, not after.

Where they form, and what the definition actually says

A pressure injury is localised damage to skin and underlying soft tissue, usually over a bony prominence or related to a medical or other device, caused by intense or prolonged pressure, or pressure combined with shear. It may be intact skin or an open wound, and it may be painful.

• The sites: sacrum, heels, ischial tuberosities, trochanters, lateral malleoli — sacrum and heels commonest. But a pressure injury can form on any soft tissue under pressure, so do not let location alone decide the aetiology
• The NPIAP modifiers: tissue tolerance for pressure and shear is altered by microclimate — the heat and moisture at the skin surface — and by nutrition, perfusion, comorbidities and the condition of the soft tissue itself. That list is why two patients on the same mattress for the same hours do not fare the same
• The name matters. Pressure ulcer, bedsore and decubitus ulcer all imply confinement to bed; these injuries form in chairs and under devices too, so pressure injury is the term the chapter uses throughout

Staging, as a depth

The staging system is not a severity scale — it is a statement about which tissue plane has been lost. Reading it that way makes the whole table fall out, including why you cannot reverse-stage and why an obscured base cannot be staged at all.

• Stage 1 blanches? Then it is reactive hyperaemia, not a pressure injury. The test is whether it fails to blanch. In darker skin the colour change is unreliable — use temperature, firmness and pain instead
• Stage 2 has no slough and no granulation. If either is present it is deeper than stage 2 by definition, whatever it looks like
• Deep tissue injury reflects intense pressure about 48 hours earlier, and the damage starts in muscle, deep, before the skin shows it. Stage 1 changes are reversible but may take 1 to 3 weeks to settle; stage 2 heals in 2 to 4 weeks
• Deep tissue injury can declare either way — it may resolve without breaking, or open into a full-thickness injury within days. Documenting it early protects the patient and you
• Mucosal injuries cannot be staged at all, because mucosa has no keratinised layer to stage through

Risk assessment

Braden is the instrument to know: six subscales — sensory perception, moisture, activity, mobility, nutrition, and friction and shear. The first five score 1–4, friction and shear scores 1–3, giving 6–23, with lower being worse and 18 or below commonly taken as at risk. Norton is the older alternative. Clinical judgement plus a documented scale beats either alone.

Prevention

• Repositioning on a schedule matched to the support surface and the patient, not a fixed two hours for everyone — ch 46 cites a trial randomising residents to turning every 2, 3 or 4 hours on a viscoelastic mattress that found no significant difference in pressure-injury incidence — 942 residents, and the rates were 2.5% at two-hourly, 0.6% at three-hourly and 3.1% at four-hourly, with no significant difference between them. Limit head-of-bed elevation to 30° where possible — higher generates shear at the sacrum
• Seated patients need weight shifted every 15 minutes where they can do it themselves, and repositioning at least hourly in a chair — sitting concentrates more pressure over a smaller area than lying does
• Prophylactic multilayer foam at the sacrum and heels in critical care cut injuries at the heel (3% vs 12%) and sacrum (1% vs 5%), number needed to treat 10
• Offload heels completely, floating them off the bed; heel devices rather than pillows that slip
• Support surfaces — reactive or alternating-pressure, selected by risk and by whether the patient can reposition themselves
• Skin and moisture — incontinence-associated dermatitis is a separate diagnosis from a pressure injury and is treated with barrier products, not repositioning
• Nutrition — adequate protein and calories. Routine high-dose vitamin C or zinc is not supported in the absence of deficiency
• Prophylactic multilayer foam dressings over sacrum and heels in high-risk patients

Dressings, chosen by the wound

Dressing changes are the main source of pressure-injury pain. Analgesia before the change, soak adherent dressings off, and consider topical lidocaine.

Debridement and infection

• Sharp, enzymatic, autolytic or larval, according to urgency and setting
• Do not debride stable, dry, adherent eschar on an ischaemic heel — it is acting as a biological cover. Assess perfusion first. Debride it if it becomes fluctuant, draining, erythematous or malodorous
• Infection is a clinical diagnosis. All open pressure injuries are colonised, so a surface swab answers a question you should not be asking. Deep tissue or bone culture if needed
• If it probes to bone, suspect osteomyelitis — MRI, and a plan that involves surgery, not another antibiotic course

Odour, when the patient cannot be debrided — the palliative stage-4 wound the 2024 paper described — is managed by ground metronidazole applied to the wound bed; for infection needing parenteral therapy the chapter's empiric choices are cefazolin, ceftriaxone or clindamycin, vancomycin for MRSA, and metronidazole, a carbapenem or a beta-lactamase inhibitor where anaerobes are suspected. Antiseptics belong on full-thickness wounds: cadexomer iodine, silver, honey or PHMB.

When healing is not the goal

In advanced illness the aim shifts to comfort: odour control, exudate management, pain relief at dressing changes, and repositioning only as far as the patient tolerates. A Kennedy terminal ulcer — sudden, pear-shaped, purple, sacral, in the last days — is a sign of dying, not of failed care, and naming it to the family matters.

Israeli layer: pressure-injury prevalence is a reported quality indicator in Israeli long-term care, which is why a grade 4 ulcer that will not heal turns up in questions about תום שיקום and placement rather than only in wound care.

Nutrition, and what the evidence actually supports

The targets ch 46 gives: 30–35 kcal/kg and 1.3–1.5 g/kg of protein daily, which has been shown to improve healing significantly. Supplements produced a relative reduction in incidence of about 25%.

Malnutrition is a risk factor and an impediment to healing, so screen (MNA-SF) and correct energy and protein intake, with a higher protein target while a wound is open. What is not supported is reflexive supplementation of vitamin C or zinc in the absence of deficiency — a recurrent distractor — and neither albumin nor prealbumin is a nutritional marker in an inflamed patient; they fall as acute-phase reactants and tell you about the illness, not the diet.

The ones that are not pressure injuries

Key Clinical Points — ch 46

Hazzard ch 59 p893 — 6 q · plus the four named 2024 papers

The diagnostic half. With ch 60 and 63 this block ranks 3rd at 4.8%; on its own, ch 59 sits below the behavioural and movement chapters. Questions here cluster on telling the subtypes apart from a vignette, the biomarker papers, the genetics, and capacity. Behaviour, the drug tables and the non-Alzheimer discriminators are in 60·63 Dementia — behaviour and treatment.

Diagnosis

Major neurocognitive disorder requires decline from a previous level in one or more cognitive domains, sufficient to interfere with independence in everyday activities, not occurring exclusively during delirium and not better explained by another mental disorder. Mild neurocognitive disorder (MCI) is the same decline without loss of independence — the functional criterion is what separates them, not the test score.

Subtypes — the discriminating features

Mixed pathology is the rule at post-mortem, not the exception, and a vignette with both vascular burden and an amnestic profile is usually testing whether you say so.

What to exclude before settling on a label

B12 deficiency, hypothyroidism, depression (the classic pseudodementia, though the distinction is less clean than teaching suggests — depression is also a prodrome), hyponatraemia, hypercalcaemia, obstructive sleep apnoea, chronic subdural haematoma, NPH, and — the one you can actually fix that day — drugs. Anticholinergics, benzodiazepines, Z-drugs, opioids and antiepileptics. Cross-reference Beers Table 7 and STOPP Section M before ordering an MRI.

Cognitive testing

The numbers behind the diagnosis

• Alzheimer disease is 50–90% of all dementia in the United States. Prevalence by age: about 3% at 65–74, 17% at 75–84, and 32% at 85 and over
• Median survival from diagnosis to death is 7 to 10 years
• 12–15% of people with MCI progress each year to Alzheimer disease or another dementia — the annual rate, distinct from the 5-to-8-year figure in 57 Cognitive change
• Familial Alzheimer disease is only 1–5% of cases, linked to APP, PSEN1 and PSEN2. Everything else is sporadic, whatever the family history suggests

APOE, in numbers you can quote

Read the right-hand column before quoting the left: a 12-fold relative risk sounds deterministic, and yet 45% of ε4 homozygotes had not developed Alzheimer disease by 85. That gap is why the chapter treats APOE as a risk factor and not a diagnostic test, and why the Lancet Commission's finding that risk is modifiable irrespective of APOE status is the clinically useful half of the genetics.

The 2024–25 diagnostic criteria papers

Revised Alzheimer's Association criteria (2024)

Shifts diagnosis from syndrome to biology: Alzheimer's disease is defined by its biomarkers, whether or not symptoms are present. Biomarkers are grouped by when they become abnormal, which is what the exam asked.

• Core 1 — the early-changing markers, and abnormality here is sufficient to establish the diagnosis: amyloid markers (amyloid PET, CSF Aβ42/40) together with the phosphorylated and secreted tau markers that track alongside amyloid (p-tau217, p-tau181, p-tau231)
• Core 2 — markers that turn abnormal later and are used for staging rather than diagnosis: tau PET, and CSF MTBR-tau243 and p-tau205
• Non-specific — neurodegeneration (NfL, MRI atrophy, FDG-PET) and inflammation (GFAP), which support but do not define

IWG recommendation, JAMA Neurology 2024

The counter-position, and the axis the exam tests. The International Working Group argues against calling biomarker-positive asymptomatic people “Alzheimer's disease”, because most will not progress within their lifetime. It reserves presymptomatic for those with a deterministic cause — autosomal dominant mutation, or Down syndrome — and calls everyone else biomarker-positive but asymptomatic at risk. Know both positions and which one the stem is citing.

VasCog-2 / WSO criteria, JAMA Neurology 2024

Revised criteria for vascular cognitive impairment and vascular dementia, asked twice in 2026 — once as “who meets the criteria” and once applied to a patient with lacunar infarcts who was also CSF-tau positive and APOE ε4 positive. The examinable tension is mixed pathology: vascular burden and Alzheimer biology commonly coexist, and the criteria have to say what to call that.

Dementia prevention narrative review

Asked as risk reduction advice to a worried 68-year-old with an affected parent. The modifiable risk factors are the examinable content — hearing loss, hypertension, physical inactivity, diabetes, smoking, obesity, depression, low education, social isolation, excess alcohol, head injury, air pollution, and more recently untreated vision loss and high LDL.

Verify every biomarker detail above against the papers themselves before relying on it. This is orientation written from the question stems and general knowledge, not extracted from the primary sources.

Staging, and why it is asked

Because almost every management decision — driving, capacity, tube feeding, hospice eligibility, which drug — is a function of stage rather than diagnosis. The two systems in use are the Clinical Dementia Rating (0, 0.5, 1, 2, 3 — none, questionable, mild, moderate, severe, scored across six domains) and the FAST scale of seven stages with substages through 7f. Advanced dementia in the sense used by the palliative literature means profound memory loss, minimal verbal output, inability to walk independently, total ADL dependence and incontinence — FAST 7. That is the population in which the tube-feeding evidence applies.

The anti-amyloid antibodies, in proportion

Lecanemab and donanemab slow decline modestly in early symptomatic Alzheimer disease with confirmed amyloid, and the effect size is small relative to the demands they make. Know the shape of the argument rather than the trial numbers:

• Eligibility is narrow: MCI or mild dementia, biomarker-confirmed, with the cognitive and imaging infrastructure to run it — which excludes most of the patients a geriatrician sees
• ARIA — amyloid-related imaging abnormalities, oedema and haemorrhage — is the characteristic harm, commoner in APOE ε4 homozygotes, requiring serial MRI
• Anticoagulation and these drugs together raise the haemorrhage question sharply
• They do not apply to moderate or severe disease, to non-Alzheimer dementias, or to prevention

Meanwhile the things with the best evidence for the patient in front of you remain unglamorous: treat the reversible contributors, deprescribe the anticholinergics, treat pain, support the carer, plan while capacity remains, and use cholinesterase inhibitors and memantine where they are indicated — which the Lancet Commission restates for Alzheimer and Lewy body disease.

One number on prevention

SPRINT-MIND: intensive blood-pressure lowering against a standard target of 140 mmHg produced a statistically significant 19% reduction in the incidence of MCI. It is the closest thing to a positive prevention trial in this chapter, it is a vascular intervention, and it sits directly behind the Lancet Commission's instruction to hold systolic pressure at or below 130 from the age of 40.

Capacity, autonomy and the Israeli legal layer

• Capacity is decision-specific and time-specific, not global. The four abilities: understand the information, retain it long enough to decide, weigh it against alternatives, and communicate a choice. A diagnosis of dementia does not itself remove capacity
• ייפוי כוח מתמשך — the durable power of attorney, executed while capacity is retained, and now the preferred route over guardianship
• אפוטרופסות — guardianship, a court appointment and a more restrictive instrument; מקבל החלטות זמני and supported decision-making are the less restrictive alternatives
• Driving. Fitness to drive is a defined medical reporting question in Israel, routed through המכון הרפואי לבטיחות בדרכים. Dementia is not automatically disqualifying at the mild stage, but moderate dementia is, and the assessment is not a cognitive score alone
• Caregiver burden is examinable in its own right — depression, physical morbidity, financial strain, and the point at which home care becomes unsafe. In Israel this is where חוק ביטוח סיעוד entitlement and the תשושי נפש versus סיעודי classification enter

Hazzard ch 60 p919 · 63 p981 — 17 q combined

The treatment half. Ch 60 and 63 both appear in every sitting and between them outscore the diagnostic chapter. Questions cluster on what not to prescribe, the starting doses, and separating the non-Alzheimer dementias from a vignette. Diagnosis, biomarkers, APOE and the anti-amyloid antibodies are in 59 Dementia — diagnosis and biology.

Behavioural and psychological symptoms

Look for the cause before reaching for a drug: pain, constipation, urinary retention, infection, a new medication, sensory deprivation, an over- or under-stimulating environment, or a caregiver interaction that is being experienced as a threat. The named framework is DICE — Describe, Investigate, Create, Evaluate.

How universal the behavioural symptoms are

• Neuropsychiatric symptoms affect almost everyone with dementia at some point — ch 60 puts it at 98% of people with Alzheimer disease across the course of illness. Cognitive deficit is the hallmark; behaviour is the near-universal accompaniment
• The commonest three are depression, apathy and anxiety — not agitation, which is merely the one that generates referrals
• They appear before dementia does. Depression and irritability are common even prior to MCI and are the first symptoms in well over half of those who later develop dementia
• Mild behavioural impairment (MBI) is the term for late-life onset of neuropsychiatric symptoms of any severity, in someone without dementia, lasting over six months, not attributable to another psychiatric disorder. It is a prodromal category, and the six-month duration is the examinable part
• Severity rises with cognitive decline, in the community and in nursing homes alike, and the symptoms carry worse quality of life, increased mortality, accelerated progression, higher costs and greater carer burden

The carer is the second patient

Carer burden predicts institutionalisation more strongly than the patient's cognitive score does. What has evidence: multicomponent interventions combining education, skills training, support and respite — not education alone, which reliably fails. Screen the carer for depression, ask about their sleep and their own health, and name the eligibility for services rather than mentioning that services exist. This is examinable as a management answer, not as a courtesy.

Key Clinical Points — ch 60

The five that matter, side by side

The discriminators ch 63 states outright

• Mixed pathology is the norm in old age. The chapter says plainly that with advancing age it becomes increasingly common for more than one pathology to be producing the dementia syndrome. A clean single-diagnosis answer is often the wrong instinct
• Lewy body dementia: cognitive impairment with profound fluctuation, spontaneous parkinsonism, REM sleep behaviour disorder, and visual hallucinations
• Parkinson disease dementia: about 30% of people with Parkinson disease develop dementia, and — unlike Alzheimer disease — the motor symptoms predate the dementia by many years
• Behavioural-variant FTD: behavioural and personality change dominant, in the relative absence of memory and perceptual symptoms. Behaviour appears in advanced dementia of any cause; it is its prominence as the presenting feature that points to FTD
• Progressive supranuclear palsy: falls and dysphagia with abnormal vertical gaze but a preserved oculocephalic reflex — the doll's-eye manoeuvre overcomes the gaze palsy, which places the lesion above the brainstem nuclei. That triad is the exam's PSP question; the imaging answer the 2025 committee wanted (Table 63-5) was atrophy of the dorsal midbrain
• The proteins, since the exam asks them: FTLD pathology is tau in nearly half (including Pick bodies and corticobasal degeneration) and TDP-43 in a similar share (semantic-variant PPA is TDP-43 type C, tau-negative), FUS under 10%. The 2024 stem — a 60-year-old with disinhibition, sweet cravings, MMSE 28 and a failed clock — wanted tau. Many FTD patients are misdiagnosed as Alzheimer disease in life

The chapter also insists the cognitive history be comprehensive — memory, language, visuospatial function, executive function, behaviour and attention — because a history that only asks about memory finds only Alzheimer disease.

Key Clinical Points — ch 63

Pharmacological treatment

• Cholinesterase inhibitors — donepezil, rivastigmine, galantamine — for mild to moderate Alzheimer's. Rivastigmine is the named agent for Lewy body and Parkinson's disease dementia (START D4)
• Memantine for moderate to severe disease, alone or added to a cholinesterase inhibitor
• Their adverse effects generate the exam questions: bradycardia and syncope (STOPP D17 and D18 — avoid with bradycardia under 60, heart block, unexplained syncope, or alongside other rate-limiting drugs), nausea and weight loss, vivid dreams, and urinary urgency — which is the start of the classic prescribing cascade when it is met with oxybutynin
• Memantine with a seizure history is a STOPP criterion (D19)
• Nootropics are dismissed wholesale by STOPP D20 — ginkgo, piracetam, phosphatidylserine, modafinil, omega-3, ginseng, creatine
• Anti-amyloid monoclonals (lecanemab, donanemab) are covered with the biomarker criteria in 59 Dementia — diagnosis and biology

The drugs, at the doses actually used

Table 59-7 and tables 60-8 and 60-9 are all cited in recent papers, and they are cited because they contain numbers.

Cholinesterase inhibitors and memantine are both restated by the Lancet Commission for Alzheimer and Lewy body disease. They do not modify the disease; they buy months of function. Stopping them is a legitimate decision at the point where there is no function left to protect, and it is taken with the family rather than announced to them.

If an antipsychotic is unavoidable: the doses

Everything in 60 Behavioural symptoms says to exhaust the non-drug route first — DICE, analgesic trial, environment, the contributing causes. When a drug is genuinely required, the starting doses are far below the psychiatric ones, and every agent here carries the boxed warning for increased mortality in dementia and can prolong the QT.

Long-acting injectables are not used for neuropsychiatric symptoms in dementia. Start low, review at days rather than months, and document the plan for stopping at the same visit you start.

Antidepressants, starting doses

Hazzard ch 58, p879 — 15 q, in all four sittings

Rank 5, 4.6%. The highest-yield single fact on the topic: the commonest subtype is the one that gets missed.

The four numbers this chapter is built on

• Delirium affects as many as half of all hospitalised people aged 65 and over
• It is preventable in up to 50% of cases — the single most important fact on the topic, and the reason prevention outranks treatment everywhere below
• It goes unrecognised in up to 70% of older patients. Not missed at the margin: missed as the rule, mostly because the hypoactive form does not announce itself
• People aged 65 and over account for nearly 40% of all days of hospital care, which is why a condition of hospitalised elders is a system-level problem

Definition and what each element means

• Acute onset and fluctuating course — hours to days, waxing and waning within a day. Collateral history is how you establish this; the patient cannot
• Inattention — the cardinal feature. Test it: months of the year backwards, serial sevens, digit span. If attention is intact, it is probably not delirium
• Disorganised thinking or altered level of consciousness
• Evidence of a medical cause — an illness, an intoxication, a withdrawal, a drug

The three subtypes

Screening instruments

CAM requires feature 1 (acute onset and fluctuation) and feature 2 (inattention), plus either feature 3 (disorganised thinking) or feature 4 (altered consciousness). The and/or structure is itself examinable. CAM-ICU for ventilated patients, 3D-CAM for a structured three-minute version, and 4AT as a rapid bedside tool needing no special training, which also functions when the patient is too unwell to complete cognitive testing.

Inouye's model: vulnerability multiplied by insult

The practical consequence is that a highly vulnerable patient needs only a trivial insult — a catheter, a night of poor sleep — while a robust patient needs a major one. That is why the same operation causes delirium in one 80-year-old and not another, and why prevention works by lowering both sides.

Risk, in multiples

Two-thirds is the number to carry: a first episode of delirium should prompt cognitive follow-up once the illness has settled, because the odds are the dementia was already there.

What it costs, in outcomes

• Delirium in the ICU carries a fourfold increase in in-hospital mortality and a sixfold increase at six months
• In the emergency department, a sevenfold increase in mortality at six months
• It does not clear on discharge: symptoms generally persist a month or more, and as few as 20% have complete symptom resolution at six months
• Prevalence at admission ranges 7–80% and incidence during admission 8–82% across studies; up to 48% in nursing homes and post-acute care, and up to 83% at the end of life

Those wide ranges reflect the setting and the detection method, not real disagreement — which is itself the point about a condition missed in 70% of cases.

Prevention beats treatment

Multicomponent non-pharmacological prevention cut delirium by 40% against standard care, and proactive geriatric consultation reduced delirium after hip fracture by 40% in a randomised trial — two separate 40% figures, both worth quoting. Multicomponent non-pharmacological prevention is the intervention with the best evidence on this entire topic. The named programme is HELP (Hospital Elder Life Program), cited by name in the Beers Alternatives paper as AGS CoCare: HELP. Its components map directly onto the precipitants: repeated orientation, therapeutic activity, early mobilisation, sleep protocols without drugs, vision and hearing aids restored, hydration, and pain control.

If a drug is genuinely unavoidable: the chapter's doses

Only for severe symptoms threatening the patient's safety or the medical treatment itself — never for disturbance.

Management

1. Find and treat the cause, and expect more than one. Review every drug started in the preceding days — a medication chart is the highest-yield investigation on the round
2. Non-pharmacological measures first, every time: glasses, hearing aids, light, family presence, mobilisation, bladder and bowel, remove the catheter and the restraint
3. If a drug is unavoidable: low-dose antipsychotic short-term. Avoid entirely in Lewy body dementia and Parkinson's — quetiapine or clozapine if truly unavoidable. Benzodiazepines are only first-line for alcohol or benzodiazepine withdrawal, and otherwise make delirium worse
4. Follow up. Delirium is not a transient event that resolves at discharge in this population — it predicts mortality, institutionalisation and persistent cognitive decline, and it both unmasks and accelerates dementia. Document it, tell the family what it means, and arrange cognitive review after recovery

Cross-reference: Beers Table 3 lists the delirium precipitants to avoid, with opioids added in 2023. STOPP D14 covers anticholinergics in delirium, and the Beers ≥3 CNS-active drug rule catches the combination that produces most hospital delirium.

What happens afterwards

Delirium is not a self-limiting episode with no sequel. It is independently associated with longer stay, institutionalisation, functional decline and death, and it accelerates the trajectory of cognitive decline in patients with dementia. In a third or more it persists for weeks to months, so a patient still muddled at discharge is not evidence that the diagnosis was wrong. Document it, tell the family what it was, arrange cognitive follow-up at a distance from the illness, and do not diagnose a dementia during it.

Delirium, dementia, depression, psychosis — the grid

Table 58-4, reduced to the discriminators. The stem will give you two or three of these features and expect the diagnosis.

Perceptual disturbance occurs in only about 15% of delirium, so its absence proves nothing: acute onset, inattention, altered consciousness and global cognitive impairment are what identify it. The reason to separate these carefully is pharmacological — treating the “depression” or the “psychosis” with an anticholinergic drug makes an unrecognised delirium worse.

Why the same insult spares one patient and floors the next

The model that organises the whole chapter: delirium is what happens when an acute stressor meets an already vulnerable brain. The two trade off against each other, so there is no fixed amount of insult that causes delirium — only an amount relative to the reserve of the person receiving it.

Two consequences worth holding. Prevention works on both axes: you cannot undo the dementia, but you can remove the catheter, restore the glasses, fix the sleep and stop the anticholinergic — each of which moves the patient left along the insult axis. And delirium in response to a trivial insult is itself diagnostic information: it tells you the baseline vulnerability was already high, which is why a first episode of delirium earns cognitive follow-up at a distance from the illness.

What sits underneath vulnerability, per the chapter: structural lesions, vascular change, altered connectivity, neuroinflammation, neurodegeneration and the other changes of age. The EEG shows global slowing of the dominant posterior alpha rhythm — asked in 2024 twice — with a false-negative rate of 17% and false-positive of 22% against non-delirious patients, so it is reserved for suspected subclinical seizures and for separating delirium from psychiatric illness; imaging is for focal signs, head trauma or no other cause. Imaging and EEG point to dysfunction in the prefrontal cortex, thalamus, basal ganglia and the temporoparietal and fusiform cortex of the non-dominant hemisphere. No single pathway explains it; the final common picture is a large-scale network failure. The neurotransmitter picture ch 58 gives, and the 2025 committee asked: raised dopaminergic activity, reduced cholinergic activity, or an imbalance of the two, with serotonin either way; cholinergic deficiency has the most evidence, alongside inflammation and abnormal glucose metabolism. The validated precipitating factors in hospital, each conferring a two- to fourfold risk and cumulative: physical restraints, malnutrition (low albumin), more than three drugs added in a day, an indwelling catheter, and any iatrogenic event — the 2023 question wanted the albumin, not the depression, disability or severe illness, which are predisposing.

Key Clinical Points — ch 58

Hazzard ch 65 p1007 — 12 q across the three 8th-edition sittings

The chapter with the most questions and no section until today. Ch 65 is unusual: it is less a pharmacology chapter than an argument about what antidepressants do and do not do in the old, and the committee has asked that argument directly — twice with the same question. The drug tables are in 60·63 Dementia — behaviour and treatment.

The PHQ-9, and what “minor” means

• The chapter's instrument is the PHQ-9, built on DSM-5 major-depression criteria, and it insists on symptom frequency — depression is not a bad day. 0–4 no or clinically insignificant symptoms; 5–9 mild, generally subsyndromal (“minor”); 10 or higher correlates with major depression and warrants evaluation and intervention. It doubles as the measurement-based-care tool, and item 9 asks about death wishes and suicidal ideation. Both the 2024 and 2025 papers built a stem on the 5–9 band
• Subsyndromal depression has about the same effect on health and quality of life as major depression, but antidepressants do not help it. The chapter's answer is not “you do not merit treatment” but non-pharmacological approaches and, where symptoms are mild, watch and wait with check-ins — never a first choice if there is any safety concern. The 2025 committee answer for the active 90-year-old with new low mood was exactly this: minor depression, follow up, no drug
• Real-world care: about 5% of older adults in primary care have clinically significant depression, half are recognised, a fifth of those get guideline-concordant care — about 10% treated adequately. And adequate care does not guarantee remission: in collaborative care, under a third remit at six months

What the chapter says about antidepressants — and the exam has asked twice

• The placebo effect is powerful. In many studies antidepressants offered no significant benefit over placebo and placebo response rates were large, especially in less severe cases; some meta-analyses find the drug's own effect only in more severe depression. Telling the patient that believing in the drug increases its chance of working is honest. Most antidepressants take 6 to 8 weeks to show a measurable effect
• They treat major, not minor, depression. “A little antidepressant for a little depression” does not hold up, and starting one for a condition that may resolve on its own commits the patient to a drug and to the problem of stopping it
• They do not generally work in dementia. A large number of negative studies; consensus is that their usefulness is questionable, and patients with dementia cannot usually engage in psychotherapy either. Carer support and behavioural interventions lift mood; an antidepressant is continued only if it yields a measurable benefit. The 2024 (“efficacy in dementia is doubtful”) and 2026 (“avoid the antidepressant, use behavioural approaches”) answers both come from this paragraph
• Slow up-titration improves adherence — patients started low and titrated over visits adhered far better than those started at the full dose, reaching the same dose; probably the attention, and fewer side effects
• There is no standard effective dose. The dose–response relationship is small, placebo is substantial at every dose, and the right dose is the one that works for that patient, capped by safety
• Patients do not volunteer side effects, sexual ones above all; they stop the drug instead. Name them and ask. The chapter's stated risks: falls, SIADH and mortality

Stopping, and what discontinuation looks like

• Withdrawal from SSRIs and SNRIs: sweating, chills, dizziness, flu-like symptoms, GI upset, insomnia, vivid and excessive dreaming, low mood, vertigo, numbness, shock-like sensations — the 2024 answer was the dreams and sleep disturbance. The danger is mistaking withdrawal for relapse
• The risks of stopping are “surprisingly low”: relapse trajectories on drug and on placebo look alike, and continued treatment offers only modest protection. Indefinite continuation is reasonable only where depression returned on stopping (not merely during withdrawal) or after multiple prior episodes; otherwise consider a deliberate, monitored taper — SSRIs, SNRIs and mirtazapine over 2 to 4 weeks, tricyclics over months; fluoxetine needs no taper and can smooth the last step. Decide the end game when you start

ECT, collaborative care, psychotherapy

• ECT remains by far the most effective treatment: response over 80%, remission over 60%; no trial has found a drug more effective; older patients respond even better than younger ones; in its current form it is safe with few side effects, and the short-term memory loss almost always resolves within a few weeks. Three to six treatments a week for 2 to 4 weeks; the benefit decays, so maintenance ECT is sometimes needed. Early option for severe or psychotic depression and for poor medication response. Asked in 2024 (remission in more than half) and 2025 (safe, few side effects)
• Collaborative care — care manager, primary care prescriber, mental-health consultant, measured symptoms — is the most effective system-level approach; still, fewer than a third remit at six months in real clinics
• Psychotherapy: CBT, interpersonal therapy, behavioural activation, reminiscence groups all work; structured and short, about 3 months at most, and if nothing has moved in 1 to 2 months change modality. In the largest geriatric trial 51% preferred counselling and 38% medication. Exercise is not a treatment for depression but the active have less onset and recurrence. Carers of a spouse with dementia may need permission to look after themselves

Suicide

Every depressed older patient is assessed for suicide risk: thoughts, intent to act, plans, and access to means. Suicidal ideation may be a communication of suffering or a genuine wish to die; the chapter's stance is a problem-solving one, keeping the patient engaged and focused on what can be changed. Risk and protective factors are Figure 65-3. In Israel the practical routes and the reporting layer are in Ethics & EOL law.

Hazzard ch 104 p1667 · 105 p1679 · 106 p1699 — 9 q

Rank 4, 4.7%. Two themes carry most of the marks: infection presents differently, and most positive urine cultures should not be treated.

Atypical presentation

The febrile response is blunted with age, so a normal temperature does not exclude infection. In long-term care the thresholds come from a document worth naming, because a stem may name it: the revised McGeer surveillance criteria (Stone et al., Infect Control Hosp Epidemiol 2012;33(10):965–977). Fever is a single oral temperature above 37.8°C, or repeated oral readings above 37.2°C, or a rise of more than 1.1°C above the patient's own baseline — the last being the point, because the baseline is often low.

Presentation is frequently through one of the geriatric syndromes instead: delirium, a fall, functional decline, anorexia, incontinence that is new, or simply “off legs”. Hypothermia and a normal white count are consistent with severe sepsis in this group, and qSOFA is insensitive here.

Asymptomatic bacteriuria — the single highest-yield item

• Pyuria is not infection. It is near-universal in bacteriuric older adults and tells you nothing about whether to treat
• Catheterised patients are bacteriuric by definition within days. A culture from a long-standing catheter answers a question you should not be asking
• Cloudy or malodorous urine is not an indication. Neither is a positive dipstick in an asymptomatic patient
• The hard case: a patient with dementia, delirium, and a positive culture but no localising urinary symptoms. Delirium alone does not establish UTI. Evaluate for other causes first; treat empirically only if the patient is systemically unwell or no alternative explanation emerges. This is what the Loeb and McGeer minimum criteria exist to discipline

Pneumonia

Distinguish community-acquired from aspiration, which is the one the geriatric vignette is usually describing: dysphagia after stroke, Parkinson's disease, advanced dementia, sedating drugs, poor dentition. Severity scores such as CURB-65 were not built for frailty and tend to understate risk in this group. The healthcare-associated pneumonia (HCAP) category has been abandoned — it drove broad-spectrum overtreatment without benefit. Assess swallowing before the first oral dose in anyone with a plausible aspiration story.

C. difficile

• Risk factors that are yours to control: antibiotic exposure — particularly fluoroquinolones, clindamycin, broad-spectrum cephalosporins — and proton pump inhibitors, which is why Beers caps scheduled PPI use at eight weeks
• Treatment has moved: oral vancomycin or fidaxomicin are first-line; metronidazole is no longer preferred except where neither is available
• Do not test formed stool, and do not re-test to prove cure — carriage persists
• Recurrent disease is where faecal microbiota transplantation and bezlotoxumab enter

The rest of the ground

• Skin and soft tissue: cellulitis is over-diagnosed — venous stasis, lipodermato- sclerosis and dependent oedema mimic it, and bilateral lower-limb “cellulitis” is usually neither. Pressure ulcers are colonised by definition; infection is a clinical diagnosis, and a swab of an open ulcer is not one. Suspect underlying osteomyelitis when an ulcer probes to bone
• Herpes zoster and post-herpetic neuralgia — hence the START criterion for zoster vaccination, and the topical lidocaine patch for established post-herpetic pain
• Endocarditis presents as weight loss, anaemia and confusion rather than fever and murmur; degenerative valve disease and intracardiac devices raise the base rate
• Tuberculosis reactivates with age and immunosenescence, and presents non-specifically. Relevant to the Israeli population mix
• Antimicrobial dosing by renal function — cross-reference Beers Table 6 and STOPP Section E. Nitrofurantoin is the criterion the two tools disagree about; ciprofloxacin needs reduction under 30 and carries CNS effects, tendon rupture and QTc prolongation; trimethoprim- sulfamethoxazole causes hyperkalaemia with an ACE inhibitor, ARB or ARNI
• Vaccination is a START section in its own right: influenza annually, pneumococcal at least once, varicella-zoster, SARS-CoV-2

The Israeli layer

Local resistance patterns drive empirical choice and differ from the American texts — ESBL-producing Enterobacterales are common in Israeli long-term care, which changes what “empirical cover for urosepsis” means. Use the institutional antimicrobial guidelines (at SZMC, the DAG guidelines) over a textbook regimen. Influenza and pneumococcal vaccines are covered from 65.

Two things that are not what they look like

• Bilateral “cellulitis” is almost never cellulitis. Bilateral erythema is venous stasis dermatitis, lymphoedema or contact dermatitis; cellulitis is unilateral, tender and acute. This is the commonest infectious-disease misdiagnosis on the general ward
• A swollen, red, warm calf in a patient who has been in bed is a deep vein thrombosis until imaged. Cellulitis and DVT are treated in opposite directions, and the one that kills is the one without the fever

Herpes zoster

Incidence and severity both rise with age, as does post-herpetic neuralgia. Antivirals within 72 hours of rash onset reduce acute pain and shorten the course; they reduce but do not abolish the neuralgia. For established post-herpetic neuralgia the options are a gabapentinoid, a tricyclic (with the anticholinergic cost, so nortriptyline over amitriptyline), and the topical 5% lidocaine patch, which is START K2 for localised neuropathic pain. The recombinant zoster vaccine is the one that works in this age group, is not live, and is a START criterion.

Hazzard ch 43, p633 — 5 q, in all four sittings · Lord & Menant. Tables 43-1, 43-2 and 43-3 read from the page.

Rank 6, 3.7%. It is also the syndrome that most other topics feed into — medication review, cognition, vision, sarcopenia and bone health all arrive here. The chapter is written by two balance physiologists, and it reads that way: falls are what happens when balance control fails, and everything else is a route to that failure.

What a fall is, and how big the problem is

Ch 43 opens with the WHO definition — an event which results in a person coming to rest inadvertently on the ground or floor or other lower level — and it is the definition a stem will use. It says nothing about cause, so a syncopal collapse, a trip on a kerb and a slow slide off a chair are all falls; and it says nothing about injury, so the fall that hurt nobody still counts and still predicts the next one. The incidence figures come straight from the first paragraph and from the Key Clinical Points, which is where this paper draws from.

The pattern to understand rather than memorise: the rate rises with age and with the frailty of the setting. The hospital pair is the counter-intuitive one and the one to hold — 27% on an acute geriatric ward against 2% in a general hospital. The general-hospital figure is diluted by every young surgical and obstetric patient in the denominator; the geriatric-ward figure is what happens when delirium, new drugs, unfamiliar rooms and weak legs are concentrated in one place. Your ward is the highest-incidence setting in the chapter, and a stem contrasting settings is asking whether you know that.

Most falls do not cause serious harm — 37–56% cause a minor injury and 10–15% a major one — but because falls are so common, the small injurious fraction adds up to falls being the leading cause of injury-related hospitalisation in people 65 and over, 14% of emergency admissions and 4% of all hospital admissions in that age group. The chapter is explicit that a fall without physical injury still has serious consequences: significant fear of falling, which leads to reduced mobility and frailty through the avoidance of daily activities, and falls are a key predisposing factor for needing institutional care.

Why older people fall: the balance system, and what erodes it

The chapter's organising idea is that staying upright needs the continuous integration of sensory information about where the body is (vision, the vestibular system, peripheral sensation from the feet), central processing to work out what to do about it, and a motor system fast and strong enough to do it. Any of these can be damaged by a specific disease or worn down by ageing itself, and older people usually have several marginal at once. That is why a fall is a symptom, not a diagnosis; why the causes are typically multiple and additive in the same patient; and why single-factor fixes disappoint in the frailest.

Vision. It supplies a continually updated reference frame for where the body is relative to the surroundings. The two components that matter for falls are contrast sensitivity and depth perception — not acuity as such — because they are what let you see a kerb edge or a pavement misalignment. Table 43-1 rates contrast sensitivity strong evidence and acuity only moderate.

Muscle strength and reaction time. Reduced knee extension strength raises fall risk in large community studies; in residential care, knee extension and ankle dorsiflexor weakness are the major risk factors. Strength is what corrects a perturbation once it is sensed; reaction time is how quickly the correction starts, and both are strong-evidence factors.

Gait. The older gait is slower, with shorter steps, more step-time variability, a wider base and more time in double support. Some of this is adaptation for safety, but these patterns still predict falls, and high-risk older people are worse at adapting gait on short notice to an obstacle, at recovering from a trip or slip, and at controlling the accelerations of the trunk and head while walking, which unsettles the gaze.

Cognition. Walking is not fully automatic in old age; it draws on attention and executive function. A substantial fraction of residential-care residents cannot answer a simple question while continuing to walk, and those who cannot are at higher risk. Fallers do badly on tests of executive function and attention, and imaging during stepping shows increased reliance on the prefrontal cortex with task complexity, with older people showing neural inefficiency — more cortical activity for no better performance. This is the mechanism behind the dual-task stems and behind cognitive impairment undermining every intervention that requires remembering to do something.

Fear and mood. Fear of falling and depression are both strong-evidence risk factors. They restrict daily activity and pastimes, which leads to inactivity and social isolation, and high fear or generalised anxiety produces inadequate postural strategies — excessive stiffening in threatening conditions such as standing at height. Fear, in other words, is not just a consequence of falling; it changes the way the person stands.

Environment. Extrinsic factors are involved in 35–45% of injurious falls. The chapter's point is the interaction: a person with good physical ability copes with most hazards but can still fall on an icy footpath, and a person with very poor ability falls in a hazard-free room. Depending on the population, a quarter to a half of falls occur away from home, where stairs, slips and trips, and temporary unanticipated hazards are more often involved; a fall on stairs doubles the risk of injury.

The risk factors, graded — Table 43-1

The table is a stem-writer's list, because each row carries a strength-of-evidence grade and the surprises are examinable. The strong-evidence (***) rows are the ones to know cold; the rows that are weaker than you would expect are the ones the committee has asked about.

Three rows are traps. Orthostatic hypotension and vestibular disorders are commonly thought to be strong fall-risk factors; the chapter says the community cohorts have not found them to be, though they may matter in clinical populations — its example is orthostatic hypotension in Parkinson disease. Alcohol has not been shown to be a fall-risk factor in several large cohorts; the chapter suspects under-reporting and selection bias rather than safety, notes that long-term heavy intake causes osteoporosis, cerebellar atrophy and neuropathy, and that traumatic brain injury is twice as prevalent among fall presentations with an indication of alcohol use. And NSAIDs, in the most recent comprehensive review, do not raise fall risk in adjusted analysis.

Two demographic points are worth a line. Women fall more than men in the community — attributed to living alone, being underweight, more psychotropic use and less muscle strength, and they fracture more because of osteoporosis — but in hospitals and institutions the rates are similar, or higher in men. And being physically active does not always prevent falls, because active people are exposed to more fall-risk situations and frail people starting exercise overestimate their balance early on; the chapter's answer is that exposure risk is outweighed by the gains in function and independence.

Medications, in the chapter's own terms

The consistent association is with psychoactive medication — benzodiazepines, antidepressants, antipsychotics — and with multiple medication use. Psychoactive drugs carry a two- to threefold increase in falls and a twofold increase in hip fracture. Two points the committee could turn into a stem:

• Antihypertensives must be examined by class, not lumped. Some classes — the chapter names angiotensin-system blockers — appear protective against falls, mechanism unclear. What raises risk is the 24 hours after starting an antihypertensive or changing its dose: an acute effect, hence the “acute effects” grading in Table 43-1.
• The FRID list used on the ward is STOPP Section K in full — benzodiazepines, Z-drugs, antipsychotics, antidepressants, antiepileptics, opioids, first-generation antihistamines, vasodilators with a postural drop, alpha blockers, centrally-acting antihypertensives, antimuscarinics — and Beers Table 3 adds anticholinergics as a class, new in 2023. That list is from the prescribing tools, not from ch 43, which sticks to the psychoactive classes.

Screening and assessment

The chapter's screen is simpler than the guideline one you may know. Ask about falls in the past 12 months and assess balance and mobility; previous falls are among the strongest predictors of falling again in the following year, so every older person should be asked. (The three-question screen used in practice — fallen in the past year, feel unsteady, worry about falling — comes from the CDC STEADI and World Falls Guidelines, not from ch 43; a single yes triggers assessment.) Whoever screens positive — a fall in the past year and poor performance on a simple balance or gait test — gets a full assessment to build an individualised care plan, using either a multifactorial tool covering the range of risk factors or functional mobility assessments of vision, strength, coordination, balance and gait.

The setting changes the rule. In hospital, NICE says patients aged 65 and over, and those aged 50–64 with a predisposing condition such as recent stroke, are assumed to be at risk and go straight to multifactorial assessment, targeting what can be fixed during the stay: postural instability, cognition, vision, continence, medications. In residential care most residents are at risk, so it is pragmatic to skip screening and assess everyone regularly; the CaHFRiS screen (MMSE below 17, impulsivity, reduced standing balance, reliance on a frame, a fall in the previous year, antidepressants or hypnotics/anxiolytics) takes six-month fall probability from zero with no factors to 100% with six or more, and doubles as a list of things to fix.

The bedside measures — Table 43-2, and what each one tests

Best-practice guidelines recommend the Timed Up and Go — rise from a chair, walk 3 m at normal pace with the usual aid, turn, return, sit — as the screen for who needs a fuller balance and gait assessment, recording the quality of the walk and transfers as well as the time. But the chapter then reports a systematic review of 25 studies finding its predictive value for falls in community-dwellers limited, with no cut-point that could be recommended. The alternatives it endorses as simple and predictive are the Alternate Step Test, the Sit to Stand Test and the SPPB.

What works: the single interventions, as listed in Key Clinical Point 4

The chapter's own list of single interventions proven in RCTs is the list to reproduce: exercise, enhanced podiatry, occupational-therapy home interventions, expedited cataract extraction, single-lens glasses for regular multifocal wearers, cardiac pacing for carotid sinus hypersensitivity, and vitamin D in people with low levels living in residential care. Each has a figure and a caveat.

Exercise has the strongest evidence, and the reason is mechanical: what stops a stumble becoming a fall is a fast, strong corrective step, and that is trainable. Across 64 RCTs, community-dwellers allocated to well-designed programmes had 23% fewer falls. The effect is larger with a high challenge to balance and three or more hours a week; it holds at 75 and older, in groups or individually, delivered by a health professional or a trained leader. Walking alone is not balance training. Programmes aimed specifically at rapid stepping — the key balance-recovery response — cut falls by 52% in a meta-analysis of seven RCTs. Table 43-3's four exemplars are individually prescribed home exercise three times a week (tandem stance, heel and toe walking, stepping over objects, chair rises, stair climbing, ankle-weight strengthening plus walking — the Otago pattern), the LiFE programme (balance and strength folded into daily activity rather than a set of exercises), a hybrid group-plus-home programme, and tai chi. Exergames on step mats or balance boards match traditional exercise on risk factors; whether they prevent falls is still under study.

The caveat is the examinable part. Exercise alone works in Parkinson disease and cognitive impairment but not in long-term stroke survivors or people recently discharged from hospital — one home-based programme after discharge increased falls by 43%. In those groups it has to sit inside a multifactorial plan. In subacute hospitals three RCTs show balance and mobility training prevents falls; in acute hospitals the effect of exercise alone is not known. In residential care the single-intervention evidence is mixed, but twice-weekly progressive resistance and balance training for 25 weeks (Sunbeam) cut falls by 55%, and a multifaceted programme with staff training, resident education, environmental changes, hip protectors and twice-weekly group exercise cut them by 45%. The 2023 committee answer to “single most effective” was exercise.

Vision. Cataract surgery for the first eye cut falls by 34%; the second eye by 32%, not significant. A cohort study adds that the benefit holds when the lens-power update after surgery is cautious (under ±0.75 dioptres). Simply updating glasses is a different story: one RCT of vision correction did not reduce falls and another increased them by about 50% — new lenses change the visual world the person has adapted to. The intervention that worked was single-lens distance glasses for people who regularly wore multifocals outdoors, with counselling to show how the lower lens blurs ground-level hazards: 40% fewer falls, fewer outdoor and injurious falls, in those who regularly went out. For uncorrectable low vision, a targeted home-safety programme cut falls by 41%.

Medication management. The chapter is uncomfortable reading for deprescribers. Early trials found withdrawing centrally acting drugs reduced falls, but a recent systematic review of five trials and 1309 participants found FRID withdrawal strategies did not significantly reduce falls (rate ratio 0.98, 95% CI 0.63–1.51) over 6–12 months. The chapter's conclusion is that the effort belongs in preventing initial uptake: benzodiazepines, Z-drugs and other psychotropics for insomnia should be avoided unless there is established dependence or a failed withdrawal, and insomnia managed non-pharmacologically. Vitamin D is the same shape: current mega-trials show supplementation does not improve physical function or prevent falls in community-dwelling older adults, even when deficient; it may reduce falls in residential care, which is the only setting the chapter endorses it for. Two sittings have asked exactly this (2023, 2024).

Home safety. Effective in high-risk groups and when delivered by an occupational therapist: 21% fewer falls overall, 39% in high-risk groups. The mechanism the chapter emphasises is not the grab rail but the behaviour change and safe-mobility advice that comes with it — successful programmes reduced outdoor falls too, which no rug removal can explain. Home modification also cuts fall injuries (26% overall in a cluster RCT of 842 households).

Feet and footwear. Multifaceted podiatry — orthoses or safe footwear, footwear advice, foot and ankle exercises, routine podiatry — cuts falls by 23% in community-dwellers, and multifactorial programmes that include podiatry by 27%. In residential care a pilot showed a small effect at 3 months that was gone by 6.

Psychological. CBT programmes as sole interventions have not prevented falls in people with fear of falling; adding CBT principles to a multifactorial programme with exercise may beat exercise alone. Depression, anxiety and sleep disorders are risk factors and CBT treats them, but that has not been tested as fall prevention.

Cardiovascular. Not all falls are balance failures. Recurrent falls with no obvious gait or balance problem earn a specialist assessment and specific investigation. Three studies of pacemakers for cardioinhibitory carotid sinus hypersensitivity — an abnormal haemodynamic response to carotid sinus massage, presenting as unexplained dizziness or syncope — show fewer drop attacks, less syncope and fewer falls. (The orthostatic-hypotension definition, 20/10 mmHg within 3 minutes, and its work-up belong with syncope, ch 45, not ch 43.)

Multifactorial interventions, and the two big trials that failed

A multifactorial intervention assesses the individual's risk factors and intervenes on the ones found; common components are medication adjustment, home safety, exercise and education. A Cochrane review of 43 community trials found 23% fewer falls — the same size as exercise alone. In residential care, where frailty and cognitive impairment are the norm, single interventions have generally not worked, and multifactorial programmes delivered by multidisciplinary teams with staff education are what reduce falls. In hospital, the successful combinations include supervised exercise and balance training, patient and staff education, medication review, vitamin D, environmental review, walking aids and hip protectors.

After the fall

Ch 43 is a prevention chapter and does not cover the post-fall work-up; what follows is ward practice, not chapter-sourced. The post-fall assessment is a search for the injuries that hide. Head injury on an anticoagulant needs a low threshold for CT, and a subdural can present weeks later. A long lie causes rhabdomyolysis from muscle compressed under body weight, acute kidney injury from the myoglobin, pressure injury over the same bony points, hypothermia and dehydration; ask how long they were down and whether they could get up. An occult fracture is the hip fracture with a normal radiograph: if the patient cannot weight-bear, MRI or CT; the pubic rami, vertebrae and proximal humerus break the same way. Then bone health: a fragility fracture is itself the indication for osteoporosis treatment, and START H4 covers it. And fear of falling, addressed explicitly, or the exercise programme will not be done.

Restraints, bedrails and alarms

The Israeli layer

Inpatient falls-risk scoring is routine on Israeli wards and drives documentation as much as care. Community fall-prevention and physiotherapy run through the kupot; home adaptation and personal-care hours come through חוק ביטוח סיעוד via ביטוח לאומי, which is often the practical answer to “what would you actually arrange for this patient?”. Recurrent unexplained falls also bear on fitness to drive.

Key Clinical Points — ch 43

Hazzard ch 67–70, p1045 — low yield as a source; the law section carries the EOL questions

Rank 7, 3.1%. The one topic where the rest of this file stops applying — the Beers Criteria explicitly exclude hospice and end-of-life care, and saying so is often the answer.

Illness trajectories

Knowing which curve a patient is on tells you when to have the conversation.

The surprise question — would you be surprised if this patient died in the next twelve months? — is crude but outperforms clinician gestalt at identifying who needs planning. For dementia, functional staging (FAST) is the conventional instrument.

Symptom control

Artificial nutrition and hydration in advanced dementia

Deprescribing at the end of life

The Beers Criteria do not apply here, and preventive drugs whose time-to-benefit exceeds the remaining life expectancy are now pure burden.

• Stop: statins (discontinuation in advanced illness is safe and improves quality of life), bisphosphonates, vitamins and supplements, antihypertensives as blood pressure falls, tight glycaemic control — relax targets and accept hyperglycaemia over hypoglycaemia, antiplatelets and anticoagulants once bleeding risk dominates, AChEIs and memantine when no longer meaningful
• Keep and optimise: analgesia, antiemetics, laxatives, anxiolytics, antisecretories, and anything treating a symptom the patient can feel
• Change the route: when swallowing fails, convert to subcutaneous rather than abandoning symptom control

Palliative sedation

Proportionate sedation for a refractory symptom — one that cannot be relieved by any tolerable means — titrated to comfort rather than to unconsciousness as an end in itself. The distinguishing feature from euthanasia is intent and proportionality: the aim is symptom relief, the dose is the lowest that achieves it, and death is not the goal or the measure of success. It requires multidisciplinary agreement, documentation, and family discussion. The doctrine of double effect covers giving opioids sufficient to relieve suffering even if a foreseen but unintended consequence is a shortened life.

The Israeli layer

See the ethics section for חוק החולה הנוטה למות in full — it is the legal frame for everything above, and the withholding versus withdrawing distinction it draws is stricter than the bioethics textbooks. הוספיס בית and community palliative services are covered through the kupot and are the usual route for dying at home; הוספיס אשפוזי exists but capacity is limited. Cultural and religious expectations around disclosure and around withdrawing treatment vary sharply across the populations you will treat in Jerusalem, and the examinable skill is asking rather than assuming.

Hazzard ch 47, p699 — 12 q, in all four sittings

Rank 8, 3.1%. Under-reported by patients, under-asked by doctors, and a direct driver of falls, pressure injury, social isolation and institutionalisation.

How common, and in whom

Defined as the complaint of any involuntary leakage of urine. Prevalence rises with age and with frailty, and is 1.3 to 2.0 times greater in older women than older men.

• Severity runs from rare dribbling to continuous leakage with concurrent faecal incontinence
• Many older people who do not leak at all have bothersome lower urinary tract symptoms — urgency, frequency, nocturia — that affect their lives. Absence of leakage is not absence of a problem
• Incontinence can be cured or greatly improved, particularly where mobility and cognition are adequate; and where it cannot, it can always be managed to improve comfort, reduce carer burden and cut the cost of care
• Patients are embarrassed and often unaware treatment exists, so the clinician has to ask, periodically, and record it as a problem. That is a chapter instruction, not a courtesy

The neural control, asked in 2023: during filling, sympathetic tone closes the bladder neck and inhibits the parasympathetic supply, relaxing the dome, while somatic tone holds the pelvic floor; voluntary pelvic-floor contraction also inhibits the parasympathetic. For emptying, sympathetic and somatic tone fall and parasympathetic cholinergic impulses contract the detrusor. That is why an antimuscarinic quiets urgency and an alpha-blocker opens the outlet.

Reversible causes first — DIAPPERS

The types, and how the vignette signals each

Assessment

History with a bladder diary over three days — volumes, times, leaks, fluid intake — which is the single most informative investigation. Then urinalysis, post-void residual, a cough stress test with a full bladder, pelvic examination for atrophy and prolapse, and rectal examination for impaction and prostate. Urodynamics only when the diagnosis is unclear or surgery is contemplated.

Treatment

• Behavioural measures are first-line for both urge and stress, and outperform drugs in trials. Pelvic floor muscle training with proper instruction; bladder retraining with progressively spaced voiding for urge; timed and prompted voiding for the cognitively impaired, which works through the carer rather than the patient; weight loss; caffeine reduction; treating constipation
• Urge, pharmacological: the whole antimuscarinic class is a Beers PIM and STOPP flags it in dementia, glaucoma, constipation and BPH with a residual over 200 mL. β-3 agonists (mirabegron) are preferred on cognitive grounds — but STOPP flags mirabegron in labile or severe hypertension and lists it among the QTc-prolonging drugs, so it is not free
• Refractory urge: intradetrusor onabotulinumtoxinA (with a risk of retention and the need to self-catheterise), percutaneous tibial nerve stimulation, sacral neuromodulation
• Stress: pelvic floor training, then pessary or surgery (midurethral sling). Duloxetine is for stress, not urge — STOPP I7 — and is avoided below eGFR 30
• Atrophy-related: topical vaginal estrogen, which is also first-line prevention for recurrent UTI. Systemic oral or transdermal estrogen worsens incontinence and is a Beers avoid in women with incontinence of any type
• Overflow from BPH: alpha-1 blocker, 5-alpha reductase inhibitor, or both — both are START criteria where prostatectomy is not appropriate. Prefer silodosin or tamsulosin; STOPP flags the less selective alpha blockers with orthostatic hypotension or syncope, and Beers bans doxazosin, prazosin and terazosin as antihypertensives outright
• Catheters are for retention that cannot be managed otherwise, sacral wound healing, accurate output measurement in critical illness, and comfort at the end of life. Not for convenience, not for incontinence, not to save nursing time. Every catheter day adds bacteriuria, and a long-standing catheter makes urine culture uninterpretable

Israeli practical layer: incontinence products are funded through the kupot subject to eligibility criteria, and the entitlement assessment is often what the family actually needs from the consultation. Severity of incontinence also feeds the ביטוח סיעוד dependency assessment.

The assessment, in order

1. History that establishes the type — leak with cough or effort, leak preceded by urgency, constant dribble, or inability to reach the toilet — plus duration, volume, pad use, fluid and caffeine, bowels, and the drug list
2. A bladder diary for three days: times, volumes, leaks and their triggers. It is the single most informative investigation in the chapter and it costs nothing
3. Examination: abdominal for a palpable bladder, pelvic for atrophy, prolapse and pelvic floor contraction, rectal for tone, impaction and prostate, plus a neurological screen and a functional assessment — can they walk to the toilet, undress, and see it at night
4. Urinalysis to exclude haematuria, glycosuria and symptomatic infection — not to hunt asymptomatic bacteriuria
5. Post-void residual, because retention with overflow and DHIC both change the management completely. A cough stress test where stress incontinence is suspected

Urodynamics is not routine. It belongs where surgery is planned, where the diagnosis remains unclear after the above, or where empirical treatment has failed — and its findings in the frail often do not change what can actually be offered.

The post-void residual, in three numbers

• Under 100 mL, in the absence of straining to void, reflects adequate emptying
• Over 200 mL is abnormal. Between 100 and 200 is indeterminate and has to be interpreted against the patient rather than the number
• Where catheterisation is being used to manage retention, the goal is to keep residual under about 300–400 mL
• Measure it within a few minutes of a spontaneous void, or the number means nothing. Gross distension — 500 mL or more — should be detectable on examination

Two physiological numbers that make the diary interpretable: the first urge to void comes at 150–350 mL, and normal bladder capacity is 300–600 mL, with voiding every three to four hours while awake.

Drug doses, from ch 47 Table 47-12

Expect a 60–70% reduction in incontinence episodes from drug therapy in selected older adults — a reduction, not continence. And the signal worth weighing before you start: bladder antimuscarinic therapy is associated with a 50% relative increase in the odds of developing dementia, which is the argument for mirabegron.

The drugs, with their costs

Behavioural therapy, with the actual protocols

• Pelvic floor exercises: three sets daily of 15 contractions, building from 3 seconds to 10 seconds each
• Prompted voiding: prompt the patient to try to void up to three times, every two hours during the day. 25–40% of nursing home residents respond very well, and responders are identified by a three-day trial — so the trial is the test, not a guess
• At night, do not prompt on a schedule. Check every hour or two and toilet only if the patient is found awake — waking someone to keep them dry costs more sleep than it saves laundry
• The cough stress test: at a comfortably full bladder, about half capacity (200 mL). Immediate leakage is stress incontinence; delayed leakage, after about three seconds, suggests a cough-provoked bladder contraction instead

When drugs are not the answer

• Refractory urgency: intravesical botulinum toxin — effective, with a real rate of retention needing intermittent catheterisation, which must be a realistic option for that patient — percutaneous tibial nerve stimulation, or sacral neuromodulation
• Stress incontinence: supervised pelvic floor training first and for at least three months; then a midurethral sling or urethral bulking, weighed against life expectancy and anaesthetic risk
• Male outflow obstruction: medical therapy, then TURP or a minimally invasive alternative. Chronic retention with a rising creatinine is a urological referral, not a catheter and a review in six weeks
• Containment done well is a legitimate goal in advanced dementia or at the end of life: the right absorbent product, skin barrier protection, a scheduled toileting routine, and a commode in the room. Incontinence-associated dermatitis is a separate diagnosis from a pressure injury and is treated with barriers, not repositioning

The drug list, which is where the fix usually is

Table 47-4 is cited three times in the recent sittings, alongside the reversible-causes table. Between them they say the same thing: before typing the incontinence, look at what the patient is taking.

The chapter frames continence itself as a set of requirements, all of which must hold: a bladder that stores at low pressure and empties, an outlet that closes at rest and opens on command, the cognition to recognise and interpret the sensation, the mobility and dexterity to reach a toilet and undress, and the motivation and environment to do so. Incontinence is a failure of any one of them, which is why the drug list and the walking aid matter as much as the urology.

Mechanistically there are only two faults: failure to store — an overactive or poorly compliant bladder, or too little outflow resistance — and failure to empty — a poorly contractile bladder, or too much outflow resistance. Every type maps onto one or both, and one patient often has two at once.

Key Clinical Points — ch 47

Hazzard ch 73–79 — mostly deferred tier; ch 75 Valvular carries 5 q

Rank 9, 3.0%. Two examinable themes: it does not present the way the textbook says, and age alone is not a reason to withhold treatment.

Presentation

Dyspnoea is the commonest anginal equivalent in older adults — more common than chest pain in the very old. Other presentations: confusion or delirium, syncope or near-syncope, fatigue, weakness, epigastric or shoulder pain, nausea, a fall, or new heart failure. Silent infarction is more frequent with age and with diabetic neuropathy.

Acute coronary syndrome: strategy

• Older patients have more to gain in absolute terms because their event rate is higher — and more to lose, because their bleeding rate is higher too. That tension, not age, is the decision
• An invasive strategy in NSTEMI reduces recurrent infarction and revascularisation; the mortality signal in the very old and frail is less clear and the evidence continues to evolve. Frailty, cognition, life expectancy and the patient's own goals should drive the choice, and a frailty assessment before catheterisation is the geriatrician's actual contribution
• Radial access materially reduces access-site bleeding and is preferred
• Contrast nephropathy risk rises with age and CKD — hydrate, minimise contrast volume, and hold nephrotoxins

Antithrombotics — where most of the iatrogenic harm happens

Secondary prevention, and what START demands

The omissions are as examinable as the excesses. START requires, in documented coronary disease: a statin unless end-of-life or moderate-to-severe frailty, an ACE inhibitor, a β-blocker for symptomatic disease, and antiplatelet therapy. Add cardiac rehabilitation, which is effective and systematically under-referred in this age group. Blood pressure and lipid targets are relaxed with frailty, not abandoned.

Where it collides with the rest of the syllabus

• β-blocker with verapamil or diltiazem — STOPP B3, risk of complete heart block; and any rate-limiting drug with bradycardia under 50
• Severe symptomatic aortic stenosis coexists often; STOPP B20 warns against antihypertensives there, and the answer to symptomatic severe AS is valve intervention, with TAVI having changed what is offerable at 85
• Anaemia and CKD both worsen ischaemia and both complicate antithrombotic choice
• Nitrates plus a PDE5 inhibitor — STOPP B14, cardiovascular collapse
• Digoxin, amiodarone and dronedarone all carry Beers and STOPP restrictions; see the Alternatives section for the rate and rhythm control alternatives

Israeli layer: catheterisation is widely and rapidly available, which makes the restraint question — should this frail 92-year-old go to the lab — a live one rather than theoretical. שיקום לב is covered and under-used.

Atrial fibrillation and the anticoagulation decision

The commonest sustained arrhythmia in this population, frequently asymptomatic, and the stroke it causes is larger and more disabling than other strokes. Score with CHA₂DS₂-VASc — age 75 or over scores 2, age 65–74 scores 1, so essentially every geriatric patient reaches the treatment threshold on age alone. Anticoagulate at 2 or more in men, 3 or more in women.

• A DOAC, not warfarin, unless there is a mechanical valve or moderate-to-severe mitral stenosis. Dose by the licensed criteria, not by instinct: apixaban is reduced when two of three are present — age 80 or over, weight 60 kg or less, creatinine 1.5 mg/dL or more. Under-dosing a DOAC to feel safer is common, is not safer, and loses the benefit
• Aspirin is not an alternative. It provides little stroke protection in AF at nearly the same bleeding cost
• Falls are not a contraindication. The modelling behind that statement is worth remembering: a patient would need to fall an implausible number of times a year for the haemorrhage risk to outweigh the stroke prevented. Recurrent head injury on anticoagulation is a real problem; a falls history on its own is not a reason to withhold
• HAS-BLED identifies modifiable bleeding risk — blood pressure, alcohol, aspirin and NSAIDs alongside, labile INR — and is not a tool for declining treatment
• Rate control is enough for most; beta-blocker first, digoxin as an adjunct with the Beers dose cap. Rhythm control for symptoms, and remember that amiodarone brings a thyroid, lung, liver and interaction burden that lasts for months after it is stopped

Hazzard ch 76 p1165 — 12 q across the three 8th-edition sittings

Split out of the cardiac block because the committee keeps returning to it: HFpEF pharmacology, device criteria, high-output failure and the newer drugs. Coronary disease, valves and the Beers/STOPP cardiac criteria stay in 73–79 Cardiac.

What it is in the old, and what precipitates it

• More than half of older patients with heart failure have preserved ejection fraction; 70–80% of them are hypertensive, and cardiac amyloid is an increasingly recognised cause. Age-related diastolic dysfunction — impaired active relaxation and reduced compliance — explains the failure when nothing else is found; the 2024 vignette with a soft late-diastolic sound at the apex and pulmonary congestion wanted a relaxation problem, not contractility
• High-output failure is overlooked: chronic anaemia, hyperthyroidism, thiamine deficiency, arteriovenous shunting (dialysis fistula, AV malformation), morbid obesity. Asked in 2024 with B1 as the answer
• The commonest precipitant of decompensation is non-adherence to drugs or diet — up to two-thirds of exacerbations (64% in one study), with emotional and environmental factors behind 26% of readmissions. Cognitive impairment, depression, poor mobility and thin social support make complex regimens fail; the 2023 committee answer for “commonest cause of worsening” was non-adherence. Atypical presentations (Table 76-5) are often the only sign after 80
• BNP is the single most useful test but rises with age (especially in women), renal impairment and anaemia, and is lower with obesity, so its specificity falls with age: over 500 pg/mL strongly suggests active failure; under 100 in a non-obese older adult makes it unlikely. Echocardiography for every new diagnosis, because HFrEF and HFpEF cannot be separated clinically

HFrEF drugs — the ones the exam has asked

Devices — the criteria, and the age caveats

• ICD: primary prevention in NYHA II–III HFrEF with EF 35% or less, and secondary prevention after resuscitated ventricular arrhythmia. Guidelines ignore age, but few trial patients were over 75, benefit declines with age from competing mortality, life expectancy under 12–18 months means no benefit, and the over-80s have twice the implantation complications. Shared decision; and agree in advance when the shocks are switched off, and whether the generator gets replaced
• CRT: improves symptoms, exercise tolerance, quality of life and survival — including in octogenarians — where ICDs only prevent sudden death. Criteria: NYHA II–IV HFrEF, EF 35% or less, QRS 150 ms or more; greatest benefit with left bundle branch block (present in 20–30% of HFrEF) and in women. The 2024 answer was “HFrEF, NYHA III”; the QRS distractor was 120
• MitraClip (COAPT): fewer admissions and deaths in HFrEF with severe secondary mitral regurgitation despite maximal therapy — but only after optimisation, because much regurgitation is load-dependent and disappears with diuretics and afterload reduction. Heart-team decision with a geriatrician on it
• LVAD: the recurring stem (2020, 2025) is the woman in her seventies with EF 20% and six admissions in six months — the keyed answer is the left-ventricular assist device as the intervention that can both lengthen and improve life, not inotropes, not transplant, and not palliation alone

HFpEF: what actually has evidence

Treatment begins with aggressive blood-pressure control — SPRINT-derived targets under 130/90 for ambulatory community-dwelling older adults, personalised for falls risk — and diuretics for congestion. Beta-blockers may worsen the chronotropic incompetence that limits exercise in HFpEF; nitrates are not recommended after a negative crossover trial; digoxin is not beneficial; ACE inhibitors and ARBs are reasonable, especially with hypertension; spironolactone (TOPCAT) may help and the FDA has considered it. The one drug with a positive outcome trial is the SGLT2 inhibitor above.

The non-drug programme, which is where the readmissions are

Disease-management programmes in a meta-analysis of 33 trials: heart-failure readmissions down 42%, all-cause readmissions down 24%, mortality down 20%. In one cohort of over-70s, 47% were readmitted within 90 days, and the reasons were social: non-adherence (drugs 15%, diet 18%), inadequate support (21%), poor discharge planning (15%), inadequate follow-up (20%), and patients not seeking help when symptoms returned (20%). Influenza and pneumococcal vaccination reduce the pneumonia and viral decompensations, and influenza vaccination is associated with lower mortality — the 2023 committee answer for HFrEF (Table 76-8). Exercise training is safe and improves function in both HFrEF and HFpEF (Tables 76-9, 76-10). Heart failure often worsens after discharge — ch 53's warning for the hip-fracture patient sent to subacute care — and the geriatrician's job is the transition.

Hazzard ch 68, p1055 — 11 q, three of four sittings

Rank 10, 2.9%. Older adults are systematically under-treated for pain, and patients with dementia most of all — because they do not report it, not because they do not feel it.

Definitions and scale, before anything else

Nociceptive pain comes from stimulation of pain receptors by tissue injury, inflammation or mechanical distortion — trauma, burns, infection, arthritis, ischaemia, tissue distortion — and responds to analgesics; neuropathic pain is a lesion or disease of the somatosensory system, peripheral or central. The 2026 pairing question keyed angina as nociceptive.

• Acute pain runs with the tissue healing, roughly 3 months or less. Chronic pain persists or recurs beyond 3 months, and the IASP updated its chronic-pain classification in 2019 — that date and that threshold are both examinable
• Chronic pain in older adults: 25–75%, and higher in residential care
• High-impact chronic pain — pain limiting life or work activities on most days — affects 11% at 65–84 and 16% at 85 and over
• Neuropathic pain affects 7–10% of the population
• Complexity is the norm: most older adults with chronic pain report multisite pain, and a majority report more than one type of pain at once

Assessment, including in the non-verbal patient

Self-report remains the standard wherever it is obtainable. Ch 68 asks for simple, transparent questions — “does this hurt?” beats “how are you doing?” — and points out that in chronic pain the impact on mobility, mood and fatigue is more meaningful than the intensity score. Unidimensional scales are Table 68-3, multidimensional Table 68-4. Where the patient cannot self-report, ask the family or carer and use a nonverbal observational instrument — PAINAD (the one the committee asked for, from Table 68-3), Abbey or DOLOPLUS-2 — which score breathing, vocalisation, facial expression, body language and consolability.

The ladder, adapted

• Paracetamol first-line for persistent musculoskeletal pain. STOPP L6 caps it at under 3 g/24h where BMI is under 18 or there is chronic liver disease
• NSAIDs: avoid chronic oral use — GI bleeding, AKI, hypertension, heart failure decompensation — and avoid entirely below CrCl 30 (Beers) or eGFR 50 (STOPP). Topical NSAIDs are the geriatric answer for localised osteoarthritis, with a far better systemic profile
• Avoid the combinations: NSAID with an anticoagulant or antiplatelet; NSAID with a corticosteroid; two NSAIDs

Starting doses, from ch 68 Table 68-6

The chapter gives starting doses for an opioid-naive older adult. These are starting points to titrate from, not targets.

Opioid-induced hyperalgesia — asked in 2025. A sensitisation in which the opioid itself raises pain sensitivity: pain rises with dose, and allodynia spreads beyond the original site. Risk is highest with a history of opioid use disorder, chronic-pain patients on opioids, and high doses of potent opioids — not the opioid-naive patient being titrated, which is the distractor. It is distinguished from tolerance by that dose–pain paradox. Management is gradual reduction or withdrawal, non-opioid co-analgesia, buprenorphine, or an NMDA antagonist.

Opioids

• Start low, go slow, but do start. Roughly 25–50% of the usual adult starting dose, with unhurried titration and a short-acting agent for breakthrough
• Every regular opioid gets a laxative — STOPP L2 — and an antiemetic for the first days if needed
• Do not use a long-acting opioid without short-acting cover for breakthrough (STOPP L3); do not use a strong opioid first-line for mild pain (STOPP L1); do not use long-term opioids for osteoarthritis (STOPP H9)
• Beers Table 5 combinations to avoid: opioid with a benzodiazepine, and opioid with a gabapentinoid — both risk profound sedation and respiratory depression. Opioids were added to the delirium-precipitant list in 2023, and appear in the falls criteria of both tools

Co-analgesic doses, from Table 68-8

Neuropathic pain

• Gabapentinoids — effective but sedating and ataxic; reduce the dose below eGFR 60 for both gabapentin and pregabalin, and never alongside an opioid
• Duloxetine — useful, but avoid below eGFR 30 and note the hyponatraemia caution
• TCAs — effective and essentially unusable here: strongly anticholinergic, orthostatic, cardiotoxic. Beers avoid, STOPP D1
• Topical 5% lidocaine patch — START K2 for localised neuropathic pain such as post-herpetic neuralgia. STOPP L4 rejects the same patch for chronic osteoarthritis pain, which is exactly the kind of paired criterion a stem is built on
• Capsaicin and other topicals; diclofenac 1% gel is the one ch 68 names as the safe NSAID form. The methyl-salicylate one-week caution is label advice, not in the chapter
• Gabapentinoids for non-neuropathic pain lack evidence — STOPP L5

Paracetamol: two ceilings that are not the same

Ch 68 sets the maximum total daily dose at 4000 mg, falling to 2000–3000 mg with hepatic impairment or a history of alcohol use. STOPP L6 flags paracetamol at 3 g/24 h or more where BMI is under 18 or there is chronic liver disease. The chapter gives a drug ceiling; STOPP gives a patient-specific flag. They do not conflict, and a stem may cite either.

Non-pharmacological, which is not an afterthought

Exercise therapy of any type, physical therapy, needling therapies, psychological approaches including CBT, heat and cold, and education. The Alternatives paper lists these ahead of drugs for chronic pain and makes function, not the pain score, the target of treatment.

Israeli layer: strong opioids require a special prescription form, which is a real friction on discharge planning and a common source of gaps in continuity. Pain clinics (מרפאות כאב) exist through the kupot with waiting times that matter for planning.

Ranks 11 and 12 combined, 5.6% — which makes this the second heaviest subject on the paper after polypharmacy, and the one where Israeli law diverges most from the international textbooks.

Capacity

Covered in the dementia section: decision-specific, time-specific, and defined by four abilities — understand, retain, weigh, communicate. Three additions that get examined on their own:

• Capacity is presumed until demonstrated otherwise, and the demonstration must be specific to the decision at hand
• Refusing recommended treatment is not evidence of incapacity. A capacitous patient may make a choice the team considers unwise
• Fluctuating capacity obliges you to assess at the best moment, not the most convenient one, and to optimise the conditions — hearing aids, interpreter, time of day, absence of delirium

Surrogate decision-making

Where capacity is absent, decisions follow, in order of preference: the patient's prior expressed wishes; then substituted judgement — what would this patient have chosen, which is a question about their values, not the surrogate's; then best interests, as a last resort. A common error in vignettes is a family member asserting their own preference; the correct move is to redirect to what the patient would have wanted.

Withholding versus withdrawing

חוק החולה הנוטה למות — the Dying Patient Act, 2005

The governing statute, and the single most Israel-specific thing on this syllabus.

Verify the current wording and any amendments before you rely on a specific clause in an answer — this is a statute, not a guideline, and the detail is what is examined.

The other Israeli instruments

• ייפוי כוח מתמשך — durable power of attorney, executed while the person has capacity, covering health, property or personal matters. Now the preferred route, and the thing to recommend at the time of a mild dementia diagnosis rather than later
• אפוטרופסות — guardianship, court-appointed, restrictive, and increasingly a last resort. מקבל החלטות זמני and supported decision-making are the less restrictive alternatives the reformed law prefers
• Mandatory reporting — suspected abuse or neglect of a helpless person carries a legal duty to report. This is a legal obligation, not a clinical judgement call
• Fitness to drive — a defined medical reporting pathway through המכון הרפואי לבטיחות בדרכים

Communication, and what the stems actually test

• Truth-telling. Western bioethics defaults to full disclosure to the patient; several communities you will treat expect disclosure to the family first. The examinable answer is to ask the patient how much they want to know and whom they want told, which respects autonomy without imposing a disclosure norm
• Goals-of-care conversations should start from values and function, not from a menu of interventions. “Would you want CPR?” asked cold is a badly formed question
• DNR is not do-not-treat. A resuscitation decision says nothing about antibiotics, admission, or surgery, and conflating them is a recognised error
• Futility is contested territory: physiological futility is a medical judgement, while qualitative futility is a values judgement and cannot be declared unilaterally
• Religious consultation is legitimate and often requested — rabbinic, and the equivalents in the Muslim and Christian communities. Facilitating it is part of the job; deferring the medical facts to it is not

Eight real sittings, every question with the committee's own answer key and its מראה מקום reference. This is the exam itself rather than anything written about it, and it is the best material in this file.

Mock paper

One hundred questions drawn from the 2023–2026 sittings and weighted to the measured source mix — about 81 Hazzard, 5 Israeli law and MoH, 5 named articles, 4 Harrison, 2 Brookdale, and the rest — with the Hazzard share weighted by the pooled chapter frequency. No feedback until the end, as on the day.

The official duration is not stated anywhere in the papers you have, so the per-question allowance here is an assumption, not a rule. Change it if you learn the real figure.

How you are doing

Weakest chapters, from real questions

Only the 2023–2026 papers carry usable chapter references. Four wrong in one chapter is a chapter to re-read.

One card at a time, shuffled. Write or say the answer before you reveal it — reading the answer first feels like learning and is not.

Where the misses are

Grouped by chapter, from the cards you have marked missed. Four or five in one place is a chapter to re-read, not a set of cards to re-drill.

Hazzard ch 44, p643 — 6 questions across four sittings, and one of the nine that appear every time. The examinable core is what ageing does to sleep, what it does not do, and why almost every drug answer here is wrong.

What ageing actually changes

• Less slow-wave sleep, more arousals, lower sleep efficiency
• A phase advance — sleepy early, awake early. This is circadian, not insomnia
• Sleep need falls little; the ability to consolidate it falls a lot

The vocabulary, before the picture

• Sleep cycles roughly every 90 minutes through a fixed progression of stages
• Two states: REM — brain activity resembling wake, rapid eye movements, complete loss of skeletal muscle tone, and the vivid dreaming people report on waking from it — and NREM, divided into N1, N2 and N3
• N3 is slow-wave sleep. Arousal gets harder from N1 through N3, so N1 is light sleep and N3 deep sleep
• Sleep efficiency is time actually asleep divided by time spent in bed. It falls with age, and it is the number that makes sense of the patient who is in bed ten hours for six hours of sleep
• What changes with age: longer to fall asleep, more fragmentation, and less N3 from early adulthood onward — more markedly in men. Whether REM changes with age is less clear, which is the honest answer if a stem asks

Two framings from the chapter worth carrying: poor sleep in older adults is more closely associated with comorbidity than with chronological age, and healthy older adults actually show a lower propensity to fall asleep when offered naps — a lower sleep drive — even though unplanned napping is commoner. Sleepiness in an older patient is a finding, not an expectation.

The architecture, in one picture

Two clinical consequences. A patient who is in bed for ten hours to get six hours of sleep does not need a hypnotic — they need sleep restriction, which is the component of CBT-I that works and the one patients resist. And because the falls in slow-wave sleep and the rise in fragmentation happen to everyone, neither is by itself a disease; what makes it a disorder is daytime consequence.

Chronic insomnia

CBT-I is first-line, by a trained provider or digitally. Its active ingredients are sleep restriction and stimulus control — the two patients find hardest. Sleep hygiene alone is not effective for chronic insomnia and is the classic distractor.

Obstructive sleep apnoea

Prevalence rises with age and the presentation is less typical — nocturia, cognitive complaints, falls and refractory hypertension rather than a bed partner's complaint about snoring. STOP-BANG screens; polysomnography or home testing confirms; the apnoea–hypopnoea index grades it. CPAP remains the treatment and adherence, not age, is the limiting factor. Untreated OSA is a reversible contributor on every dementia workup list.

Restless legs and periodic limb movements

Diagnosis is clinical and five-part: an urge to move, worse at rest, relieved by movement, worse in the evening, and not better explained by something else. Check ferritin — ch 44 gives the numbers: oral iron when ferritin is under 75 µg/L, and IV iron in moderate-to-severe RLS with ferritin under 300 — a far higher bar than the anaemia threshold. Review the drugs that provoke it: antidepressants, antipsychotics, antihistamines, metoclopramide.

REM sleep behaviour disorder

Loss of REM atonia — the patient acts out dreams, often violently. It is a strong prodromal marker of the synucleinopathies: Parkinson disease, dementia with Lewy bodies, multiple system atrophy, often by years. Manage the bedroom first — clear the bedside, lower the mattress, protect the partner. Melatonin is the usual first drug; clonazepam works but is a benzodiazepine in a patient who already has a falls and cognition problem coming.

Circadian and institutional sleep

Advanced sleep phase responds to timed bright light in the evening and to keeping the patient out of bed during the day. In long-term care the usual cause of broken sleep is the institution: night-time noise, light, observations and toileting rounds. Sundowning is an evening behavioural disturbance in dementia, not an insomnia diagnosis, and it is managed as behaviour — light, routine, activity, pain — not as a sleep disorder.

Sleep in hospital, which is where you will be asked about it

Inpatients are woken for observations, drugs and blood tests, in light and noise, on a ward with no daytime activity. The result is fragmented sleep, then a hypnotic, then delirium and a fall. The answer to the three-in-the-morning bleep is almost never a drug.

• Cluster observations and bloods; stop the four-hourly set that nobody is acting on
• Light and activity by day, dark and quiet at night; earplugs and eye masks have trial evidence in the intensive care unit
• Warm drink, orientation, pain and bladder attended to; no caffeine after the afternoon
• Review the drugs that fragment sleep: diuretics late, steroids late, beta-blockers with nightmares, SSRIs and SNRIs, donepezil, levodopa, theophylline, and alcohol, which shortens sleep latency and wrecks the second half of the night

A newly-started hypnotic in hospital is the drug most likely to still be on the discharge list for no reason — and once it is home, it will be there in a year.

Key Clinical Points — ch 44

Hazzard ch 55, p817 — 7 questions across four sittings, in every sitting. It is also the chapter with the highest density of numbers you either know or do not: FIM, Barthel, and the thresholds that decide where a patient goes after the acute bed.

The instruments

FIM measures assistance, not impairment and not diagnosis. A hemiplegic patient who has adapted may score better than a frailer patient with milder impairment — which is the point of the instrument, not a flaw in it.

Who benefits

• A new and potentially reversible functional loss, not a long-standing plateau
• Enough cognition to participate, or a deficit that rehabilitation itself can work around — dementia is not an automatic exclusion, and short repetitive task-specific practice works
• Medical stability enough to tolerate the sessions
• A realistic discharge destination, which is a question about the home and the family as much as about the patient

Hip fracture

Outcome is predicted, in order, by pre-fracture function, cognition and delirium, comorbidity, nutrition, depression, pain control, and the home going back to. Delirium in the first days is both a predictor and a target — it is the one on the list you can still change. Orthogeriatric co-management, surgery within about 48 hours where the patient is fit, early mobilisation, and bone-health follow-through: a fragility fracture is itself the indication to treat osteoporosis, and it is missed in the majority of cases.

Stroke

• Timing. Early mobilisation, but very early high-dose out-of-bed activity within 24 hours reduced favourable outcomes at three months in the AVERT trial. Early, frequent, short and graded — dose matters as much as timing
• Intensity and repetition drive motor recovery; constraint-induced movement therapy for a partially useful arm
• Most recovery occurs in the first three months, but gains continue and a plateau in one domain does not close the others
• Screen swallow before the first oral anything; screen for post-stroke depression, shoulder subluxation and pain, spasticity, and neglect, which is the deficit that most often defeats rehabilitation

Complications that end a rehabilitation admission

Delirium, falls, pressure injury, deconditioning from a weekend without therapy, deep vein thrombosis, constipation and urinary retention, malnutrition, and undertreated pain — which looks like refusal to participate and is treated as motivation when it should be treated as pain.

Hospital-associated disability

A substantial share of older patients leave hospital less able than when the illness started, and much of that is the admission rather than the disease. Muscle strength falls measurably within days of bed rest, and the losses are fastest in the already weak. The examinable countermeasures are unglamorous and mostly not drugs.

• Out of bed and dressed daily; walk to the toilet rather than being taken to it
• Remove the tethers: catheters, infusion lines, telemetry, continuous oxygen — each is a reason to stay in the chair
• Delirium prevention, sleep protection, hearing aids and glasses within reach
• Eat sitting up, with help if needed; nil-by-mouth orders reviewed daily
• Plan the discharge from day one and measure function on admission, or you will not know that it has fallen

Spasticity, contracture and the team

Spasticity is treated when it interferes with function, hygiene or comfort — positioning and stretching first, then focal botulinum toxin, which is preferred to oral agents because baclofen and tizanidine sedate and both appear in the prescribing criteria. Contractures are prevented, not treated. The multidisciplinary team is examined as a structure: physiotherapy, occupational therapy, speech and language therapy, dietetics, social work, nursing and medicine, with goals set jointly, written down, and reviewed against a date.

What is different about rehabilitating an older adult

The purpose is to restore what disease or injury took away and reach the highest attainable function, independence and quality of life. The techniques do not differ by age. The outcomes and the approach do, and the chapter is organised around why.

• A younger adult usually has one acute event. An older adult more often has multiple comorbidities that produce disability over time — and where there is an acute event, a hip fracture or a stroke, the comorbidities still shape the outcome
• Subclinical impairment surfaces under challenge. Mild cognitive impairment is often first recognised during rehabilitation, when the patient cannot learn to use a new walking aid. Difficulty learning the frame is a cognitive finding, not a motivational one
• Recovery takes longer and the functional outcome is usually worse than in a younger patient. Saying so at the outset is part of goal-setting, not pessimism
• The goals differ in kind. Younger patients aim at work or study; older patients at self-care and mobility — and at whatever they actually value. The chapter's own examples are golf and fishing. A patient who has never cooked will not be served by kitchen practice

Stroke recovery, from the handbook the exam cites

The EBRSR Stroke Rehabilitation Clinician Handbook is named directly in the מראה מקום tables, so this is source material rather than commentary. Its first distinction is the one everything else hangs on.

Functional improvement can occur with no neurological recovery at all, and can continue for months after neurological recovery has finished. A patient who is walking better without a better Fugl-Meyer is not a paradox.

The time course, in weeks

Peak neurological recovery falls in the first one to three months, slowing thereafter, with about 5% still recovering at a year. From the Copenhagen Stroke Study of 1,197 patients, which is where the numbers in this chapter come from:

• Neurological recovery precedes functional recovery by about two weeks
• 80% reach their best neurological recovery by 4.5 weeks, and their best ADL function by 6 weeks
• 95% reach best neurological recovery by 11 weeks and best ADL by 12.5 weeks
• Best walking function: 4 weeks with mild paresis, 6 weeks moderate, 11 weeks severe
• Two-thirds of survivors have mild-to-moderate strokes and achieve ADL independence. Only about 10% of moderate-to-severe strokes reach “full recovery”

The best single predictor of outcome is the initial clinical severity, and it predicts the time to maximal recovery as well as its extent. Three mechanisms account for the early improvement and none of them is therapy: resolution of peri-lesional oedema, reperfusion of the ischaemic penumbra, and resolution of diaschisis.

Swallowing and the complications that follow it

• Aspiration occurs in 16–52% of acute strokes, and silent aspiration in 8–27% — passage below the vocal cords with no cough and no outward sign. That is the whole argument for a formal swallow screen before the first oral anything, and for videofluoroscopy, the gold standard, when clinical assessment is not enough
• Malnutrition is reported in 6–62% after stroke, the range reflecting that there is no agreed definition, and it worsens during the admission. Albumin is a marker of inflammation here, not of diet
• Encourage self-feeding where it is safe — being fed carries its own aspiration risk

Stairs, canes and the hip precautions — ch 55 verbatim

• “Up with the good, down with the bad.” Ascending: the uninvolved leg goes up first while the involved leg and the cane stay on the lower step. Descending: the involved leg and the cane go down first, then the uninvolved leg joins them. The 2024 answer for a healed right-ankle fracture was “descend right foot first”. The cane is held in the hand opposite the affected leg, and the elbow is 15–20° flexed — the 2026 answer for a left-sided weakness was a single-point cane in the right hand
• After a total hip replacement by the posterolateral approach, for the first months: no crossing the legs, no hip flexion beyond 90°, no rolling the leg in and out, and a raised toilet seat. The 2024 “forbidden action” was crossing the legs; a high chair and full weight-bearing are allowed

Choosing the walking aid

Table 55-5 is cited three times across the recent sittings, which makes it one of the highest-yield tables in the book, and it is the sort of thing that is never taught explicitly. The logic is a ladder: more support means less freedom, more weight on the arms and, past a point, more risk rather than less.

Key Clinical Points — ch 55

Hazzard ch 42, p615 — 9 questions across four sittings. Two competing definitions, and the exam wants you to know that they are different constructs, not rival measurements of one thing.

The two models

The Clinical Frailty Scale is the judgement-based 1–9 ordinal scale used at the bedside — quick, reproducible enough, and the one that actually gets recorded. It is scored on the baseline two weeks before the acute event, not on the patient in front of you.

Sarcopenia is not frailty

Sarcopenia is loss of muscle mass and function. The examinable point is the order: low strength raises the suspicion, low muscle quantity or quality confirms it, and low physical performance marks it as severe. SARC-F is the screening questionnaire; grip dynamometry and the chair-stand test are the bedside measures; gait speed under 0.8 m/s is the usual performance cut-off. Frailty is broader — a state of reduced reserve across systems, of which sarcopenia is the muscular component.

How common, and in which populations

The wide community range — 4% to 59% for the same construct — is itself the examinable point: it reflects which definition was used, not which population. And frailty is not a one-way state: using the Fried definition, nearly 60% of people over 70 make at least one transition between frailty states over 4.5 years, though transitions toward greater frailty outnumber those away from it.

What to do about it

• Progressive resistance exercise is the only intervention with consistent benefit, and it works in the tenth decade and in nursing homes. Multicomponent programmes add balance and gait
• Protein of roughly 1.0–1.2 g/kg/day for healthy older adults and more in acute illness — above the 0.8 g/kg general recommendation, which is a common distractor. Note the tension: the ADA 2026 older-adults standards say only “at least 0.8 g/kg/day, potentially higher and individualised”, and state there is no current evidence that going beyond 0.8–1.5 g/kg improves glycaemic or cardiovascular outcomes. The geriatric nutrition societies are more assertive than the diabetes one; answer to whichever the stem names
• Deprescribing, correction of vitamin D deficiency, treating depression, and addressing vision, hearing and dentition
• Comprehensive geriatric assessment is the intervention with the best evidence at the system level — it increases the chance of being alive and in your own home
• Testosterone, growth hormone and appetite stimulants: not recommended. Megestrol and dronabinol are Beers items

Frailty as a decision tool

It changes what you offer, not merely what you predict: surgical risk, chemotherapy tolerance, dialysis initiation, anticoagulation, intensity of glycaemic and blood-pressure control, and where the time-to-benefit argument bites. STOPP B16 encodes one version of it, and all three clauses are in the criterion: “Statins for primary cardiovascular prevention in persons aged ≥ 85 and established frailty with expected life expectancy likely less than 3 years”. Age alone is not the rule — a fit 88-year-old on a primary-prevention statin does not trip B16.

The comprehensive geriatric assessment

Frailty is the problem; CGA is the instrument with the best evidence against it. It is a multidimensional, multidisciplinary process producing a coordinated plan, and in hospitalised older people it increases the chance of being alive and in your own home at follow-up. Know the domains, because a stem describing a “thorough history” that omits three of them is testing exactly this.

Weight loss and the anorexia of ageing

Unintentional loss of 5% in a month or 10% in six is significant and is the entry point of the Fried phenotype. Screen with MNA-SF. Before investigating a cancer, work through the mechanisms that are commoner and fixable: drugs (digoxin, metformin, SSRIs, opioids, anticholinergic dry mouth), dentition and swallowing, depression, dementia at the stage where eating is forgotten, constipation, pain, thyroid disease, and the practical questions — who buys the food, who cooks it, can the patient reach the kitchen and afford the shopping.

Oral nutritional supplements increase intake and weight; they do not reliably change function or mortality, and they work best given between meals, not with them. Appetite stimulants are not recommended — megestrol and dronabinol are Beers items. Protein and resistance exercise together beat either alone.

The deficit index, and its ceiling

The deficit index is a proportion — deficits present divided by deficits counted — so it is the fraction that matters, not which particular deficits a person has. Three properties worth knowing: it rises non-linearly with age in a Gompertz-like curve, it runs higher in women, and it has a universal ceiling at about 0.7. Nobody is observed above it, which is read as the limit of deficits a human can accumulate and remain alive. It correlates strongly with mortality and is unwieldy at the bedside, which is why the Clinical Frailty Scale is what actually gets recorded.

The Fried criteria, operationalised

Table 42-2 is the single most-cited table in the recent papers — four times across four sittings — which means the committee treats it as the definition. It is worth knowing not just as five words but as the five measurements the Cardiovascular Health Study actually used.

Three or more is frail; one or two is pre-frail; none is robust. Notice what the cut-points are: population quintiles adjusted for sex, and for BMI or height — not absolute numbers. That is why a grip strength in kilograms is not by itself a Fried criterion, and it is the sort of detail a table-cited question is built on.

What the chapter says is wrong with it

• Cognition is excluded, which sits badly against the clinical experience that cognitive impairment usually travels with frailty. The subgroup with dual decline — mobility and cognition together — is at very high risk of disability, and the phenotype does not capture them
• Weight loss is a problem as a criterion as obesity becomes commoner: sarcopenic obesity is frailty with a rising weight, and the phenotype misses it
• The phenotype is deliberately physical, so it excludes disability and comorbidity in order to predict them — a strength for research and a limitation at the bedside, which is why the Clinical Frailty Scale is what actually gets recorded

What the chapter will not claim

Three layers, not one thing

The chapter frames frailty as a construct with three overlaying dimensions, like the layers of an onion. It is worth holding because it explains why the definitions disagree: each of them is measuring a different layer.

So the Fried phenotype samples the outer layer and nothing else; the deficit index samples the outer layer more broadly; and everything being researched — inflammatory markers, anabolic hormones, mitochondrial function, senescence — belongs to the layers underneath, which is why no blood test diagnoses frailty and why none is likely to soon.

Hazzard ch 57, p853 — 8 questions across four sittings. The chapter before dementia, and the one that decides whether the patient in the stem has a disease at all.

Normal ageing

Subjective decline, MCI, dementia

• Subjective cognitive decline — the complaint without objective impairment. It carries some risk, and the patients who complain and test normally worry more than those brought in by a worried family, who test worse
• Mild cognitive impairment — objective decline in one or more domains, independence preserved. Amnestic subtypes more often progress to Alzheimer disease, non-amnestic to the other pathologies. Progression runs at roughly 10–15% a year in clinic populations, and a meaningful minority revert to normal — usually those whose deficit was drugs, depression, sleep or thyroid all along
• Major neurocognitive disorder — the decline interferes with independence in everyday activities. Not the score; the interference. Neither is diagnosed during delirium

The instruments, and what they are for

These thresholds are the ones in common use rather than diagnostic criteria. Education, language and sensory impairment shift all of them, and testing a patient without their hearing aids produces a number about the hearing aids.

Normal ageing, split more finely than most notes split it

Ch 57 Table 57-1 makes two distinctions worth having, because both are the difference between a worried patient and a patient with disease:

• Executive function: real-world executive function is preserved; it is novel executive tasks that decline. The patient who still runs the household but fails a lab task they have never seen is showing normal ageing
• Visuospatial: simple construction and simple copying are preserved; mental rotation, complex copy and mental assembly decline

The chapter's framing of the whole table: normal ageing produces mild changes in circumscribed aspects of cognition, and those changes are much smaller than the effects of age-related medical conditions or neurodegenerative disease. If the change is not mild and not circumscribed, stop calling it ageing.

Before you label anything

B12, thyroid, sodium, calcium, depression, obstructive sleep apnoea, alcohol, subdural haematoma, normal-pressure hydrocephalus — and drugs, which is the one you can fix that day: anticholinergics, benzodiazepines, Z-drugs, opioids, antiepileptics. Check Beers Table 7 and STOPP Section M before ordering the MRI.

What to offer in MCI

No drug is approved or recommended for MCI — cholinesterase inhibitors do not prevent progression and carry their own harms, and that is a recurring stem. What is defensible: treat the reversible contributors, deprescribe, aerobic and resistance exercise, hearing and vision correction, vascular risk-factor control, alcohol reduction, social and cognitive engagement, and planned review. The Lancet Commission's modifiable risk factors — hearing loss, education, hypertension, smoking, obesity, depression, physical inactivity, diabetes, alcohol, head injury, air pollution, social isolation, and the 2024 additions of untreated vision loss and high LDL cholesterol — are cited by name in this exam and are worth holding as a list.

Numbers from ch 57 worth carrying

• Visual hallucinations with an MMSE above 20 are highly suggestive of dementia with Lewy bodies. Hallucinations that early, with cognition that intact, do not fit Alzheimer disease
• Post-operative cognitive dysfunction after cardiac surgery is reported in 30–70% at discharge and remains at 15–25% at twelve months
• Normal-pressure hydrocephalus is about 6% of dementia cases. After shunting, at six months 83% improved in gait and 46% in memory; among survivors at five years, 39% still had better gait and fewer than 10% still showed cognitive improvement. Gait is what shunting buys, and it buys it far more reliably than cognition
• Behavioural-variant FTD is 50–60% of all FTLD diagnoses
• Late-onset depression may itself be a prodromal symptom of neurodegenerative disease, not merely a mimic of it or a reaction to it

The rarer dementias the exam has drawn from ch 57

• Prion disease (Creutzfeldt–Jakob the best known): profound, rapidly progressive dementia, sporadic cases in the 60s, early symptoms vague — poor memory and concentration, apathy, lability, sleep and appetite loss — with hyperreflexia, incoordination, saccadic abnormalities and incontinence as it advances over months; EEG changes, MRI hyperintensities and CSF 14-3-3 protein. The 2023 examination finding was the hyperreflexia; the 2024 stem (a retired engineer, months, hyperreflexia, MRI hyperintensity) wanted palliative referral, because diagnosis comes late and the course is fast
• Corticobasal syndrome: tau pathology, asymmetric frontoparietal atrophy, onset in the early 60s; progressive asymmetric tremor, incoordination, rigidity, myoclonus, alien limb and unilateral ideomotor apraxia, with memory relatively preserved early — the 2025 answer for asymmetric tremor plus apraxia
• Neurosyphilis: dementia 5–25 years after infection with hallucinations, delusions, mood change, personality change, strokes, ataxia and global decline; classed as reversible but with only limited cognitive benefit from penicillin; think of it with HIV and in any dementia of unclear cause (2025)
• Normal ageing (Table 57-1): processing speed, new learning and retrieval, word-finding and fluency all decline; syntax, vocabulary and semantic knowledge are preserved — the 2024 “which is not a normal change” answer was the grammatical ability

Driving

A dementia diagnosis is not by itself a bar to driving, and early Alzheimer disease with mild impairment may be compatible with it; moderate or severe impairment is not. What predicts risk better than any score is the informant's account, a history of near-misses, getting lost on familiar routes, and executive and visuospatial deficit rather than memory. Practical tools: clock drawing and Trail Making B as flags, then an on-road assessment, which is the only real test.

Telling the patient

Disclosure is the default: a person with capacity has a right to their diagnosis, and hiding it removes their chance to plan while they still can — an enduring power of attorney (ייפוי כוח מתמשך), advance directives, finances, driving, and the conversation about where they want to live. When family ask you not to tell, the correct move is to ask the patient how much they want to know and whom they want told, which respects autonomy without imposing disclosure. Pair the diagnosis with what happens next — follow-up, support services, and the reversible contributors you are still treating.

What counts as abnormal, and what MCI actually does next

• On formal neuropsychological testing, scores 1 to 1.5 standard deviations below the mean for age- and education-matched peers are treated as abnormal — though the chapter notes there is no universally agreed cut-off. Age and education matching is the part that matters: a raw score means nothing without them
• 80% of people diagnosed with MCI go on to develop Alzheimer disease within 5 to 8 years
• They convert at roughly 10–15% a year, against a general-population conversion rate of 1–2%. That contrast is the reason the diagnosis is worth making
• But 15–40% of those diagnosed with MCI later test in the normal range — reversion is inherent to the construct, not a sign someone got it wrong
• Alzheimer pathology is present years to decades before symptoms are noticed, which is what makes the prodromal categories meaningful at all

The fourteen risk factors, with the numbers

The Lancet standing Commission's 2024 report adds untreated vision loss and high LDL cholesterol to the twelve named in 2020. Together the fourteen carry a weighted population attributable fraction of 45% — nearly half of dementia worldwide could in theory be prevented or delayed. These are the Commission's own weighted figures, which is what a question citing this paper is likely to be built on.

The Commission groups the fourteen by life stage, and a stem can ask which stage a factor belongs to: one early-life factor (education), ten midlife, and three late-life (social isolation, air pollution, untreated vision loss). The prevalence column is the other half of the arithmetic — PAF is driven by it, not by the relative risk.

What the Commission actually tells you to do

• Make hearing aids accessible; reduce harmful noise exposure
• Keep systolic blood pressure at or below 130 mmHg from the age of 40
• Detect and treat high LDL cholesterol from midlife
• Treat depression effectively; screen and treat vision loss
• Helmets in contact sports and on bicycles; exercise; smoking cessation; weight and diabetes control; reduce heavy drinking; reduce air pollution
• Age-friendly housing and communities, and reduce social isolation by facilitating living with others and joining in

Because risk clusters in the same individuals, the Commission argues that interventions should usually be multicomponent rather than single-factor, and should start early and continue across life. After diagnosis it is equally explicit: cholinesterase inhibitors and memantine for Alzheimer disease and Lewy body dementia; multicomponent coping interventions for family carers; activity interventions for neuropsychiatric symptoms — and no evidence for exercise as a treatment for neuropsychiatric symptoms, which is the trap in that sentence.

Hazzard ch 45, p665 — 5 questions across three of four sittings. In an older patient the presentation is usually a fall, and the amnesia for the event is why the distinction matters.

The three that matter

The physiology behind the stems: standing pools 500–1000 mL in the legs and splanchnic bed; the fall in pressure unloads the carotid and aortic baroreceptors, whose reflex raises sympathetic outflow, cuts vagal tone, and lifts heart rate, resistance and venous return. Ageing blunts baroreflex sensitivity — the 2026 “normal autonomic ageing” answer — and a disrupted baroreflex causes OH. Postprandial hypotension is splanchnic pooling with a failure to raise sympathetic activity, heart rate and resistance, plus vasodilatory gut peptides — asked in 2024 and 2026.

Definitions worth being exact about

• Orthostatic hypotension: a fall of 20 mmHg systolic or 10 mmHg diastolic within three minutes of standing. Measure supine after five minutes' rest, then at 1 and 3 minutes standing
• Postprandial hypotension: a 20 mmHg systolic fall within two hours of the start of a meal. Prevalence 24–36% of nursing-home residents, 23% of older adults admitted with syncope or falls, and 50% of those with unexplained syncope — usually missed because nobody measures after lunch. It may or may not coexist with OH
• Carotid sinus hypersensitivity: on 5–10 seconds of carotid sinus massage, asystole of 3 seconds or more (cardioinhibitory), a systolic fall of more than 50 mmHg (vasodepressor), or both (mixed) — with or without symptoms. When massage reproduces syncope it is carotid sinus syndrome, and that is the distinction the chapter draws. Most susceptible: coronary disease, hypertension, digoxin, beta-blockers, methyldopa; a third also have vasovagal syncope or OH. Do not massage over a bruit or soon after stroke or infarction. Dual-chamber pacing abolishes syncope in 85–90% of cardioinhibitory cases; midodrine helped the vasodepressor form in one small trial

The workup that is actually indicated

History from a witness, examination including lying and standing pressures, and a 12-lead ECG answer most of it. Add echocardiography where structural disease is suspected, prolonged monitoring or an implantable loop recorder where an arrhythmia is suspected but intermittent, and carotid sinus massage in unexplained syncope or falls in the older patient. Tilt testing is for recurrent unexplained episodes, not first-line. Routine head CT, carotid Dopplers, EEG and blood panels have a very low yield in undifferentiated syncope and are a favourite wrong answer.

Treatment of orthostatic hypotension, in order

1. Withdraw the offending drug — antihypertensives, alpha blockers, nitrates, diuretics, tricyclics, antipsychotics, dopaminergics
2. Non-pharmacological: squat or cross the legs when symptoms start, hydration and salt, waist-high graded stockings or an abdominal binder, sleep 30–45° head-up, avoid heat, straining and alcohol. For the postprandial form: antihypertensives between meals rather than with them, no diuretics or prolonged sitting before eating, small frequent meals high in protein and fat, cold rather than warm, and a walk afterwards
3. Then, for severe OH or PPH, the chapter's four: caffeine 250 mg (two brewed cups) in the morning, fludrocortisone 0.1–0.5 mg daily, midodrine 2.5–10 mg three times daily, and octreotide 50 µg subcutaneously 30 minutes before meals — watching supine hypertension, oedema and heart failure. Droxidopa is not in ch 45 — its source is Kaufmann et al., Neurology 2014;83:328 (NOH301, read from the paper; 162 responders randomised for one week): OHQ composite improved 0.90 units over placebo, standing systolic up 11.2 vs 3.9 mmHg, supine systolic over 180 in 4.9% vs 2.5%, headache the commonest adverse effect

Driving: an episode of syncope has licensing consequences under the Ministry of Health circular on fitness to drive (כשירות לנהיגה), which the law block examines. The reporting duty and the thresholds are in Ethics & EOL law.

Dizziness is four complaints, not one

The word means whatever the patient means by it, and the first job is to find out which. In older patients the causes are frequently multiple, which is itself the answer to some stems.

• BPPV is the commonest and the most rewarding: seconds of vertigo on turning in bed or looking up. Dix–Hallpike provokes it — upbeating torsional nystagmus with a latency of 1–5 seconds, a burst of 10–20 seconds, and fatigability — and the Epley manoeuvre treats it, repeated until the Dix–Hallpike is negative. Duration sorts the episodic causes: under a minute BPPV; 20–120 minutes TIA or migraine; 2 hours to 2 days Ménière or recurrent vestibulopathy. Vestibular suppressants — meclizine, scopolamine, benzodiazepines — are dangerous in the old, cause falls and confusion, and with prolonged use paradoxically prolong the dizziness by blocking central compensation; low dose, acute episodes only
• Acute vestibular syndrome — continuous vertigo over days — is where you separate neuritis from stroke. The HINTS examination (not in ch 45; Kattah et al., Stroke 2009;40:3504, read from the paper — 101 high-risk patients, 76 central: a normal head impulse, direction-changing nystagmus or skew was 100% sensitive and 96% specific for stroke, and early DWI-MRI was falsely negative in 12%, all under 48 hours) does it better than early imaging: a normal head-impulse test, direction-changing nystagmus, or skew deviation each point central. Remember that a posterior-circulation stroke can present with dizziness and almost nothing else, and that an early CT will be normal

Hazzard ch 51, p759 — 5 questions across three of four sittings. The examinable content is who to treat, with what, and when to stop — and the one drug you can never simply stop.

Who gets treated

• A hip or vertebral fragility fracture — treat, whatever the T-score. No DXA required to make the decision
• T-score −2.5 or below at femoral neck, total hip or lumbar spine
• Osteopenia with a FRAX 10-year probability above the national intervention threshold

T-score compares with a young adult; Z-score compares with age-matched peers and a Z below −2.0 should send you looking for a secondary cause. Degenerative change, aortic calcification and vertebral fracture all falsely raise the lumbar reading in the old, which is why the hip site is preferred.

Secondary causes to exclude first

Vitamin D deficiency, hyperparathyroidism, coeliac disease, myeloma, hyperthyroidism and over-replacement, hypogonadism, CKD, and drugs — glucocorticoids above all, plus aromatase inhibitors, androgen deprivation, anticonvulsants, long-term heparin, and the proton pump inhibitors and SSRIs whose association is weaker but real.

The drugs

Duration and the holiday

The chapter's rule: a holiday may be considered after 3 to 5 years of antiresorptive therapy in moderate or low fracture risk, because a short break costs little. In high-risk patients — prior hip or vertebral fracture, T-score still very low, continuing glucocorticoids — benefit continues beyond 5 years and the timing needs care. The evidence behind it: in pooled extension trials, 6 or more years of bisphosphonate gave fracture rates of 9.3–11% against 8.0–8.8% on placebo. The holiday rests on the skeletal retention of bisphosphonates and it does not apply to denosumab, which is the trap. The oral-five / IV-three split is a common convention, not ch 51.

The rare harms, in proportion

Atypical femoral fracture — risk rises with duration of any bisphosphonate (and of denosumab), and falls rapidly once it is stopped. Radiograph any groin or thigh pain on treatment, then bone scan or MRI if needed; the chapter notes that the fracture precedes the pain in most cases. Osteonecrosis of the jaw — exposed necrotic maxillofacial bone that fails to heal; at osteoporosis doses the incidence is 0.001–0.01%, against the frequent high doses of oncology. Both are rare enough that fear of them causes more fractures than the drugs do, and saying so in proportion is usually the answer being tested.

Vertebral fracture

Most are silent: the presentation is height loss, progressive kyphosis, and chronic back pain rather than an event, and the diagnosis is often made on a chest film. They matter because they predict the next one — half of patients refracture within 10 years, and half of those within 2 years of the first, a risk equal to that of a woman 10 years older — which is what “imminent risk” means and why the chapter says most older patients with a prior fracture are treatment candidates now, not after the DXA.

• Management is analgesia, early mobilisation, and starting bone treatment. Bed rest costs more than it relieves
• Vertebroplasty and kyphoplasty did not beat sham in the better trials and are not recommended routinely — a favourite question, though ch 51 does not discuss them
• Suspect something other than osteoporosis if there is fever, a known cancer, neurological signs, thoracic pain in a young-old patient, or a fracture above T4

Calcium and vitamin D, in proportion

Ch 51 holds both halves of this at once, and the exam can ask either. On one side: calcium and vitamin D are “an essential component of any treatment”, recommended for all, and every drug trial supplemented both. On the other: calcium from diet or supplements produced only small BMD gains; vitamin D supplements have no clear effect on fracture; and high-dose bolus vitamin D — 500,000 IU yearly or 60,000 IU monthly — consistently increased falls and fractures. Dietary calcium is associated with lower all-cause mortality; very high supplemental calcium carries a possible cardiovascular signal. So: diet first, daily doses in the ranges above, measure the level in the frail, and never the bolus.

Hazzard ch 87, p1353 — 7 questions across three of four sittings. A cheap chapter with two expensive traps: fibre in the wrong patient, and diarrhoea that is actually impaction.

Define it properly

Not frequency alone. Rome-type criteria require two or more of: straining, lumpy or hard stools, sensation of incomplete evacuation, sensation of anorectal blockage, manual manoeuvres, and fewer than three spontaneous bowel movements a week — in at least a quarter of defecations. Colonic transit changes little with age; what changes is mobility, fluid, drugs, and the pelvic floor.

Red flags before treating

New onset in someone previously regular, weight loss, rectal bleeding or iron-deficiency anaemia, a change in stool calibre, a family history of colorectal cancer, or an abdominal or rectal mass. These earn a colonoscopy, weighed against life expectancy. Everyone gets a rectal examination — tone, impaction, mass, and the paradoxical contraction of dyssynergia.

Secondary causes

How common, and how slow the gut actually gets

• Prevalence 12–19% in North America, rising with age. Asked directly about recurrent constipation: 26% of women and 16% of men, and at 84 and over, 34% and 26%
• Total gut transit is normally under 72 hours. In ambulant older patients it runs to about 3 days; in bedridden patients, 3 weeks; and in institutionalised or bedridden patients with constipation, 6 to more than 14 days. In immobile older people with diabetes it reaches 200 ± 144 hours
• Just one week of bed rest slows transit in healthy young volunteers — the point being that immobility alone does this, before any drug or disease is involved

The chapter's diagnostic definition is stricter than the symptom list alone: two or more of the symptoms, present on more than 25% of occasions, for at least 12 weeks in the last 12 months. Hard stool means Bristol type 1–2. Normal perineal descent is under 4 cm.

Treatment, in the order it is examined

1. Non-drug: mobility, fluid, a toilet routine exploiting the gastrocolic reflex after breakfast, footstool for posture, privacy and time. Removing the commode-behind-a-curtain problem is a real intervention
2. Osmotic: polyethylene glycol carries the chapter's only level 1 evidence and out-performs lactulose on frequency and straining, at the cost of more liquid stools — but ch 87 places it, for older people, at acute disimpaction (1 L or 8 sachets daily for 3 days cleared 89% of impactions) and at regular use only in high-risk patients resistant to milder, cheaper agents. Lactulose and sorbitol are equally effective and sorbitol is much cheaper; that is the exam answer on cost. Docusate has no effect on motility and little or no laxative action even at 300 mg/day
3. Stimulant: senna or bisacodyl, alone or added. Senna acts in 8–12 hours, so at bedtime; 20 mg daily for six months in the over-80s lost no protein or electrolytes, and in patients with weak sphincters senna alone may treat the constipation without the soiling that softeners cause. Bisacodyl 10 mg daily is the alternative; bisacodyl carries an ischaemic-colitis risk in the old. Long-term senna is not known to damage the bowel — the “lazy colon” is folklore and a common distractor
4. Secretagogues — lubiprostone, linaclotide, prucalopride — for refractory chronic constipation
5. Opioid-induced constipation that fails laxatives: a peripherally acting mu-opioid receptor antagonist (naloxegol, naldemedine, methylnaltrexone). This is the specific answer, and every regular opioid needs a laxative from day one — STOPP L2

Opioid-induced constipation, by the numbers

The PAMORA numbers from ch 87: naldemedine raised the response rate to 52% against 35% on placebo (RR 1.51, 95% CI 1.32–1.72); naloxegol to 42% against 29% (RR 1.43, 1.19–1.71), response defined as three or more spontaneous bowel movements a week. Methylnaltrexone improved frequency across five trials but the evidence was graded low quality. For refractory constipation the chapter offers transanal irrigation, with a pooled response rate of about 50%.

Three mechanism answers from ch 87. Transit slows because of a loss of colonic enteric neurons producing nitric oxide and acetylcholine (2024). Rectal dysmotility — the loaded rectum that cannot contract, elicitable in only 14% with a history of impaction — is attributed to diminished parasympathetic outflow from sacral cord damage, or to years of suppressing the urge (2026). And acute colonic pseudo-obstruction (Ogilvie), the caecum at 10 cm or more in the frail, acutely ill or postoperative patient: nil by mouth, flatus tube, correct potassium and magnesium, then neostigmine if the heart allows, then endoscopic decompression, with PEG afterwards to prevent recurrence (2025).

Faecal incontinence, briefly

It is on the official exam topic list with no frequency behind it. Commonest mechanism in the old is overflow from impaction; then sphincter and pudendal injury, diarrhoeal states, cognitive and mobility causes. Treat impaction, firm the stool, schedule toileting, and consider biofeedback where there is a functioning sphincter and cognition to use it.

Hazzard's stepwise ladder, which does not start with PEG

Note what the ladder does with a bulking agent: it is first in the ambulant patient and skipped entirely where fluid intake is doubtful — which is the same judgement the file makes elsewhere about fibre in the immobile, arrived at from the other side. Targets from the chapter: fibre 20–35 g/day, starting at 3–4 g and building; fluid 1.5–2 L/day.

Dyssynergic defecation

A quarter or more of refractory constipation is not slow transit at all but a failure to coordinate the pelvic floor: the patient strains, the puborectalis and external sphincter contract instead of relaxing, and no laxative will fix it. You detect it with your finger — ask the patient to bear down and feel for paradoxical contraction and inadequate perineal descent — and confirm it with anorectal manometry and a balloon-expulsion test. Biofeedback is first-line and outperforms laxatives in this group, which is a question that rewards knowing the mechanism rather than the drug ladder.

Bowel care in the frail and the dying

• The regimen follows the opioid, not the complaint: start the laxative with the opioid, and titrate it with the opioid
• Osmotic plus stimulant is the usual combination; bulk-forming agents are wrong here
• Rectal measures — suppository or enema — when the rectum is loaded, and manual disimpaction when it is hard and the patient cannot evacuate. Distressing, and more distressing if it is delayed for three days
• Constipation is a common and reversible cause of delirium, of agitation in advanced dementia, of urinary retention and of overflow incontinence. It belongs on every one of those checklists

Hazzard ch 99, p1559 — 12 questions across three of four sittings, the highest count outside the core nine, and the exam also cites the ADA older-adults standards by name. The whole chapter is one argument: in this population the danger is overtreatment.

Targets by health status

A1c is also less reliable here: anaemia, transfusion, CKD and altered red-cell survival all distort it. Where it is doubtful, use glucose patterns or continuous monitoring.

Drug choice

Simplify insulin: once-daily basal rather than premixed or multiple daily injections where a carer administers it; match the regimen to who is actually giving it and to whether meals are eaten.

Complications, weighted for age

Albuminuria (over 30 mg/g creatinine) is the first clinical sign of diabetic kidney disease, a marker of cardiovascular risk, and — the 2024 committee answer — a prognostic tool and a measure of response to treatment; a raised creatinine means the damage is already irreversible. Confirm it more than once. Combined albuminuria and reduced GFR is far commoner in the old, and about half of older people with diabetes have kidney disease.

Screen feet and vision, treat blood pressure and lipids with the same time-to-benefit reasoning used everywhere else, and remember the geriatric complications the chapter singles out: falls, cognitive impairment, depression, incontinence, persistent pain, and polypharmacy. Diabetes doubles the risk of cognitive decline, and cognitive impairment is itself a reason to loosen targets and simplify the regimen.

Ageing physiology the exam keeps asking (ch 99): in people without diabetes there is a slight rise in fasting glucose and a much more marked slowing of the return to normal after an oral glucose load (2026); insulin sensitivity falls, and the rise in central adiposity accounts for most of that insulin resistance, with inactivity adding to it (2024). Cardiovascular protection: statins have a prognostic benefit within a few years and are recommended for most (2024); the SGLT2 inhibitor is the change with a prognostic effect in the patient with coronary disease, albuminuria and an eGFR of 40 (2024 stem: start empagliflozin, not stop the statin, not more insulin). Prediabetes: follow-up screening every 2 to 3 years, yearly in the high-risk; lifestyle change beats metformin at delaying progression, including over 60 (DPP). In the nursing-home patient with limited life expectancy, a restrictive low-salt, low-cholesterol diet is a cause of weight loss — liberalise the diet and prioritise appetite and enjoyment (2025).

Why hypoglycaemia recovery fails in this population

Table 99-15 lists the risk factors, and they are worth reading as four groups rather than one list, because the therapeutic group is the only one you prescribe:

• Therapeutic: insulin, sulphonylurea
• Physiologic: impaired autonomic function, diminished glucagon secretion, kidney or liver failure
• Functional and cognitive: cognitive disorder, sensory impairment (vision, hearing), impaired mobility or hand dexterity
• Behavioural and other: poor or irregular nutrition, alcohol or another sedating agent, polypharmacy

The counter-regulatory failure is the part to understand rather than memorise: the catecholamine response that should raise glucose rapidly is blunted — by autonomic neuropathy from long-standing diabetes, and by the β-blockers so many of these patients take for cardiovascular disease. The drug that masks the warning also impairs the recovery.

The threshold shifts, which is why they feel it early

A patient with diabetes may begin to feel hypoglycaemic at a glucose of 80–100 mg/dL — well above the symptomatic threshold in people without diabetes, which is 50–60 mg/dL. Long-standing hyperglycaemia resets the alarm upward, so a patient reporting symptoms at a normal glucose is not imagining it, and a patient comfortable at 55 has lost the alarm altogether.

Hypoglycaemia is the outcome that matters

• Under 70 mg/dL (3.9 mmol/L) is the alert value; under 54 is clinically significant; severe is defined by needing another person's help, whatever the number
• Autonomic warning symptoms are blunted by age, by duration of disease and by beta-blockers, so the presentation is confusion, a fall, a seizure or a collapse — and it is routinely admitted as delirium or stroke
• Sulfonylurea and insulin are the only common causes. A long-acting sulfonylurea in renal impairment can produce hypoglycaemia lasting a day or more, so those patients are admitted and observed, not treated and sent home
• Treat with 15 g of fast carbohydrate, recheck at 15 minutes, then give a longer-acting snack. Intramuscular glucagon works poorly in the malnourished, in liver disease and in sulfonylurea-induced hypoglycaemia
• One episode is a reason to loosen the target and simplify the regimen, not to add a monitoring instruction

How common it is, and what it brings with it

• Among people 65 and over in the US, total diabetes prevalence is about 27% and prediabetes nearly 50% — both higher than in younger groups
• About 20% of people who meet the criteria are undiagnosed
• Cardiovascular burden rises steeply with age: of US adults with diabetes reporting heart disease or stroke, 28% of those aged 35–64, 43% at 65–74 and 55% at 75 or over
• Diabetic kidney disease is the leading cause of CKD in the United States

Diagnosis, and the drugs that fake it

Ch 99 Table 99-1 gives three routes to the diagnosis, and they apply regardless of age — there is no separate older-adult threshold.

Fasting means no caloric intake for at least eight hours, and any criterion should be confirmed on a separate day unless the patient is frankly symptomatic. About 90% of older adults with diabetes have type 2, where asymptomatic hyperglycaemia is the hallmark and ketoacidosis is not part of the syndrome.

Drugs that raise glucose and can manufacture a diagnosis (Table 99-2): diuretics, α-adrenergic agonists, β-blockers, alcohol, calcium-channel blockers, clozapine — alongside the corticosteroids everyone remembers.

Hazzard's own framework, which is not the ADA's

Table 99-6 sorts patients into three health groups and gives glucose ranges for each, with a second, higher set of targets where sulphonylureas or insulin are in use — a distinction the ADA table does not make, and the sort of thing a chapter-referenced stem is built on.

Note the lower bounds in the right-hand column. On a sulphonylurea or insulin the chapter sets a floor as well as a ceiling — an A1c of 6.8% in group 2 is a target breached downward, which is the point the ADA makes in prose and Hazzard makes in a number.

The ADA targets, exactly as written

The exam cites this section by name. The 2026 edition numbers the recommendations, and the table below is its Table 13.2 reduced to what gets examined.

Glucose in mg/dL. “Multiple” chronic illnesses means at least three. Statin unless contraindicated in the first two rows; in the third, consider the likelihood of benefit — the time to benefit for a statin is 2.5 years, which is the number that decides it. A1c is unreliable after transfusion and in conditions altering red-cell turnover, which in this population means kidney failure, recent blood loss and erythropoietin.

Prevention, and the exercise prescription

• In the Diabetes Prevention Program, caloric restriction plus exercise cut progression from prediabetes to type 2 diabetes despite only 5–7% weight reduction being achieved — the modest number is the point, because it is attainable
• The activity prescription, stated exactly: at least 150 minutes a week of moderate-intensity aerobic activity, spread over at least 3 days, with no more than 2 consecutive days without exercise, plus resistance training at least twice a week

Screening: the tools and their cut-points

Recommendation 13.2 is to screen at least annually for geriatric syndromes. Table 13.1 names the instrument and the number for each, and it doubles as a revision list for the whole syllabus.

Screening for mild cognitive impairment or dementia is recommended from age 65, at the first visit and annually — and again whenever self-care slips: insulin dose errors, skipped meals, failure to recognise a hypo.

Simplifying insulin, by the algorithm

• Move the basal dose from bedtime to morning, and titrate on fasting glucose over a week to a target of 90–150 mg/dL: up by 2 or more units if half the fasting values are above goal, down by 2 or more if more than two a week fall below 80
• Premixed insulin: take 70% of the total premixed dose and give it as basal only, in the morning
• Prandial above 10 units a dose: halve it and add a non-insulin agent, then titrate the prandial down as the other agent goes up. Prandial at 10 units or less: stop it and add a non-insulin agent
• Add metformin 500 mg daily if eGFR is 45 or above; otherwise go to a second-line agent chosen on comorbidity
• No rapid- or short-acting insulin at bedtime, and stop the correction scale once it is not needed daily

Where Hazzard and the ADA differ — answer to the stem's source

Neither is wrong; they are different documents written at different times. If the מראה מקום points at ch 99, answer 140/90. If it points at the ADA standards, answer 130/80. Knowing that the two disagree is worth more than knowing either number.

Doses the chapter states outright: metformin 500–1000 mg twice daily, maximum 2500 mg/day (ER 500–2000 mg once daily); low-dose aspirin 75–162 mg/day for secondary prevention after assessing bleeding risk; SGLT2 inhibitors need renal dose adjustment and are avoided below eGFR 25–30.

Drugs, with the 2026 caveats

• Metformin down to eGFR 30, reduced dose at 30–45, eGFR every 3–6 months if kidney function is falling. Check B12 periodically, particularly where there is anaemia or peripheral neuropathy — that is the ADA's wording; the four-year mark often quoted alongside it comes from observational data (DPPOS) rather than from the Standards. Watch the appetite suppression in someone already losing weight
• Sulfonylureas: glyburide avoided outright; glipizide preferred for its shorter action, though a recent analysis found the highest rate of major adverse cardiovascular events with glipizide. Antimicrobials — fluoroquinolones and co-trimoxazole above all — raise the effective dose and precipitate hypoglycaemia, so reduce or hold the sulfonylurea
• DPP-4 inhibitors: weak but safe; linagliptin needs no renal adjustment
• GLP-1 receptor agonists: not preferred where there is unexplained weight loss, undernutrition, problematic constipation, gastroparesis or prior bowel obstruction; monitor for dehydration and loss of muscle mass
• SGLT2 inhibitors: volume depletion in the frail or orthostatic, genital mycotic infection, more urinary frequency — ask about incontinence before and after starting — and euglycaemic ketoacidosis, most often triggered by infection in long-term care
• Pioglitazone: heart failure, fluid retention, weight gain, osteoporosis, macular oedema

In long-term care, and at the end of life

Ch 99's end-of-life goals (Table 99-16) add a target the ADA section leaves in prose: be respectful and supportive, match the level of intervention to the stage of illness, support self-management as long as possible, and aim for a painless and symptom-free death — while avoiding metabolic decompensation, that is hypoglycaemia, ketoacidosis, hyperosmolar hyperglycaemic state and persistent symptomatic hyperglycaemia. The chapter names a working glucose range rather than an A1c: 6–15 mmol/L, that is 108–270 mg/dL. It also names two goals that are easy to forget at this stage — avoiding foot complications in the frail, bed-bound patient, and avoiding symptomatic volume depletion.

The long-term-care alert thresholds above come from ADA 2026 §13, not from ch 99. Both documents cover this ground and a stem may cite either.

The alert thresholds are specified, and the reason they exist is that a resident may be reviewed only every 30 to 60 days. Call immediately for a glucose below 70 — but treat first. Call as soon as possible for values of 70–100, for two readings above 250 within 24 hours alongside a change in clinical status, for values consistently above 250 over 24 hours or above 300 over two days, for an unreadably high result, or for illness with vomiting, fever, lethargy or poor intake.

At the end of life the 2026 text sets out three states. The stable patient continues their plan with hypoglycaemia prevention as the priority and no role for A1c. The patient in organ failure has hypoglycaemia-causing agents reduced and dehydration prevented — in type 1 diabetes insulin is reduced as intake falls but never stopped. The dying patient with type 2 diabetes can reasonably have all agents discontinued; in type 1 there is no consensus, and a small amount of basal insulin may still be given to prevent ketoacidosis and symptoms. Withdrawal of statins in advanced life-limiting illness improved quality of life in a randomised trial.

Hazzard ch 98, p1545 — 5 questions across three of four sittings. Two examinable ideas: the reference range moves with age, and the subclinical states are decided by the number and the age together.

The reference range moves

TSH rises with normal ageing, so a TSH of 5–6 in an 85-year-old may be age-appropriate rather than disease. Treating it to a young adult's range is a recognised harm, and higher TSH in the very old has been associated with better, not worse, survival.

Subclinical hypothyroidism — raised TSH, normal free T4

• Prevalence about 10% at 65 and over, more in women. Repeat first: in 1 to 3 months if the TSH is 4.5–14.9, in 1 to 2 weeks if 15 or higher — recovery from silent thyroiditis or non-thyroidal illness explains many of them
• Figure 98-4, the chapter's rule at 65 and over: treat at a persistent TSH of 10 or higher, consider treating at 7.0–9.9, observe below 7. TSH of 10 or more predicts cardiovascular mortality, coronary disease and heart failure independent of age; 7 or more predicts cardiovascular mortality. It is not associated with fractures, dementia or depression, and one study found lower mortality in older people with it. The randomised trial in the over-60s showed no benefit in fatigue or symptoms, and no trial has clinical endpoints
• Start at 25 or 50 µg — the smallest tablets, and the chapter's stated starting dose; “start low, go slow” in cardiac disease is convention never tested in a trial. Recheck TSH at 7 to 8 weeks in the old (6 elsewhere), increments of 25 µg early and 12 µg above 75–100 µg. Target TSH toward the upper limit of normal, as high as 4–6 mU/L over 80 — the 2025 committee answer. Subclinical replacement averages 1 µg/kg/day; full replacement in the old is below the young adult's 1.6. No T3, no desiccated thyroid

Absorption: take it fasting, separated from calcium, iron, proton pump inhibitors and soya. A rising dose requirement usually means an interaction or adherence, not progression — Table 98-7 groups the culprits as drugs that reduce absorption (calcium carbonate, iron, PPIs), drugs that increase metabolism (carbamazepine, phenytoin, rifampicin), and conditions that raise TBG; the 2024 answer was “all of the above”.

Subclinical hyperthyroidism — suppressed TSH, normal free hormones

Prevalence 1–6% on a single test and it persists in more than half on repeat at 3 to 6 months. Ch 98's position is unqualified: treat older people with subclinical hyperthyroidism to laboratory euthyroidism, because it raises atrial fibrillation, heart failure, overall cardiovascular mortality, fracture and dementia. The 0.1 versus 0.1–0.4 split is guideline convention, not the chapter. The causes are those of overt disease: autonomous nodules in a nodular gland (commoner with age, and the one an iodine load unmasks), Graves, over-replacement, amiodarone. Methimazole is preferred over PTU for half-life and safety; agranulocytosis in 0.1–0.5% with either; radioiodine is definitive for an autonomous nodule and often leaves the rest of the gland euthyroid.

Presentation is atypical, and that is the question

New atrial fibrillation, unexplained weight loss or a change in cognition all justify a TSH. Myxoedema coma — hypothermia, hyponatraemia, hypoventilation, bradycardia, depressed consciousness, usually precipitated by infection, cold, alcohol, or narcotics, sedatives and antipsychotics — is treated by drawing a cortisol, giving intravenous hydrocortisone first (thyroid hormone without it can precipitate adrenal crisis; stop it once cortisol is adequate), then levothyroxine 200–400 µg followed by 75–100 µg daily. T3 is discouraged in the old for cardiac risk, and sedatives are avoided because every drug is cleared slowly. The committee's 2020 answer was the hydrocortisone. Hypoglycaemia and a raised CK are part of the picture.

Do not screen thyroid function during acute illness if it can wait: non-thyroidal illness syndrome produces low T3, sometimes low T4 and a variably low TSH, and it recovers with the illness. Thyroid nodules are common with age and most are benign; ultrasound and selective fine-needle aspiration, with less aggression as life expectancy shortens.

Amiodarone, which is a separate disease

Ch 98 says the two cannot be told apart at presentation — radioiodine scanning is blocked by the drug's own iodide (37% by weight) — so patients are often started on both corticosteroid and methimazole. The Doppler-flow distinction is guideline teaching, not the chapter — the 2018 European Thyroid Association guideline (Bartalena et al., Eur Thyroid J 2018;7:55, read from the paper), which also says glucocorticoids can be added from the start when the type is uncertain and that stopping amiodarone is a joint cardiology–endocrine decision. Expect a benign transient TSH rise and a chronically higher free T4 with lower T3 that need no treatment — the committee's 2021 answer was “common side effect, observe”. In most cases amiodarone is not stopped: the effects take months to wear off and thyroid dysfunction is managed alongside it. Hypothyroidism arises in those with Hashimoto disease and is often reversible on cessation.

Nodules, and how hard to look

Ultrasound documents size in three dimensions and the high-risk features: hypoechogenicity, irregular margins, taller-than-wide shape, intranodular vascularity and microcalcifications (the 2025 answer); spongiform or mixed cystic–solid appearance is low risk. Lymph nodes are assessed and a TSH drawn.

Thyroid nodules become the rule rather than the exception with age and the great majority are benign. Check TSH first: a suppressed TSH points to an autonomously functioning nodule, which is scanned rather than biopsied. Otherwise ultrasound decides who needs fine-needle aspiration, by size and by sonographic pattern. Anaplastic carcinoma is a disease of the old — a rapidly enlarging, hard, fixed neck mass with hoarseness or stridor is an emergency, not a routine referral. Against that, a small papillary carcinoma in an 88-year-old may reasonably be watched: the same life-expectancy reasoning that governs screening governs this too.

Hazzard ch 81, p1235 — 5 questions across three of four sittings. In this population the examinable content is less about the drugs than about diagnosis at the margins and about whether the patient can physically use the device.

Diagnosis, with the age caveat

Post-bronchodilator FEV₁/FVC below 0.70 confirms persistent airflow limitation. But that fixed ratio overdiagnoses in the old, because the ratio falls with normal ageing and a ratio under 0.70 is frequent in healthy never-smokers over 65; ch 81 calls this GOLD's first flaw, and expressing FEV₁ as percent-predicted its second. The lower limit of normal — the fifth percentile, or a Z-score — halves the prevalence: 37.7% by GOLD against 15.4% by Z-score in older persons, and the excess is not associated with worse outcomes. A breathless older patient without a smoking history is more often in heart failure, deconditioned, anaemic or obese — and the two commonly coexist, which is why an unexplained response to diuresis does not exclude COPD and vice versa.

Assessment and grouping

Grade symptoms with mMRC or CAT, then group by exacerbation history. Ch 81 was written to the older A–D scheme (it puts the eosinophil rule in “Group D”); the current GOLD scheme is A, B and E, with E for two or more moderate exacerbations or one hospitalisation in a year, whatever the symptom burden. Spirometric grade and symptom group are separate axes and questions exploit the confusion. The chapter adds its own screen: blood eosinophils and an α1-antitrypsin level at diagnosis.

Treatment

• Long-acting bronchodilators are the backbone — LAMA, LABA, or the combination. Inhaled anticholinergics are not a Beers problem; systemic ones are. Theophylline is third or fourth line at best — narrow window, arrhythmia and seizures, and ciprofloxacin raises the level (asked in 2024)
• Inhaled corticosteroids where exacerbations persist, especially with blood eosinophils above about 300, or with asthma overlap. Against that: pneumonia risk, oral candidiasis, bruising and bone loss, all of which matter more here
• Pulmonary rehabilitation is the intervention with the largest effect on dyspnoea, exercise capacity and quality of life, and it is under-referred at this age. After an exacerbation it also reduces readmission
• Long-term oxygen (Table 81-11, asked twice) for resting hypoxaemia — PaO₂ at or below 55 mmHg, or at or below 59 with cor pulmonale, right heart failure or a haematocrit above 55%. Worn more than 16 hours a day it improves survival in severe resting hypoxaemia. For isolated moderate resting or exertional desaturation an RCT found no benefit in lung function, exercise, quality of life, admissions or mortality — the chapter permits it only for exertional desaturation under 88% or PaO₂ under 55 and a demonstrated gain in dyspnoea or exercise capacity. Nocturnal-only oxygen without daytime hypoxaemia: unclear. Home NIV added to oxygen for persistent hypercapnia (PaCO₂ over 53) within four weeks of a hypercapnic exacerbation lengthens time to readmission and cuts death over the next year
• Vaccinate. START v3 names four and only four: seasonal influenza annually, pneumococcal, varicella-zoster, and SARS-CoV-2. RSV and a pertussis-containing booster are recommended for older adults by public-health bodies but are not START criteria — on a criteria-specific stem, picking them loses the mark
• Smoking cessation is the only thing that changes the slope of decline, at any age

The device is the intervention

Exacerbations and what follows

Intensified bronchodilator first, withholding the long-acting agents temporarily if the patient is on frequent nebulised doses. Systemic corticosteroid for severe exacerbations — the five-day oral course is GOLD, not the chapter; what ch 81 says is that delirium is more common above 60 mg daily. Antibiotics where sputum purulence or volume increases, especially in moderate or severe exacerbations, for 3 to 7 days in the absence of pneumonia or bronchiectasis. Sputum culture is not needed in a routine exacerbation — it cannot separate colonisers from pathogens — but it earns its place after failed outpatient treatment or with Pseudomonas risk: admission in the past 3 months, four or more antibiotic courses in a year, or severe obstruction (the 2025 answer for the winter admission). Expect resistant organisms (MRSA, Pseudomonas) in nursing-home residents, anyone in acute or subacute care within 90 days, and FEV₁ under 30%. Non-invasive ventilation for acute hypercapnia reduces intubation, length of stay, ventilator pneumonia and death, and is not withheld on age alone. Discharge means pulmonary rehabilitation referral, an oxygen and NIV decision, and a step-up in maintenance therapy. In advanced disease, low-dose opioids relieve refractory breathlessness, oxygen helps only the hypoxaemic, and a handheld fan genuinely works. Advance care planning belongs at the second admission, not the last one.

Breathlessness in an older patient is rarely one disease

COPD, heart failure, deconditioning, anaemia, obesity, pulmonary hypertension, thromboembolism and anxiety coexist, and treating one to the exclusion of the others is why patients stop improving. Two practical consequences: natriuretic peptides and echocardiography earn their place when the response to bronchodilators is poor, and a six-minute walk or a sit-to-stand test tells you more about function than any spirometric number.

The company COPD keeps

• Cardiovascular disease — the commonest cause of death in mild-to-moderate COPD. Cardioselective beta-blockers are not contraindicated and are systematically withheld
• Muscle wasting and weight loss — independent predictors of mortality; resistance training and nutrition, not just inhalers
• Osteoporosis — from steroids, smoking, inactivity and low body weight
• Anxiety and depression — present in a large minority, under-treated, and a major driver of readmission
• Lung cancer — the risk is concentrated in exactly this group, and screening decisions turn on life expectancy and fitness for treatment rather than age

Advance care planning belongs in the clinic during a good spell: what the patient wants if ventilation is needed, whether non-invasive ventilation is a ceiling or a bridge, and who decides if they cannot. Deciding it at three in the morning during a hypercapnic exacerbation is how it usually happens, and it is the wrong time.

Hazzard ch 101 p1607 and ch 100 p1589 — 6 q across three of four sittings

Three entities that overlap in the same patient and are separated by where the pain is, how fast it came, and what the serology does. RA, lupus and the drug-induced syndromes are ch 101; polymyalgia and giant cell arteritis are ch 100. Gout has no Hazzard chapter of its own — the committee's gout questions cite the AGS Geriatrics Review Syllabus.

Late-onset rheumatoid arthritis, and what the chapter says is different

The mimics ch 101 lists for a new polyarthritis in an old patient: PMR, osteoarthritis (including the erosive inflammatory form), crystal arthritis, malignancy-associated arthritis, late-onset spondyloarthropathy, hypothyroidism, hyperparathyroidism, diabetic cheiroarthropathy, amyloid, oedematous scleroderma. ESR, not RF titre, tracks disease activity; anti-CCP titre is not used for follow-up.

Polymyalgia: the steroid course, by the numbers

• Start at the lowest effective dose in the range 12.5–25 mg prednisone daily — the higher end for those at high relapse risk. Reduce to 10 mg within 4 to 8 weeks, then 1 mg every 4 weeks once in remission. Symptoms, not the ESR, guide the taper; alternate-day dosing is inferior
• A course takes at least 12 to 18 months; three-quarters are off steroids by two years; 10–30% need a low chronic dose; relapse during taper in about a quarter
• Methotrexate is the only steroid-sparing agent with randomised evidence, considered early where relapse or steroid toxicity is likely; tocilizumab is second line for individual cases. NSAIDs do not lower the ESR and do not suppress a subclinical GCA
• Screen for GCA at every visit; a rising ESR after initial fall, or failure of low-dose steroid, points to it or to another diagnosis — and biopsy is still worth doing on steroids

Giant cell arteritis — the emergency

Headache in a temporal or occipital distribution (65%), visual disturbance (30%), permanent visual loss 10–15%, jaw claudication 45%, swallowing or tongue claudication, arm claudication, cough. A decreased temporal pulse (46%) is a commoner sign than scalp tenderness (27%). Constitutional symptoms can be the only manifestation, and the fever can meet FUO criteria. PMR in 40–60% and the presenting feature in 25%. Large-vessel disease in a quarter; aortic aneurysm years later, so a yearly chest radiograph.

• Tocilizumab (GiACTA, FDA-approved) early for new-onset disease at high risk of steroid harm, relapse or prolonged treatment, and at relapse. Methotrexate: modest steroid-sparing effect; infliximab ineffective
• Low-dose aspirin reduces cranial ischaemic events, particularly anterior ischaemic optic neuropathy
• Bone protection from day one; Pneumocystis prophylaxis at prednisone 20 mg or more. Twenty to fifty percent develop steroid toxicity; life expectancy is otherwise normal unless there is severe aortitis

Late-onset lupus, and the drugs that imitate it

• About 10% of lupus begins after 55, and the delay to diagnosis averages 2 to 3 years. Female-to-male 4:1 against 9:1 in the young. Milder: less rash, photosensitivity, alopecia, ulcers, nephritis and lymphadenopathy; more cytopenias, serositis and interstitial pneumonitis. Death is from treatment — infection, cardiovascular disease, stroke
• Low-titre ANA is present in up to a third of healthy older adults; anti-dsDNA and anti-Sm are specific. The 2019 EULAR/ACR criteria are unvalidated in the old
• Drug-induced lupus: over 100 drugs, onset 50–70, more often men; antihistone antibodies typical of procainamide and hydralazine, less so of minocycline and anti-TNF. A third on procainamide over a year become symptomatic and nearly all ANA-positive by two; under 20% on hydralazine, dose-dependent above 200 mg/day. Fewer renal, neuropsychiatric and skin features; pleuropulmonary symptoms common with procainamide. It resolves on withdrawal, by definition
• Antimalarials for arthritis, serositis, mucositis and skin, long term; avoid NSAIDs with renal involvement

Before you start an immunosuppressant

• Fungal, tuberculous and atypical mycobacterial infection is the chief risk of anticytokine therapy in the old, and infection rates were higher in older RA patients in the CORRONA registry. TB skin test or interferon-gamma release assay plus a chest radiograph before starting; yearly rescreening for those at risk, including nursing-home residents. Hepatitis B reactivation with rituximab
• Immunise before the DMARD or biologic, especially before rituximab. Influenza and pneumococcal as for the general population; holding methotrexate for 2 weeks after influenza vaccination improves the response without a meaningful flare. Tetanus immunoglobulin rather than toxoid within 24 weeks of rituximab. No live vaccines on biologics; the recombinant zoster vaccine is the safe option, and RA activity, biologics, prednisone and hydroxychloroquine all raise zoster risk
• Methotrexate is weekly, with folate; parenteral if dyspepsia or dysphagia; extreme caution with any chronic or active infection

Crystal and septic arthritis — not in Hazzard

Neither ch 100 nor ch 101 covers gout, pseudogout or septic arthritis beyond listing them as mimics; the committee's gout questions cite the GRS. What follows is conventional teaching, not textbook-cited, and the one sitting that asked it wanted: steroids for polyarticular gout given for 5 to 10 days with a taper.

• Gout in the old is more often polyarticular, tophaceous, upper-limb and diuretic-related, and mistaken for infection or RA. NSAIDs are usually the wrong answer — renal impairment, heart failure, anticoagulation, GI bleeding; low-dose colchicine, a short steroid course or an intra-articular steroid instead
• Allopurinol started low and titrated, with colchicine cover for the first months, not stopped during an attack and not started without cover. Calcium pyrophosphate: chondrocalcinosis, wrists and knees, positively birefringent rhomboids, symptomatic treatment only
• Septic arthritis: often afebrile, often on a damaged or prosthetic joint. Aspirate before treating; Gram stain, culture, cell count; empirical antibiotics after the sample; drainage

Hazzard ch 27, p371 — this is the chapter the analysis could not name. It appeared in every one of the 2024–2026 sittings and had no section here because the number had never been resolved to a title. It is Perioperative Care, and it sits alongside ch 28 Anesthesia (p383) and ch 29 Surgical Quality and Outcomes (p399).

What the preoperative assessment is actually for

Not to clear the patient. To estimate risk, modify what can be modified, and set expectations — and in a geriatric patient the outcome that matters is often function and independence rather than 30-day mortality. Ask what the operation is meant to achieve and whether the patient's goals survive the likely recovery.

Also assess: nutrition (albumin and unintentional weight loss), the medication list, alcohol and tobacco, falls history, sensory impairment, social support and the discharge destination. Routine preoperative testing — ECG, chest radiograph, full panels — in an asymptomatic patient having low-risk surgery does not improve outcomes and generates false trails.

Drugs, held and continued

Anticoagulation: most minor procedures do not need interruption; where a DOAC is stopped, timing follows renal function and bleeding risk, and bridging with heparin is usually unnecessary and harmful outside a mechanical valve or very high thrombotic risk. Antiplatelets after a coronary stent: elective surgery is deferred until the mandatory dual-antiplatelet period is complete; aspirin is usually continued through.

The numbers ch 27 gives, and the exam has used

• 25–30% of all perioperative deaths are cardiac — the 2020 committee answer. LVEF under 40% is an independent MACE risk; heart failure with preserved EF still raises it, and two-thirds of older adults with heart failure have normal systolic function. A prior stroke is a MACE risk factor (2023)
• In the over-80s after major non-cardiac surgery, any complication multiplies 30-day mortality from 4% to 26%. Postoperative atrial fibrillation in about 25%; a beta-blocker or amiodarone reduces it but is not routinely recommended
• Delirium in 10–15% after general surgery, 30% after cardiac surgery, up to 50% after hip fracture repair; it predicts worse recovery and long-term decline. Postoperative cognitive dysfunction in 26% at one week in the over-60s after abdominal or orthopaedic surgery (ISPOCD), with older age, lower education and prior cognitive impairment as risks
• Anticoagulants: warfarin — hold four doses to an INR under 1.5, check on the day, vitamin K 1 mg IV or 2.5 mg oral if needed; LMWH bridging only for the high-thrombotic patient, stopped 24 h before and restarted 24–48 h after; DOACs need no bridging, stopped 1 to 4 days before by renal function and bleeding risk. Elective surgery waits 3 months after a DVT or PE (asked 2024)
• Transfuse for symptoms, and possibly below 7 g/dL; preoperative IV iron in the anaemic reduces transfusion

Postoperative delirium is the complication of this chapter

The commonest surgical complication in older patients, and largely preventable by the same multicomponent programme used on the ward. Prevention beats treatment here more decisively than anywhere else: orientation, glasses and hearing aids back on immediately, early mobilisation, sleep protection, hydration, bowel and bladder care, adequate analgesia — undertreated pain is a delirium precipitant, not a protection against one.

• Avoid meperidine, benzodiazepines, anticholinergics including scopolamine and promethazine, and high-dose antihistamines
• Regional versus general anaesthesia has not been shown to change the rate of delirium or of functional recovery after hip fracture in the large trials — a favourite question with a counterintuitive answer
• Depth-of-anaesthesia monitoring and avoiding excessively deep planes is reasonable, and the evidence for it is softer than it is often presented
• Remove catheters and lines as early as possible; every one of them is a tether and a source of sepsis

Hazzard ch 31, p437 — and הפרעות בליעה is on the official exam topic list in its own right. The chapter that connects stroke, Parkinson disease, dementia, pneumonia and the feeding-tube question.

Presbyphagia is not dysphagia

Ageing slows the swallow, reduces tongue pressure and lengthens the pharyngeal phase — presbyphagia, a reduced reserve that is compensated in health and decompensates under any insult. Dysphagia is the symptomatic disorder. The distinction matters because a stroke or a pneumonia unmasks presbyphagia rather than creating a new disease.

Drug causes are worth a separate sweep: anticholinergics and diuretics through dry mouth, antipsychotics and metoclopramide through drug-induced parkinsonism, benzodiazepines and opioids through sedation, and bisphosphonates, doxycycline, potassium chloride and NSAIDs through direct pill oesophagitis — which is why those are taken upright with a full glass of water.

Assessment

• Screen before the first oral intake after a stroke. A simple water-swallow screen with a trained observer reduces pneumonia
• Videofluoroscopy is the gold standard, and is the only way to see silent aspiration — passage below the cords with no cough. FEES is the bedside alternative and needs no radiation, useful in a patient who cannot be transported
• Look for the consequences as well as the cause: weight loss, dehydration, recurrent chest infection, food avoidance, and the social loss of eating with other people

Management, and what the evidence does not support

• Postural and manoeuvre strategies — chin tuck, head turn to the weak side, supraglottic swallow — taught by a speech and language therapist, and effective in the patients who can follow them
• Texture modification and thickened fluids reduce visible aspiration but the evidence that they reduce pneumonia is weak, and they reduce palatability, intake and hydration. They are not a free intervention
• Oral hygiene is one of the few measures with real evidence against aspiration pneumonia — it is the oral flora that is aspirated, and this is the cheapest thing on the list
• Upright positioning for the meal and for 30 minutes after; unhurried, small mouthfuls; self-feeding where safe; minimise sedating drugs at mealtimes
• In advanced dementia, tube feeding does none of what it is used for — no better survival, no fewer aspiration pneumonias, no pressure-ulcer benefit, no comfort. Careful hand feeding, offered as an active choice, is the alternative. See 67–70 Palliative. Do not generalise that to all dysphagia: the evidence is specific to advanced dementia, where the swallow fails as part of a terminal illness. In acute stroke with a swallow expected to recover, and in motor neurone disease, enteral feeding is standard and gastrostomy is offered deliberately — the question is whether the underlying course is recoverable

Hazzard ch 48 p719 — and the MoH procedure the exam names

In Israel this is not only a clinical topic but a statutory reporting duty, and the MoH procedure on abuse of the helpless (נוהל התעמרות בחסרי ישע) is named directly in the exam's source tables. The clinical half below is ch 48; the legal half is the procedure.

The categories, and how often each is reported

Laboratory markers proposed but not validated: malnutrition, dehydration, hypo- or hyperthermia, rhabdomyolysis, toxicology. The chapter's point is that fracture, depression and malnutrition all have a chronic disease explanation and an abuse explanation, and mistreatment sits in the differential of almost every geriatric presentation.

Risk, recognition and outcome

• Up to 10% of community-dwelling older adults a year; nearly half of people with dementia are mistreated by carers in smaller studies. As few as 1 in 24 cases reaches the authorities — victims are isolated, ill, demented, afraid of reprisal or of the institution, or protecting the abuser
• The most consistent risk finding is the relationship: spouses and adult children are the usual perpetrators, sons and daughters equally. Two dependencies run in opposite directions — the victim's on the carer, and the perpetrator's financial and housing dependence on the victim. The “stressed caregiver” is one pattern among several; the schizophrenic adult child and the spouse with progressive dementia are different problems needing different plans
• Physicians are among the least helpful of 17 occupational groups in reporting to protective services, and only a small fraction of reports come from them
• Outcomes: more ED visits, admissions, dementia, depression and nursing-home placement, and a threefold mortality against non-abused controls. There are no randomised trials of intervention

The Israeli duty, which is what gets examined

• There is a duty to report a reasonable suspicion of harm to a helpless person (חסר ישע) — someone who cannot care for themselves through age, illness or disability. It falls on the professional, and it is not discharged by telling a relative or a colleague
• The report goes to the police or to the social worker appointed under the law (עובד סוציאלי לפי חוק); the MoH procedure נוהל התעמרות בחסרי ישע sets out the institutional pathway
• The duty exists irrespective of the patient's wishes, which is what distinguishes it from most of the consent-based material elsewhere in the syllabus, and it is where a vignette will try to catch you with a competent patient who asks you not to report — and with the patient whose designated חוק סיעוד carer is the one neglecting them
• Document what was observed rather than what was concluded: injuries described and measured, the account given, who gave it, and who was present

Hazzard ch 53 p793 — and the NEJM review the exam cites by name

שברים is an official exam heading, and this is the fracture that carries the syllabus. Written from ch 53, with the perioperative numbers ch 53 defers to ch 27 and the rehabilitation numbers it defers to ch 55.

The numbers that frame it

• Average age at fracture is 82. Of those who reach 85, about 19% of women and 12% of men will fracture a hip; by 90, 30% and 20%. A quarter to 30% of hip fractures are in men
• One-year mortality: about 18% of women and 36% of men, concentrated in the first few months and in those in the poorest health. The fracture itself accounts for nine extra deaths per 100 patients over four years, and the excess persists well beyond the first year even in the fittest
• Delirium occurs in up to 60% after repair; dementia is the major risk factor, and admission from an institution and heart failure are the hip-specific ones. Proactive geriatric consultation cuts the incidence by a third and severe delirium by half in one study
• The body pays for the fracture: 3–6% of muscle mass lost within two months, fat mass up 3–4% over the year, nearly 3% of total-hip and more than 4.5% of contralateral femoral-neck bone density lost in the first year — 12.5 times the expected loss. Vitamin D deficiency is very common at presentation
• More than 75% have at least four comorbid conditions before the fracture

The operation follows the blood supply

About half of hip fractures are femoral-neck (intracapsular) and half intertrochanteric (extracapsular); subtrochanteric fractures, within 5 cm below the lesser trochanter, are less common. Intertrochanteric patients tend to be older with more comorbidity. The femoral neck and head are fed by fine retinacular branches of the medial circumflex femoral artery that lie against the neck, so a neck fracture threatens non-union and osteonecrosis; the intertrochanteric region has a plentiful, redundant supply and heals after fixation.

Non-operative management is a real option for a defined group: those who could not walk before the fracture, advanced dementia with contractures, little pain from the fracture, very high surgical risk, or very short life expectancy. The chapter puts the geriatrician in charge of that “big picture” decision and of the palliative pathway that follows it.

The perioperative decisions that are actually examined

• Timing. Ch 53 says only that the geriatrician's job is to get the patient to theatre rapidly and medically stabilised, and that delay for testing that will not change the operation is the thing to avoid. The Management of Acute Hip Fracture review (Bhandari, NEJM) supplies the figure the committee asks for: operating within 24 hours reduces mortality
• Anticoagulant clearance is ch 27's material, and it is what stops a same-day operation: warfarin — withhold four scheduled doses to bring the INR under 1.5, measure it on the day, and reverse with a small dose of vitamin K (1 mg IV or 2.5 mg orally, acting within 24 hours) or fresh frozen plasma if it cannot wait. DOACs need no bridging and are stopped 1 to 4 days beforehand according to thrombotic risk, bleeding risk and renal function. LMWH is stopped 24 hours before and restarted 24 to 48 hours after, by bleeding risk
• General versus spinal anaesthesia: no difference in mortality, and the choice is the surgeon's and anaesthetist's by local practice. That is the whole of what ch 53 says. The trial answer — not in Hazzard, cited from the trial — is REGAIN (Neuman et al., N Engl J Med 2021;385:2025, doi 10.1056/NEJMoa2113514, read from the paper; 1,600 patients aged 50 and over): spinal against general made no difference to the composite of death or inability to walk unassisted at 60 days, and the incidence of delirium was similar
• A preoperative regional nerve block reduces pain, can be placed by an anaesthetist or a trained emergency physician, often makes intravenous opioids unnecessary after surgery — and may make the urinary catheter unnecessary. Catheters are for patients who cannot roll or use a bedpan for pain, and if placed they come out the morning after surgery
• Transfusion threshold is ch 27's again: reserve transfusion for symptoms, and possibly a haemoglobin below 7 g/dL; preoperative intravenous iron in anaemic patients reduced postoperative transfusion in small trials. The hip-fracture trial — not in Hazzard, cited from the trial — is FOCUS (Carson et al., N Engl J Med 2011;365:2453, doi 10.1056/NEJMoa1012452, read from the paper; 2,016 patients over 50 with cardiovascular disease or risk factors and a haemoglobin under 10 after hip-fracture surgery): a restrictive trigger of symptoms or 8 g/dL was no worse than a liberal 10 g/dL for death or inability to walk across a room at 60 days — 34.7% vs 35.2%, odds ratio 1.01; a median of two units given in the liberal arm and none in the restrictive. So the textbook says 7, the hip-fracture trial used 8, and both say symptoms first; the exam has not yet asked which
• Mobilise on the day of surgery, and certainly by day one. Deep-breathing exercises with or without an incentive spirometer — postoperative pulmonary complications are among the most lethal. DVT prophylaxis for 21 to 31 days after surgery, continued into the rehabilitation setting. Pressure-injury vigilance: stage 2 or worse in about 5% within 21 days of the fracture. A bowel regimen from the start, because opioids plus immobility make constipation routine
• Orthogeriatric co-management. A meta-analysis of 18 studies: reduced in-hospital mortality, reduced long-term mortality, shorter stay. A randomised trial of care on a geriatric ward did not reduce delirium but improved mobility at four months. Watch for the postoperative pair the chapter names, atrial fibrillation and heart failure — and expect the heart failure to worsen after discharge

What to tell the family, in the order recovery happens

Recovery follows a sequence. Depression, upper-limb ADL and cognition peak by about 4 months; balance and gait plateau at about 9 months; lower-limb activities and instrumental tasks recover by about a year (Figure 53-5, the Magaziner eight-domain study). Most patients plateau by six months. Depressive symptoms and cognitive limitation are common and often resolve with little intervention within two months, but anyone who reports depressive symptoms even transiently recovers less well, and mild depression is worth treating.

• At one year, of those who managed alone before the fracture, 20% need help putting on trousers, 40% to cross a room, 66% to use the toilet, 90% to climb five stairs. Housework, getting out of walking distance and shopping: 62%, 53% and 42% newly dependent
• Up to half of previously independent patients need a cane or walker at one year; fear of falling then limits activity further. A fall injury makes admission to a nursing facility ten times more likely, and the over-85s have longer rehabilitation, worse lower-limb recovery, more long-term care and more persistent pain
• Slower-than-expected recovery has a checklist: persistent delirium, depression, polypharmacy, vitamin D and other nutritional deficits, social isolation

Rehabilitation eligibility, and after

Services are offered when two conditions hold: a functional loss, and the ability to participate. Severe cognitive impairment, unresolved delirium, severe pre-fracture disability such as being bed-bound, and terminal illness may exclude standard rehabilitation. Acute inpatient rehabilitation demands 3 hours a day and at least two disciplines; subacute care is for those who cannot manage that. Continued exercise after discharge improves function through the first year, progressive-resistance training most of all. Reassess at discharge from rehabilitation and again at the end of home physiotherapy. Hip protectors offer little or no benefit in randomised studies and are not a substitute for falls and bone work. The Israeli framework for who is admitted where is in 55 Rehabilitation.

Hazzard ch 62 p953 — with the post-stroke numbers from ch 55 p817

The acute chapter behind 55 Rehabilitation, and the one that decides how much there is to rehabilitate. Fifth leading cause of death, first cause of disability; incidence rises with age and a growing share of strokes occur at 85 and over. Prognosis is worse in the old, yet the mortality decline of the last decades reaches every age group including the over-85s — because the treatments below are given to them.

TIA and minor stroke: the definitions and the score

• A TIA is focal symptoms resolving within 24 hours with no lesion on diffusion MRI. If DWI is positive it is a stroke, however brief the symptoms — a minor stroke. The distinction does not change the workup; both demand urgent, complete investigation
• Recurrence after TIA or minor stroke is about 10% in 90 days, highest in the first 48 hours, and as high as 10% in the first week depending on mechanism
• ABCD2: Age over 60 (1) · Blood pressure over 140/90 (1) · Clinical — speech alone (1), unilateral weakness (2) · Duration — under 10 min (0), 10–59 min (1), 60 min or more (2) · Diabetes (1). A score of 4 or more is elevated risk, and is the entry criterion the DAPT trials used

Acute treatment: the windows and the thresholds ch 62 actually gives

There is no NIHSS threshold in ch 62 — the scale appears once, as the median of a trial. Cut-offs such as “5 or under is minor” are conventions from the trials themselves, not from the textbook; the chapter's thresholds are ASPECTS under 6, ABCD2 of 4 or more, and disabling versus nondisabling. Stroke-unit care improves outcomes over a general ward; avoid hyperthermia; heparin or pneumatic compression until walking.

What causes it here

Secondary antiplatelet therapy: how long, and the two trials

• Aspirin 50–325 mg/day as soon as possible; most of its benefit is in the first days. Clopidogrel is an equal first-line alternative (CAPRIE: no difference in the stroke subgroup); ticagrelor was not superior to aspirin at 90 days (SOCRATES). Antiplatelet monotherapy remains the cornerstone for non-cardioembolic stroke
• CHANCE (China, 5,170 patients): aspirin plus clopidogrel for 3 weeks then aspirin alone, started within 24 hours of a high-risk TIA or minor stroke. Recurrent stroke 8.2% vs 11.7% — a 32% relative reduction — without a significant rise in bleeding
• POINT (4,881 patients, North America, Europe, Australasia): the same combination for 90 days. Ischaemic stroke, MI or vascular death 5.0% vs 6.5%, but major haemorrhage 0.9% vs 0.4% — attributed to the higher clopidogrel loading dose and the longer course. The benefit was concentrated in the first 30 days; the bleeding accrued after the first week
• Hence the American Heart Association / American Stroke Association recommendation: DAPT with aspirin and clopidogrel, started within 24 hours, for 21 days, after high-risk TIA or minor stroke, provided the patient did not receive thrombolysis. THALES reproduced the benefit with ticagrelor for 30 days but at the cost of more major, fatal and intracranial haemorrhage
• Atrial fibrillation: warfarin (INR 2–3) cuts recurrent embolism by close to 70% against aspirin; DOACs are non-inferior with significantly less intracranial haemorrhage and more gastrointestinal bleeding, net benefit in their favour, no monitoring. Not in infective endocarditis. Heart failure in sinus rhythm gets no anticoagulation beyond the acute phase (WARCEF)
• Blood pressure remains the most effective single measure, lowered after the acute phase; a statin unless life expectancy makes it pointless

Post-stroke depression and the other late complications

Ch 62 lists the late complications as spasticity, contracture, pressure ulcers, shoulder dislocation and depression, and gives the numbers over to ch 55: depression in 15–70% depending on the study, with a significant negative effect on rehabilitation and a response to several antidepressants; bladder incontinence in 50–70% in the first month, returning to population levels by six months with timed voiding and residual monitoring; hemiplegic shoulder pain in 34–84%, managed by positioning, early range of motion, and electrical stimulation with benefit up to six months. Two prescribing points the exam can build on: fluoxetine does not improve functional recovery — the early signal was not reproduced in three trials, and it increased fractures, so it is for mood only; and for central post-stroke pain nortriptyline is preferred over amitriptyline in older adults. Botulinum toxin helps spasticity. Fatigue, sleep-disordered breathing and anxiety are to be sought, not waited for.

The rehabilitation handoff

Every patient with stroke-related disability is assessed by physiotherapy, occupational therapy and speech therapy as appropriate, and a substantial number benefit from inpatient rehabilitation; patients treated in dedicated stroke units by experienced physician, nursing and rehabilitation staff do better than on a general ward. Recovery takes six months to a year or longer, paced by neuroplasticity. Screen the swallow before anything by mouth (31 Disorders of Swallowing). Almost half of stroke deaths in older adults occur in hospital or rehabilitation facilities, so goals of care — which recovery matters to this patient: walking, communicating — are set early and primary palliative care runs alongside. The Israeli admission criteria, the FIM, the Copenhagen recovery numbers and the EBRSR handbook are in 55 Rehabilitation; vascular cognitive impairment is graded by the VasCog-2-WSO criteria in 59 Dementia.

Hazzard ch 102, p1629 — four pooled questions, present in three of four sittings, and Table 102-3 is cited three times, which puts it level with the frailty and continence tables at the top of the cited list. This was the tier-2 chapter with the thinnest coverage in this file.

The differential, split the way the chapter splits it

Referred hip osteoarthritis and a leaking aortic aneurysm are on the same page for a reason: the commonest miss and the one that kills. Non-specific low back pain is a diagnosis made after working through this list, not before.

Red flags, and when to image

The chapter uses the American College of Radiology criteria to decide who gets imaged. Each of these marks a higher probability of systemic disease:

• Age over 70 — which is most of your clinic, and is why the threshold to image is genuinely lower here than in a younger patient
• Recent significant trauma, or milder trauma over the age of 50
• Unexplained weight loss or fever
• Immunosuppression, including diabetes
• History of cancer
• Osteoporosis, or prolonged glucocorticoid use
• Focal neurological deficit that is progressive or disabling
• Symptoms lasting six weeks or more
• Intravenous drug use

Vertebral compression fracture

• Severe mechanical pain when the fracture is lower thoracic or lumbar
• High and mid-thoracic fractures are often asymptomatic and found incidentally — which is why two-thirds of them are never diagnosed at the time
• The pain typically resolves over four to six weeks. Pain that does not is a reason to look again, not to escalate the analgesia
• It sharply raises the risk of the next fracture — the “imminent risk” window. It is the indication to treat osteoporosis. See 51 Osteoporosis
• Vertebroplasty and kyphoplasty did not beat sham in the better trials and are not routine

Lumbar spinal stenosis — the table that gets cited

The whole clinical picture follows one mechanical fact. Flexion opens the canal: the disc bulges less, the ligamentum flavum stretches thin, the cross-sectional area available to the roots increases. Extension closes it: the disc bulges in, the ligamentum flavum buckles and folds into the canal, the roots thicken. Everything else is a consequence.

• The presenting complaint is usually leg pain, not back pain: heaviness, numbness, cramping or burning, usually both legs though often worse on one
• Neurogenic claudication, also called pseudoclaudication, and often an incomplete sciatica rather than a dermatomal one
• True weakness is uncommon; examination is frequently normal at rest, which is a trap — examine after walking if you can

Both coexist often enough that finding one does not excuse you from looking for the other — and a third possibility, hip osteoarthritis, refers into the groin and thigh and is worse on weight-bearing regardless of posture.

Management

• Conservative first for stenosis without progressive deficit: exercise with a flexion bias, physiotherapy, weight reduction, and activity modification. Many patients are stable or improve over years
• Analgesia is the geriatric problem. NSAIDs collide with renal function, heart failure, anticoagulation and the GI tract; opioids are not first-line for chronic musculoskeletal pain and STOPP names them; paracetamol is modest; a gabapentinoid for the neuropathic component buys sedation and falls. Say so rather than reaching for the next drug
• Epidural steroid injection gives short-term relief at best for stenosis and does not change the course
• Surgery — decompression, with fusion only where there is instability — for disabling symptoms after conservative treatment, and urgently for cauda equina or a progressive deficit. Age alone is not the contraindication; frailty and the ability to rehabilitate afterwards are what decide it

Primary documents

Beers 2023. By the 2023 AGS Beers Criteria Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc 2023;71(7):2052–2081. doi:10.1111/jgs.18372. Free full text at agsjournals.onlinelibrary.wiley.com.

Beers Alternatives. AGS Beers Criteria Alternatives Panel, Steinman MA. Alternative Treatments to Selected Medications in the 2023 American Geriatrics Society Beers Criteria. J Am Geriatr Soc 2025;0:1–21. doi:10.1111/jgs.19500. Released July 2025, first update since 2015. Thirty of the 36 Table 2 criteria, covering 21 conditions.

STOPP/START v3. O'Mahony D, Cherubini A, Guiteras AR, et al. STOPP/START criteria for potentially inappropriate prescribing in older people: version 3. Eur Geriatr Med 2023;14(4):625–632. doi:10.1007/s41999-023-00777-y. Open access. Read the correction with it — Eur Geriatr Med 2023;14:633, doi:10.1007/s41999-023-00812-y.

Syllabus. ההסתדרות הרפואית בישראל, המועצה המדעית. התמחות בגריאטריה — תכנית ההתמחות, Syllabus. Versions 1.0 (2011), 2.0 (2021), 3.0 (2024), 4.0 (in force 07.06.2026). The 46 topics in this plan derive from its נושאי הבחינה list, weighted by frequency across 3,833 past items.

Hazzard's, and the five papers read in full

Hazzard's. Halter JB, Ouslander JG, Studenski S, et al., eds. Hazzard's Geriatric Medicine and Gerontology. 8th ed. New York: McGraw Hill; 2022. Every content section carries in its footer the chapter it was checked against and the date; 39 chapters were read directly from the PDF on 13 Sep 2026. The 7th-edition page references in the 2020–2022 papers are ignored; only the 8th edition is used.

Kattah JC, Talkad AV, Wang DZ, Hsieh YH, Newman-Toker DE. HINTS to diagnose stroke in the acute vestibular syndrome: three-step bedside oculomotor examination more sensitive than early MRI diffusion-weighted imaging. Stroke 2009;40(11):3504–3510. doi:10.1161/STROKEAHA.109.551234. PMC4593511, free. Read in full; cited in 45 Syncope.

Kaufmann H, Freeman R, Biaggioni I, et al. Droxidopa for neurogenic orthostatic hypotension: a randomized, placebo-controlled, phase 3 trial (NOH301). Neurology 2014;83(4):328–335. doi:10.1212/WNL.0000000000000615. PMC4115605, free. Read in full; cited in 45 Syncope.

Bartalena L, Bogazzi F, Chiovato L, et al. 2018 European Thyroid Association (ETA) guidelines for the management of amiodarone-associated thyroid dysfunction. Eur Thyroid J 2018;7(2):55–66. doi:10.1159/000486957. PMC5869486, free. Read in full; cited in 98 Thyroid.

Carson JL, Terrin ML, Noveck H, et al. Liberal or restrictive transfusion in high-risk patients after hip surgery (FOCUS). N Engl J Med 2011;365(26):2453–2462. doi:10.1056/NEJMoa1012452. Paywalled. Read in full; cited in 53 Hip fracture.

Neuman MD, Feng R, Carson JL, et al. Spinal anesthesia or general anesthesia for hip surgery in older adults (REGAIN). N Engl J Med 2021;385(22):2025–2035. doi:10.1056/NEJMoa2113514. Paywalled; the NEJM summary slides were read, not the full paper. Cited in 53 Hip fracture.

What the syllabus mandates, against what the papers actually cite

Mandatory textbooks: Hazzard's Principles of Geriatric Medicine; the AGS Geriatrics Review Syllabus; Harrison's geriatrics chapter. Mandatory journals: Journal of the American Geriatrics Society, Age and Ageing, Lancet, New England Journal of Medicine, Harefuah reviews.

Knowledge requirement beyond the 42 exam headings

The syllabus's הידע הנדרש section maps a wider content scope than the exam headings. Most of it lands under a heading that is already in the plan; a minority has no heading at all. The distinction below decides what gets reading hours.

Lands under a heading already in the plan

No exam heading at all

Currency, as of 12 September 2026

• Beers 2023 is still the current edition. No 2026 update published. AGS updates every 3–4 years and the next is anticipated 2026–2027 — check once in spring 2027.
• STOPP/START v3 is still current. There is no v4.
• One genuinely new adjacent document: the Australian Deprescribing in Older People clinical practice guideline, MJA 2026, doi:10.5694/mja2.70174, open access, 185 recommendations across when to deprescribe, ongoing need, how, and monitoring. Excellent for practice; not what this exam is written from. Read it after the other three.

Not from these documents

The five topic sections — cancer, dementia, delirium, infections and falls — are written from standard geriatric medicine, cross-referenced to the three documents above wherever a Beers or STOPP/START criterion applies. Where a threshold is conventional rather than sourced from those papers (MoCA cut-offs, TUG seconds, long-term-care fever thresholds, G8 scoring) it is described as commonly used rather than as a citation. Verify any figure you intend to quote in an exam answer against Hazzard’s or the relevant society guideline.

The ageing-pharmacology material (volume of distribution, albumin binding, the phase I versus phase II distinction behind preferring lorazepam over diazepam), the ACB threshold of 3, and the five-step deprescribing framework come from Hazzard's and the deprescribing literature, not from the three papers above. Flagged so you know which claims have a document behind them in this file and which do not.

The named sources the papers actually cite

Every one of these appears by name in a מראה מקום table, and none of them is in Hazzard's. This is the material with no textbook behind it, which is why it is listed separately and why it is worth obtaining rather than reading around. Citations below were checked against the publisher record rather than quoted from memory; where a detail could not be confirmed it is left out rather than guessed.

Dementia — diagnosis and prevention

Lancet standing Commission, 2024 report. Livingston G, Huntley J, Liu KY, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. Lancet 2024;404(10452):572–628. doi:10.1016/S0140-6736(24)01296-0. Obtained — 57 pages, and its Table 1, the weighted attributable fractions, is now in 57 Cognitive change. Untreated vision loss and high LDL cholesterol are the additions to the twelve from 2020, and the fourteen come to 45%. Cited by name in 2021 and 2025. Read Table 1 and the executive summary's action list; the rest is policy.

The 2020 report is the one the older papers cite: Livingston G, Huntley J, Sommerlad A, et al. Lancet 2020;396(10248):413–446. Twelve risk factors, about 40% of dementia worldwide. Know which list a stem is built on — twelve or fourteen.

Revised Alzheimer criteria, Alzheimer's Association workgroup. Jack CR Jr, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer's disease. Alzheimers Dement 2024;20(8):5143–5169. doi:10.1002/alz.13859. Biological definition, the A/T/N-derived biomarker categories including blood-based markers, and the integrated biological–clinical staging.

The dissenting position — IWG. Dubois B, Villain N, Schneider L, et al. Alzheimer disease as a clinical-biological construct: an International Working Group recommendation. JAMA Neurol 2024;81(12):1304–1311. doi:10.1001/jamaneurol.2024.3770. Diagnosis requires a clinical syndrome and biomarkers; biomarker-positive, cognitively normal people are at risk, not diseased. The exam has cited JAMA Neurology three times in 2026 alone, and a question that pits these two documents against each other is the obvious use of them.

Vascular cognitive impairment. The VasCog-2-WSO Criteria Consortium. Revised diagnostic criteria for vascular cognitive impairment and dementia — the VasCog-2-WSO criteria. JAMA Neurol, published online 16 September 2025. doi:10.1001/jamaneurol.2025.3242. VCID as the umbrella term with preclinical, mild and major levels; three subtypes — ischaemic, haemorrhagic, mixed; six cognitive domains; the first revision since VasCog-1 in 2014.

Advanced dementia. Mitchell SL. Advanced dementia. N Engl J Med 2015;372(26):2533–2540. doi:10.1056/NEJMcp1412652. Free in PubMed Central. The source of the tube-feeding answer, the pneumonia and eating-problem trajectories, and the comfort-feeding alternative.

Fracture and rehabilitation

Hip fracture. Bhandari M, Swiontkowski M. Management of acute hip fracture. N Engl J Med 2017;377(21):2053–2062. doi:10.1056/NEJMcp1611090. Fracture location decides the operation; arthroplasty over internal fixation for displaced femoral-neck fractures from 65 onward. Pair it with 55 Rehabilitation and 51 Osteoporosis.

Stroke rehabilitation. Teasell R, Hussein N, et al. Stroke Rehabilitation Clinician Handbook, 2020 edition, Evidence-Based Review of Stroke Rehabilitation (EBRSR), ebrsr.com — free, and not a journal article, which is why it is easy to miss when hunting the citations. Obtained: seven chapters — clinical consequences; brain reorganisation, recovery and organised care; lower extremity; upper extremity; cognitive impairment; medical complications; depression and community reintegration. Chapters 2 and 6 are worked into 55 Rehabilitation; read chapter 2 first, because the recovery time-course numbers are the examinable part.

Diabetes

ADA Standards of Care, Section 13: Older Adults. American Diabetes Association Professional Practice Committee. 13. Older adults: Standards of Care in Diabetes—2026. Diabetes Care 2026;49(Suppl. 1):S277–S296. doi:10.2337/dc26-S013. Obtained, and worked into 99 Diabetes. Revised every January, so cite the year you read; note that this section carries an explicit endorsement from the American Geriatrics Society. The three health-status categories and their targets, deintensification, and the geriatric syndromes section. Diabetes Care is not on the syllabus's mandatory journal list and has supplied questions anyway.

What this tells you about the mandatory journal list

Still to obtain, and the honest gaps

• The 2020 Lancet Commission DOI is not printed here because the volume and pages were confirmed and the identifier was not. Retrieve it by title.
• Several מראה מקום entries give a title without a journal. Advanced Dementia and Management of Acute Hip Fracture are both NEJM and both recur, which is strong but not proof that those tables mean these papers.
• A citation carried in an earlier version of this file was wrong: the IWG recommendation was recorded as JAMA Neurol 2024;81(12):1245–1257. That page range belongs to an accompanying editorial; the paper itself is 1304–1311. Corrected above.
• The Brookdale/JDC yearbook (קשישים בישראל — שנתון סטטיסטי) supplies one to two questions a sitting. It is free, it is updated annually, and the questions are figures — proportions over 65 and over 80, living arrangements, institutionalisation rates. Download the current edition and read the summary tables only.
• The Israeli legal and MoH block has no document list of its own. The statutes and circulars named in the papers are set out in Ethics & EOL law; they are the most reliable four to seven questions on the paper and the ones most often left to last.