Biostatistics & Population Health

Life expectancy and life table interpretation

Clinical Overview and When to Suspect Limited Life Expectancy

— Considering cancer screening (mammography, colonoscopy, PSA, lung CT) in adults ≥65–75

— Considering tight glycemic, BP, or lipid targets in older adults

— Deciding on AAA repair, CEA, ICD placement, dialysis initiation, or major elective surgery

— Hospice eligibility (Medicare requires physician certification of LE ≤6 months if disease runs usual course)

— Advance care planning, POLST/MOLST completion, deprescribing

— Period life table: uses current cross-sectional mortality rates; assumes today's rates persist (most common, e.g., CDC/SSA tables)

— Cohort life table: follows a real birth cohort over time; reflects actual experience but only knowable retrospectively

Life expectancy (LE) = average additional years a person of a given age is expected to live, derived from a life table that applies age-specific mortality rates to a hypothetical cohort

Step 3 uses LE less as a number to memorize and more as a decision-making tool to individualize screening, preventive care, and goals-of-care conversations

When to actively estimate LE in practice:

Two complementary frameworks:

Lag time to benefit is the operational concept that flows from LE: a screening test or preventive intervention only helps if the patient lives long enough to realize benefit (e.g., colonoscopy ~10 yr, mammography ~10 yr, tight glycemic control ~8 yr)

Clues a patient has LE <10 years despite chronologic age: dependence in ≥2 ADLs, advanced CHF/COPD/CKD, dementia, unintentional weight loss, recurrent hospitalizations, metastatic cancer

Validated calculators: ePrognosis (Lee, Schonberg, Walter indices) integrate age, sex, comorbidities, function

Board pearl: On Step 3, the right answer for cancer screening in a 78-year-old is almost never "screen everyone the same"—it is "estimate life expectancy and compare to lag time to benefit." If LE < lag time, stop screening and shift to symptom-directed care and advance care planning.

Presentation Patterns and Key History

— 72-year-old woman, independent, walks 2 miles daily, no chronic disease, asking about mammography

— Estimated LE often >15 years → continue screening; USPSTF supports mammography to age 74 and individualized after

— 82-year-old man with moderate dementia (MMSE 18), needs help bathing/dressing, prior CVA, CHF NYHA III

— Estimated LE often <5 years → stop routine cancer screening, focus on symptom control, fall prevention, advance directives

— 60-year-old with metastatic pancreatic cancer or ESRD on dialysis with multiple comorbidities; chronologic age misleads—disease-specific LE drives decisions

— Functional status (ADLs/IADLs) — single strongest predictor in older adults

— Unintentional weight loss >10 lb in 6 months

— Recent hospitalization, especially for pneumonia, hip fracture, HF exacerbation

— Dementia (median survival ~4–8 years from diagnosis; less if advanced)

— Comorbidity burden (Charlson index, number of meds)

— Social: living alone, food insecurity, no caregiver

Step 3 vignettes rarely ask "what is LE?"—they embed LE reasoning in a clinical scenario where you must decide screen vs stop, treat aggressively vs deprescribe, operate vs palliate

Classic stem 1 — the healthy older adult:

Classic stem 2 — the frail older adult:

Classic stem 3 — the "younger" patient with severe disease:

Key history elements that shift LE estimates downward:

History elements that shift LE upward: independent ADLs, regular exercise, normal cognition, absence of smoking, strong social support

Key distinction: Chronologic age vs physiologic age vs life expectancy—a robust 80-year-old may outlive a frail 70-year-old. Step 3 expects you to recognize that function and comorbidity trump the number on the chart when applying preventive guidelines.

Physical Exam Findings and Functional Assessment

— Measure usual walking speed over 4 meters

— <0.6 m/s → markedly reduced LE, high fall risk

— 0.8 m/s → near-median LE for age

— >1.0 m/s → above-average LE

— Sarcopenia, temporal wasting, loss of subcutaneous fat → cachexia, poor prognosis

— BMI <22 in older adults paradoxically associated with higher mortality than mild overweight

— Resting tachycardia, narrow pulse pressure (advanced HF)

— Hypotension on standing (autonomic dysfunction, polypharmacy)

— Cheyne-Stokes respirations (end-stage HF/neuro)

— JVD, ascites, anasarca in advanced HF/cirrhosis

— Asterixis, scleral icterus → hepatic decompensation (MELD predicts 3-month mortality)

— Uremic frost, pericardial rub → end-stage uremia

Unlike most Step 3 topics, "exam" for LE estimation is the geriatric functional assessment—the physical signs that predict mortality

Gait speed — the "sixth vital sign":

Grip strength (dynamometer): low grip correlates with sarcopenia and all-cause mortality

Timed Up and Go (TUG): >12 seconds suggests increased fall risk and reduced LE

Mini-Cog or MoCA: cognitive impairment independently predicts shortened LE; advanced dementia (FAST stage 7) qualifies for hospice

Body habitus:

Vital sign patterns suggesting limited LE:

Skin and pressure points: stage 3–4 pressure injuries, especially sacral, strongly predict 6-month mortality

Hemodynamic and end-organ clues:

Step 3 management: When a stem describes an older adult with slow gait, weight loss, dependence in bathing, and recent hospitalization, the expected action is to recalibrate preventive care—stop screening colonoscopies, deprescribe statins started for primary prevention, intensify advance care planning—rather than order another screening test. Function-based prognosis trumps age-based protocols.

Diagnostic Workup — Reading and Interpreting a Life Table

— x = exact age

— q(x) = probability of dying between age x and x+1 (age-specific mortality rate)

— l(x) = number alive at exact age x

— d(x) = number dying between x and x+1 = l(x) × q(x)

— L(x) = person-years lived between x and x+1

— T(x) = total person-years lived from age x onward = Σ L(x) from x to end

— e(x) = life expectancy at age x = T(x) / l(x)

— Given l(x) at two ages, compute survival probability: l(70)/l(60) = probability a 60-year-old survives to 70

— Recognize that e(x) rises slightly with age in older adults (a 75-year-old has higher LE than a newborn's LE minus 75, because they've already survived risky early years)

— e(0) (LE at birth) is heavily influenced by infant mortality; e(65) is the more useful number for adult medicine

— e(0) ≈ 77 years (M 74, F 80)

— e(65) ≈ 18–20 years

— e(75) ≈ 11–13 years

— e(85) ≈ 6–7 years

A life table is a structured table with columns that let you compute LE at any age

Standard columns (cohort of 100,000 hypothetical births):

Key interpretation skills tested:

Sample US data points (approximate, period table):

Median vs mean LE: life tables give mean; for skewed survival (cancer, dementia), median survival is more clinically useful

Conditional life expectancy: LE given survival to a certain age—what changes screening logic in 80+ patients

Board pearl: If a question gives you l(x) values at two ages, the probability of surviving the interval = l(later)/l(earlier), and the probability of dying in the interval = 1 − l(later)/l(earlier). Don't subtract ages—divide survivors.

Diagnostic Workup — Advanced Concepts: HALE, DALY, and Prognostic Indices

— Years expected to live in full health, discounting time lived with disability

— US e(0) ≈ 77; HALE ≈ 66 → ~11 years of life lived with significant morbidity

— Used in policy and shared decision-making ("you may live 10 more years but 4 with significant disability")

— DALY = YLL (years of life lost) + YLD (years lived with disability)

— 1 DALY = 1 year of healthy life lost

— Used by WHO/Global Burden of Disease to compare disease impact

— Key distinction: QALY (quality-adjusted life year) is a benefit measure used in cost-effectiveness (higher = better); DALY is a burden measure (higher = worse)

— Premature mortality metric: sum of (reference age − age at death) for deaths before a cutoff (often 75)

— Highlights diseases killing young (injury, suicide, overdose) vs old (Alzheimer's)

— Lee index — community-dwelling older adults, 4-, 10-, 14-yr mortality

— Schonberg index — similar, validated in women

— Walter index — post-hospital discharge 1-year mortality

— Charlson Comorbidity Index — comorbidity-weighted 10-yr survival

— Karnofsky / ECOG — cancer-specific functional prognosis

Beyond raw LE, Step 3 may test quality-adjusted mortality metrics:

HALE (Health-Adjusted Life Expectancy):

DALY (Disability-Adjusted Life Year):

YPLL (Years of Potential Life Lost):

Validated clinical prognostic indices (ePrognosis platform):

Disease-specific tools: MELD (liver), Seattle Heart Failure Model, BODE (COPD), ADAD for dementia

Board pearl: When a vignette gives age + comorbidities + functional status and asks whether to screen or treat, the implied tool is a multivariable prognostic index (Lee/Schonberg), not the population life table. Population LE is the starting point; individualized prognosis is the decision tool.

Risk Stratification — Applying LE to Screening and Prevention Decisions

— Mammography: ~10 years

— Colonoscopy: ~10 years (FOBT/FIT: ~5 years)

— PSA screening: ~10–15 years

— Lung CT (LDCT): ~5–7 years

— Tight glycemic control (A1c <7): ~8 years for microvascular benefit

— Statin primary prevention: ~2–5 years

— BP control: ~1–2 years (stroke benefit can appear within months)

— Mammography: 40–74 (biennial 50–74 strong)

— Colorectal: 45–75 routine, 76–85 individualized, stop ≥85

— Cervical: stop at 65 if adequate prior screening

— PSA: individualized 55–69, against ≥70

— AAA one-time US: men 65–75 who ever smoked

— LDCT lung: 50–80, 20 pack-yr, current/quit <15 yr; stop when LE limited or unable/unwilling to undergo curative surgery

— LE >10 yr (e.g., healthy 75-yr-old) → all standard screening + aggressive risk factor control

— LE 5–10 yr → individualize, shared decision-making, deprescribe primary prevention statins

— LE <5 yr → stop routine screening, focus on symptom control, ACP

— LE <6 mo → hospice eligible

Core Step 3 framework: Match lag time to benefit against estimated life expectancy

If LE > lag time → intervention likely helps → proceed

If LE < lag time → intervention unlikely to benefit, may harm → stop/defer

Lag times to benefit (approximate):

USPSTF age cutoffs (with individualization clause):

Risk stratification by LE tiers:

Step 3 management: A 78-year-old with multiple comorbidities and LE estimated at 4 years asks about a screening colonoscopy. Correct answer: counsel that lag time to benefit (~10 yr) exceeds estimated LE; recommend against screening and document shared decision. This beats "order colonoscopy" or "order FIT" on the exam.

Pharmacotherapy — Deprescribing Through the LE Lens

— Statins for primary prevention: lag to benefit 2–5 yr; in LE <1 yr, statins do not improve QoL or survival (Kutner trial supports discontinuation in advanced illness)

— Bisphosphonates: 3–5 yr lag; after 5 yr of therapy, consider drug holiday especially if LE limited

— Tight glycemic control: A1c target loosens to <8% or even <8.5% in frail older adults; avoid sulfonylureas and insulin when possible due to hypoglycemia

— Intensive BP control: SPRINT-style <120 may not apply; AGS suggests SBP 130–150 in frail elders

— Anticoagulation for AF: weigh CHA₂DS₂-VASc vs HAS-BLED and remaining LE; still often beneficial because stroke prevention lag is short

— Pain control (opioids, acetaminophen)

— Diuretics for symptomatic HF

— Bronchodilators for symptomatic COPD

— SSRIs for active depression

— Anticoagulation if recent VTE or mechanical valve

— Benzodiazepines, anticholinergics (diphenhydramine, oxybutynin), long-acting sulfonylureas (glyburide), NSAIDs in CKD, muscle relaxants

LE estimation directly drives deprescribing decisions—a core Step 3 outpatient skill

Medications to reconsider when LE <5–10 years:

Medications that should usually continue even with limited LE because benefit is rapid:

Beers Criteria drugs to avoid in older adults regardless of LE:

STOPP/START criteria: complementary tool for inappropriate prescribing

Polypharmacy threshold: ≥5 chronic meds → review at every visit

Board pearl: In a Step 3 vignette featuring an 82-year-old with dementia, falls, and LE ~3 years on atorvastatin for primary prevention, the best answer is usually discontinue the statin—the lag to benefit exceeds remaining life, and pill burden/myalgia cause real harm. Continue secondary prevention statins (post-MI, post-stroke) longer because lag is shorter.

Applying LE — Procedural and Surgical Decision-Making

— Elective repair if aneurysm ≥5.5 cm (men) or ≥5.0 cm (women)

— Benefit requires LE >2–3 years; in frail patient with LE <2 yr, surveillance or palliation preferred

— Symptomatic 70–99% stenosis: clear benefit if LE >5 yr

— Asymptomatic: requires LE >5 yr and low surgical risk; otherwise medical therapy alone

— Indicated for EF ≤35% on GDMT

— Requires reasonable expectation of meaningful survival >1 year per ACC/AHA

— Avoid in NYHA IV non-transplant candidates, advanced dementia, terminal illness

— TAVR has expanded options for high/intermediate-risk elderly

— STS score + frailty + LE >1 yr supports TAVR; LE <1 yr → palliative balloon valvuloplasty or comfort care

— In patients >75 with high comorbidity, dialysis may not extend LE meaningfully vs conservative management

— Shared decision-making essential; "time-limited trial" is acceptable

— Lag to benefit ~1 yr; reasonable even with moderate LE if pain limits QoL

— Curative vs palliative intent depends on LE and performance status (ECOG 0–2 generally tolerates therapy; ECOG 3–4 → palliation)

Major procedures require explicit LE estimation because perioperative risk is front-loaded while benefit accrues over years

AAA repair:

Carotid endarterectomy (CEA):

ICD placement (primary prevention):

TAVR vs SAVR vs medical therapy in severe AS:

Dialysis initiation:

Joint replacement:

Cancer-directed therapy:

CCS pearl: On a CCS case of an 84-year-old with severe AS, dementia, dependent ADLs, and recurrent syncope, the correct order set is palliative care consult, advance directive discussion, symptom-directed diuresis—not "consult cardiothoracic surgery." Move the clock forward and watch the case end with appropriate goals-of-care alignment, not aggressive intervention.

Special Populations — Elderly, Renal, and Hepatic Impairment

— LE estimation must incorporate function, cognition, comorbidity—not age alone

— Use Lee/Schonberg indices; "eyeball" estimate: dependent in ADLs + ≥3 chronic conditions ≈ LE <5 yr

— Heterogeneity is enormous: top-quartile 80-year-olds may have LE >13 yr; bottom-quartile <3 yr

— CKD stage 3–4: LE roughly halved compared to age-matched peers

— ESRD on dialysis: 5-yr survival ~40% overall, much lower in elderly with comorbidities (5-yr survival for incident dialysis at age 75+ ~20–25%)

— Conservative (non-dialytic) management may yield similar LE with better QoL in highly comorbid elderly

— Step 3 management: When a 78-year-old with diabetes, HF, and CKD stage 5 asks about dialysis, present dialysis vs conservative care as comparable options, involve nephrology + palliative care, and document shared decision—do not default to "start dialysis"

— MELD-Na predicts 3-month mortality and drives transplant priority

— MELD 15 ≈ 6%, MELD 25 ≈ 20%, MELD 35+ ≈ 50%+ 3-month mortality

— Child-Pugh historical but still used: A (1-yr 100%), B (80%), C (45%)

— Decompensation events (variceal bleed, SBP, HRS, HE) drop median survival to ~2 yr without transplant

— Seattle HF Model and MAGGIC risk score quantify LE

— Advanced HF (NYHA IV, frequent hospitalizations, inotrope-dependent): 1-yr mortality 25–50%

— Triggers palliative care consult, LVAD/transplant evaluation, or hospice depending on candidacy

Elderly (≥65, especially ≥80):

CKD and ESRD:

Cirrhosis/hepatic failure:

Heart failure:

Board pearl: Organ-specific scores (MELD, Seattle HF, BODE for COPD, FAST for dementia) outperform generic LE estimates in patients with that organ disease. Use disease-specific tools when available, generic life tables when not.

Special Populations — Pediatrics, Pregnancy, and Disparities

— LE at birth dominated historically by infant mortality (deaths <1 yr); now dominated by injury and congenital conditions in HICs

— Infant mortality rate (IMR) = deaths <1 yr / 1000 live births; sensitive indicator of population health

— US IMR ~5.4/1000—worse than peer high-income countries, with stark racial disparities (Black infants ~2× white)

— Neonatal mortality (<28 d) vs post-neonatal (28 d–1 yr) — neonatal driven by prematurity/congenital, post-neonatal by SIDS, injury, infection

— Pediatric LE projections rarely tested directly; epidemiologic concepts (IMR, U5MR) more common

— Maternal mortality ratio: deaths per 100,000 live births; US ~22, highest among HICs, with 3–4× disparity for Black women

— Pregnancy-related LE loss = major contributor to disparities in YPLL

— Step 3 may frame as quality/safety: postpartum follow-up at 1–3 wk, screen for hypertension, depression, cardiomyopathy

— Non-Hispanic Black e(0) ~5 yr lower than non-Hispanic White

— American Indian/Alaska Native e(0) significantly reduced (worsened post-COVID)

— Income gradient: top 1% vs bottom 1% males differ by ~15 yr in LE

— Geography: rural Appalachia, Deep South ~10 yr below coastal urban

— Drivers: structural inequities, healthcare access, chronic disease prevalence, violence, overdose

Pediatrics:

Pregnancy and maternal mortality:

Race, ethnicity, and socioeconomic disparities in US LE:

COVID-19 effect: US LE dropped ~2.7 yr from 2019 to 2021—largest decline since WWII—then partial rebound

Key distinction: Life expectancy disparity vs healthcare disparity—LE gaps reflect upstream social determinants (housing, education, food, environment) plus medical care. Step 3 expects recognition that addressing disparities requires more than clinic-based interventions.

Complications — Errors and Misuse of LE in Clinical Practice

— Screening colonoscopy in 85-yr-old with dementia → complications, no benefit

— Aggressive A1c <7 with insulin in frail diabetic → hypoglycemia, falls, fractures

— Statin started for primary prevention in patient with metastatic cancer → pill burden, myalgia

— Anticoagulation in high-fall-risk demented patient without atrial fibrillation history misjudged

— Defibrillator placed in patient with <1 yr LE → end-of-life shocks, distressing death

— Withholding mammography from healthy independent 76-yr-old → missed early breast cancer

— Not offering AAA repair to robust 78-yr-old with 6 cm aneurysm

— Ageism: assuming all 80+ patients are "too old" for chemotherapy when ECOG 0–1 patients often tolerate it

— Failure to control BP in elderly because of fear of side effects—stroke benefit accrues quickly

— Quoting population median LE as if it applied to an individual ("you have 8 years left")

— Failing to distinguish median from mean in skewed distributions

— Not framing as range with uncertainty

— Chronologic age anchoring (treating 80 as a cutoff)

— "Surprise question" — "Would I be surprised if this patient died in the next year?"—a NO suggests palliative care consult; useful counter-anchor

Misapplication of LE causes both undertreatment and overtreatment

Overtreatment errors (LE overestimated):

Undertreatment errors (LE underestimated):

Communication failures:

Anchoring biases:

ICD shock at end of life: failure to deactivate ICDs in dying patients is a recognized quality gap; addressed by goals-of-care conversation + device deactivation order

Board pearl: When the stem asks "what is the most appropriate next step" in an older adult with multiple comorbidities, the answer often involves deactivating an ICD, stopping a statin, or transitioning to hospice—not adding another intervention. Recognize the cue: dependence in ADLs + advanced disease + recent hospitalization.

When to Escalate — Palliative Care and Hospice Triggers

— Triggered by serious illness regardless of prognosis

— Symptom burden, complex decision-making, family distress, frequent ED visits

— Should be early and concurrent, not "end-stage only"

— Physician certifies LE ≤6 months if disease runs usual course

— Patient elects comfort-focused care, foregoes curative therapy for terminal diagnosis

— Covers nursing, aides, meds related to terminal diagnosis, equipment, bereavement

— Recertification at 90 days, 90 days, then 60-day periods; can extend indefinitely if still eligible

— Patients can revoke at any time

— Cancer: metastatic, declining performance status, ECOG ≥3

— Dementia: FAST stage 7, comorbid pneumonia/UTI/pressure ulcer, weight loss, dysphagia

— HF: NYHA IV, optimal therapy, EF ≤20%, recurrent admissions

— COPD: dyspnea at rest, O₂-dependent, FEV1 <30%, frequent ED visits

— ESRD: not on dialysis, CrCl <10, symptomatic uremia

— Stroke: poor functional recovery, dysphagia, weight loss, recurrent infections

Palliative care consult (concurrent with disease-directed therapy) — appropriate at any LE:

Hospice (Medicare hospice benefit):

Common hospice eligibility criteria by disease:

The Surprise Question: "Would you be surprised if this patient died within 12 months?" — a powerful screen for palliative referral

CCS pearl: On a CCS case of metastatic pancreatic cancer with ECOG 3, weight loss, uncontrolled pain, and frequent admissions, the high-yield orders are palliative care consult, hospice evaluation, opioid titration, advance directive completion, family meeting—and withholding further chemo/scans/aggressive labs. Order what changes management given the prognosis, not reflexive workup.

Key Differentials — Related Population Health Metrics

— Average remaining years from age x; derived from life table

— Period vs cohort calculation

— Time at which 50% of a defined cohort has died

— Preferred for skewed distributions (cancer); easier for patients to understand

— Proportion alive 5 yr after diagnosis

— Subject to lead-time bias (earlier diagnosis inflates apparent survival without delaying death) and length-time bias (screening catches slower-growing disease preferentially)

— Deaths per population per time (e.g., 8 per 1000 per year)

— Crude vs age-adjusted (standardized) — age-adjusted needed when comparing populations with different age structures

— Cause-specific mortality = deaths from specific cause / total population

— Case fatality rate = deaths from disease / cases of disease (severity measure)

— Observed deaths / expected deaths (based on reference population)

— SMR >1 = excess mortality; useful for occupational cohorts, hospitals

Distinguishing LE from neighboring metrics is high-yield because they're easily confused

Life expectancy (e_x):

Median survival:

Five-year survival:

Mortality rate:

Standardized Mortality Ratio (SMR):

Years of Potential Life Lost (YPLL): premature death burden

HALE/DALY/QALY: quality-adjusted (covered in chunk 5)

Key distinction: Crude vs age-adjusted mortality—a retirement community will have high crude mortality (old population) but normal age-adjusted mortality. Step 3 epidemiology stems often hinge on this distinction when comparing rates between populations or over time.

Key Differentials — Other Population Indicators and Health-System Metrics

— Crude birth rate: births per 1000 population

— Total fertility rate (TFR): lifetime births per woman; replacement ≈ 2.1

— Dependency ratio: (population <15 + >64) / population 15–64

— Aging populations (Japan, much of EU, increasingly US) → rising dependency ratio → strains on Medicare/SS

— YPLL (covered)

— Deaths before age 75 — useful policy benchmark

— Amenable mortality: deaths preventable with timely effective care (e.g., treatable cancers, appendicitis); higher in US than peer nations

— Preventable mortality: deaths preventable with public health interventions (tobacco, vaccination, injury prevention)

— HDI (Human Development Index): LE + education + income

— Bloomberg Healthiest Country Index, OECD Better Life Index

— Black-White LE gap narrows at very old ages ("crossover")—partly survival bias, partly differential mortality selection

Healthy life expectancy (HALE) vs LE: HALE subtracts years lived with disability

Disability-free life expectancy (DFLE): similar concept, threshold-based (no ADL dependence)

Active life expectancy: years without functional impairment; key in geriatrics research

Fertility and demographic measures (not LE but often co-tested):

Premature mortality indicators:

Healthcare system performance:

Composite indices:

Mortality crossover and racial disparities:

Important Step 3 trap: confusing incidence (new cases) with mortality; survival improvements can change mortality without changing incidence (cancer screening can do both, in different directions)

Board pearl: When a stem describes country/hospital comparisons of "death rates," check whether rates are age-adjusted. Crude rates mislead when populations differ in age structure. The right answer often involves recognizing this confounder and recommending age-standardization or stratification.

Long-Term Plan — Integrating LE Into the Care Plan

— Estimated LE range (e.g., "5–10 years") with brief rationale (age, comorbidities, function, ePrognosis output)

— Screening decisions: which to continue, which to stop, with reasoning

— Preventive medication review: continue, deprescribe, or shorten

— Symptom management goals

— Advance directives: healthcare proxy, living will, POLST/MOLST as appropriate

— Code status: full code, DNR, DNI, comfort measures only

— Identified caregivers and decision-makers

— Stop: primary prevention statins (LE <1–2 yr), bisphosphonates after 5 yr, aspirin for primary prevention in older adults (USPSTF now recommends against starting ≥60)

— Loosen: A1c target to <8–8.5%, BP target individualized

— Continue: symptom meds, secondary prevention with short lag (anticoagulation, BP control)

— Age 65+ wellness visit (Medicare ACP benefit, billable code)

— New serious diagnosis

— Functional decline

— Hospital discharge after major event

— Portable medical orders, not just preferences

— Travels across care settings (home, hospital, SNF, EMS)

— Distinct from living will (statement of values) and healthcare proxy (decision-maker)

LE-informed care plans should be explicit, documented, and revisited

Components of an LE-aligned longitudinal plan:

Specific deprescribing/continuing decisions in limited LE:

Advance care planning conversations: should occur with all adults, especially at:

POLST/MOLST:

Step 3 management: At the Medicare annual wellness visit for a 72-year-old, include review of advance directives, fall risk, cognitive screen, depression screen, and functional assessment — all of which inform individualized LE estimation and downstream screening/prevention decisions. The visit is a structured Step 3-favorite touchpoint for LE-informed care planning.

Follow-Up, Monitoring, and Counseling Cadence

— Annual Medicare wellness visit for 65+

— Every 1–2 yr functional/cognitive assessment in 75+

— After every hospitalization, especially for pneumonia, hip fracture, HF

— After any new diagnosis altering trajectory (cancer, dementia, advanced organ failure)

— New ADL dependence

— Unintentional weight loss >5% in 6–12 mo

— Gait speed decline

— Cognitive decline (≥2 point MoCA drop)

— Recurrent infections, pressure injuries

— Falls with injury

— Use plain language: "About half of people like you live more than X years; half less"

— Acknowledge uncertainty: ranges, not point estimates

— Frame around values: "What matters most to you in the time ahead?"

— Use Serious Illness Conversation Guide or Vital Talk frameworks

— Ask permission before sharing prognosis ("Would it be helpful to talk about what to expect?")

— Survivorship care plans should address surveillance, late effects, secondary prevention, return to function

— Cardiac rehab post-MI/CABG/HF: improves LE and QoL across age groups when functional status permits

— Pulmonary rehab in COPD: GOLD recommends across stages; improves dyspnea, exercise capacity

— PT/OT for fall prevention, ADL training in frail elders

LE estimates are dynamic—reassess at defined intervals and trigger points

Routine reassessment intervals:

Monitoring parameters that signal LE decline:

Counseling principles:

Cancer survivorship and chronic disease counseling:

Rehabilitation:

CCS pearl: On a CCS case of an older patient discharged after pneumonia or hip fracture, schedule a 1–2 week follow-up and order PT/OT, medication reconciliation, advance care planning revisit, and caregiver assessment. Post-hospitalization is when LE often resets downward—and is the optimal moment to re-individualize the plan.

Ethical, Legal, and Patient Safety Considerations

— Patients have the right to know their prognosis; physicians have a duty to disclose when patients want to know

— Avoid therapeutic privilege misuse (withholding prognosis "to spare" patient); rarely justified in modern practice

— Cultural humility: some patients/families prefer indirect or family-mediated disclosure—ask: "Are you the kind of person who likes to know all the details, or do you prefer I share with your family?"

— Decision-specific, not global

— Hierarchy when no proxy designated (state-dependent): spouse → adult children → parents → siblings

— Substituted judgment (what would patient want) preferred over best interest when proxy knows patient

— Denying treatment based on age alone is unethical and often illegal (Age Discrimination Act)

— Decisions must be based on individualized risk-benefit, not chronologic age

— Withdrawal vs withholding: ethically and legally equivalent

— Double effect: appropriate symptom management (opioids for dyspnea) is ethical even if it may hasten death

— POLST is not a DNR alone—covers full spectrum of orders

— Medical aid in dying: legal in some states with strict criteria (terminal diagnosis, LE ≤6 mo, capacity, waiting period); know it exists, know not all states permit

— Elder abuse/neglect: mandatory in most states; suspicion of abuse in declining elder requires APS report

— Hospital → SNF or home: medication errors, missed follow-up, code status not transmitted

— POLST and code status must accompany transfer

Informed consent and prognosis disclosure:

Capacity and surrogate decision-making:

Avoiding ageism:

End-of-life ethical issues:

Mandatory reporting:

Transition-of-care risks:

Step 3 management: A vignette describes a hospitalized patient with metastatic cancer whose family asks you "not to tell him" his prognosis. The correct response: explore the family's concerns, then gently ask the patient how much he wants to know. Respect autonomy by asking the patient directly, not by accepting unilateral family veto—unless the patient himself has previously delegated information receipt.

High-Yield Associations and Rapid-Fire Facts

— Mammo / colon / PSA ≈ 10 yr

— Statin primary prevention ≈ 2–5 yr

— Tight glucose ≈ 8 yr

— BP control ≈ 1–2 yr

— Anticoagulation for AF ≈ months

US LE at birth (2023): ~77.5 yr (M ~74.8, F ~80.2)

US LE at 65: ~19 yr; at 75: ~12 yr; at 85: ~6–7 yr

US infant mortality: ~5.4/1000; maternal mortality: ~22/100,000

COVID-19 dropped US LE ~2.7 yr from 2019 to 2021

Top causes of death US adults: heart disease, cancer, unintentional injury, COVID/respiratory, stroke, Alzheimer's

Top causes age <45: unintentional injury (incl. overdose), suicide, homicide

Black-White LE gap: ~5 yr; AI/AN gap larger

Income LE gap (top vs bottom 1%): ~15 yr (M), ~10 yr (F)

Lag-to-benefit memory aids:

Hospice eligibility: LE ≤6 mo if disease runs usual course; physician certification

Surprise question: "Would I be surprised if this patient died in next year?" — if NO, consider palliative care

ePrognosis indices: Lee (community), Schonberg (women), Walter (post-hospital), Charlson

Disease-specific scores: MELD-Na (liver), Seattle HF, BODE (COPD), FAST (dementia), ECOG/Karnofsky (cancer)

Crude vs age-adjusted mortality: always age-adjust when comparing

Lead-time vs length-time bias: inflate apparent survival from screening

Period vs cohort life table: period uses current rates; cohort follows real birth cohort

DALY = YLL + YLD; QALY is benefit, DALY is burden

Beers Criteria drugs: benzodiazepines, anticholinergics, glyburide, NSAIDs in CKD

AGS frailty BP target: SBP 130–150 acceptable; A1c <8–8.5%

Aspirin primary prevention: USPSTF recommends against initiating ≥60

Board pearl: When a Step 3 stem includes age + functional dependence + advanced organ disease and asks the "next best step," scan answer choices for deprescribe / stop screening / hospice / palliative consult / advance directive before considering any new test or aggressive intervention.

Board Question Stem Patterns

Pattern 1 — Stop screening: Frail 84-year-old with dementia, ADL-dependent, asks about screening colonoscopy → answer: recommend against screening; discuss goals of care

Pattern 2 — Continue screening: Healthy active 72-year-old, normal cognition, no chronic disease, last mammogram 2 yr ago → continue mammography

Pattern 3 — Deprescribing: 80-year-old with mild dementia, falls, on atorvastatin for primary prevention, simvastatin, glyburide, diphenhydramine → identify inappropriate meds (Beers) and stop primary prevention statin given limited LE

Pattern 4 — A1c target: Frail 82-year-old diabetic with falls, on insulin, A1c 6.8% → loosen target to <8%, reduce insulin to prevent hypoglycemia

Pattern 5 — Hospice eligibility: Patient with metastatic cancer, ECOG 3, weight loss, declining function, prognosis ~3 mo → refer to hospice

Pattern 6 — Advance care planning: Hospitalized patient post-second HF admission this year, NYHA IV, EF 15%, not transplant candidate → palliative care consult, goals-of-care discussion, consider hospice

Pattern 7 — Life table calculation: Given l(60)=85,000 and l(70)=70,000, probability of dying between 60 and 70 = (85,000−70,000)/85,000 = 17.6%; survival probability = 70,000/85,000 = 82.4%

Pattern 8 — Mortality comparison: Two populations have different crude mortality rates → answer involves age-adjustment to compare fairly

Pattern 9 — Screening bias: Improved 5-yr survival after screening introduced → may reflect lead-time bias, not true mortality benefit; mortality rate (denominator population) is gold standard

Pattern 10 — Disparities: Lower LE in a population subgroup → answer involves social determinants, structural inequity, healthcare access—not biologic difference

Pattern 11 — Surrogate decisions: Patient lacks capacity, no advance directive, family disagrees → use substituted judgment via legal proxy hierarchy

Board pearl: The Step 3 "wrong" answers in LE vignettes are usually more testing, more meds, more intervention. The "right" answers are usually individualization, deprescribing, conversation, and alignment with values and prognosis.

One-Line Recap

In Step 3 practice, life expectancy is the lens through which every screening, prevention, and major-intervention decision in older or seriously ill adults must be filtered—matched against lag time to benefit—so that care aligns with how much time and function the patient actually has, not with chronologic age or reflexive protocol.

Life table mechanics: e(x) = T(x)/l(x); period tables use current mortality, cohort tables follow real cohorts; survival between ages = l(later)/l(earlier)

Lag time vs LE: screening helps only when LE > lag (mammo/colon/PSA ~10 yr; statin primary prevention ~2–5 yr; tight glucose ~8 yr); when LE is shorter, stop screening and deprescribe rather than add care

Prognostic individualization: function (gait speed, ADLs), cognition, comorbidity, and disease-specific tools (MELD, Seattle HF, BODE, FAST, ECOG) outperform age and population life tables for clinical decisions; use ePrognosis indices (Lee, Schonberg, Walter)

End-of-life triggers: Surprise Question NO → palliative care; physician-certified LE ≤6 mo → hospice; advance care planning, POLST, code status, and ICD deactivation are the high-yield "do this" answers when vignettes describe frailty, advanced organ failure, or metastatic disease

Equity lens: US LE disparities by race, income, and geography reflect social determinants; age-adjusted rates and standardized comparisons are required to compare populations fairly—and Step 3 expects you to recognize when crude rates, lead-time bias, or chronologic-age anchoring are misleading the clinical or epidemiologic answer