← Master Index· Appendices & Drug-Induced Disease

Geriatric Assessment and Pharmacotherapy

GeriatricsBeers CriteriaPolypharmacyAppendix

30-Second Snapshot

What it is:A reference framework, not a disease. It's the set of physiologic changes, pharmacokinetic shifts, and screening tools you need to layer on top of every otherchapter once your patient is old. Age itself isn't a diagnosis, but it changes how every drug you've already learned behaves in the body.

The core problem:Aging organs handle drugs differently (mostly they clear them slower and are more sensitive to them), while aging bodies accumulate more diagnoses, which means more prescribers, more drugs, and more chances for those drugs to fight each other or the patient's other diseases. Polypharmacy is the natural output of correct guideline-following medicine applied one condition at a time without anyone standing back and looking at the whole list.

What you do about it:Before you add a drug to an older adult, run it through the physiology (will they clear this?), the screening tools (is this a Beers/STOPP flag?), and the Medication Appropriateness Index (does this drug still earn its spot on the list?). Start low, go slow, but don't undertreat just because someone is old.

Worth knowing

This appendix is the connective tissue for every geriatric patient scattered through the other chapters. The exam trap is treating it as trivia. It isn't. It's the lens: same disease, same drug class, different rulesonce the patient is 75 instead of 45.

Physiologic Changes with Aging

Four organ systems drift in predictable directions with age. None of these changes are diseases by themselves, they're just the new baseline you're prescribing against.

SystemWhat changes
Renal↓ Glomerular filtration rate, ↓ renal blood flow, ↓ number of functioning nephrons, ↓ tubular secretory function, ↓ renal mass
SensoryPresbyopia (can't focus on near objects), ↓ night vision, presbycusis (high-pitch/high-frequency hearing loss), ↓ smell and taste
Musculoskeletal↓ Skeletal bone mass (osteopenia), ↓ muscle mass, joint stiffening from reduced tendon/ligament/cartilage water content, altered gait and posture
Skin / HairThinning stratum corneum, ↓ melanocytes, ↓ subcutaneous fat depth, atrophy of sweat glands, thinning and graying hair
Why this matters beyond the exam

Every row on that table has a downstream prescribing consequence. Renaldecline is the big one, it's the reason half this appendix exists. Sensoryloss means a patient literally can't read a small pill bottle label or hear you explain a taper over the phone, which shows up as "nonadherence" when it's really an access problem. Musculoskeletalchanges drive fall risk, which is exactly why CNS-active drugs get flagged so hard in Beers. Skinchanges mean thinner skin bruises more easily on anticoagulants, and topical or transdermal absorption can shift because the barrier is different.

Pharmacokinetic Changes with Aging

This is the section that actually changes your dosing decisions. Walk it phase by phase, in order: absorption, distribution, metabolism, excretion.

ABSORPTION

Mostly unchanged

Passive diffusion and bioavailability are stable for most drugs. Active transport drops, lowering bioavailability for a subset of drugs. First-pass metabolism falls, which raisesbioavailability for some drugs but lowersit for prodrugs that need that first pass to activate.

DISTRIBUTION

Depends on the drug's solubility

↓ Body water shrinks the volume of distribution for water-soluble drugs, so plasma concentration rises. ↑ Body fat expands the volume of distribution for lipid-soluble drugs, which stretches out the terminal half-life.

METABOLISM

Slower, unevenly

↓ Clearance and ↑ half-life for capacity-limited drugs (Phase I hit harder than Phase II). Drugs with high hepatic extraction ratios (flow-limited metabolism) also clear slower because hepatic blood flow itself declines.

EXCRETION

The one that bites you most often

↓ Clearance and ↑ half-life for renally eliminated drugs and their active metabolites. This is the single most common reason a "normal" adult dose becomes toxic in an older patient.

The distinction that gets tested

Water-soluble drug → smaller volume of distribution → higher plasma level per dose.Think digoxin: a frail, dehydrated older adult has proportionally less body water to dilute the same milligram dose into, so levels run higher than you'd predict from weight alone. Lipid-soluble drug → bigger volume of distribution → longer terminal half-life.Think a benzodiazepine: more body fat to sequester the drug means it keeps leaching back into circulation long after you'd expect it to be gone, which is exactly how an older adult ends up sedated and falling three days after a "one-time" dose.

Serum creatinine lies in older adults

↓ Muscle mass means less creatine turned into creatinine, so serum creatinine can sit inside "normal" range even while the actual GFR has dropped substantially. A SCr of 0.9 mg/dL in a frail 88-year-old woman is not reassuring the way it would be in a muscular 30-year-old. Always estimate renal function with a formula, never eyeball the raw creatinine, and remember that formula itself is only as good as the weight and age inputs you feed it.

Pharmacodynamics shifts too, though the appendix's own data here is thinner: older adults are generally more sensitive at a given plasma concentration to CNS depressants, anticholinergics, and antihypertensives, and less able to compensate hemodynamically for a given drop in blood pressure. That sensitivity is on top ofthe higher plasma levels the PK changes above already produce, not instead of them. The two stack.

Screening Tools: Beers Criteria vs STOPP

Two major frameworks exist to flag potentially inappropriate medications (PIMs) in older adults. They overlap heavily but aren't organized the same way, and mixing up which is which is a classic distractor.

CriteriaHow it's organized
Beers Criteria
(American Geriatrics Society)
Five separate tables:
1. PIMs for all older adults, by organ system and drug category
2. PIMs for older adults with specific drug-disease/drug-syndrome interactions (syncope, delirium, dementia, falls/fractures history, GI ulcer history, urinary incontinence, BPH, heart failure)
3. PIMs to use with caution (e.g., aspirin for primary prevention, dabigatran/rivaroxaban in patients >75 for GI bleed risk)
4. Clinically important drug-drug interactions to avoid (opioids, lithium, phenytoin, warfarin; avoid ≥3 concurrent CNS-active drugs to cut fall risk)
5. Medications to avoid or dose-reduce by kidney function
STOPP
(Screening Tool of Older Persons' Prescriptions)
Organized by physiological system across 11 sections, covering drugs to avoid, drug-drug interactions, dosing for reduced renal function, and drug-disease interactions. A separate general-concepts section flags prescribing without an evidence-based indication, prescribing beyond the recommended duration, and therapeutic duplication.
The distinction that gets tested

Beers is organized by drug/drug category first.You look up the drug and it tells you the concern. STOPP is organized by physiologic system first, and it also catches prescribing-process problems Beers doesn't explicitly structure for, like duplicate therapy or a drug that's outlived its indication. STOPP has a companion tool, START(Screening Tool to Alert doctors to Right Treatment), that flags under-prescribing, the mirror problem to PIMs. Beers doesn't have a formal "you're missing a drug" counterpart built the same way.

Worth knowing

Rule number 4 of Beers (the ≥3 concurrent CNS-active drug ceiling) is one of the highest-yield single facts in this appendix. It's not about any one drug being bad, it's that sedative burden is additive across an antidepressant, a muscle relaxant, a sleep aid, and an opioid even when each one individually looks "fine" on its own Beers check.Falls risk in an older adult is often a polypharmacy problem, not a single-drug problem.

Medication Appropriateness Index - The 10 Questions

Where Beers and STOPP are lists of specific drugs and interactions to check against, the Medication Appropriateness Index (MAI)is a process you run on every single medicationon the list, one at a time, regardless of what the drug is. Ask all ten questions for each drug before you accept it belongs on the regimen.

#Question
1Is there an indication for the medication?
2Is the medication effective for the condition?
3Is the dosage correct?
4Are the directions correct?
5Are the directions practical?
6Are there clinically significant drug-drug interactions?
7Are there clinically significant drug-disease interactions?
8Is there unnecessary duplication with another medication?
9Is the duration of therapy acceptable?
10Is this the least expensive alternative compared with others of equal utility?
How to actually use the MAI

Notice the order. Questions 1 and 2 ask whether the drug should exist on this list at all. Questions 3 through 5 ask whether it's dosed and directed right for thispatient. Questions 6 through 9 look outward at how the drug interacts with everything else going on. Question 10 is the tiebreaker once everything else checks out. Run it in that orderand you catch "why is this still here" problems before you get lost tweaking a dose on a drug that shouldn't be prescribed in the first place.

High-Risk Drug-Disease Interactions

These are expert-consensus pairings where a drug class predictably worsens a specific geriatric condition, independent of any single-drug Beers flag. Know the mechanism behind each one, not just the pairing, because the mechanism is what lets you generalize to a drug not explicitly on this list.

Disease / ConditionDrug or class to avoid
Heart failureNSAIDs, COX-2 inhibitors, thiazolidinediones, dronedarone
HFrEF specificallyNon-dihydropyridine calcium channel blockers
SyncopeAcetylcholinesterase inhibitors, peripheral nonselective α1-blockers, tertiary tricyclic antidepressants, antipsychotics
DeliriumAnticholinergics, antipsychotics, benzodiazepines, corticosteroids, H2-receptor antagonists, zolpidem/zaleplon ("z-hypnotics")
Cognitive impairmentAnticholinergics, antipsychotics, benzodiazepines, nonbenzodiazepine hypnotics, z-hypnotics
History of fallsAntiepileptics, antipsychotics, benzodiazepines, antidepressants, z-hypnotics, opioids
Parkinson diseaseAntipsychotics (except quetiapine, clozapine, pimavanserin), metoclopramide, prochlorperazine, promethazine
Peptic ulcer diseaseAspirin ≥325 mg/day, NSAIDs
CKD stage IV and higherNSAIDs
Urinary incontinence (women)Peripheral α1-blockers
The pattern underneath the list

Four rows in that table (syncope, delirium, cognitive impairment, falls) are all hit by the same core group of drugs: anticholinergics, antipsychotics, benzodiazepines, and other sedating/CNS-active agents. That's not four separate facts to memorize, it's one mechanism (CNS depression and anticholinergic burden) showing up as four different clinical facesdepending on which older adult you're looking at. Learn that cluster once and you've covered most of the table.

Why NSAIDs show up twice

NSAIDs block renal prostaglandin synthesis, which is protective vasodilation the kidney leans on more heavily once it's already got fewer nephrons and lower renal blood flow. That's why NSAIDs are dangerous in both heart failure(sodium and water retention worsen congestion) and advanced CKD(further drops renal perfusion in a kidney with no reserve left). Same drug, same underlying mechanism, two different disease entries on the table.

Atypical Disease Presentation in Older Adults

Classic textbook presentations were mostly written from younger patients. Older adults blunt or skip the "obvious" symptom and present with something nonspecific instead, which is exactly why geriatric disease gets missed or diagnosed late.

DiseaseHow it actually shows up
Acute MIOnly about 50% present with chest pain. More often: weakness, confusion, syncope, abdominal pain. ECG findings, notably, still look like they do in younger patients, so the ECG stays reliable even when the story doesn't.
Heart failureInstead of dyspnea, may present with hypoxic symptoms, lethargy, restlessness, and confusion.
GI bleedMortality runs around 10%, but the presenting symptoms are nonspecific, from altered mental status to syncope with hemodynamic collapse. Abdominal pain is frequently absent.
Upper respiratory infectionLethargy, confusion, anorexia, and decompensation of a preexisting condition. Fever, chills, and productive cough may or may not show up at all.
Urinary tract infectionDysuria, fever, and flank pain may be absent. More commonly: incontinence, confusion, abdominal pain, nausea/vomiting, and azotemia.
The trap this sets clinically

Look at the pattern: MI, HF, GI bleed, URI, and UTI all can present as "confused" or "just not themselves."A family member saying grandma is "more confused than usual" is not automatically a dementia or delirium workup, it can be the only presenting sign of a heart attack, an infection, or a bleed. New confusion in an older adult is a symptom that needs a cause, not a diagnosis by itself.

Putting It Together: A Prescribing Approach

None of the tools above work in isolation. Layer them every time you're starting, continuing, or reviewing a drug in an older adult.

  1. Estimate real renal functionbefore you dose anything renally cleared. Don't trust a "normal-looking" serum creatinine in a low-muscle-mass patient, it can hide a substantially reduced GFR.
  2. Check the drug against Beers and STOPP.Is it a PIM outright, a caution-flag drug, or a drug-disease interaction for a condition this patient actually has?
  3. Check it against Table A2-5's drug-disease listspecifically for delirium, falls, syncope, or cognitive impairment risk, since those four are the highest-frequency traps.
  4. Run the drug through all ten MAI questions.Especially question 8 (duplication) and question 9 (duration), since those catch the slow accumulation that produces polypharmacy in the first place.
  5. Start low, go slow, but don't stop short.The PK changes above mean a lower starting dose and a slower titration interval, not necessarily a lower final target. Undertreating pain, depression, or heart failure because someone is old is its own harm.
Worth knowing

"Start low, go slow" gets repeated so often it becomes background noise. The part people forget is the second half: go slow doesn't mean stop early.A geriatric patient undertreated for a genuinely treatable condition because titration stalled at a subtherapeutic dose is just as much of a prescribing failure as one who's over-medicated.

Deprescribing & Managing the Whole List

Polypharmacy isn't caused by any one bad prescribing decision, it's the sum of correct, guideline-driven decisions made by different prescribers treating different diagnoses without ever comparing notes. Deprescribing is the deliberate, periodic process of reviewing the whole list and removing what no longer earns its place.

A framing worth keeping

Deprescribing is not the opposite of good prescribing, it's the other half of it. Every drug you leave on a list unreviewed carries the same interaction, adherence, and toxicity risk as one you'd never have started fresh today.

Monitoring - What, When, Why

ParameterWhenWatching for
Estimated renal functionBaseline, then whenever a renally cleared drug is started, dose-adjusted, or the patient is acutely illA GFR drop the raw serum creatinine won't show because of reduced muscle mass
Full medication list, all prescribersEvery visit, and at every care transition (admission, discharge, new specialist)Duplication, prescribing cascades, drugs whose indication has expired
CNS-active drug countEvery visitCrossing the ≥3 concurrent threshold that sharply raises fall risk
Mental status / new confusionAny acute change reported by patient or caregiverAn underlying MI, infection, bleed, or new drug-disease interaction hiding behind "just confusion"
Falls historyEvery visitA new or worsened flag for the Table A2-5 falls-risk drug classes
Adherence and functional capacityEvery visitVision, hearing, and dexterity limits that make a "correct" regimen impractical to actually take

Patient Counseling - What You'll Actually Say

  • Face the patient, speak clearly and a bit slower, don't shout.Presbycusis is high-pitch loss first, so dropping your pitch slightly can help more than raising your volume.
  • Write it down in large printin addition to saying it out loud. "Take one tablet twice daily" printed in a normal 9-point font is genuinely hard to read with presbyopia.
  • Ask about the whole list, not just today's prescription:"Can you walk me through everything you're taking, including anything over the counter or from another doctor?" This is how you actually catch duplication and interacting OTCs.
  • Explain the start-low-go-slow plan out loudso a lower-than-expected starting dose doesn't read as the prescriber being unsure: "We're starting you lower than usual because your body processes this differently now, and we'll increase it carefully."
  • Ask directly about falls:"Have you had any falls or near-falls in the last year?" Patients underreport this unprompted.
  • Set the expectation that new confusion needs a call, not a wait-and-see:"If you or your family notice you're suddenly more confused than usual, that's worth calling us about right away, it can be a sign of something else going on."
  • Involve a caregiver when one exists,especially for regimens with complex directions or when sensory decline limits independent management.

High-Yield Recall Sheet

  • Water-soluble drugs→ ↓ volume of distribution with age → higher plasma concentration per dose (think digoxin).
  • Lipid-soluble drugs→ ↑ volume of distribution with age → longer terminal half-life (think benzodiazepines).
  • Serum creatinine can look normaldespite a real drop in GFR, because ↓ muscle mass means less creatinine production. Always estimate, don't eyeball.
  • Beers Criteriais organized around five tables of drug/drug-category flags. STOPPis organized by 11 physiologic-system sections and also flags process problems (no indication, wrong duration, duplication).
  • STARTis STOPP's under-prescribing counterpart; Beers doesn't have a formally paired equivalent.
  • Beers rule of thumb:avoid ≥3 concurrent CNS-active drugs to reduce fall risk.
  • MAI has 10 questions, run in order: indication and effectiveness first, then dose/directions, then interactions/duplication/duration, then cost.
  • Anticholinergics, antipsychotics, benzodiazepinescluster across syncope, delirium, cognitive impairment, and falls, that's one mechanism showing up four ways, not four separate facts.
  • NSAIDsare flagged in both heart failure and CKD stage IV+ through the same mechanism: blocked renal prostaglandins removing protective vasodilation from an already-compromised kidney.
  • Non-dihydropyridine CCBsare specifically a HFrEF flag, not a general heart failure flag.
  • Only ~50% of older adults with an MIpresent with chest pain; the ECG remains reliable even when symptoms don't fit the classic story.
  • "More confused than usual"is a symptom needing a cause (MI, infection, GI bleed, drug interaction), not an automatic dementia workup.
  • UTI in an older adultoften presents with confusion and incontinence rather than dysuria, fever, or flank pain.
  • Deprescribing is not the opposite of prescribing,it's the review step every list needs at every reconciliation.
  • Start low, go slowmeans adjust the titration pace, not the eventual treatment goal. Don't undertreat just because the patient is old.