Medical Pharmacology Question Bank

Chapter 1: General Pharmacology — Module 6: Special Populations
Tier: Tier 1 — Foundational Recall

1. The WHO six-step rational prescribing process begins with which of the following as its first and most fundamental step?

A) Selecting a drug from the national formulary that matches the patient's diagnosis code
B) Calculating the correct dose based on the patient's weight and renal function before identifying the clinical problem
C) Defining the patient's problem — establishing an accurate diagnosis or clinical problem statement that drives all subsequent prescribing decisions
D) Writing the prescription legibly and completely, including drug name, dose, route, frequency, and duration
E) Verifying that the selected drug is listed on the WHO Essential Medicines List before initiating therapy

ANSWER: C

Rationale:

The WHO six-step rational prescribing process, as described in the "Guide to Good Prescribing" (de Vries et al., WHO 1994), proceeds in a deliberate sequence: Step 1 — Define the patient's problem; Step 2 — Specify the therapeutic objective; Step 3 — Verify that the chosen treatment is suitable for this patient; Step 4 — Start treatment; Step 5 — Provide information, instructions, and warnings; Step 6 — Monitor treatment and stop if necessary. Step 1 is foundational because all subsequent decisions flow from an accurate problem definition. An imprecise or incorrect problem statement leads to inappropriate therapeutic objectives, wrong drug selection, and preventable harm. Defining the patient's problem requires integrating the presenting complaint, history, examination findings, and relevant investigations into a clinical problem statement that is specific enough to guide drug selection — not merely assigning a diagnostic code. Option A is incorrect — formulary selection is a component of Step 3, not Step 1. Option B is incorrect — dose calculation is part of Step 4 and requires prior completion of Steps 1–3. Option D is incorrect — writing the prescription is part of Step 4 after the drug, dose, and regimen have been determined through Steps 1–3. Option E is incorrect — the Essential Medicines List is a resource tool; while it informs Step 3, consulting it is not the first step of the prescribing process, and prescribing is not limited to Essential Medicines in all clinical contexts.

2. Therapeutic drug monitoring (TDM) is most clinically indicated for which of the following drug characteristics?

A) Drugs with wide therapeutic windows, predictable pharmacokinetics, and linear dose-response relationships where plasma concentrations reliably correlate with clinical outcome
B) Drugs with narrow therapeutic indices, significant interindividual pharmacokinetic variability, and where plasma concentration correlates with both therapeutic effect and toxicity — particularly drugs where clinical endpoints of efficacy or toxicity are difficult to assess in real time
C) All drugs administered orally, because gastrointestinal absorption variability makes plasma concentration monitoring universally necessary for dose adjustment
D) Drugs that are highly protein-bound, because free drug concentration cannot be estimated from total plasma concentration without direct measurement in all clinical contexts
E) Drugs that are renally eliminated, because TDM is exclusively used to adjust dosing intervals in patients with chronic kidney disease

ANSWER: B

Rationale:

Therapeutic drug monitoring — the measurement of drug plasma concentrations to guide dosing decisions — is most clinically valuable when specific pharmacological characteristics make empirical dosing unreliable and plasma concentration measurement actionable. The classical indications for TDM are: (1) narrow therapeutic index — a small difference between effective and toxic concentrations means that even modest pharmacokinetic variability can produce toxicity or therapeutic failure (e.g., digoxin, lithium, phenytoin, vancomycin, aminoglycosides, cyclosporine, tacrolimus, warfarin via INR); (2) significant interindividual pharmacokinetic variability — due to genetic polymorphisms (CYP2D6, CYP2C19), disease states (renal failure, hepatic cirrhosis, heart failure), drug interactions, age extremes, or body composition differences that make standard dosing unreliable; (3) correlation between plasma concentration and clinical effect — TDM is only useful if there is an established pharmacokinetic-pharmacodynamic relationship that allows concentration data to guide dose adjustment; (4) clinical endpoints difficult to assess in real time — when the therapeutic endpoint (e.g., seizure suppression, transplant rejection prevention) cannot be continuously monitored at the bedside, plasma concentration serves as a surrogate endpoint. Option A describes drugs where TDM provides no additional value — wide therapeutic index drugs with predictable PK require no routine monitoring. Option C is incorrect — oral route alone does not mandate TDM; most oral drugs with predictable absorption and wide therapeutic windows require no plasma concentration monitoring. Option D is incorrect — high protein binding is one factor considered in TDM interpretation (free drug concentration may differ from total), but it is not itself a sufficient indication for TDM. Option E is incorrect — while renal impairment is a common clinical context triggering TDM (for renally eliminated drugs), TDM indications extend far beyond renal dosing and are not exclusively tied to CKD.

3. The American Geriatrics Society Beers Criteria identify medications that are potentially inappropriate for use in older adults. Which of the following best describes the pharmacological basis for why many drugs appear on the Beers Criteria for elderly patients?

A) The Beers Criteria list drugs that are exclusively contraindicated in patients over 65 years due to regulatory agency withdrawal of approval for geriatric use — prescribing these drugs in elderly patients constitutes off-label use in all circumstances
B) Drugs on the Beers Criteria are listed because elderly patients have impaired gastrointestinal absorption, making oral drug bioavailability universally reduced and standard doses subtherapeutic — the criteria identify drugs requiring dose increases in the elderly
C) The Beers Criteria identify drugs whose pharmacokinetic or pharmacodynamic profiles create disproportionate risk in elderly patients — age-related reductions in renal clearance, hepatic metabolism, lean body mass, plasma protein binding, and altered pharmacodynamic sensitivity (particularly CNS sensitivity to anticholinergic, sedative, and opioid drugs) combine to increase drug accumulation and adverse effect risk; polypharmacy amplifies these risks through drug interactions; drugs on the list are those where the risk-benefit ratio is unfavorable specifically in elderly populations, often because safer alternatives exist
D) The Beers Criteria exclusively identify drugs with renal dose adjustment requirements — any drug requiring eGFR-based dose reduction is automatically included in the criteria regardless of its specific risk profile in elderly patients
E) The Beers Criteria apply only to patients in long-term care facilities and have no relevance to community-dwelling elderly patients or acute hospital settings

ANSWER: C

Rationale:

The American Geriatrics Society Beers Criteria (updated most recently in 2023) represent an evidence-based, expert consensus-developed list of potentially inappropriate medications (PIMs) in adults aged 65 and older. The pharmacological rationale for inclusion reflects the intersection of age-related pharmacokinetic and pharmacodynamic changes with specific drug risk profiles. Age-related pharmacokinetic changes: reduced GFR (average decline approximately 1 mL/min/1.73m² per year after age 40) increases accumulation of renally eliminated drugs; reduced hepatic blood flow and mass reduces first-pass extraction and systemic CYP-mediated metabolism of high-extraction drugs; reduced serum albumin increases free fraction of highly protein-bound drugs; reduced lean body mass reduces Vd for drugs that distribute into muscle (digoxin); increased body fat increases Vd and prolongs half-life for lipophilic drugs (benzodiazepines, many CNS drugs). Age-related pharmacodynamic changes: increased CNS sensitivity to anticholinergic drugs (cognitive impairment, delirium, urinary retention, constipation, falls); increased sensitivity to opioid-mediated respiratory depression; increased sensitivity to sedative-hypnotics (falls, fractures, delirium); altered baroreceptor reflexes increasing orthostatic hypotension risk with antihypertensives, alpha-blockers, and antipsychotics. Polypharmacy (most elderly patients take five or more medications) amplifies drug interaction risk. Beers Criteria categories include: drugs always to avoid regardless of indication in older adults (e.g., first-generation antihistamines, muscle relaxants, meperidine); drugs to avoid in specific disease states (e.g., NSAIDs in CKD, benzodiazepines with respiratory disease); drug-drug interactions to avoid; drugs to use with caution and dose adjustments in elderly. Option A is incorrect — Beers Criteria drugs are not regulatory withdrawals of approval; many are approved for use in elderly patients but carry unfavorable risk-benefit ratios compared to safer alternatives. Option B is incorrect — gastrointestinal absorption is generally well-preserved in elderly patients; the primary pharmacokinetic changes are in renal and hepatic clearance and distribution, not absorption. Option D is incorrect — Beers Criteria encompass pharmacodynamic as well as pharmacokinetic concerns and include many drugs not requiring renal dose adjustment. Option E is incorrect — Beers Criteria apply across all healthcare settings — community, hospital, and long-term care — and are widely used for medication review in all clinical contexts involving elderly patients.

4. The FDA Pregnancy and Lactation Labeling Rule (PLLR) replaced the former A, B, C, D, X letter category system in 2015. Which of the following best describes a key improvement the PLLR provides over the former letter system?

A) The PLLR assigns numerical risk scores from 1–10 to each drug based on quantitative teratogenicity data from animal studies, providing a more precise risk stratification than the former letter categories
B) The PLLR eliminates all risk classification for drugs in pregnancy — it mandates that clinicians consult the drug manufacturer directly for pregnancy risk information rather than relying on standardized labeling
C) The PLLR provides structured narrative summaries in three sections — Pregnancy (including a Pregnancy Exposure Registry where available), Lactation, and Females and Males of Reproductive Potential — replacing the letter categories that were widely misinterpreted as absolute safety rankings; the narrative format requires specific data on human and animal studies, risk summaries, and clinical considerations, enabling more nuanced prescribing decisions in pregnancy and lactation
D) The PLLR applies only to drugs approved after 2015 — drugs approved before this date retain their original A, B, C, D, X letter categories indefinitely, creating a dual-system that will persist until all pre-2015 drugs are reformulated
E) The PLLR replaced the letter system with trimester-specific contraindication codes that specify whether a drug is safe in the first, second, or third trimester independently — a system that more accurately captures the timing-dependent teratogenicity of many drugs

ANSWER: C

Rationale:

The FDA Pregnancy and Lactation Labeling Rule (PLLR), implemented in 2015, fundamentally restructured how pregnancy and lactation risk information is communicated in drug prescribing information — addressing well-documented shortcomings of the former five-letter (A, B, C, D, X) categorization system. The principal failures of the letter system included: oversimplification — letters implied a linear safety ranking (A safest, X most dangerous) that did not reflect the complexity of teratogenicity data; cross-drug comparisons — clinicians and patients inappropriately compared drugs within the same letter category as equivalent in risk; misinterpretation of "C" — the most common letter, assigned when animal data showed adverse effects but human data were insufficient, was frequently misread as implying moderate safety rather than genuine uncertainty; and absence of lactation and reproductive potential information in a standardized format. The PLLR mandates three structured narrative subsections in prescribing information: (1) Pregnancy — including human and animal data summary, a risk summary statement, and a Pregnancy Exposure Registry reference where available; (2) Lactation — including data on drug presence in breast milk, effects on the breastfed infant, and effects on milk production; (3) Females and Males of Reproductive Potential — including pregnancy testing requirements, contraception recommendations, and effects on fertility. This narrative format requires specific data disclosure and clinical context, enabling more nuanced individualized prescribing decisions than a letter category could provide. Transition: drugs approved on or after June 30, 2015 must comply immediately; drugs approved between 2001 and June 2015 must be phased in over several years; drugs approved before 2001 are exempt from the new rule but many manufacturers have voluntarily updated labeling. Option A is incorrect — the PLLR uses narrative descriptions, not numerical scores. Option B is incorrect — the PLLR does not eliminate risk classification; it provides more detailed structured narrative risk information. Option D is incorrect — the transition timeline applies to when compliance is required, but the PLLR applies broadly across all approved drugs on a phased schedule. Option E is incorrect — the PLLR does not use trimester-specific contraindication codes; it uses narrative summaries that may include trimester-specific information where relevant.

5. The Child-Pugh score is used to assess hepatic function in the context of drug dosing. Which of the following most accurately describes what the Child-Pugh classification captures and its pharmacological relevance?

A) The Child-Pugh score measures hepatic CYP enzyme activity directly through a breath test, providing a quantitative assessment of a patient's metabolic capacity for CYP3A4 substrate drugs
B) The Child-Pugh score is a composite clinical and laboratory scoring system that stratifies the severity of hepatic dysfunction into classes A (mild), B (moderate), and C (severe) using five variables — serum bilirubin, serum albumin, prothrombin time/INR, ascites, and hepatic encephalopathy; higher Child-Pugh class correlates with greater impairment of hepatic drug metabolism (CYP enzyme activity, conjugation, oxidation), reduced first-pass extraction of high-clearance drugs, impaired hepatic blood flow, and reduced plasma protein synthesis — collectively predicting the degree of pharmacokinetic alteration requiring dose adjustment for hepatically eliminated drugs
C) The Child-Pugh score applies exclusively to alcohol-related liver disease — it is not validated for drug dosing decisions in patients with viral hepatitis, autoimmune hepatitis, or other causes of hepatic dysfunction
D) The Child-Pugh score measures hepatic blood flow using Doppler ultrasound parameters — its pharmacological relevance is exclusively to high-extraction ratio drugs whose clearance is blood flow-dependent
E) Child-Pugh Class A indicates severe hepatic impairment requiring the most aggressive dose reductions, while Class C indicates mild impairment requiring no dose adjustment — the scoring system is organized inversely to what clinicians typically assume

ANSWER: B

Rationale:

The Child-Pugh score (originally Child-Turcotte score, modified by Pugh) is a validated composite scoring system that integrates five clinical and laboratory parameters to estimate the overall functional reserve of the liver in patients with chronic liver disease. The five parameters and their scoring (each scored 1–3 points, total range 5–15): (1) Serum bilirubin (mg/dL): <2 = 1, 2–3 = 2, >3 = 3; (2) Serum albumin (g/dL): >3.5 = 1, 2.8–3.5 = 2, <2.8 = 3; (3) Prothrombin time prolongation (seconds) or INR: <4s/<1.7 = 1, 4–6s/1.7–2.3 = 2, >6s/>2.3 = 3; (4) Ascites: none = 1, mild = 2, moderate-severe = 3; (5) Hepatic encephalopathy: none = 1, grade 1–2 = 2, grade 3–4 = 3. Class A (5–6 points) = mild hepatic dysfunction; Class B (7–9 points) = moderate; Class C (10–15 points) = severe. Pharmacological relevance: hepatic drug clearance depends on hepatic blood flow (Q), intrinsic metabolic clearance (CLint, reflecting functional hepatocyte mass and enzyme activity), and plasma protein binding (fu). Child-Pugh class correlates broadly with all three: Q is reduced by portal hypertension; CLint is reduced by hepatocyte loss and reduced CYP enzyme expression; fu is increased by hypoalbuminemia. For high-extraction drugs (hepatic extraction ratio >0.7, e.g., morphine, propranolol, lidocaine), clearance is primarily blood flow-dependent and Child-Pugh-related reductions in hepatic blood flow produce the greatest proportional clearance impairment. For low-extraction drugs (e.g., warfarin, diazepam), clearance depends on CLint and fu; Child-Pugh-related hypoalbuminemia and reduced CLint both affect these drugs. Prescribing information for hepatically eliminated drugs typically specifies dose adjustments by Child-Pugh class. Option A is incorrect — the Child-Pugh score uses clinical and standard laboratory parameters, not breath tests measuring CYP activity directly (though 13C-aminopyrine or erythromycin breath tests exist as research tools for CYP activity). Option C is incorrect — Child-Pugh was originally developed for cirrhosis from any cause and is applied across etiologies of chronic liver disease. Option D is incorrect — while hepatic blood flow is one relevant parameter, the Child-Pugh score also captures hepatocellular synthetic function (albumin, prothrombin time), bilirubin excretion, and clinical complications — its pharmacological relevance extends beyond blood flow-dependent drugs. Option E is completely incorrect — Class A (lowest score, 5–6) represents the mildest dysfunction with the best-preserved function; Class C (highest score, 10–15) represents severe dysfunction.

6. Which of the following best defines the number needed to treat (NNT) and correctly interprets an NNT of 20 for a cardiovascular preventive therapy?

A) NNT of 20 means that 20% of patients will benefit from the therapy — the remaining 80% will experience harm from adverse drug effects
B) NNT of 20 means that 20 patients must be treated with the therapy for one patient to experience a specified adverse drug reaction — it is a measure of drug toxicity rather than efficacy
C) NNT of 20 means that 20 patients must be treated with the therapy (compared to placebo or comparator) for a defined period to prevent one additional adverse event (e.g., one additional myocardial infarction prevented) — a lower NNT indicates greater treatment benefit; NNT is always calculated for a specific outcome, population, and time horizon and cannot be generalized across different clinical contexts
D) NNT of 20 means that 20% of patients receiving the therapy will achieve the primary endpoint, compared to 0% in the control group — it directly reports the absolute event rate in the treatment group
E) NNT of 20 is a relative measure of treatment effect that accounts for baseline risk — it can be directly compared across different diseases and patient populations without adjustment for absolute risk difference

ANSWER: C

Rationale:

The number needed to treat (NNT) is an absolute measure of treatment benefit derived from the absolute risk reduction (ARR) in a clinical trial: NNT = 1 / ARR. It answers the clinically intuitive question: "How many patients do I need to treat to prevent one additional bad outcome?" For example, if 10% of patients in the placebo group experience a myocardial infarction over five years and 5% in the treatment group experience an MI, ARR = 10% − 5% = 5% = 0.05, and NNT = 1/0.05 = 20. This means 20 patients must be treated for five years to prevent one additional MI compared to placebo. Key properties and limitations: (1) NNT is specific to the population, outcome, comparator, and time horizon of the trial — an NNT of 20 for MI prevention in high-risk post-ACS patients cannot be applied to low-risk primary prevention patients; (2) NNT is an absolute measure that accounts for baseline risk, unlike relative risk reduction (RRR) which can be misleadingly similar across different baseline risk levels; (3) NNT should always be interpreted alongside NNH (number needed to harm) for a balanced risk-benefit assessment; (4) A lower NNT indicates greater benefit — an NNT of 5 indicates more clinically meaningful benefit than an NNT of 200. The companion measure, NNH (number needed to harm), = 1/absolute risk increase for a specified adverse effect. Option A is incorrect — NNT of 20 does not mean 20% benefit rate; NNT describes the number of patients treated, not the percentage experiencing benefit. Option B is incorrect — NNT measures efficacy (events prevented), not toxicity; NNH measures harm. Option D is incorrect — NNT is not the percentage of patients achieving the primary endpoint; it is derived from the ARR (difference in event rates between treatment and control). Option E is incorrect — NNT is an absolute measure that depends on baseline event rate and therefore cannot be directly compared across different populations without accounting for baseline risk; it is not a relative measure.

7. Which of the following best describes Bayesian therapeutic drug monitoring (TDM) and how it differs from standard population-based TDM?

A) Bayesian TDM uses genetic testing (CYP genotyping) as the primary input for dose adjustment — it differs from standard TDM in that plasma concentration measurement is replaced by genotype-predicted pharmacokinetic parameters
B) Bayesian TDM uses a mathematical framework that combines population pharmacokinetic (popPK) prior information — derived from a model built on data from many patients — with individual patient concentration measurements to generate patient-specific pharmacokinetic parameter estimates; these individualized estimates then allow prediction of the dose required to achieve a target concentration with greater precision than population-based dosing alone, particularly in patients with unusual pharmacokinetics (extremes of renal function, obesity, critical illness, drug interactions)
C) Bayesian TDM is a statistical method used exclusively in clinical trials to adjust for confounding in pharmacokinetic-pharmacodynamic studies — it has no application in routine clinical drug monitoring
D) Bayesian TDM differs from standard TDM in that it eliminates the need for blood sampling — it uses patient demographics (weight, age, sex, renal function) alone to calculate individualized doses without any plasma concentration input
E) Bayesian TDM applies only to drugs with linear (first-order) pharmacokinetics — it cannot be used for drugs exhibiting Michaelis-Menten saturation kinetics such as phenytoin or mycophenolate mofetil

ANSWER: B

Rationale:

Bayesian TDM represents a methodologically sophisticated evolution of drug concentration monitoring that leverages the mathematical framework of Bayes' theorem to combine two sources of information: prior knowledge (what we know about how a drug behaves in a population, encoded in a validated population pharmacokinetic model — the "prior") and new information (the individual patient's measured drug concentration at a specific time point — the "likelihood"). Bayes' theorem combines these mathematically to generate a posterior estimate — the most probable pharmacokinetic parameter values for this specific individual, updated by their measured concentration. The key advantage of Bayesian TDM over traditional population-based TDM is individualization: rather than assuming a patient has average population PK parameters, Bayesian analysis uses their actual concentration data to estimate their individual clearance, volume of distribution, and half-life. For example, in vancomycin TDM, Bayesian software (e.g., DoseME, InsightRx, PrecisePK) uses a population PK model for vancomycin alongside the patient's measured trough or AUC-based concentration, renal function, weight, and age to generate individual PK parameter estimates and calculate the dose required to achieve a target AUC24/MIC. Bayesian TDM is most valuable when: patients deviate substantially from population average PK (critically ill patients, extremes of body size, severe organ dysfunction, drug interactions); sparse sampling is available (limited blood draws); target attainment is time-critical (once daily aminoglycosides, vancomycin in severe infections). Bayesian TDM is now endorsed by ASHP/IDSA/SIDP guidelines for vancomycin monitoring. Option A is incorrect — Bayesian TDM uses plasma concentration measurement, not genetic testing, as its primary individual data input; genotyping is a complementary tool. Option C is incorrect — Bayesian TDM is actively implemented in clinical practice for vancomycin, aminoglycosides, tacrolimus, and other drugs. Option D is incorrect — demographics alone constitute the "prior" in population dosing; Bayesian TDM specifically requires at least one actual concentration measurement to update the prior and individualize estimates. Option E is incorrect — Bayesian approaches can accommodate non-linear pharmacokinetic models; Bayesian methods are applicable to phenytoin's saturable kinetics and have been used for mycophenolate and other drugs with complex PK.

8. The CAST (Cardiac Arrhythmia Suppression Trial) is cited as a landmark lesson in the hierarchy of clinical evidence and the distinction between surrogate and clinical endpoints. Which of the following best summarizes the pharmacological and methodological lesson of the CAST trial?

A) The CAST trial demonstrated that antiarrhythmic drugs (encainide, flecainide, moricizine) successfully suppressed ventricular ectopy (surrogate endpoint) but paradoxically increased mortality (clinical endpoint) compared to placebo in post-MI patients with asymptomatic ventricular arrhythmias — illustrating that improvement in a surrogate endpoint does not guarantee benefit for the clinical endpoint that matters to patients, and that pharmacological plausibility alone is insufficient to justify prescribing without randomized controlled trial evidence for clinical outcomes
B) The CAST trial demonstrated that suppression of ventricular ectopy with antiarrhythmic drugs reduced mortality in post-MI patients — validating the surrogate endpoint approach and supporting the use of Holter monitor arrhythmia burden as a reliable surrogate for cardiovascular mortality in drug approval
C) The CAST trial demonstrated that encainide and flecainide were ineffective at suppressing ventricular ectopy in post-MI patients — the trial was stopped early because neither drug reduced arrhythmia burden, and the lesson is that antiarrhythmic drugs should be reserved for symptomatic arrhythmias only
D) The CAST trial is a landmark example of publication bias — the original positive results were published but the negative mortality data were suppressed for five years before disclosure; the primary lesson is about research ethics and data transparency rather than surrogate endpoint validity
E) The CAST trial demonstrated that the hierarchy of evidence (RCT > observational study > case series) reliably predicts clinical benefit — drugs that reduce surrogate endpoints in RCTs consistently reduce mortality in subsequent larger trials, validating the FDA's accelerated approval pathway for surrogate endpoints

ANSWER: A

Rationale:

The CAST trial ( Cardiac Arrhythmia Suppression Trial, Echt et al., NEJM 1991) is one of the most consequential and pharmacologically instructive clinical trials in modern medicine — a cautionary paradigm for the entire field of evidence-based pharmacotherapy. The pathophysiological reasoning that motivated CAST was compelling: post-MI patients with frequent ventricular ectopy (premature ventricular contractions, PVCs) have increased mortality risk; class IC antiarrhythmic drugs (encainide, flecainide) effectively suppress PVCs on Holter monitoring (validated surrogate endpoint); therefore, suppressing PVCs should reduce mortality. CAST was designed to test whether this surrogate-endpoint benefit translated to mortality reduction. The results were profoundly counterintuitive and fatal: the trial was terminated early because patients randomized to encainide or flecainide had significantly higher all-cause mortality (primarily sudden cardiac death) than those randomized to placebo — despite the drugs successfully suppressing PVCs. The drugs were doing exactly what they were supposed to do pharmacologically (suppress ectopy) but the pharmacological mechanism of PVC suppression (sodium channel blockade, slowing of conduction velocity) also increased the risk of lethal re-entrant arrhythmias in the structurally abnormal post-MI myocardium. CAST delivered multiple critical lessons: (1) Surrogate endpoints can be misleading — improvement in a measurable intermediate does not guarantee clinical benefit; (2) Pharmacological plausibility without RCT evidence for clinical outcomes is insufficient justification for prescribing; (3) Drugs can have mechanistically coherent explanations for their harm as well as their benefit; (4) The importance of adequately powered RCTs with clinical endpoints (mortality, MI, stroke) rather than surrogate endpoints for drug approval and prescribing decisions. CAST directly influenced FDA policy on antiarrhythmic drug approval and contributed to the recognition of surrogate endpoint limitations across therapeutic areas. Option B is the opposite of what CAST found — mortality increased, not decreased, with antiarrhythmic therapy. Option C is incorrect — encainide and flecainide were effective at PVC suppression; the trial was stopped because of increased mortality, not because of inefficacy at the surrogate endpoint. Option D is incorrect — CAST is not primarily a publication bias story; the trial was conducted transparently and published promptly; its lesson is about surrogate vs clinical endpoints. Option E is incorrect — CAST demonstrates the opposite: a drug that improved the surrogate endpoint (PVC suppression) worsened the clinical endpoint (mortality), directly challenging the validity of surrogate endpoint-driven approvals.