PGIM MD Family Medicine common mistakes
PGIM MD Family Medicine · Understand the traps and how to avoid them.
Committing to a single diagnosis and then selecting the option that fits that diagnosis, instead of generating a short prioritized differential and choosing the option that best fits the whole vignette (age, tempo, red flags, negatives).
Family medicine vignettes are deliberately ambiguous and mix common presentations of uncommon disease with uncommon presentations of common disease; candidates anchored by pattern recognition from one striking feature (e.g., a single abnormal lab) and stop reasoning.
What to do instead
Practise a fixed 30-second routine before answering: state the syndrome, list 3 diagnoses with pretest likelihood, identify the discriminating feature the stem gives (or withholds), then eliminate options that ignore a stated negative or red flag. Drill with vignettes that have a plausible 'attractive distractor' and write down why it is wrong.
Missing the single best next step in the management sequence, choosing definitive treatment instead of the immediate stabilizing or diagnostic step (or vice versa).
Candidates conflate 'what is the diagnosis' with 'what do you do now'; exam stems reward the next action in the algorithm (e.g., ABCs, confirm the diagnosis, exclude a dangerous alternative) rather than the final therapy.
What to do instead
For every management question, explicitly classify the option as: resuscitate/stabilize, confirm/diagnose, treat, refer, or follow up. Learn the standard first-step algorithms for the high-yield emergencies (chest pain, stroke, anaphylaxis, sepsis, DKA, preeclampsia, acute asthma, GI bleed, suicidal ideation) and rehearse them as ordered lists.
Applying adult thresholds and drug doses to children, pregnant patients, and older adults without adjusting for weight, renal function, or physiology.
Candidates memorize one set of numbers and default to it; exam writers exploit this with pediatric dosing, neonatal fever criteria, pregnancy-safe antihypertensives/antibiotics, and Beers-type cautions in the elderly.
What to do instead
Build a one-page 'special populations' sheet: weight-based pediatric doses, age-specific vital sign and fever cutoffs, contraindicated drugs in pregnancy and lactation, and drugs to avoid in renal impairment or in the elderly. Re-answer any pediatric or obstetric question by first asking 'does the age/weight change the threshold or the dose?'
Answering screening and prevention questions with the test or interval that seems clinically sensible rather than the one supported by the guideline the exam is testing, and ignoring the difference between screening, case-finding, and diagnostic testing.
Screening recommendations are counterintuitive (harms, overdiagnosis, stopping ages, non-screening for some cancers), and candidates rely on 'more testing is better' intuition or on outdated advice.
What to do instead
Learn each major screening program as a grid: who, what test, start age, stop age, interval, and the one exception (e.g., high-risk groups). Explicitly note where guidelines diverge and which body the question implies (national task force vs specialty society). Practise items where the correct answer is 'no screening indicated'.
Confusing study designs and their inherent biases, e.g., labeling a case-control study as cohort, or attributing selection bias to a design that mainly risks recall or measurement bias.
Design names are learned as vocabulary rather than as structures with direction of sampling and timing; candidates then cannot map a described flaw to the correct bias category.
What to do instead
Draw the 2x2 sampling diagram for each design (cohort: exposure to outcome; case-control: outcome to exposure; cross-sectional: simultaneous; RCT: allocation then follow-up) from memory until automatic. For each design, list its two characteristic biases and one classic example, and practise classifying novel scenarios rather than memorizing definitions.
Misreading the output of a study: treating a p-value as the probability the hypothesis is true or the probability of a false positive, and treating a confidence interval that crosses the null as proof of no effect.
Statistical significance is confused with clinical importance and with Bayesian probability; candidates also ignore the difference between a ratio measure and a risk difference when judging impact.
What to do instead
Rehearse the correct sentence for each statistic: p-value = probability of data this extreme if the null were true; 95% CI = range of effects compatible with the data; 'no significant difference' = inconclusive, not equivalent. Practise converting the same result into relative risk, absolute risk reduction, NNT, and likelihood ratio, and decide which one the clinical question needs.
Answering critical appraisal items by judging whether the study's conclusion sounds right, rather than checking the internal validity steps: randomization and allocation concealment, blinding, completeness of follow-up, intention-to-treat, and applicability to the patient in the stem.
Candidates read the abstract conclusion and work backwards; they skip the methods-level checks that determine whether the result can be trusted at all.
What to do instead
Use a fixed appraisal checklist in order: population and comparability, allocation and concealment, blinding, follow-up and attrition, analysis set, effect size with CI, then applicability. Practise writing a two-sentence verdict on short method excerpts, naming the single most serious threat to validity before considering the result.
Handling consent, capacity, confidentiality, and end-of-life questions by choosing the option that preserves life or follows the family's wishes, instead of the option that follows the patient's expressed wishes and assessed decision-making capacity.
Candidates import personal or cultural values and default to paternalism or family authority; they also conflate incapacity with disagreement with medical advice.
What to do instead
Apply a fixed sequence: does the patient have capacity for this specific decision, what are their previously expressed wishes, is there an advance directive or substitute decision-maker, and what does the law in the jurisdiction require. Remember that capacity is decision-specific, that refusing treatment is not itself evidence of incapacity, and that a capable patient's refusal must be respected. Drill cases involving adolescents, minors, and confidential care.
Answering health-systems and resource-allocation items with the clinically optimal intervention while ignoring cost-effectiveness, equity, accessibility, and the level of the health system (primary care vs hospital) the question is about.
Clinical training rewards the best possible care for the individual; population health questions reward the intervention with the greatest aggregate benefit per resource, which often looks 'less thorough'.
What to do instead
For each preventive or system question, ask: what is the population, what is the resource constraint, and which option improves the most outcomes per unit cost while remaining equitable and feasible at the stated level of care? Learn the concepts of primary, secondary, and tertiary prevention, sensitivity/specificity trade-offs in population screening, and the distinction between individual and population benefit.
Rushing the last third of the paper and changing answers on review without a stated reason, leaving no time to check the long vignettes that carry the most discriminating detail.
Fatigue and time pressure cause skim-reading of stems; candidates then revise correct first answers based on anxiety rather than new evidence, and miss the 'EXCEPT' or 'most likely' qualifier.
What to do instead
Simulate full-length timed papers and adopt a pacing rule (e.g., average time per item with a hard mark-and-return threshold). On review, change an answer only if you can name the specific stem word you missed. Before selecting, re-read the final sentence of the stem to confirm what is being asked, and flag any question containing a negation.