ABFM Family Medicine common mistakes
ABFM Family Medicine · Understand the traps and how to avoid them.
Committing to a diagnosis after reading only the first one or two sentences of the vignette, then selecting the answer that fits that initial impression without re-reading the remaining data.
Board stems are written to reward pattern recognition, so candidates who recognize a familiar syndrome early stop processing the contradicting details (normal WBC, atypical age, absence of the expected risk factor) that are deliberately placed to disqualify the first diagnosis.
What to do instead
Read the entire stem once without looking at the options, then write down the single detail that does not fit the leading diagnosis. Only after that, compare the top two choices against every abnormal finding; if one finding is unexplained, switch to the diagnosis that explains all of them.
Selecting the most sensitive test (e.g., a screening study or a broad panel) when the question asks for the next step in confirming a diagnosis in a stable outpatient.
Candidates conflate 'best test' with 'test that catches the most disease' and forget that in a symptomatic patient the question is usually about diagnostic accuracy, not screening yield.
What to do instead
Before answering, classify the question intent as screen, diagnose, or monitor. For diagnose, choose the test with the highest specificity/likelihood ratio for the suspected condition; reserve sensitive tests for screening or ruling out in high-risk presentations.
Applying adult screening intervals and adult numeric thresholds to adolescents and children (e.g., using adult lipid or blood pressure cutoffs, or omitting adolescent depression and STI screening).
Most studying is done from adult-oriented material, and adult numbers are more frequently rehearsed, so pediatric age bands get mentally collapsed into 'young adult.'
What to do instead
Build and review a one-page age-band table (0-2, 3-10, 11-17, 18-21) listing screening items, immunization timing, and the specific pediatric reference thresholds before the exam.
Reassuring and discharging a febrile infant younger than 2-3 months, or deferring the sepsis evaluation to outpatient follow-up, because the infant appears well.
Outpatient training conditions clinicians to trust the well-appearing child, and the rarity of serious bacterial illness makes the low prior feel like a reason to skip the workup.
What to do instead
Memorize the hard age cutoff as an absolute rule: any febrile infant under 2-3 months gets a full sepsis evaluation and empiric coverage or admission regardless of appearance, and let the answer choice reflect that.
Treating an elevated blood pressure in pregnancy as an isolated number to lower, without classifying severity features or initiating magnesium sulfate for severe preeclampsia.
Candidates focus on the blood pressure value and forget that seizure prophylaxis and delivery timing, not just antihypertensives, drive management in severe disease.
What to do instead
For any hypertension in pregnancy, run a fixed checklist: gestational age, proteinuria/platelets/transaminases/creatinine/headache/visual symptoms, then decide severe vs non-severe. Severe features trigger magnesium sulfate plus delivery planning, not just a BP medication.
Intensifying glycemic or blood pressure therapy in a frail older adult to reach the same targets used for a healthy middle-aged patient, or prescribing a Beers-listed medication without checking.
Guideline numbers are learned in the context of healthy adults, and the competing risk of hypoglycemia, falls, and polypharmacy is not spontaneously weighted under time pressure.
What to do instead
Before selecting a target or drug, state the patient's function and life expectancy in one line. If frail or limited life expectancy, choose the relaxed target and screen every proposed drug against the Beers list.
Attributing depressive symptoms, poor concentration, or weight loss to a chronic medical condition and never administering a PHQ-2/PHQ-9 or asking about suicidality.
Somatic complaints have an obvious medical explanation in these patients, so the mood disorder is masked and the visit agenda is already full.
What to do instead
Treat PHQ-2 as a mandatory vital sign in any patient with fatigue, insomnia, unexplained weight change, or new functional decline, and ask directly about suicidal ideation whenever the screen is positive.
Ordering imaging (radiograph or MRI) for uncomplicated acute low back pain before a trial of conservative care, or failing to ask about red flags when the pain is atypical.
Patient expectation and the desire for a definitive answer push toward imaging, and the red-flag history is easy to skip when the presentation 'looks mechanical.'
What to do instead
For acute low back pain, first document the red-flag review (fever, cancer history, night pain, neurologic deficit, bowel/bladder change, IV drug use, trauma). If none are present, choose conservative care and avoid imaging in the answer.
Dismissing a possible acute coronary syndrome because the patient is a woman, has diabetes, or describes indigestion or exertional dyspnea rather than classic crushing chest pain.
The textbook vignette trains a single prototypical presentation, and atypical features feel like evidence against ischemia.
What to do instead
For any acute chest, epigastric, or exertional symptom in a patient with vascular risk factors, obtain an ECG and troponin before assigning a musculoskeletal or GI cause, regardless of whether the pain is 'typical.'
Choosing a screening test or intervention based on enthusiasm for the disease rather than the USPSTF grade, or misapplying sensitivity/specificity when the question asks about a positive result.
Candidates memorize test names but not the grade and population for which each is recommended, and they confuse the probability of disease given a positive test with the test's sensitivity.
What to do instead
For each preventive service, memorize four items: grade, target population, interval, and starting/stopping age. When a question gives prevalence and test characteristics, compute the post-test probability with a 2x2 table rather than reasoning from sensitivity alone.