ABPN Psychiatry common mistakes
ABPN Psychiatry · Understand the traps and how to avoid them.
Selecting a psychotropic and dose that would be appropriate for a healthy adult while ignoring the constraint buried in the stem (eGFR 28, Child-Pugh B cirrhosis, CYP2D6 poor metabolizer, QTc 510 ms, pregnancy, or documented prior nonresponse to that same drug).
The stem is long, so candidates read it for the diagnosis and then jump to the options; the qualifier is usually one clause near the end and gets skimmed. Drug-name recognition feels like sufficient competence, so the metabolic context is never actively checked.
What to do instead
Before answering any medication item, run a fixed 10-second constraint scan and physically list: organ function, age/extremes of life, pregnancy/lactation, interactions and CYP genotype, ECG/electrolyte risk, prior failed trials, and any stated contraindication. Eliminate any option that fails one constraint before comparing efficacy, and prefer the option that survives all constraints.
Committing to a diagnosis after the first two sentences of a vignette and then reading the remaining options only to confirm it, without re-opening the differential when a new temporal, exclusion, or severity criterion appears.
Vignettes are engineered so the opening symptoms are deliberately prototypical of one disorder and the discriminating detail arrives later; closure feels efficient under time pressure, and confirmation bias makes the later detail seem like a minor detail rather than the crux.
What to do instead
Force a written differential of at least three diagnoses with the single feature that would rule each in or out, then read the rest of the stem as a test of that list. When a new fact appears, ask explicitly which differential entry it eliminates. Answer only after the final sentence has been re-read against the chosen option.
Labeling developmentally expectant phenomena as pathology (magical thinking and separation distress at age four, risk-taking and sleep-phase shift at fifteen, concrete thinking at seven) or, conversely, dismissing genuine impairment as 'just a phase.'
Adult-derived criteria and rehearsed adult vignettes dominate study time, so the candidate has no calibrated developmental timeline to compare against and substitutes gut impression of what seems abnormal.
What to do instead
For every child or adolescent item, state the age-expected norm, the onset and course, and the functional impairment (home, school, peers) before touching the options. Pre-build a one-page chart of normal milestones and red-flag deviations across preschool, school-age, and adolescent stages and review it weekly.
Misassigning neuroanatomy or neurotransmitter systems in items about a lesion, syndrome, or drug mechanism (for example confusing dorsal and ventral striatal circuits, attributing neuroleptic malignant syndrome to dopamine excess, or mixing up frontal versus parietal findings).
Neuroscience is often studied as vocabulary lists detached from clinical syndromes, so the material is memorized at recognition level only and collapses when the item requires mapping structure to function to bedside sign.
What to do instead
Build and rehearse a three-column table: structure/transmitter, function when intact, syndrome when disrupted (with the bedside exam finding). Practice in reverse, starting from the sign and naming the circuit. Anchor each row to one memorable patient or drug example so the knowledge is retrievable in both directions.
Jumping straight to the most restrictive or most aggressive intervention in safety and ethics items (involuntary commitment, seclusion, security, terminating the relationship) without first demonstrating immediate danger, capacity assessment, or a least-restrictive effective alternative.
Fear of missing a dangerous situation drives over-treatment, and restraint or commitment feels like the 'safe' answer. Candidates also conflate the question 'what is legally permissible' with 'what is the next appropriate step.'
What to do instead
Apply a fixed sequence to every safety item: is there imminent risk, can it be managed with the least restrictive effective measure, has decision-making capacity been assessed, is collateral/consultation needed, and how is it documented. Choose the earliest step in that sequence that the stem actually supports, and reserve restrictive measures for stems that state imminent danger or failed lesser measures.
Matching a psychotherapy modality to a vague impression of the case rather than to the specific disorder and target symptom (for example selecting psychodynamic or supportive therapy for obsessive-compulsive disorder, or insight-oriented therapy as first-line for a specific phobia).
Modalities are learned as broad theories rather than as disorder-specific indications, and the options are written so that each sounds plausible and humane; without an indication table, the choice defaults to whichever label feels most sophisticated.
What to do instead
Memorize a disorder-to-first-line-modality grid with the active ingredient named (exposure and response prevention for OCD, exposure-based therapy for phobias, CBT for panic, dialectical behavior therapy for chronic suicidality and emotion dysregulation, behavioral activation for depression). On each item, name the target symptom and the mechanism that addresses it before reading the options.
Settling on a primary psychiatric diagnosis without first excluding a medical, neurologic, pharmacologic, or substance-induced cause, even when the stem supplies onset, fluctuating course, vital-sign, or laboratory clues pointing to a mimic.
Case familiarity rewards pattern recognition of common psychiatric presentations, and delirium, hypothyroidism, steroid-induced mania, and stimulant intoxication are taught as separate topics rather than as mandatory exclusion steps inside the diagnostic algorithm.
What to do instead
Insert a mandatory 'medical and substance screen' checkpoint into every diagnostic item: age of onset, acuity, fluctuating attention, abnormal vitals, recent medication or substance change, and any laboratory or imaging clue. If any checkpoint is positive, the organic or substance-related answer must be actively excluded rather than assumed unlikely.
Answering evidence-based-medicine and study-design items by intuition, confusing sensitivity with positive predictive value, misreading a confidence interval or p-value as a probability that the hypothesis is true, or ignoring base rate when interpreting a screening test.
Statistical content appears in a small fraction of items, so it receives the least rehearsal and is read as a memory test of definitions rather than as a calculation that can be performed on the page.
What to do instead
Rehearse a small set of computations until automatic: two-by-two table construction, sensitivity, specificity, predictive values with base rate varied, number needed to treat, and correct interpretation of confidence intervals. In each item, identify the design and the measure being asked for before reading the options, and write the numbers down rather than reasoning verbally.
Proposing an individual-level remedy for a systems problem in quality-improvement, professionalism, and patient-safety items (counseling or remediating one clinician, adding a reminder to 'be more careful') when the correct answer is a process or system intervention such as a checklist, standardized handoff, redundancy, or error reporting.
Clinical training instills a personal-responsibility frame, and the individually focused option sounds decisive and blameless-free. Candidates also misjudge which level of intervention a QI item is asking about.
What to do instead
Classify the item first: is the failure a knowledge or skill deficit in one person, or a predictable process vulnerability? For recurrent or system-embedded errors, choose the intervention that changes the process and builds in redundancy, and reserve individual remediation for isolated competence deficits. Rehearse the hierarchy: forcing function, standardization, checklist, redundancy, education, then admonition.
Making arithmetic or unit errors on quantitative items such as half-life and steady-state timing, therapeutic drug monitoring interpretation, sodium correction, or lethal-dose comparisons, and then selecting the option closest to the miscalculated figure.
These items are treated as recall of a fact ('about five half-lives') rather than as a calculation, so the candidate estimates verbally under time pressure and never writes intermediate steps, allowing a single slip to propagate.
What to do instead
Always write the calculation: state the starting value, the number of half-lives or the conversion factor, and the unit at each step, then confirm the answer's order of magnitude is physiologically plausible. Memorize a short anchor set (roughly five half-lives to steady state, the milliequivalent conversions, the common sodium formulas) and practice them with units attached.