ACG CxA common mistakes
ACG CxA · Understand the traps and how to avoid them.
Choosing the single most likely diagnosis and stopping there when the vignette asks what should be done next, rather than identifying the one answer that changes management (e.g., ordering the test that excludes the dangerous alternative before treating the presumed benign cause).
Candidates pattern-match the classic presentation and commit to a diagnosis, then pick the answer that 'fits' the diagnosis instead of answering the actual question stem.
What to do instead
Before reading the options, restate the stem as a task ('next step', 'best initial test', 'most appropriate management'). Reject any option that confirms a diagnosis already established by the vignette; prefer the option that resolves the remaining uncertainty or addresses the highest-acuity possibility.
Treating every positive celiac serology as diagnostic and moving straight to a gluten-free diet, skipping the recommendation for duodenal biopsy while on a gluten-containing diet (or skipping the biopsy only in the narrowly defined no-biopsy pathways).
Serology is emphasized in study material, and the confirmatory-biopsy requirement plus the need to remain on gluten are easy to forget under time pressure.
What to do instead
Build a fixed sequence for suspected celiac disease: serology first (tTG-IgA with total IgA), then duodenal biopsies with adequate sampling while the patient is still eating gluten; only invoke no-biopsy criteria when the specific serologic and clinical thresholds are met.
Selecting an endoscopic intervention or sedation plan without first addressing anticoagulant/antiplatelet status, bowel preparation adequacy, or the indication-specific timing window (e.g., urgent vs elective colonoscopy for lower GI bleeding).
Peri-procedural medication management and timing rules are numerous and specialty-specific, and candidates default to the procedure itself as the 'answer'.
What to do instead
For every endoscopy question, run a pre-procedure checklist in order: indication and urgency, preparation, antithrombotic hold/continue decision, and sedation risk. Eliminate options that ignore a checklist item the vignette explicitly raises.
Managing a GI bleed by transfusing to a routine hemoglobin target or giving a PPI before resuscitation and risk stratification, rather than following the ABCs, restrictive transfusion strategy, and risk-stratified timing of endoscopy.
The dramatic bleeding presentation pulls attention to pharmacologic and endoscopic specifics, and restrictive transfusion thresholds are counterintuitive to candidates trained on higher targets.
What to do instead
Anchor bleed questions on resuscitation first, then restrictive transfusion thresholds, then risk stratification to set endoscopy timing; treat acid suppression and other drugs as adjuncts that never precede stabilization.
Escalating therapy or calling a flare 'refractory' without first excluding infection (notably C. difficile) and other mimics, and without confirming objective inflammation before changing immunomodulators or biologics.
The word 'flare' in a stem triggers treatment escalation, and candidates skip the step of verifying that symptoms reflect active inflammation rather than infection or a functional overlap.
What to do instead
For any IBD flare, first test for enteric infection including C. difficile, then confirm objective inflammation (endoscopy, calprotectin, imaging) before changing therapy; reserve 'refractory' for patients who meet those criteria.
Treating corticosteroid therapy as a maintenance strategy or stopping steroids abruptly, instead of using them for induction and planning a defined taper with a steroid-sparing agent for maintenance.
Steroids reliably control symptoms, so they feel like the 'right' answer, and the induction-versus-maintenance distinction is blurred in recall.
What to do instead
Classify every IBD drug option as induction, maintenance, or both before choosing. Reject any answer that uses steroids for long-term maintenance or omits a taper and a maintenance plan.
Applying a single 'abnormal liver enzyme' algorithm to every pattern instead of first classifying the injury as hepatocellular, cholestatic, or mixed, and then pursuing the pattern-specific next test.
Candidates memorize individual diseases rather than the branching evaluation framework, so they jump to a specific serology or imaging study that matches a remembered disease.
What to do instead
For abnormal liver tests, first compute and classify the pattern, then select the next test dictated by that pattern (viral/metabolic/autoimmune workup for hepatocellular; biliary imaging for cholestatic). Reject options that skip the classification step.
Interpreting a low or borderline platelet count or mildly abnormal synthetic function as trivial in a patient with cirrhosis, rather than recognizing decompensation, portal hypertension, and the need for variceal surveillance and HCC screening at guideline intervals.
Mild laboratory abnormalities are common and easy to dismiss, and surveillance intervals are easy to conflate with diagnostic testing.
What to do instead
In any cirrhosis vignette, explicitly list decompensation markers, confirm variceal surveillance and HCC screening are up to date, and check for precipitating factors of decompensation before choosing management.
Labeling a motility or functional complaint as purely psychological or dismissing it, instead of applying the symptom-based diagnostic criteria, excluding organic disease where indicated, and offering a structured biopsychosocial treatment plan.
Normal tests create pressure to conclude 'nothing is wrong', and candidates under-use formal symptom criteria and structured dietary/neuromodulator/behavioral options.
What to do instead
For functional GI complaints, verify the symptom criteria are met, document the limited and indication-driven organic workup, then choose an evidence-based multimodal plan (dietary, pharmacologic, behavioral) rather than a reassurance-only answer.
Choosing a screening or preventive strategy without accounting for the patient's prior screening history, family history, or risk category, and defaulting to average-risk intervals for everyone.
Interval and age-threshold tables are memorized in isolation, and candidates overlook the modifying details embedded in the stem.
What to do instead
Before answering any screening question, extract risk category, prior results, and family history, then apply the corresponding interval or modality. If an option uses an average-risk interval for a higher-risk patient, eliminate it.
Selecting a quality metric or improvement action that measures volume or satisfaction rather than a validated process or outcome measure (e.g., adenoma detection rate, cecal intubation, adequate bowel prep, appropriate surveillance intervals), or proposing an intervention without a measurement plan.
Quality and practice-management content feels administrative, so candidates pick plausible-sounding but non-specific actions instead of the defined metric or a plan-do-study-act loop.
What to do instead
For quality questions, name the specific validated metric and its benchmark, then choose the intervention paired with remeasurement. Reject options that lack a measurable indicator or a feedback loop.
Ignoring access-related drivers in a clinical scenario (cost, insurance, language, transportation, health literacy) and choosing the guideline-ideal option that the patient cannot realistically complete, rather than the equity-informed equivalent that achieves the same outcome.
Guideline-ideal answers feel 'most correct', and social determinants are treated as background rather than as constraints that change the right answer.
What to do instead
When a stem includes access barriers, screen each option for feasibility: prefer the effective option the patient can actually obtain (interpreter services, lower-cost equivalent, navigation support) and reject plans that depend on resources the vignette says are unavailable.