Unofficial guide · not affiliated with the AAMC, NBME, USMLE, LCME, or NRMP Part of the SI MedEd family · SI MedEd portal →

Elsie Item-Writing Academy

Academy › Distractor Engineering › Step 2 CK

Distractor Engineering: Step 2 CK — Management Options That Behave Like Medicine

About 45 minutes · Academy module: Distractor Engineering: Step 2 CK — Management Options That Behave Like Medicine

Unofficial faculty-development resource. Not affiliated with, endorsed by, or representing the AAMC, NBME, USMLE, LCME, or NRMP. Supports faculty-development documentation.

Teaching: what makes a distractor work at the Step 2 CK level

CK distractors are management options, and management options carry a special hazard: students treat the key as a clinical recommendation. A sloppy distractor set does not just weaken an item — it can reinforce a wrong practice pattern. Engineer accordingly.

Common misconceptions as distractors. The richest CK distractors are the errors clinicians actually make: reaching for steroids alone in Kawasaki disease, giving magnesium as definitive therapy for severe preeclampsia, choosing prophylaxis when treatment is indicated. These are not trick options; they are the decisions your learners will face, and the item analysis will show you precisely how many would make them. Draft each distractor from a real, nameable clinical error.

The "second-best answer" trap. Management items naturally produce near-miss options — the right drug without the indicated adjunct (antibiotics without albumin in SBP), the right destination reached the wrong way. The second-best distractor must be wrong for a reason you can state in one sentence, and it must survive content review without becoming defensible. Every rewrite below passed two independent cold solves with no defensible alternative named.

Avoiding the throwaway option. In a five-option management item, the throwaway is usually the absurd one — "discharge home" for an unstable patient, an invasive procedure for a medical problem. It does active harm: it tells the student which options you took seriously, narrowing the decision to a subset. If an option would never be chosen by a thinking clinician, replace it with a plausible-but-wrong management path.

Homogeneity of length and detail. Management options vary naturally in complexity ("cefotaxime plus albumin" vs. "observation"), so the discipline here is about specificity, not word count: no option should carry unique qualifiers, doses, or time frames that mark it as the author's favorite. Parallel structure across all five.

Distractor ordering. Management options often have a natural escalation order (observe → medical → procedural); when they do, keep it — scrambling a clinical ladder looks artificial. When there is no natural order, keep the key position varied across the exam.

Feeding pooled data back into revision. CK item analysis is clinical-education intelligence. A distractor with heavy pull is a practice gap in your cohort — feed it back to the clerkship director, not just the item bank. Watch the p-value for items that drift too easy (everyone knows it) or too hard (nobody does); both are revision candidates. And the standing rule holds at every level: negative discrimination triggers faculty review — content re-examination for a second key or miskey — never automatic rekeying or deletion. Statistics flag; faculty decide. This posture also supports faculty-development documentation: the review trail shows the item was examined deliberately, not discarded reflexively.

---

Drill 1 — Distractors that miss the real clinical error

Flawed item (intended key: A)

A 2-year-old girl has had 5 days of fever to 39.5°C, bilateral conjunctival injection without exudate, cracked red lips, a polymorphous rash, and a 1.8-cm tender cervical lymph node. Echocardiogram is pending. What is the most appropriate treatment?

A) Intravenous immunoglobulin and high-dose aspirin B) High-dose vitamin C C) Amoxicillin D) Bed rest E) Observation with repeat examination

Flaw taxonomy label: Distractors not built from the common clinical error (with a throwaway option).

Correction: The error clinicians actually make in Kawasaki disease is reaching for corticosteroids alone — and it is nowhere in the options, while option B (vitamin C) is padding no one would choose. The item therefore cannot detect the misconception that matters, and its distractor pull will be uninformative. The rewrite installs the real error as a live distractor (corticosteroids alone — the "steroids treat inflammation" reflex), keeps the plausible-but-wrong symptomatic options, and drops the padding. Now a pull on option B tells the clerkship director exactly which gap to teach.

Model rewrite — cold-solve verified: 2/2 (solver 1: A, solver 2: A)

A 2-year-old girl has had 5 days of fever to 39.5°C, bilateral conjunctival injection without exudate, cracked red lips, a polymorphous rash, and a 1.8-cm tender cervical lymph node. Echocardiogram is pending. What is the most appropriate treatment?

A) Intravenous immunoglobulin and high-dose aspirin B) Corticosteroids alone C) Amoxicillin D) Ibuprofen alone E) Observation with repeat examination

Key: A

Drill 2 — The missing second-best trap

Flawed item (intended key: A)

A 55-year-old man with known cirrhosis presents with fever, abdominal pain, and tense ascites. Paracentesis shows 640 neutrophils per microliter. Blood pressure is 118/72 mm Hg. What is the most appropriate management?

A) Cefotaxime plus albumin B) Oral ciprofloxacin prophylaxis only C) Large-volume paracentesis D) Diagnostic laparoscopy E) Observation

Flaw taxonomy label: Missing second-best distractor — the near-miss option is absent.

Correction: The student who knows spontaneous bacterial peritonitis needs antibiotics but forgets the albumin adjunct has no option to choose — the item cannot distinguish full understanding from partial understanding, which is precisely the gap that determines outcomes. The near-miss ("antibiotics without albumin") is the most instructive distractor a management item can carry: it must be wrong for a reason statable in one sentence (albumin reduces mortality in SBP with high-risk features) while remaining the natural choice of the almost-ready learner. The rewrite installs it as option A and keeps the remaining options parallel.

Model rewrite — cold-solve verified: 2/2 (solver 1: C, solver 2: C)

A 55-year-old man with known cirrhosis presents with fever, abdominal pain, and tense ascites. Paracentesis shows 640 neutrophils per microliter. Blood pressure is 118/72 mm Hg. What is the most appropriate management?

A) Albumin infusion alone B) Oral ciprofloxacin prophylaxis only C) Cefotaxime plus albumin D) Large-volume paracentesis E) Diagnostic laparoscopy

Key: C

Drill 3 — Heterogeneity of specificity (the "author's favorite" tell)

Flawed item (intended key: A)

A 30-year-old woman with no significant medical history presents for a routine visit. Blood pressure is 148/96 mm Hg on two separate occasions. She does not smoke, and her BMI is 24 kg/m². Which of the following is the most appropriate initial management?

A) Lifestyle modification with repeat blood pressure measurement in 3 to 6 months, including DASH diet counseling, sodium reduction to less than 2.3 grams daily, weight loss if overweight, regular aerobic exercise for 150 minutes per week, and moderation of alcohol intake B) Start lisinopril immediately C) Start hydrochlorothiazide immediately D) Order renal ultrasound E) Start amlodipine immediately

Flaw taxonomy label: Heterogeneity of length and detail — the key carries unique qualifiers.

Correction: Option A is not just longer; it is the only option with doses, time frames, and a detailed plan — the author's favorite, marked by specificity. Students read specificity as correctness. The discipline for management options is parallel structure at equal specificity: every option the same kind of answer (an action, with or without the same level of detail), so the choice is clinical rather than stylistic. The rewrite cuts all five to the same shape; the lifestyle option keeps its time frame because the time frame is the clinical content (reassess in 3–6 months), but nothing else gets one.

Model rewrite — cold-solve verified: 2/2 (solver 1: A, solver 2: A)

A 30-year-old woman with no significant medical history presents for a routine visit. Blood pressure is 148/96 mm Hg on two separate occasions. She does not smoke, and her BMI is 24 kg/m². Which of the following is the most appropriate initial management?

A) Lifestyle modification with repeat blood pressure measurement in 3 to 6 months B) Start lisinopril immediately C) Start hydrochlorothiazide immediately D) Order renal ultrasound E) Start amlodipine immediately

Key: A

---

Drill key — reasoning

  • Drill 1: Installing the real clinical error (corticosteroids alone) as a live distractor makes the item detect the misconception that matters. Both solvers selected A (IVIG plus high-dose aspirin for Kawasaki disease) with no alternative defended.
  • Drill 2: The albumin-alone option is the second-best trap: the natural choice of the learner who knows SBP needs treatment but misses the adjunct. Both solvers selected C (cefotaxime plus albumin) with no alternative defended — the trap is instructive without being defensible.
  • Drill 3: Cutting all options to parallel structure removes the specificity tell. Both solvers selected A (lifestyle modification first for stage 1 hypertension in a low-risk patient) with no alternative defended — on clinical content, not formatting.

Takeaway checklist

  • [ ] Every distractor is a real clinical error — the decisions learners will actually face, not padding.
  • [ ] The second-best trap is present and wrong for one statable reason; it never survives as a disputed key.
  • [ ] No option carries unique doses, time frames, or qualifiers that mark the author's favorite.
  • [ ] Natural escalation orders (observe → medical → procedural) are preserved, not scrambled.
  • [ ] Heavy-pull distractors are reported to the clerkship as practice gaps, not just bank statistics.
  • [ ] Negative discrimination means faculty review for a second key or miskey — never auto-rekey, never auto-delete.

Finished this module?

Recording it adds the module to your Academy completion record on this device — module, track, level, and date, ready to download from your account page for faculty-development documentation.