Read the case first, not the shape

By design, most board questions use four homogeneous options — all diagnoses, all regimens, all the same kind of thing — because that's the NBME standard for a well-written item. When the options are homogeneous, there is no structural shortcut: the right move is to reason the clinical case.

A minority of questions do have a recognizable structure that's worth a small tactical cue — numbers, a conservative-to-aggressive spread, a timing decision, or one option standing apart. Use the tips below as a double-check, never as the reason for your answer. Picking an option because of its shape rather than the medicine is a test-taking tell, and it's exactly what well-written items are designed to defeat.

NBME Item-Writing Guide →

1. Numeric / Threshold

The options are doses, intervals, durations, or cutoff values — they differ by a number, not a concept. These reward knowing the specific figure and applying it in context.

Tip:

  • Commit to the number before you look, then find it — if you're guessing between two, you don't know it yet
  • Extremely high or low values are usually distractors
  • Watch the stem for renal/hepatic adjustment

Example: A patient with giant cell arteritis presents with new vision changes. Most appropriate initial treatment?

A) Prednisone 20 mg daily
B) Prednisone 60 mg daily
C) Methylprednisolone 1,000 mg IV daily for 3 days ✓
D) Tocilizumab

GCA with vision symptoms is an emergency requiring IV pulse steroids. Oral prednisone (A, B) is insufficient when vision is threatened; tocilizumab (D) is steroid-sparing, not initial therapy for acute vision loss.

2. Conservative vs Intervene

The options run from watchful waiting or supportive care up to a drug, procedure, or surgery. The question tests whether you match the aggressiveness of the response to how sick the patient actually is.

Tip:

  • Anchor on disease severity and acuity first
  • Red flags → act decisively; no red flags → conservative is often right
  • The most aggressive option is not automatically the safest answer

Example: A patient with scleroderma renal crisis and BP 196/118. Initial management?

A) Start ACE inhibitor and titrate aggressively ✓
B) Initiate hemodialysis immediately
C) Discontinue steroids and start pulse methylprednisolone
D) Start therapeutic plasma exchange

Despite the dramatic presentation, the ACE inhibitor (A) is the proven first-line therapy. The more aggressive-sounding options are either harmful or not first-line.

3. Sequencing / Timing

More than one option is a reasonable thing to do — they differ by order or timing(induction vs maintenance, hold vs resume, now vs after a step). The question tests temporal reasoning: not just what to do, but when. Lead-ins like "at this time" or "next step" are the tell.

Tip:

  • Place the patient in the disease course: newly diagnosed, in remission, or relapsing?
  • Decide what to do right now, not eventually — two defensible options is the signal
  • Induction → maintenance → escalation: match the phase to the scenario

Example: ANCA-associated vasculitis in remission after 6 months of cyclophosphamide and prednisone. Most appropriate next step?

A) Continue cyclophosphamide for another 6 months
B) Switch to rituximab for maintenance ✓
C) Stop all immunosuppression
D) Add methotrexate to cyclophosphamide

After induction achieves remission, the patient transitions to maintenance (rituximab). Continuing cyclophosphamide adds toxicity; stopping everything risks relapse. The drugs aren't the question — the timing is.

4. Convergence / Odd-One-Out

Three options cluster around one idea and one stands apart. This is the tip to use most cautiously: an option being "different" is a test-taking signal, not proof. Sometimes the outlier is the key (it's the disease-specific or exception answer); just as often it's bait. Reason the case forward, then use the structure only to sanity-check — never to decide.

Tip:

  • The outlier tends to be correct when the lead-in asks for an exception, contraindication, or distinguishing feature
  • When the lead-in asks for the "most appropriate" routine action, the outlier is more often the distractor
  • If you can't justify it clinically, the shape alone isn't enough

Example: Which biomarker combination best monitors anti-MDA5 dermatomyositis?

A) Lymphocyte count and CRP levels
B) Anti-MDA5 antibody titer and ferritin levels ✓
C) ESR and creatine kinase levels
D) Complement levels and ANA titer

A, C, and D are generic inflammatory markers; B is the disease-specific combination for MDA5-DM. Here the outlier is correct — but only because the medicine supports it, not because it looks different.

Gierl MJ, et al. Developing, Analyzing, and Using Distractors for Multiple-Choice Tests. Rev Educ Res. 2017 →

References

← Back to Question Analysis