CLIN-03 · Base rates in the waiting room

Same test, same threshold, two Virginia clinics with genuinely different prevalence. Using public Virginia health data, recalculate predictive values for both settings — and say plainly which clinic should not be using this result to rule a condition in.

Public prevalence or incidence figures for one Virginia condition, cited to a public source — Virginia Department of Health data portals (https://www.vdh.virginia.gov/data/) or CDC surveillance (https://data.cdc.gov/). Two clinic profiles the student constructs and labels as hypothetical: a referral or symptomatic setting and a general or asymptomatic setting. Test characteristics come from a cited source, never from the student.

Using a statewide rate as a clinic rate without saying that is a substitution. Confusing incidence with prevalence. Building the two clinics to produce a dramatic gap and calling that a finding. Naming a real clinic — the profiles are hypothetical and must say so on the page. Reading a county rate as a waiting-room rate when the clinic's referral pattern is the thing that actually sets prevalence.

A two-setting comparison: cited source and figure for the condition; both clinic profiles with the stated reasoning for each prevalence estimate and its uncertainty; both 2×2s at fixed test characteristics; PPV/NPV for each; and a one-paragraph recommendation naming which setting can use a positive result to rule in, which cannot, and what the second setting should do instead.

CT: distinguishing a population rate from a clinic rate. CR: constructing defensible clinic profiles. CO: a peer challenges one prevalence estimate and the student either defends it or revises it in writing. CM: a paragraph a clinic manager could act on. CZ: what over-testing costs the low-prevalence population.

A clinician, public-health analyst, or clinic administrator reviews the prevalence reasoning. Standing instruction: reject any profile that reads as a real named facility. School-supervised.

R1: one condition, one test, two named settings. R2: every rate cited with source and year. R3: comparator is the statewide rate, with the substitution stated. R4: uncertainty in the prevalence estimates carried into the conclusion. R5: recommendation is one actionable paragraph. R6: refuses to name real facilities or to present hypothetical profiles as observed.

(a) A statewide number is dropped into both clinics, so the two columns are identical and the brief has no content. (b) The student invents the prevalence gap, then reports the resulting PPV gap as a discovery.

Lane A immediately (health-science seminar, HOSA, or a community-health project). No verified credit claim.