I85PREP
⌘ K
NMLE Integrated study materials
NMLE Integrated study resources

NMLE Integrated: Preventive Medicine and Public Health: Numbers and Bias

NMLE Integrated · Work through your exam, one topic at a time.

All guide sections

Preventive Medicine and Public Health: Numbers and Bias

Preventive medicine and public health items test measures of disease frequency and association, study design, screening, outbreak investigation, immunisation and health policy. The questions are numerical or logical rather than memorised lists, and most errors come from mixing up which denominator belongs to which measure.

Core rules

  • Frequency. Incidence counts new cases per population at risk per unit time; prevalence counts all existing cases. Prevalence ≈ incidence × average duration, so longer survival raises prevalence without changing incidence.
  • Association. Relative risk compares incidence between exposed and unexposed and is used in cohort studies. The odds ratio comes from case-control studies and approximates relative risk only when the outcome is rare.
  • Impact. Attributable risk is the difference in incidence between exposed and unexposed; number needed to treat is 1 divided by the absolute risk reduction.
  • Screening. Sensitivity and specificity are properties of the test; positive predictive value depends on prevalence and falls as prevalence falls.
  • Bias. Lead-time bias makes earlier diagnosis look like longer survival; length bias over-samples slowly progressive disease; selection bias distorts who enters the study.
  • Outbreak. Define the case, describe person, place and time, form a hypothesis, test it, and implement control before waiting for final confirmation.

Worked example

In a screening study of 1,000 people, 100 have the disease. Of these, 90 test positive. Of the 900 without disease, 180 test positive. Sensitivity is 90/100 = 90%; specificity is 720/900 = 80%; positive predictive value is 90/270 = 33%; negative predictive value is 720/730 ≈ 98.6%. If the same test were used in a population with 1% prevalence, sensitivity and specificity would be unchanged, but positive predictive value would collapse. The reasoning: predictive values move with prevalence because the number of false positives changes while the test characteristics do not.

Traps that cost marks

  • Reporting sensitivity as positive predictive value, or vice versa.
  • Using an odds ratio as if it were a relative risk when the outcome is common.
  • Concluding that screening prolongs survival when only lead time has increased.
  • Confusing incidence with prevalence when evaluating a preventive intervention.