CPR APPLICATIONS FOUNDATION · CHAPTER 20 · NORMAL
Normal Foundations of Biostatistics & Epidemiology
Organize the chapter before drilling into individual CPR concepts.
Supplemental foundation map: this page orients the detailed CPR Applications for Chapter 20. It is not part of the canonical CPR object inventory.
Move through Chapter 20
Normal Foundations · You are here
Abnormal Foundations
Treatment Foundations
20.1 · SOURCE CPR FOUNDATION
Concept
Biostatistics is clinical reasoning under uncertainty. Step 1 rarely tests math in isolation. It tests whether you can interpret a test result, study abstract, confidence interval, graph, or public-health scenario without false reassurance, false alarm, or causal overreach.
Pattern
Step 1 biostatistics is usually embedded inside clinical stems. The question may show a screening result, a new diagnostic test, an RCT abstract, an outbreak table, a survival curve, or a conclusion that sounds stronger than the data support. The task is to decide what the result really means.
Rule
- Ask what kind of uncertainty is present. Diagnostic uncertainty, statistical uncertainty, causal uncertainty, and public-health uncertainty are different.
- Match the metric to the question. Test performance, predictive value, inference, bias, risk, or decision-making each uses different logic.
- Do not overinterpret the result. A significant p-value, positive screen, or impressive relative risk may still mislead.
Applications
- Positive screen in rare disease → PPV trap
- Wide confidence interval → imprecise estimate
- Nonsignificant underpowered study → may be inconclusive; assess confidence interval and statistical power rather than assuming equivalence
- Statistically significant tiny effect → clinical irrelevance
- Subgroup-only apparent benefit → consider genuine effect modification but evaluate prespecification, interaction testing, and multiplicity
- Observational association → causal inference caution
- Tip: Biostatistics questions usually ask what the result means, not what number you can calculate
CPR APPLICATIONS · ACTIVE REASONING
Understand predictive value, uncertainty, and absolute risk
Commit to an answer before requesting feedback. Full worked explanations remain available if scripts are unavailable. Answers are not transmitted or stored.
Reveal the worked answer and changed-variable challenge
Best answer: About 8%. Among 1,000 people, about 10 have disease: 9 true positives and 99 false positives, so PPV = 9/108 ≈ 8.3%.
Transfer: How does PPV change if prevalence rises to 10% while sensitivity and specificity remain unchanged?
Reveal the worked answer and changed-variable challenge
Best answer: 5 percentage points. Absolute risk reduction is 20% minus 15% = 5 percentage points; NNT over this study period is 1/0.05 = 20.
Transfer: How would the NNT change if baseline risk were much lower but relative risk reduction stayed fixed?
Reveal the worked answer and changed-variable challenge
Best answer: The estimate is compatible with modest benefit, no effect, and modest harm. The interval includes a risk ratio of 1.0 and thus cannot exclude those possibilities at the reported confidence level.
Transfer: What would a narrower interval entirely below 1 imply, assuming a well-designed unbiased study?
Change the mechanism
Read all comparisons without scripts
- Low prevalence (1%): Sensitivity 90% / specificity 90% gives PPV ≈8.3% at prevalence 1%.
- Moderate prevalence (10%): With the same test, PPV = 50%; changing prevalence changes the mix of true and false positives.
- Higher prevalence (25%): With the same test, PPV = 75%.
Screening test simulator: prevalence changes PPV
Assume sensitivity 90%, specificity 90%, and 1,000 screened people. Change prevalence while keeping the test characteristics fixed.
Worked values if the calculator is unavailable
| Prevalence | True positives | False positives | PPV |
|---|---|---|---|
| 1% | 9 | 99 | 8.3% |
| 10% | 90 | 90 | 50.0% |
| 25% | 225 | 75 | 75.0% |
Explain it back: Identify the mechanism, name the most important competing explanation, and predict what changes when the context shifts.
Unscored educational reasoning. Not a validated assessment or individualized clinical recommendation.
How to use this foundation map
- Identify the organizing mechanism.
- Recognize the pattern.
- Restate the rule without looking.
- Test transfer with applications and question practice.
Pair CPR with the 5P Approach™ when solving cases and questions.