A researcher is conducting a study to compare fracture risk in male patients above the age of 65 who received annual DEXA screening to peers who did not receive screening. He conducts a randomized controlled trial in 900 patients, with half of participants assigned to each experimental group. The researcher ultimately finds similar rates of fractures in the two groups. He then notices that he had forgotten to include 400 patients in his analysis. Including the additional participants in his analysis would most likely affect the study's results in which of the following ways?

Increased probability of rejecting the null hypothesis when it is truly false

Wider confidence intervals of results

Increased probability of committing a type II error

Decreased significance level of results

Increased external validity of results

Primary vs secondary outcomes for USMLE Step 2 CK: High-Yield Biostatistics Lessons

Primary vs Secondary Outcomes - The Main Event

Feature	Primary Outcome	Secondary Outcome
Purpose	The main question the study is designed to answer; the focus of hypothesis testing.	Additional outcomes of interest that are also observed.
Specification	Must be pre-specified in the study protocol before data collection begins.	Usually pre-specified, but may be exploratory; not the main focus.
Power	The study's sample size is calculated based on the power to detect a difference in this outcome.	The study is not typically powered to detect differences for these outcomes.
Example	A composite endpoint like "all-cause mortality or hospitalization" in an RCT.	Individual components of the primary endpoint, or other effects like "quality of life".

The Primary Outcome - Planning for Power

📌 Primary outcome determines the study's Power and Purpose.

Primary Outcome: The single, pre-specified endpoint used to determine the study's success and calculate the required sample size. It must be clinically meaningful.
Statistical Power: The probability of finding a true effect, if one exists. It avoids a Type II error.
- Formula: $Power = 1 - \beta$
- Standard goal: Power is typically set at ≥80%.
Errors in Hypothesis Testing:
- Type I Error (α): False positive. Rejecting a true null hypothesis. The significance threshold (p-value) is usually <0.05.
- Type II Error (β): False negative. Failing to reject a false null hypothesis (i.e., missing a real effect).
Endpoints:
- Single: Measures one specific outcome (e.g., all-cause mortality).
- Composite: Combines multiple endpoints (e.g., MACE: CV death, MI, stroke). ↑ statistical power by ↑ event rate.

⭐ A study with a non-significant primary outcome is considered 'negative,' even if multiple secondary outcomes are statistically significant.

Secondary & Exploratory Outcomes - Handle with Care

Secondary Outcomes:
- Pre-specified outcomes that are not the primary endpoint of the study.
- Often lack sufficient statistical power for definitive conclusions.
- Results are considered hypothesis-generating, not confirmatory.
- 📌 Secondary results are Suggestive and need Separate studies for confirmation.
Exploratory & Post-Hoc Analyses:
- Analyses that were not pre-specified in the trial protocol.
- Carry a high risk of finding spurious associations due to "p-hacking" or data dredging.
Multiplicity & Type I Error Inflation:
- Testing multiple outcomes increases the probability of finding a statistically significant result purely by chance.
- This is the problem of multiple comparisons, which inflates the overall Type I error rate.

⭐ Testing multiple secondary outcomes increases the risk of finding a 'significant' result by chance alone (Type I error inflation), a problem known as multiplicity.

Type I and Type II Errors in Hypothesis Testing

High-Yield Points - ⚡ Biggest Takeaways

The primary outcome is the single, pre-defined endpoint used to determine a study's success, directly testing its main hypothesis.

Statistical power and sample size are calculated based on the primary outcome.

Secondary outcomes are additional endpoints that are exploratory and can generate new hypotheses.

A trial's main conclusion rests solely on the primary outcome's results.

Finding significance in secondary outcomes alone can be a result of p-hacking or data dredging.

Continue reading on Oncourse

CONTINUE READING — FREE

or get the app

App Store|Google Play

Primary vs secondary outcomes