Biostatistics & Research · patient case question
A cohort study reports a relative risk of 3.1 with a 95% confidence interval of 2.4 to 4.0. Which statement is correct?
Patient case: The business of the practice
Patient: Practice and population-level scenarios
Chief Complaint: Not applicable.
Background: A prescription is being written for prophylaxis. The practice insurance portfolio is under review. Rising floodwater is entering the building. A crown is being placed in a better material than quoted, at no extra charge. A patient irradiated for a head and neck cancer ten years ago requests an extraction and refuses radiographs. The local council is consulting on water fluoridation. A published cohort study is being appraised.
Findings: Not applicable.
Show answer and explanation
Correct answer: B. Exposed participants developed the outcome about 3.1 times as often as unexposed participants, and the result is statistically significant
- A. That describes the ODDS RATIO, which is what a CASE-CONTROL study yields. In a case-control design participants are selected by OUTCOME, so incidence cannot be measured and true relative risk cannot be calculated. The odds ratio approximates the relative risk only when the outcome is rare.
- B. CORRECT, on both halves. RELATIVE RISK is the incidence of the outcome among the EXPOSED divided by the incidence among the UNEXPOSED, so 3.1 means the exposed group developed the outcome 3.1 times as often. Interpretation: RR greater than 1 is INCREASED risk, RR of 1 is NO difference, RR less than 1 is REDUCED risk. And the confidence interval doubles as the significance test — because the interval 2.4 to 4.0 EXCLUDES the null value of 1, the result is statistically significant at the 5% level. A wide interval crossing 1 would not be, however impressive the point estimate looked.
- C. This inverts the rule. EXCLUDING the null value is exactly what makes a ratio measure statistically significant.
- D. 3.1 is a RATIO of incidences, not a percentage of participants, and a ratio on its own does not establish causation.
Note: Know which design yields which measure — COHORT studies, whether prospective or retrospective, give RELATIVE RISK because incidence is measurable; CASE-CONTROL studies give the ODDS RATIO. For RATIO measures the null value is 1; for DIFFERENCE measures such as risk difference or mean difference the null value is 0.
Practice Biostatistics & Research questions Biostatistics & Research study notes