Trial result applied to a rural clinic
A randomized trial of a new antihypertensive drug enrolled 8,000 adults with a mean age of 58 years from urban academic centers. Over 4 years, the primary composite outcome (stroke, myocardial infarction, cardiovascular death) fell from 9.2% to 7.4% (relative risk 0.80, 95% CI 0.70–0.92, p = 0.002). Absolute risk reduction was 1.8%, number needed to treat 56. The drug caused dizziness in 12% versus 6% in controls; 4% discontinued because of it. You run a rural clinic where the average patient is 71 years old, many have chronic kidney disease stage 3 and take multiple medications, and the clinic has limited laboratory monitoring.
1. Calculate and interpret the key effect measures. What does the NNT of 56 mean in your setting?
Read the worked answer
Effect measures
- Relative risk 0.80 means a 20% relative reduction in the composite outcome.
- Absolute risk reduction = 9.2% − 7.4% = 1.8 percentage points.
- NNT = 1 / 0.018 ≈ 56 over 4 years.
Interpretation
- NNT 56 is modest but clinically meaningful for a common outcome. It means 56 patients treated for 4 years prevent one event.
- The NNT is setting-dependent: if your patients have higher baseline risk (older, CKD), the ARR and benefit may be larger, but so may harms.
- Conversely, if adherence is lower and monitoring is weaker, real-world benefit shrinks.
- The 95% CI (0.70–0.92) excludes 1, so the result is statistically robust, but statistical significance is not the same as applicability.
2. Apply the trial to your population using the concepts of internal and external validity. What factors limit generalizability?
Read the worked answer
Internal validity
- Randomization and a narrow CI support a real effect within the trial population.
- However, the trial population (mean age 58, urban academic centers) differs from your clinic.
External validity / applicability
- Age: your patients are older; drug effects and fall risk may differ.
- Comorbidity: CKD stage 3 affects drug clearance and monitoring needs; the trial may have excluded such patients.
- Polypharmacy: interaction and adherence risks are higher.
- Monitoring capacity: limited labs mean you cannot replicate trial safety surveillance.
- Adherence: dizziness causing 4% discontinuation in a trial may cause higher discontinuation in routine care.
Judgment: the direction of benefit likely holds, but the magnitude and safety profile are uncertain. A reasonable approach is to start at a lower dose, monitor renal function and orthostatic blood pressure, and prioritize patients at highest cardiovascular risk.
3. How would you decide whether to adopt the drug in your formulary, and what would you monitor?
Read the worked answer
Decision framework
- Applicability: benefit is plausible but not proven in your exact population; treat the trial as supportive, not definitive.
- Harms: dizziness (12%) is a fall risk in older adults; weigh against stroke prevention.
- Cost and logistics: monitoring requirements may offset benefit.
- Shared decision-making: discuss NNT and side-effect risk with patients.
Practical adoption
- Start with a defined subgroup (e.g., high-risk patients with normal renal function).
- Use a lower starting dose and titrate.
- Monitor blood pressure (including orthostatic), renal function at baseline and after 2–4 weeks, and symptoms of dizziness.
- Review at 3 months for adherence and adverse effects.
Exam point: evidence-based medicine requires translating relative and absolute effects into the local context, not simply accepting a p-value.