Treatment of Hypertension to Prevent Stroke: Translating Evidence into Clinical Practice

Robert C. Kaplan, PhD

In This Article

Number Needed To Treat

The number needed to treat to prevent a clinical event (NNT) is more useful to the clinician than the RRR and can be obtained from data in published reports of clinical trials. This measure answers the question: How many patients like the ones studied would need to be treated to prevent one event? The NNT is computed as the reciprocal of the difference in event rates between the active treatment group and the control group. For example, in the trials involving patients with entry DBP above 115 mm Hg, the 5-year stroke incidence rate in the untreated individuals was 8.2%, or 0.082 (Table). The 5-year rate among treated patients in these trials was 4.7%, or 0.047, which may be derived by multiplying 0.082 (the rate in the untreated group) by 0.57 (1 minus the RRR of 43%). The difference in rates between the treatment and control groups is 0.035 (equal to 0.082 minus 0.047), and the reciprocal of this difference (or NNT) is 29.

In the trials conducted with patients with relatively nonsevere hypertension (entry DBP 90-110 mm Hg), the NNT to prevent one stroke is 118 (Table). The NNT estimates from trials of more severely hypertensive patients are substantially lower (NNT=52 for entry DBP at or below 115 mm Hg, and NNT=29 for entry DBP above 115 mm Hg). These figures confirm numerically what intuition tells us -- that treating the most severely hypertensive individuals is the most rewarding.

Using the NNT, we can make more informed decisions regarding the treatment of hypertensive patient populations. For example, we can decide whether it is a worthwhile "investment" to treat 118 patients with relatively nonsevere hypertension for 5 years to prevent one stroke event. Such factors as monetary costs and potential adverse effects associated with drug treatment are weighed against the benefit to be gained in terms of stroke prevention. Given the devastating consequences of stroke, most would find it sensible to treat approximately 120 individuals with nonsevere hypertension with the expectation that one stroke would be prevented over the following 5 years.

The NNT may also be used to evaluate whether antihypertensive treatment compares favorably with other interventions that may be undertaken to prevent stroke. For example, the expected benefits of antihypertensive treatment can be compared with the expected benefits of lipid-lowering drug therapy. The West of Scotland Coronary Prevention Study (WOSCOPS) results suggested that among hyperlipidemic men without prior heart disease, 642 individuals would need to be treated for 5 years with pravastatin in order to prevent one stroke.[21] These data suggest a lesser impact on stroke risk in the WOSCOPS than that obtained in the blood pressure-lowering trials.

It is important to keep in mind that the NNT is a function of the rate of clinical outcomes observed among trial participants. Patients enrolled in clinical trials are healthier, on average, than those encountered in practice.[22] If the rate of clinical outcomes in the placebo or "standard care" control arm is substantially different from what would be expected in the clinical setting, the NNT must be interpreted with caution. In recognition of this, most reports of clinical trial results present detailed information on study eligibility criteria and observed outcome rates to help readers judge the generalizability of the results.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: