A new drug is claimed to reduce the risk of stroke in high-risk patients. In its clinical trial, the new drug was compared against placebo to assess its efficacy in preventing stroke in high-risk patients. A total of 5,500 patients were included in the trial: 2,750 patients in the treatment group and 2,750 patients in the control group. In the treatment group, 55 patients experienced a stroke during the trial period while 250 patients in the control group experienced stroke during the same trial period.
What is the number needed to treat to prevent one stroke for using the new medication?
Answer with rationale:
Number needed to treat (NNT) is a statistic commonly used in healthcare to demonstrate the effectiveness of a treatment. Specifically, it is the average number of patients that need to be treated to prevent a single negative outcome (i.e., stroke in the above scenario). Therefore, the lower the NNT the more effective the treatment since fewer patients need to receive the treatment to see benefit. Many studies do not include NNT in their results. It is something the reader often needs to calculate for themselves to properly analyze the data and determine its clinical significance. Thankfully, the calculation is relatively straightforward.
To calculate NNT, you must first know the absolute risk reduction (ARR) since NNT is the inverse of ARR. ARR is the difference in the event rate of the negative outcome between the control and treatment groups and is calculated by subtracting the experimental event rate (EER) from the control event rate (CER). The ability to successfully calculate NNT depends on being able to accurately determine which data points belong to which group.
Here is the calculation in full:
The NNT is 14.1 but because you cannot treat a partial patient, NNT is always rounded up to the next whole number. Therefore the NNT would be 15 to prevent one stroke using this medication compared to the control group over the course of the evaluated time period of the study.
When evaluating efficacy, safety should always be evaluated as well. So a number needed to harm (NNH) should also be calculated when evaluating clinical studies when a treatment is associated with a side effect compared to placebo. In a study like this example, if the medication was associated with more major bleeding compared to placebo, the NNH calculation should also be performed to determine an overall risk:benefit ratio for the medication.
NAPLEX content domains covered:
1.C.8
Domain 1 Foundational Knowledge for Pharmacy Practice– C. Pharmaceutical compounding 8. Biostatistical, epidemiological, or pharmacoeconomic measures
1.E
Domain 1 Foundational Knowledge for Pharmacy Practice– E. Research design principles and biostatistics