What is NNT?

What is NNT?

• A. The number of patients with a similar condition I would need to treat to get a positive benefit for one patient
• B. The ratio of treated to control event rates
• C. The time to event in treated vs untreated
• D. The minimal clinically important difference in risk

Answer: (A)

NNT shows how many patients with the same condition you would need to treat for one to experience a net positive outcome, based on the absolute reduction in risk observed in a study. It’s the reciprocal of the absolute risk reduction: NNT = 1 / ARR, where ARR = control event rate minus treated event rate. For example, if the adverse outcome occurs in 10% of the control group and 4% of the treated group, the ARR is 0.06, so the NNT is about 17. A smaller NNT means the treatment yields a benefit more efficiently within that specific outcome and time frame. NNT is specific to the outcome and the follow-up period; changing either changes the NNT. The other concepts describe different ideas: a ratio of treated to control event rates is a relative measure, not a count of patients needed to treat; time to event is about when events occur, not how many need treatment for one to benefit; the minimal clinically important difference in risk is about the smallest change that patients perceive as meaningful, not a practical count to achieve a benefit.

NNT shows how many patients with the same condition you would need to treat for one to experience a net positive outcome, based on the absolute reduction in risk observed in a study. It’s the reciprocal of the absolute risk reduction: NNT = 1 / ARR, where ARR = control event rate minus treated event rate. For example, if the adverse outcome occurs in 10% of the control group and 4% of the treated group, the ARR is 0.06, so the NNT is about 17. A smaller NNT means the treatment yields a benefit more efficiently within that specific outcome and time frame. NNT is specific to the outcome and the follow-up period; changing either changes the NNT. The other concepts describe different ideas: a ratio of treated to control event rates is a relative measure, not a count of patients needed to treat; time to event is about when events occur, not how many need treatment for one to benefit; the minimal clinically important difference in risk is about the smallest change that patients perceive as meaningful, not a practical count to achieve a benefit.