NNT numbers needed to treat limitations

The number needed to treat (NNT) is a mathematical way of communicating the effectiveness of an intervention, in this case the effectiveness of full anticoagulation in patients with COVID-19 not in the ICU. Calculating the NNT will provide an estimate of the average number of patients who will need to be treated with full-dose heparin to prevent one additional undesired outcome (eg, the development of shock or respiratory failure). It is defined as the inverse of the absolute risk reduction (1/ARR). If 76.4% of the usual care prophylaxis group avoided organ failure or death, then 23.6% experienced this outcome. Similarly, if the primary outcome of organ failure-free survival occurred in 80.2% of the full-dose heparin group, then 19.8% of this group experienced the adverse outcome. The ARR in this case would therefore be 23.6% - 19.8% = 3.8%. For the calculation of the NNT, the ARR should be expressed as a decimal. The NNT would therefore be 1/0.038 = 26. On average, 26 patients would need to receive full anticoagulation to prevent one additional patient from developing organ failure or death (choice A is correct; choices B, C, and D are incorrect).

There are important limitations to using the NNT as a definitive answer as to whether a treatment choice is justified. First, the NNT, as calculated in this example, did not reflect the uncertainty of the treatment effect because CIs are not typically used in the NNT calculation. Second, the measured NNT is a time-specific estimate. In our case, the NNT estimated the risk reduction from using full-dose heparin for COVID-19 only during the first 21 days or during the duration of a single hospitalization, whichever came first. It is plausible that longer treatment durations could have produced different effects. Finally, the NNT provides no information by itself about other risks of a treatment. Specifically, we need to provide the pharmacy and therapeutics committee with information about the bleeding risks associated with full-dose heparin in the study described before a decision can be made about changing practice. A separate calculation, the number needed to harm (NNH), can help inform this decision.

The NNH is an estimate of the average number of patients who need to be treated with an intervention to observe one additional undesired outcome, in this case major bleeding. It is defined as the inverse of the absolute risk increase (1/ARI). Major bleeding occurred in 0.9% of the usual care group and in 1.9% of the full-dose anticoagulation group, so the ARI was 1%. Therefore, the NNH = 1/0.01 = 100. On average, 100 patients hospitalized with COVID-19 who were not critically ill would need to receive full anticoagulation with heparin to expect one additional case of major bleeding.1234

Footnotes

  1. SEEK Questionnaires ↩

  2. Number needed to treat. Wikipedia. Updated February 27, 2022. https://en.wikipedia.org/wiki/Number_needed_to_treat ↩

  3. Stang A, Poole C, Bender R. Common problems related to the use of number needed to treat. J Clin Epidemiol. 2010;63(8):820-825. PubMed ↩

  4. Wen L, Badgett R, Cornell J. Number needed to treat: a descriptor for weighing therapeutic options. Am J Health Syst Pharm. 2005;62(19):2031-2036. PubMed ↩