Andreas Stang1, Charles Poole, Ralf Bender. 1. Institute of Clinical Epidemiology, Medical Faculty, Martin-Luther-University of Halle-Wittenberg, Magdeburger Strasse 8, 06097 Halle, Germany. andreas.stang@medizin.uni-halle.de
Abstract
OBJECTIVE: To illustrate basic issues that have to be taken into account when study results are presented by means of the number needed to treat (NNT). STUDY DESIGN AND SETTING: This article presents an overview of common problems related to the NNT with corresponding explanations. RESULTS: Without stating the direction of the effect, the alternative treatment, the treatment period, and the follow-up period, information in terms of NNTs is uninterpretable. The naive use of person-time data for the calculation of NNTs is frequently inappropriate. Rounding NNTs to the next upward integer may obscure differences among therapies. CONCLUSIONS: The basic information about which treatments are compared, the treatment period, the follow-up period, and the direction of the effect should be given when study results are presented in terms of NNTs. Adequate methods should be used for point and interval estimation of NNTs. Unnecessary rounding of NNTs should be avoided. In more complicated situations of confounding or varying follow-up times, the use of more sophisticated methods is required with increasing potential for misinterpretation. Copyright (c) 2010 Elsevier Inc. All rights reserved.
OBJECTIVE: To illustrate basic issues that have to be taken into account when study results are presented by means of the number needed to treat (NNT). STUDY DESIGN AND SETTING: This article presents an overview of common problems related to the NNT with corresponding explanations. RESULTS: Without stating the direction of the effect, the alternative treatment, the treatment period, and the follow-up period, information in terms of NNTs is uninterpretable. The naive use of person-time data for the calculation of NNTs is frequently inappropriate. Rounding NNTs to the next upward integer may obscure differences among therapies. CONCLUSIONS: The basic information about which treatments are compared, the treatment period, the follow-up period, and the direction of the effect should be given when study results are presented in terms of NNTs. Adequate methods should be used for point and interval estimation of NNTs. Unnecessary rounding of NNTs should be avoided. In more complicated situations of confounding or varying follow-up times, the use of more sophisticated methods is required with increasing potential for misinterpretation. Copyright (c) 2010 Elsevier Inc. All rights reserved.
Authors: Bahman P Tabaei; Renata E Howland; Jeffrey S Gonzalez; Shadi Chamany; Elizabeth A Walker; Clyde B Schechter; Winfred Y Wu Journal: Diabetes Care Date: 2020-03-04 Impact factor: 19.112
Authors: Jeffrey F Scherrer; Joanne Salas; Patrick Lustman; Peter Tuerk; Sarah Gebauer; Sonya B Norman; F David Schneider; Kathleen M Chard; Carissa van den Berk-Clark; Beth E Cohen; Paula P Schnurr Journal: Eur J Prev Cardiol Date: 2019-05-13 Impact factor: 7.804
Authors: Jonathan C Hill; David G T Whitehurst; Martyn Lewis; Stirling Bryan; Kate M Dunn; Nadine E Foster; Kika Konstantinou; Chris J Main; Elizabeth Mason; Simon Somerville; Gail Sowden; Kanchan Vohora; Elaine M Hay Journal: Lancet Date: 2011-09-28 Impact factor: 79.321