Estimate risk difference and number needed to treat in survival analysis.
| Autor: | Zhang Z; Department of Emergency Medicine, Sir Run-Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou 310016, China., Ambrogi F; Department of Biostatistics, University of Milan, Department of Clinical Sciences and Community Health, Laboratory of Medical Statistics and Biometry 'Giulio A. Maccacaro', Campus Cascina Rosa, Via Vanzetti 5, 20133 Milano, Italy., Bokov AF; Clinical Informatics Research Division, Department of Epidemiology and Biostatistics, UT Health San Antonio, San Antonio, TX, USA., Gu H; China National Clinical Research Center for Neurological Diseases, Beijing 100050, China.; Tiantan Clinical Trial and Research Center for Stroke, Department of Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing 100050, China., de Beurs E; Department of Clinical Psychology, Faculty of Social Sciences, Leiden University, the Netherlands., Eskaf K; College of Computing and Information Technology, Arab Academy for Science, Technology and Maritime Transport, Alexandria, Egypt. |
|---|---|
| Jazyk: | angličtina |
| Zdroj: | Annals of translational medicine [Ann Transl Med] 2018 Apr; Vol. 6 (7), pp. 120. |
| DOI: | 10.21037/atm.2018.01.36 |
| Abstrakt: | The hazard ratio (HR) is a measure of instantaneous relative risk of an increase in one unit of the covariate of interest, which is widely reported in clinical researches involving time-to-event data. However, the measure fails to capture absolute risk reduction. Other measures such as number needed to treat (NNT) and risk difference (RD) provide another perspective on the effectiveness of an intervention, and can facilitate clinical decision making. The article aims to provide a step-by-step tutorial on how to compute RD and NNT in survival analysis with R. For simplicity, only one measure (RD or NNT) needs to be illustrated, because the other measure is a reverse of the illustrated one (NNT=1/RD). An artificial dataset is composed by using the survsim package. RD and NNT are estimated with Austin method after fitting a Cox-proportional hazard regression model. The confidence intervals can be estimated using bootstrap method. Alternatively, if the standard errors (SEs) of the survival probabilities of the treated and control group are given, confidence intervals can be estimated using algebraic calculations. The pseudo-value model provides another method to estimate RD and NNT. Details of R code and its output are shown and explained in the main text. Competing Interests: Conflicts of Interest: The authors have no conflicts of interest to declare. |
| Databáze: | MEDLINE |
| Externí odkaz: |
|