Decision curve analysis is a simple method for evaluating prediction models, diagnostic tests, and molecular markers.
The method was first published as:
A subsequent discussion paper gives some further details about the method:
A paper describing various extensions to decision curve analysis, such as application to survival time data, is:
Vickers AJ, Cronin AM, Elkin EB, Gonen M. BMC Medical Informatics and Decision Making. 2008 Nov 26;8(1):53. Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers.
For more on the theoretical background of decision curve analysis, see:
Steyerberg EW, Vickers AJ, Cook NR, Gerds T, Gonen M, Obuchowski N, Pencina MJ, Kattan MW. Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology. 2010 Jan;21(1):128-38.
Vickers AJ, Cronin AM. Traditional statistical methods for evaluating prediction models are uninformative as to clinical value: towards a decision analytic framework. Semin Oncol. 2010 Feb;37(1):31-8.
Two papers that extend decision curve analysis to issues of treatment response are:
Editorials recommending decision curve analysis have been published in several major journals.
Kerr KF, Brown MD, Zhu K, Janes H. Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use. J Clin Oncol. 2016 Jul 20;34(21):2534-40. doi: 10.1200/JCO.2015.65.5654.
Localio AR, Goodman S. Beyond the usual prediction accuracy metrics: reporting results for clinical decision making. Ann Intern Med. 2012 Aug 21;157(4):294-5. doi: 10.7326/0003-4819-157-4-201208210-00014.
Statistical Code for Running Decision Curve Analysis
A tutorial for running decision curve analysis using Stata, R and SAS takes the reader step by step through doing a basic decision curve analysis, including formatting the graph, interpreting the results and saving the output (download PDF ). Also included are instructions on: showing net reduction in interventions; evaluation of joint or conditional tests; evaluation of published models; application to censored data, including competing risk.
The zip file contains the data sets, Stata code, and R code used in the tutorial (download Decision Curve Analysis tutorial.zip). The example data sets used are provided as text, CSV, Stata, and R format.