Real-time imputation of missing predictor values improved the application of prediction models in daily practice

• Published in: Journal of clinical epidemiology Vol. 134; pp. 22 - 34
• Main Authors: Nijman, Steven Willem Joost, Groenhof, T. Katrien J., Hoogland, Jeroen, Bots, Michiel L., Brandjes, Menno, Jacobs, John J.L., Asselbergs, Folkert W., Moons, Karel G.M., Debray, Thomas P.A.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.06.2021; Elsevier Limited
• Subjects: Cardiovascular disease; Clinical medicine; Computerized decision support system; Confidence intervals; Datasets; Decision making; Electronic health records; Epidemiology; Missing data; Modelling; Multiple imputations; Patients; Prediction; Prediction models; Real time; Real-time imputation; Root-mean-square errors; Variables; Missing data; Real-time imputation; Multiple imputations; Computerized decision support system; Electronic health records; Prediction
• ISSN: 0895-4356; 1878-5921
• DOI: 10.1016/j.jclinepi.2021.01.003

In clinical practice, many prediction models cannot be used when predictor values are missing. We, therefore, propose and evaluate methods for real-time imputation. We describe (i) mean imputation (where missing values are replaced by the sample mean), (ii) joint modeling imputation (JMI, where we use a multivariate normal approximation to generate patient-specific imputations), and (iii) conditional modeling imputation (CMI, where a multivariable imputation model is derived for each predictor from a population). We compared these methods in a case study evaluating the root mean squared error (RMSE) and coverage of the 95% confidence intervals (i.e., the proportion of confidence intervals that contain the true predictor value) of imputed predictor values. –RMSE was lowest when adopting JMI or CMI, although imputation of individual predictors did not always lead to substantial improvements as compared to mean imputation. JMI and CMI appeared particularly useful when the values of multiple predictors of the model were missing. Coverage reached the nominal level (i.e., 95%) for both CMI and JMI. Multiple imputations using either CMI or JMI is recommended when dealing with missing predictor values in real-time settings. •Use of conditional distributions per predictor value allow for real-time imputation.•A joint distribution can be estimated directly assuming multivariate normality (JMI).•Distributions can be estimated separately via a flexible modelling strategy (CMI).•Simulations indicate both JMI and CMI to be an improvement over mean imputation.