Impact of imputation of missing data on estimation of survival rates: an example in breast cancer.
Year: 2010
Type: article
Abstract: Background: Multifactorial regression models are frequently used in medicine to \nestimate survival rate of patients across risk groups. However, their results are not \ngeneralisable, if in the development of models assumptions required are not \nsatisfied. Missing data is a common problem in pathology. The aim of this paper \nis to address the danger of exclusion of cases with missing data, and to highlight \nthe importance of imputation of missing data before development of multifactorial \nmodels. \nMethods: This study was performed on 310 breast cancer patients diagnosed in \nShiraz (Southern Iran). Performing a complete-case Cox regression model, a \nprognostic index was calculated so as to categorise the patients into 3 risk groups. \nThen, applying the Multivariate Imputation via Chained Equations (MICE) method, \nmissing data were imputed 10 times. Using imputed data sets, modelling was \nperformed to assign patients into risk groups. Estimated actuarial Overal Survival \n(OS) rates corresponding to analysis of complete-case and imputed data sets \nwere compared. \nResults: Cases with at least one missing datum experienced a significantly better \nsurvival curve. Estimates derived analysing complete-case data, relative to \nimputed data sets, underestimated the OS rate in all risk groups. In addition \nconfidence intervals were wider indicating loss in precision due to attrition in \nsample size and power. \nConclusion: Results obtained highlighted the danger of exclusion of missing data. \nImputation of missing data avoids biased estimates, increases the precision of \nestimates, and improves genralisability of results to other similar populations.
Authors: Mohammad Reza Baneshi, Abdolrasoul Talei
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2171353342