Missing not at random in end of life care studies: multiple imputation and sensitivity analysis on data from the ACTION study

Giulia Carreras,Action Consortium ,Urska Lunder,Agnes Van Der Heide,Daan Nieboer,Guido Miccinesi,Nancy Preston,Mogensm Groenvold,Luc Deliens,Michela Baccini,Andrew Wilcock

doi:10.1186/s12874-020-01180-y

Abstract

BackgroundMissing data are common in end-of-life care studies, but there is still relatively little exploration of which is the best method to deal with them, and, in particular, if the missing at random (MAR) assumption is valid or missing not at random (MNAR) mechanisms should be assumed. In this paper we investigated this issue through a sensitivity analysis within the ACTION study, a multicenter cluster randomized controlled trial testing advance care planning in patients with advanced lung or colorectal cancer.MethodsMultiple imputation procedures under MAR and MNAR assumptions were implemented. Possible violation of the MAR assumption was addressed with reference to variables measuring quality of life and symptoms. The MNAR model assumed that patients with worse health were more likely to have missing questionnaires, making a distinction between single missing items, which were assumed to satisfy the MAR assumption, and missing values due to completely missing questionnaire for which a MNAR mechanism was hypothesized. We explored the sensitivity to possible departures from MAR on gender differences between key indicators and on simple correlations.ResultsUp to 39% of follow-up data were missing. Results under MAR reflected that missingness was related to poorer health status. Correlations between variables, although very small, changed according to the imputation method, as well as the differences in scores by gender, indicating a certain sensitivity of the results to the violation of the MAR assumption.ConclusionsThe findings confirmed the importance of undertaking this kind of analysis in end-of-life care studies.

Highlights

Missing data are common in end-of-life care studies, but there is still relatively little exploration of which is the best method to deal with them, and, in particular, if the missing at random (MAR) assumption is valid or missing not at random (MNAR) mechanisms should be assumed
One of these is the Multivariate Imputation by Chained Equations (MICE) which relies on the MAR assumption [9, 10], but can be modified in order to account for MNAR mechanisms [13]
Partition of the missing values and modified MICE in the ACTION study In our analysis, we explored the violation of the MAR assumption, making a distinction between missing items due to the fact that the patient did not reply to some items in a questionnaire and missing items due to the fact that the form was completely missing, for which a violation of the MAR assumption could be hypothesised

Summary

Introduction

Missing data are common in end-of-life care studies, but there is still relatively little exploration of which is the best method to deal with them, and, in particular, if the missing at random (MAR) assumption is valid or missing not at random (MNAR) mechanisms should be assumed. When the MCAR assumption is not valid, alternative strategies can be adopted to deal with missing data: inverse probability weighting, doubly robust inverse probability weighting, maximum likelihood estimation, multiple imputation (MI). There are several ways to implement MI that could be run under MAR and MNAR [9,10,11,12] One of these is the Multivariate Imputation by Chained Equations (MICE) which relies on the MAR assumption [9, 10], but can be modified in order to account for MNAR mechanisms [13]

Objectives

Methods

Results

Discussion

Conclusion