Home > Research > Publications & Outputs > Missing not at random in end of life care studies

Text available via DOI:

View graph of relations

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

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

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

In: BMC Medical Research Methodology, Vol. 21, No. 1, 13, 09.01.2021.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

APA

Vancouver

Author

Bibtex

@article{9ea75c03c76440548421b74b4d436f1b,
title = "Missing not at random in end of life care studies: multiple imputation and sensitivity analysis on data from the ACTION study",
abstract = "BACKGROUND: 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. 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.METHODS: Multiple 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.RESULTS: Up 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.CONCLUSIONS: The findings confirmed the importance of undertaking this kind of analysis in end-of-life care studies.",
keywords = "Missing data, MAR, MNAR, Advance care planning, Oncology, Quality of life",
author = "{ACTION Consortium} and Giulia Carreras and Guido Miccinesi and Andrew Wilcock and Nancy Preston and Daan Nieboer and Luc Deliens and Mogensm Groenvold and Urska Lunder and {van der Heide}, Agnes and Michela Baccini",
year = "2021",
month = jan,
day = "9",
doi = "10.1186/s12874-020-01180-y",
language = "English",
volume = "21",
journal = "BMC Medical Research Methodology",
issn = "1471-2288",
publisher = "BIOMED CENTRAL LTD",
number = "1",

}

RIS

TY - JOUR

T1 - Missing not at random in end of life care studies

T2 - multiple imputation and sensitivity analysis on data from the ACTION study

AU - ACTION Consortium

AU - Carreras, Giulia

AU - Miccinesi, Guido

AU - Wilcock, Andrew

AU - Preston, Nancy

AU - Nieboer, Daan

AU - Deliens, Luc

AU - Groenvold, Mogensm

AU - Lunder, Urska

AU - van der Heide, Agnes

AU - Baccini, Michela

PY - 2021/1/9

Y1 - 2021/1/9

N2 - BACKGROUND: 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. 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.METHODS: Multiple 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.RESULTS: Up 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.CONCLUSIONS: The findings confirmed the importance of undertaking this kind of analysis in end-of-life care studies.

AB - BACKGROUND: 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. 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.METHODS: Multiple 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.RESULTS: Up 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.CONCLUSIONS: The findings confirmed the importance of undertaking this kind of analysis in end-of-life care studies.

KW - Missing data

KW - MAR

KW - MNAR

KW - Advance care planning

KW - Oncology

KW - Quality of life

U2 - 10.1186/s12874-020-01180-y

DO - 10.1186/s12874-020-01180-y

M3 - Journal article

C2 - 33422019

VL - 21

JO - BMC Medical Research Methodology

JF - BMC Medical Research Methodology

SN - 1471-2288

IS - 1

M1 - 13

ER -