Literature DB >> 33116899

Handling Missing Values in Interrupted Time Series Analysis of Longitudinal Individual-Level Data.

Juan Carlos Bazo-Alvarez1,2, Tim P Morris3, Tra My Pham3, James R Carpenter3,4, Irene Petersen1,5.   

Abstract

BACKGROUND: In the interrupted time series (ITS) approach, it is common to average the outcome of interest at each time point and then perform a segmented regression (SR) analysis. In this study, we illustrate that such 'aggregate-level' analysis is biased when data are missing at random (MAR) and provide alternative analysis methods.
METHODS: Using electronic health records from the UK, we evaluated weight change over time induced by the initiation of antipsychotic treatment. We contrasted estimates from aggregate-level SR analysis against estimates from mixed models with and without multiple imputation of missing covariates, using individual-level data. Then, we conducted a simulation study for insight about the different results in a controlled environment.
RESULTS: Aggregate-level SR analysis suggested a substantial weight gain after initiation of treatment (average short-term weight change: 0.799kg/week) compared to mixed models (0.412kg/week). Simulation studies confirmed that aggregate-level SR analysis was biased when data were MAR. In simulations, mixed models gave less biased estimates than SR analysis and, in combination with multilevel multiple imputation, provided unbiased estimates. Mixed models with multiple imputation can be used with other types of ITS outcomes (eg, proportions). Other standard methods applied in ITS do not help to correct this bias problem.
CONCLUSION: Aggregate-level SR analysis can bias the ITS estimates when individual-level data are MAR, because taking averages of individual-level data before SR means that data at the cluster level are missing not at random. Avoiding the averaging-step and using mixed models with or without multilevel multiple imputation of covariates is recommended.
© 2020 Bazo-Alvarez et al.

Entities:  

Keywords:  big data; electronic health records; interrupted time series analysis; missing data; mixed effects models; multiple imputation; segmented regression

Year:  2020        PMID: 33116899      PMCID: PMC7549500          DOI: 10.2147/CLEP.S266428

Source DB:  PubMed          Journal:  Clin Epidemiol        ISSN: 1179-1349            Impact factor:   4.790


  23 in total

Review 1.  Interrupted time series analysis in drug utilization research is increasing: systematic review and recommendations.

Authors:  Racquel Jandoc; Andrea M Burden; Muhammad Mamdani; Linda E Lévesque; Suzanne M Cadarette
Journal:  J Clin Epidemiol       Date:  2015-03-11       Impact factor: 6.437

2.  A methodological framework for model selection in interrupted time series studies.

Authors:  J Lopez Bernal; S Soumerai; A Gasparrini
Journal:  J Clin Epidemiol       Date:  2018-06-06       Impact factor: 6.437

3.  Identifying periods of acceptable computer usage in primary care research databases.

Authors:  Laura Horsfall; Kate Walters; Irene Petersen
Journal:  Pharmacoepidemiol Drug Saf       Date:  2012-11-04       Impact factor: 2.890

4.  Influence of a strict glucose protocol on serum potassium and glucose concentrations and their association with mortality in intensive care patients.

Authors:  Esther V Uijtendaal; Jeannette E F Zwart-van Rijkom; Dylan W de Lange; Arief Lalmohamed; Wouter W van Solinge; Toine C G Egberts
Journal:  Crit Care       Date:  2015-06-22       Impact factor: 9.097

5.  Effects of drug price reduction and prescribing restrictions on expenditures and utilisation of antihypertensive drugs in Korea.

Authors:  Ki-Bong Yoo; Sang Gyu Lee; Sohee Park; Tae Hyun Kim; Jeonghoon Ahn; Mee-Hyun Cho; Eun-Cheol Park
Journal:  BMJ Open       Date:  2015-07-15       Impact factor: 2.692

6.  Supermarket policies on less-healthy food at checkouts: Natural experimental evaluation using interrupted time series analyses of purchases.

Authors:  Katrine T Ejlerskov; Stephen J Sharp; Martine Stead; Ashley J Adamson; Martin White; Jean Adams
Journal:  PLoS Med       Date:  2018-12-18       Impact factor: 11.069

7.  Methodology and reporting characteristics of studies using interrupted time series design in healthcare.

Authors:  Jemma Hudson; Shona Fielding; Craig R Ramsay
Journal:  BMC Med Res Methodol       Date:  2019-07-04       Impact factor: 4.615

8.  Effects of long-term antipsychotics treatment on body weight: A population-based cohort study.

Authors:  Juan Carlos Bazo-Alvarez; Tim P Morris; James R Carpenter; Joseph F Hayes; Irene Petersen
Journal:  J Psychopharmacol       Date:  2019-11-14       Impact factor: 4.153

9.  The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement.

Authors:  Eric I Benchimol; Liam Smeeth; Astrid Guttmann; Katie Harron; David Moher; Irene Petersen; Henrik T Sørensen; Erik von Elm; Sinéad M Langan
Journal:  PLoS Med       Date:  2015-10-06       Impact factor: 11.069

10.  Significant reduction of antibiotic consumption and patients' costs after an action plan in China, 2010-2014.

Authors:  Lidao Bao; Rui Peng; Yi Wang; Ruilian Ma; Xianhua Ren; Wenbin Meng; Fusheng Sun; Junxia Fang; Ping Chen; Yang Wang; Qiuhong Chen; Jian Cai; Jian Jin; Jinhui Guo; Shucheng Yang; Xiaomei Mo; Enjing Zhang; Yuerong Zhang; Zhaoxin Lu; Binbin Chen; Xiujuan Yue; Meijun Zhu; Yingying Wang; Xinchao Li; Yuan Bian; Shaoshan Kong; Wenfei Pan; Qian Ding; Jun Cao; Rupin Liu; Nan Chen; Xuelian Huang; Agula B; Haijun Lyu
Journal:  PLoS One       Date:  2015-03-13       Impact factor: 3.240

View more
  1 in total

Review 1.  Current Practices in Missing Data Handling for Interrupted Time Series Studies Performed on Individual-Level Data: A Scoping Review in Health Research.

Authors:  Juan Carlos Bazo-Alvarez; Tim P Morris; James R Carpenter; Irene Petersen
Journal:  Clin Epidemiol       Date:  2021-07-23       Impact factor: 4.790

  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.