Estimating Change in Health-Related Quality of Life before and after Stroke: Challenges and Possible Solutions

Nicolas R. Thompson, Brittany R. Lapin and Irene L. Katzan
Additional contact information
Nicolas R. Thompson: Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA
Brittany R. Lapin: Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA
Irene L. Katzan: Center for Outcomes Research and Evaluation, Neurological Institute, Cleveland Clinic, Cleveland, OH, USA

Medical Decision Making, 2024, vol. 44, issue 8, 961-973

Abstract: Background Estimating change in health-related quality of life (HRQOL) from pre- to poststroke is challenging because HRQOL is rarely collected prior to stroke. Leveraging HRQOL data collected both before and after stroke, we sought to estimate the change in HRQOL from prestroke to early poststroke. Methods Stroke survivors completed the Patient-Reported Outcomes Measurement Information System Global Health (PROMIS-GH) scale at both pre- and early poststroke. Patient characteristics were compared for those who did and did not complete the PROMIS-GH. The mean change in PROMIS-GH T-score was estimated using complete case analysis, multiple imputation, and multiple imputation with delta adjustment. Results A total of 4,473 stroke survivors were included (mean age 63.1 ± 14.1 y, 47.5% female, 82.6% ischemic stroke). A total of 993 (22.2%) patients completed the PROMIS-GH at prestroke while 2,298 (51.4%) completed it early poststroke. Compared with those without PROMIS-GH, patients with PROMIS-GH prestroke had worse comorbidity burden. Patients who completed PROMIS-GH early poststroke had better early poststroke clinician-rated function and shorter hospital length of stay. Complete case analysis and multiple imputation revealed patients’ PROMIS-GH T-scores worsened by 2 to 3 points. Multiple imputation with delta adjustment revealed patients’ PROMIS-GH T-scores worsened by 4 to 10 points, depending on delta values chosen. Conclusions Systematic differences in patients who completed the PROMIS-GH at both pre- and early poststroke suggest that missing PROMIS-GH scores may be missing not at random (MNAR). Multiple imputation with delta adjustment, which is better suited for MNAR data, may be a preferable method for analysis of change in HRQOL from pre- to poststroke. Given our study’s large proportion of missing HRQOL data, future studies with less missing HRQOL data are necessary to verify our results. Highlights Estimating the change in health-related quality of life from pre- to poststroke is challenging because health-related quality-of-life data are rarely collected prior to stroke. Previously used methods to assess the burden of stroke on health-related quality of life suffer from recall bias and selection bias. Using health-related quality-of-life data collected both before and after stroke, we sought to estimate the change in health-related quality of life after stroke using statistical methods that account for missing data. Comparisons of patients who did and did not complete health-related quality-of-life scales at both pre- and poststroke suggested that missing data may be missing not at random. Statistical methods that account for data that are missing not at random revealed more worsening in health-related quality of life after stroke than traditional methods such as complete case analysis or multiple imputation.

Keywords: stroke; health-related quality of life; PROMIS Global Health; multiple imputation (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journals.sagepub.com/doi/10.1177/0272989X241285038 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:sae:medema:v:44:y:2024:i:8:p:961-973

DOI: 10.1177/0272989X241285038

Access Statistics for this article

More articles in Medical Decision Making
Bibliographic data for series maintained by SAGE Publications ().