Modeling health-related quality of life in people recovering from stroke.

backgroundThe Wilson-Cleary (W-C) model of health-related quality of life (HRQL) has not been tested in stroke, and a better understanding of the components of HRQL during recovery would lead to a more integrated and person-centered approach to health management and outcome optimization for this vulnerable population.

objectiveTo enhance our understanding for how QOL emerges from the sequelae of stroke during the recovery period, the aim was to empirically test a biopsychosocial conceptual model of HRQL for people recovering from stroke.

methodsWe present a multi-site longitudinal study of an inception cohort of 678 persons recruited at stroke onset and studied at key intervals over the first post-stroke year. As the most pronounced recovery after stroke occurs in the first 3 months, this time frame was chosen as the focus of this analysis. The measures for this study were chosen for their relevance to key constructs of stroke impact and for their optimal psychometric properties. Multiple measures for each of the W-C rubrics were available from instruments such as the Stroke Impact Scale, RAND-36, HUI, and EQ-5D, among others. A structural equation model (SEM) was fit using MPlus. To minimize potential bias arising from the missing data, multiple imputation was performed on the longitudinal data using SAS proc MI.

resultsOf the 678 subjects who entered the cohort, 618 were interviewed at 1 month post-stroke and 533 at 3 months (486 and 454 had data at 6 and 12 months, respectively). A 3-month model with paths from biological factors to symptoms and symptoms to function fits well (CFI:0.966, RMSEA:0.044), though one model with paths from function to health perception did not (CFI:0.934, RMSEA:0.058). Allowing additional paths across non-adjacent rubrics improved fit considerably (CFI:0.962, RMSEA:0.044). A final model included emotional well-being under the symptom rubric (CFI:0.955, RMSEA:0.047). Including social support as an environmental factor had little impact on the model. Total variance in health perception explained was 76.3 %.

conclusionThese results emphasize that to optimize overall HRQL during the crucial first 3 months of recovery, interventions need to continue to focus on comorbid health conditions and on reducing stroke impairments. A function-only focus too soon in the recovery process may not produce the desired impact to optimize HRQL.