Longitudinal construct validity of the minimum data set health status index

Aaron Jones; David Feeny; Andrew P Costa

doi:10.1186/s12955-018-0932-9

Longitudinal construct validity of the minimum data set health status index

Health Qual Life Outcomes. 2018 May 24;16(1):102. doi: 10.1186/s12955-018-0932-9.

Authors

Aaron Jones¹, David Feeny², Andrew P Costa^{3

4}

Affiliations

¹ Department of Health Research Methods, Evidence and Impact, McMaster University, 1280 Main St. W, Hamilton, ON, L8S 4K1, Canada.
² Department of Economics, McMaster University, 1280 Main St. W, Hamilton, ON, L8S 4M4, Canada.
³ Department of Health Research Methods, Evidence and Impact, McMaster University, 1280 Main St. W, Hamilton, ON, L8S 4K1, Canada. [email protected].
⁴ Department of Medicine, McMaster University, 1280 Main St. W, Hamilton, L8S 4K1, ON, Canada. [email protected].

Abstract

Background: The Minimum Data Set Health Status Index (MDS-HSI) is a generic, preference-based health-related quality of life (HRQOL) measure derived by mapping items from the Resident Assessment Instrument - Minimum Data Set (RAI-MDS) assessment onto the Health Utilities Index Mark 2 classification system. While the validity of the MDS-HSI has been examined in cross-sectional settings, the longitudinal validity has not been explored. The objective of this study was to investigate the longitudinal construct validity of the MDS-HSI in a home care population.

Methods: This study utilized a retrospective cohort of home care patients in the Hamilton-Niagara-Haldimand-Brant health region of Ontario, Canada with at least two RAI-MDS Home Care assessments between January 2010 and December 2014. Convergent validity was assessed by calculating Spearman rank correlations between the change in MDS-HSI and changes in six validated indices of health domains that can be calculated from the RAI-MDS assessment. Known-groups validity was investigated by fitting multivariable linear regression models to estimate the mean change in MDS-HSI associated with clinically important changes in the six health domain indices and 15 disease symptoms from the RAI-MDS Home Care assessment, controlling for age and sex.

Results: The cohort contained 25,182 patients with two RAI-MDS Home Care assessments. Spearman correlations between the MDS-HSI change and changes in the health domain indices were all statistically significant and in the hypothesized small to moderate range [0.1 < ρ < 0.5]. Clinically important changes in all of the health domain indices and 13 of the 15 disease symptoms were significantly associated with clinically important changes in the MDS-HSI.

Conclusions: The findings of this study support the longitudinal construct validity of the MDS-HSI in home care populations. In addition to evaluating changes in HRQOL among home care patients in clinical research, economic evaluation, and health technology assessment, the MDS-HSI may be used in system-level applications using routinely collected population-level data.

Keywords: Community care; Health-related quality of life; Home care; Longitudinal construct validity.

Publication types

Validation Study

MeSH terms

Accidental Falls / statistics & numerical data
Activities of Daily Living
Aged
Aged, 80 and over
Caregivers / psychology
Caregivers / statistics & numerical data
Cognitive Dysfunction / epidemiology
Cross-Sectional Studies
Female
Geriatric Assessment / methods
Health Status Indicators*
Home Care Services / statistics & numerical data*
Humans
Longitudinal Studies
Male
Ontario / epidemiology
Quality of Life*
Retrospective Studies

Grants and funding

TG2015-15P/Canadian Frailty Network