A Quantitative Evaluation of the Minnesota Long-Stay Resident Quality of Life Survey

Abstract

Purpose

The objectives were to 1) systematically evaluate the Minnesota Long-Stay Resident Quality of Life (QoL) Survey based on validity, reliability, parsimony, relevance, and ability to discriminate facility performance; and 2) arrive at a new version of the survey and composite scoring approach.

Methods

Data consisted of Minnesota nursing home resident QoL surveys, conducted through annual face-to-face interviews in 2017 (10,007 residents, 355 facilities), 2018 (9,884 residents, 352 facilities), and 2019 (9,896 residents, 347 facilities). Validity was evaluated using exploratory and confirmatory factor analyses. Reliability was assessed by Cronbach’s alpha values. Parsimony and relevance were assessed using content validity, construct validity, correlation, frequency of endorsement, and percentage missing. The ability to discriminate facility performance was assessed by examining the distributions of facility QoL scores.

Results

The current domain structure has unbalanced items ranging from 4 to 9 across the eight domains; 28 items fit as well or better empirically under an alternative domain structure; and four items are redundant and could be dropped from the survey without loss of information. The current facility QoL scores do not discriminate well in facility performance because of the lack of item balance and a ceiling effect.

Conclusion

The proposed revisions result in a shorter, more balanced, more discriminating, and more valid QoL survey, while maintaining a high level of reliability. The revised survey allows the Minnesota Department of Human Services, and others who might adopt the survey, to better assess nursing facility performance on aspects of QoL that are meaningful to residents.