Health-Related Quality of Life in Patients With a Left Ventricular Assist Device (QOLVAD) Questionnaire: Initial Psychometrics of a New Instrument.
Susan M. Joseph M.D.
Sandau, K.E., Lee, C.S., Faulkner, K.M., Pozehl, B., Eckman, P., Garberich, R., Weaver, C.E., Joseph, S.M., Hall, S., Carey, S.A., Chaudhry, S.P., Schroeder, S.E., Hoffman, R.O., 3rd, Feldman, D., Birati, E.Y., Soni, M., Marble, J.F., Jurgens, C.Y., Hoglund, B. and Cowger, J.A. (2020). “Health-Related Quality of Life in Patients With a Left Ventricular Assist Device (QOLVAD) Questionnaire: Initial Psychometrics of a New Instrument.” J Cardiovasc Nurs Dec 9. [Epub ahead of print].
BACKGROUND: Patients with a left ventricular assist device are a unique and growing population who deserve their own valid, reliable instrument for health-related quality of life. OBJECTIVE: We developed and tested the Health-Related Quality of Life with a Left Ventricular Assist Device (QOLVAD) questionnaire. METHODS: In a prospective, descriptive study, patients from 7 sites completed the QOLVAD and comparator questionnaires. Construct validity was tested using confirmatory factor analysis. Convergent validity was tested using correlations of QOLVAD scores to well-established measures of subjective health status, depression, anxiety, and meaning/faith. Reliability and test-retest reliability were quantified. RESULTS: Patients (n = 213) were 58.7 ± 13.9 years old; 81.0% were male, 73.7% were White, and 48.0% had bridge to transplant. Questionnaires were completed at a median time of 44 weeks post ventricular assist device. The 5 QOLVAD domains had acceptable construct validity (root mean square error of approximation = 0.064, comparative and Tucker-Lewis fit indices > 0.90, weighted root mean square residual = 0.95). The total score and domain-specific scores were significantly correlated with the instruments to which they were compared. Internal consistency reliability was acceptable for all subscales (α = .79-.83) except the cognitive domain (α = .66). Unidimensional reliability for the total score was acceptable (α = .93), as was factor determinacy for multidimensional reliability (0.95). Total test-retest reliability was 0.875 (P < .001). CONCLUSION: Our analysis provided initial support for validity and reliability of the QOLVAD for total score, physical, emotional, social, and meaning/spiritual domains. The QOLVAD has potential in research and clinical settings to guide decision making and referrals; further studies are needed.