||Zhang, J., Breitner, D., Dewitt, B., Hanmer, J., Saqib, M., Li, D., Edwards, N., Peipert, J., Novak, M., Mucsi, I.
||Background: PROPr is a preference-based health state summary score within the Patient-Reported-Outcomes Measurement Information System (PROMIS). It could potentially be used in cost-effectiveness analyses. We assessed the validity of PROPr among patients with end stage kidney disease (ESKD) using EQ5D5L and SF6D as “legacy” instruments.
Methods: A cross-sectional sample of adults on dialysis and KTR (kidney transplant recipients) completed questionnaires including PROMIS57 (7 domains: anxiety, depression, fatigue, physical-function, sleep-disturbance, pain-interference and ability to participate in social-roles), Patient Health Questionnaire-9, Edmonton Symptom Assessment Scale-revised, Kidney Disease Quality of Life-36 (KDQOL36), and EQ5D5L. SF6D was generated from the SF12 (part of KDQOL36). PROPr is estimated from PROMIS57 domain-scores. PROPr score ranges from -0.022 (all-worst state) to 1.0 (full health). Known-group comparisons were evaluated using age and sex-stratified median scores and calculating “clinical condition impacts”, that is the coefficient for a health condition when summary score was regressed on age, gender, and a single health condition. Convergent validity was assessed with Spearman’s correlation.
Results: Mean ([Standard deviation] SD) age of the 318 participants was 57 (17) years, 57% were male and 51% Caucasian. Median ([Interquartile range] IQR) scores were 0.38 (0.22-0.61), 0.71 (0.58-0.86) and 0.85 (0.67-0.91) for PROPr, SF6D and EQ5D5L, respectively. PROPr and SF6D scores were less subject to ceiling-effects compared to EQ5D5L. The age and sex adjusted condition impact was larger for PROPr for all conditions tested compared to the other two scores. Condition impact for PROPr was: kidney transplant recipients (KTR) vs. dialysis (-0.21, P<0.001), low vs. high comorbidity (-0.10, P<0.001), and low vs. high depression (-0.31, P<0.001). Strong correlations were observed between PROPr and EQ5D5L (rho=0.76) and SF6D (rho=0.84).
Conclusions: These results support the validity of PROPr among patients with ESKD. Moreover, PROPr may be more sensitive to differences in health states compared to other preference-based measures.
Keywords: PROMIS, PROPr, End Stage Kidney Disease, Patient-reported outcomes, Health utility, SF6D, EQ5D5L, Validation study