A Preference-Weighted Scoring Algorithm for Assessing Quality of Care for Patients With Advanced Illness in...

Most survey instruments assessing end-of-life care value different care dimensions equally instead of reflecting respondent trade-offs. We developed a preference-weighted instrument for assessing the Quality of Care for Patients with Advanced Illness among bereaved caregivers (QCPAI 2.0 bereaved caregiver) and fielded it in a multiethnic Asian population.

An online panel of 300 bereaved informal caregivers in Singapore completed QCPAI and a discrete choice experiment using a balanced-incomplete-block design. Preference weights, estimated using a 2-class latent-class logit with a "garbage" class, were rescaled to compute QCPAI preference-weighted scores (0-100, worst-best) with A to F letter grades. Socioeconomic differences were assessed using ordered probit for item scores and linear regression for preference-weighted scores, whereas reporting heterogeneity was examined via a hierarchical ordered probit model anchored on quality vignettes.

Preferences exhibited diminishing marginal utility in quality. The most important items were appropriate medical care (patients not over- or undertreated), alignment with patient goals, and management of physical symptoms. Mean QCPAI item scores were lowest for affordability, waiting times, and place of death. Although the mean QCPAI preference-weighted score was 83.2 out of 100 (grade B), low socioeconomic status was associated with significantly lower quality across 12 of 17 items and a 7.9-point lower preference-weighted score on average.

QCPAI provides a validated preference-weighted scoring algorithm suitable for routine benchmarking and evaluation of end-of-life care. Results identify items of relatively low quality but high importance as targets for intervention and highlight that affordability and socioeconomic gaps persist despite generous subsidies available to many patients.