Measuring the quality of end-of-life care

Abstract

Context: Although there is a documented need to improve end-of-life care, there are few validated and brief questionnaires that are available as outcome measures for use in improving that care.

Objectives: To examine the measurement characteristics of the Quality of End-of-Life Care (QEOLC) questionnaire.

Methods: In a multisite, cross-sectional study of a mailed questionnaire, patients with life-limiting illnesses, their families, and nurses completed the QEOLC questionnaire. Patients and nurses were identified by physicians, and families were identified by participating patients. Physicians included general internists, oncologists, cardiologists, and pulmonologists from the Southeast and Pacific Northwest of the United States.

Results: Eight hundred one patients, 310 of their families, and 885 nurses were identified by 85 physicians. Using structural equation modeling techniques corrected for clustering under physicians, we identified a patient-specific factor based on 11 items, a family-specific factor based on 22 items, a nurse-specific factor based on 11 items, and a common single-factor solution based on 10 items. Construct validity was supported by significant associations in the hypothesized direction between the identified QEOLC factors and each of the following: physician palliative care knowledge, patients' and families' ratings of overall quality of care, and patients' levels of symptom distress.

Conclusion: Although continued testing in heterogeneous samples is necessary, the current study supported the construct validity of the QEOLC questionnaire to assess physician skill at end-of-life care, thereby providing valid measures of quality end-of-life care. Furthermore, this approach is a model for development and validation of patient- and family-centered assessments of quality of care.