Dissertation Title： "Assessing Patient-Reported Quality in Medical Group Practices: An Analysis of Data from the National Field Test of the Group-Level CAHPS Instrument"

There is growing interest in measuring the quality of healthcare provided by medical group practices. In 1999, the U.S. Agency for Healthcare Research and Quality and the California HealthCare Foundation funded the development and field testing of a version of the Consumer Assessment of Health Plans Study (CAHPS©) questionnaire for use in medical group practices. CAHPS© is presently the national standard for measuring patients' experience with their health plans. This dissertation analyzes data from the national field test of the Group-Level CAHPS (G-CAHPS) instrument.

Paper #1 describes the field test design and assesses the psychometric properties of the G-CAHPS instrument, including its factor structure, scale reliability, inter-unit reliability and validity. Response rates ranged from 45% to 55% across our five study sites. Factor analysis of the items selected for the core instrument suggests five distinct dimensions of patient-reported quality including access to care, patient/doctor communication, PCP/specialist coordination, advice on preventive health, and courtesy and respect shown by the office staff. The field test demonstrated adequate levels of reliability and validity.

Using a series of ANOVA models and variance components analysis, Paper #2 assesses the ability of the instrument to discriminate among groups and evaluates the influence of nested organizational structures on G-CAHPS measures. Organizational units evaluated include: health plans, multi-group entities akin to independent practice associations and physician/hospital organizations, medical groups and individual practice sites. Results indicate that the G-CAHPS instrument has adequate power to discriminate among groups for the majority of measures, but that group-level measurement masks substantial sub-unit variation in performance.

Paper #3 examines the organizational determinants of patient-reported quality, including: 1) organizational structure (i.e., number of practice locations, specialty mix and size); 2) decision-making processes (i.e., extent of formalization and decentralization) and; 3) practice management strategies (i.e., visit planning, non-visit care and extended hours). Using path analysis and hierarchical liner modeling, this analysis suggests that decentralized decision-making has a pervasive and negative influence on patients' experience of care. Other organizational attributes with significant influences on patient-reported quality include visit planning, practices' approach to quality improvement, size, affiliation with a multi-site group and staff turnover.