Abstract
Background/Aims We evaluated changes in care processes and outcomes associated with various care-redesign initiatives toward a Patient Centered Medical Home (PCMH) at an outpatient care delivery system. Care delivery practices and the extent of the implementation of PCMH components varied widely across clinics and physicians at the organization, thereby providing a natural experiment of PCMH evaluation.
Methods We used longitudinal data from Electronic Health Records (EHR) of active primary care patients (2008–2011). The group’s patients represent diverse insurance (59% PPO, 26% HMO, 15% other insurance) and demographics. Data were aggregated into primary care physician (PCP) level (n = 256), each 6 months. We examined five indicators of patient-centeredness: 1) care continuity (%PCP’s own patients seen), 2) timely appointment (#days to 3rd next appointment), 3) access to Personal Health Record (PHR) (#enrolled patients in a PCP panel), 4) workflow efficiency (%abnormal lab test results processed within a day), and 5) patient satisfaction (%very satisfied with physician service). Clinical quality indicators were 1) monitoring chronic conditions (monitoring of HgbA1c/LDL/blood pressure/kidney function for patients with diabetes; monitoring of ACE-inhibitors/ARB/diuretics adherence), 2) preventive screening (breast/cervical/colon cancer), and 3) diabetes control (HgbA1c/ LDL/blood pressure control of patients with diabetes). Physician fixed-effects and relevant patient case-mix and sociodemographic factors were included in multivariate models.
Results As PCPs saw their own patients more often, their patients were significantly more likely to get cancer screening but were not more likely to get monitoring of chronic conditions or to get diabetes controlled. PHR enrollment was a positive predictor for monitoring of chronic conditions and preventive screening, but not for diabetes control. Workflow efficiency was a positive predictor for some preventive screening and diabetes control measures. Neither patient satisfaction nor timely appointment had statistically significant relationship with clinical quality indicators.
Conclusions Indicators of continuity of care, access to PHR, and workflow efficiency are associated with selected, mostly process-oriented, measures of clinical quality. In contrast, patient satisfaction and timely appointment, which are well-accepted indicators of patient-centeredness, are not associated with clinical quality. Construct of PCMH is multifaceted, and no single patient-centeredness characteristic can address all the diverse needs of primary care patients.
