This study seeks to evaluate which reimbursement methodologies (capitation vs. fee-for-service) promoted better care management and health outcomes, as well as more cost-effective resource utilization in the care of non-institutionalized, acutely ill, diabetic patients enrolled in 20 community health centers participating in the Public Private Partnership (PPP) program in Los Angeles County, California.
Participating community health centers were randomly assigned to capitated intervention group (n = 10) or traditional FFS payment control group (n = 10). The capitated and FFS groups received $470 per patient per quarter and $94 per visit, respectively. Patients were risk stratified and clinic visits were documented over 10 months.
Patient health data was obtained by trained healthcare providers at each clinic through electronic data registries while billing data was acquired from email communication. OSHPD 2006 annual utilization data was available online. Chronic Care Model Assessment (CCMA) scores, which are structural measures of quality, were self-assessed and reported by the clinics using the CCMA tool. Results from provider interviews and patient satisfaction and knowledge surveys were also described in this report.
Select health outcome measures (HbA1c, LDL, and blood pressure) were analyzed using SAS version 9.3 to assess improvements in patient care and health. Evaluation of the equivalence between treatment groups utilized t-test to compare means and z-scores to compare proportions.
Study groups' median CCMA scores did not differ significantly although the intervention clinics reported better CCMA scores in October of 2007 and this difference approached statistical significance (p = 0.0562). There was no evidence of harmful health effects to enrolling patients in capitated group clinics, at least under the conditions prevailing in this pilot study.
Patients in both groups appeared comparably satisfied with different dimensions of their care. However, intervention group patients showed markedly higher level of satisfaction in the time clinics took to return their phone calls and in their ability to schedule medical appointments for their diabetes when they needed one.
In terms of knowledge, patients from both study groups did not show marked difference except that more patients in the intervention group knew what their HbA1c level should be.