RELATIONSHIP OF PROVIDER VOLUME AND QUALITY OF LIPID SCREENING IN ONTARIO, CANADA: INSIGHT FROM CANHEART

Abstract

In-hospital myocardial infarction and heart failure patient volumes are associated with improved cardiovascular (CV) outcomes. Whether outpatient primary care physician (PCP) volumes are predictive of CV preventive care is unknown. In the multicenter, big data CANHEART observational study of the population of Ontario, Canada, patients 40-74 years without CV disease were evaluated for lipid screening between 2008 and 2012 with de-identified data using linked administrative healthcare databases according to their PCP’s outpatient clinical volume of concordant patients. Patient volume was defined as the mean annual number of clinic visits made to the usual care PCP. Modified Poisson regression models were used to derive the relative risk (RR) of lipid screening according to patient volumes (measured continuously and according to quintiles) with adjustment for individual clinical and sociodemographic risk factors and physician characteristics. Various sensitivity analyses were conducted restricting patients aged 50-74 years of age, physician practice location, and physician model of reimbursement. There were 4,753,994 patients seen by 10,307 usual care PCPs during the study period. Overall, 83.8% of patients underwent lipid screening at least once over the 5-year period as recommended by regional guidelines. After multivariable adjustment, there was a stepwise increase in the RR across each quintile of patient volume (Figure), with a 3.3% (95% CI, 3.2-3.5) higher lipid-screening rate for every doubling of patient volumes (P < 0.001). After restricting patients aged 50-74 and excluding physicians whose patients likely used hospital based labs (distance to hospital < 100 m and average number of yearly lab claims < 5), results remained significant. Among physicians remunerated primarily from fee for service, the trend in the relative risk of cholesterol screening across patient volumes was higher compared to physicians supported by alternative models of reimbursement (salary, capitation, etc.). Higher volume PCPs achieve higher performance on this metric of CV preventive care performance than lower volume clinicians. This suggests high-volume PCPs may provide higher, rather than lower, quality care.