Abstract
BACKGROUND: Patients with chronic diseases should meet with their primary care doctor regularly to facilitate proactive care. Little is known about what factors are associated with more regular follow-up. METHODS: We studied 70,095 patients age 40 + with one of three chronic conditions (diabetes mellitus, heart failure, chronic obstructive pulmonary disease), cared for by Leumit Health Services, an Israeli health maintenance organization. Patients were divided into the quintile with the least temporally regular care (i.e., the most irregular intervals between visits) vs. the other four quintiles. We examined patient-level predictors of being in the least-temporally-regular quintile. We calculated the risk-adjusted regularity of care at 239 LHS clinics with at least 30 patients. For each clinic, compared the number of patients with the least temporally regular care with the number predicted to be in this group based on patient characteristics. RESULTS: Compared to older patients, younger patients (age 40-49), were more likely to be in the least-temporally-regular group. For example, age 70-79 had an adjusted odds ratio (AOR) of 0.82 compared to age 40-49 (p < 0.001 for all findings discussed here). Males were more likely to be in the least-regular group (AOR 1.18). Patients with previous myocardial infarction (AOR 1.07), atrial fibrillation (AOR 1.08), and current smokers (AOR 1.12) were more likely to have an irregular pattern of care. In contrast, patients with diabetes (AOR 0.79) or osteoporosis (AOR 0.86) were less likely to have an irregular pattern of care. Clinic-level number of patients with irregular care, compared with the predicted number, ranged from 0.36 (fewer patients with temporally irregular care) to 1.71 (more patients). CONCLUSIONS: Some patient characteristics are associated with more or less temporally regular patterns of primary care visits. Clinics vary widely on the number of patients with a temporally irregular pattern of care, after adjusting for patient characteristics. Health systems can use the patient-level model to identify patients at high risk for temporally irregular patterns of primary care. The next step is to examine which strategies are employed by clinics that achieve the most temporally regular care, since these strategies may be possible to emulate elsewhere.