Towards a Fundamental Unit of Continuity of Care

<h3>Context:</h3> Continuity of care between a patient and their primary care practitioner (PCP) is a cornerstone of primary care (PC) and is associated with many positive health outcomes. Many proposed methods for measuring continuity utilize an aggregation of patient visits. <h3>Objective:</h3> To refine measurement of visit-level continuity in primary care by modeling the probability that a patient visit is with their PCP and the associations with patient and PCP characteristics. <h3>Study Design and Analysis:</h3> Retrospective cohort study of patients in 2018-19 Virginia All Payer Claims Database (APCD). Patients were attributed to the PCPs with whom they had the most visits in 2018. Clinician types included family medicine (FM), internal medicine (IM), pediatric medicine (PED), general practice (GP), obstetrics/gynecology (OB), nurse practitioner (NP), and physician assistant (PA). Independent variables, calculated using 2018 data, included patient age, gender, Charlson Comorbidity Index, any inpatient care, number of visits, payer, geographic social deprivation, rurality, and PCP type, workload, and panel size. We modeled the odds of 2019 patient visits being with their PCP using a binomial generalized linear mixed model with a random PCP intercept. <h3>Dataset:</h3> Virginia APCD <h3>Population Studied:</h3> Virginia patients <h3>Outcome Measure:</h3> Probability of visit being with PCP <h3>Results:</h3> Our sample included 1,251,551 patients attributed to 5,651 PCPs. These patients had 54% of 2019 visits with their PCP. The regression results revealed significant associations with all independent variables. Visits for patients who were younger (18-29 OR 0.64; vs 65-74 reference), female (0.78; vs male), lived in urban areas (isolated OR 1.06; vs urban), had any inpatient care (0.87; vs none), or who were attributed to NP (0.44; vs FM)/PA (0.27)/OB (0.67), PCPs with larger panels, or PCPs with fewer clinical days had significantly lower odds of being with their PCP. Conclusions: Significant variation in visit-level patient continuity underscores the importance of strategies tailored to specific populations and clinician characteristics if greater continuity is to be achieved. Patients are increasingly challenged to find appointments and achieve real continuity with their PCPs. Understanding the factors behind discontinuity at the visit level is critical to shaping meaningful payment, measurement, and system-level reforms capable of helping PCPs achieve higher continuity of care with their patients.