Impact of Longitudinal Virtual Primary Care on Diabetes Quality of Care

Background: Lack of healthcare access to due to physician shortages is a significant driver of telemedicine expansion in rural areas. Telemedicine is effective for management of chronic conditions such as diabetes but its effectiveness in primary care settings is unknown.

Objective: To evaluate differences in diabetes care before and after implementation of a longitudinal virtual primary care program.

Design: Propensity score-matched cohort study utilizing difference-in-differences analysis.

Participants: Patients with diabetes who received care at VA primary care clinics between January 2018 and December 2019 where the Virtual Integrated Multisite Patient Aligned Care Teams (V-IMPACT) program was implemented.

Exposure: Patient participation in at least one V-IMPACT visit while usual care patients did not participate in V-IMPACT.

Main measures: The primary outcome was change in hemoglobin A1C (HbA1C) and secondary outcomes included change in the proportion of patients meeting diabetes quality indicators: blood pressure control, statin use, angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (ACEi/ARB) use, and annual microalbuminuria testing.

Key results: Our propensity-matched cohort included 9010 patients split evenly between those who participated in V-IMPACT and those who remained in usual in-person care. Among individuals with diabetes who participated in V-IMPACT, the change in mean HbA1C was - 0.055% (95% CI - 0.088 to - 0.022%) while those in usual care had a - 0.047% (95% CI - 0.080 to - 0.014%) change before and after program implementation. We observed a 5.1% (95% CI 2.4 to 7.7%) absolute increase in the proportion prescribed statins in the V-IMPACT group, a 5.3% (95% CI 2.5 to 8.2%) increase prescribed ACE/ARBs, and a 4.6% (95% 1.7 to 7.5%) increase in completed yearly microalbuminuria testing. V-IMPACT was not associated with a significant difference in the proportion with controlled blood pressure at < 140/90 or < 130/90 mmHg thresholds.

Conclusions: Quality of diabetes care delivered by a longitudinal virtual primary care model was similar if not better than traditional in-person care.