UCLA

Background

The impact of telemedicine on ambulatory care quality is a key question for policymakers as they navigate payment reform for remote care.

Objective

To evaluate whether utilizing telemedicine in the first 9 months of the COVID-19 pandemic impacted performance on a diabetes quality of care measure for patients at a large academic medical center. We hypothesized care quality would reduce less among telemedicine users.

Design

Quasi-experimental design using binomial logistic regression. Covariates included age, gender, race, ethnicity, type of insurance, hierarchical condition category score, primary language at the individual level, and zip code-level income.

Participants

All adult patients younger than 75 years of age diagnosed with type 2 diabetes mellitus (N = 16,588) as of 3/19/2020 at a single academic health center.

Interventions

Completion of one or more telemedicine encounters with an institutional primary care physician or endocrinologist between 3/19/2020 and 12/19/2020.

Main measures

The components met in a five-item composite measure of diabetes quality of care, as of patients' last clinical encounter. Items were (1) systolic blood pressure less than 140 mmHg, (2) hemoglobin A1c less than 8.0%, (3) using a statin and (4) aspirin, and (5) tobacco non-use.

Key results

From the pre- to post-period, the probability of meeting any given component of the composite measure for patients only utilizing in-person care was 21% lower (OR, 95% CI 0.79; 0.76, 0.81) and for the telemedicine users 2% lower (OR 0.98; 0.85, 1.13). There was an increased likelihood of meeting any given component among telemedicine users compared to in-person care alone (OR 1.25; 1.08, 1.44).

Conclusions

Patients with diabetes utilizing telemedicine performed similarly on a composite measure of diabetes care quality compared to before the pandemic. Those not utilizing telemedicine had reductions. Telemedicine use maintained quality of care for patients with diabetes during the first 9 months of the COVID-19 pandemic.