The association between video or telephone telemedicine visit type and orders in primary care

Abstract

Introduction: Telemedicine is increasingly relied upon for care delivery in primary care, but the impact of visit type on clinical ordering behavior is uncertain.

Methods: Within Kaiser Permanente Northern California, we identified patients who self-scheduled and completed telemedicine encounters with their personal primary care provider or another available primary care provider in the same medical group, between April 1st, 2020, and October 31st, 2020, while physical distancing restrictions for COVID-19 were in place. We collected patient sociodemographic and clinical characteristics, measures of technology access, and categorized the most common primary encounter diagnoses. We measured proportions of patient-scheduled video versus telephone visits for each of eight diagnosis groups (Skin & Soft Tissue, Musculoskeletal Pain, Back Pain, General Gastrointestinal, Hypertension & Diabetes, Mental Health, Upper Respiratory, and Abdominal Pain), and compared physician orders for medications, antibiotics, lab and imaging studies by visit type within each diagnosis group.

Results: There were 273,301 included encounters, with 86,676 (41.5%) video visits and 122,051 (58.5%) telephone visits. Of the diagnosis groups, Skin & Soft Tissue conditions had the highest proportion of video visits (59.7%), while Mental Health conditions had the highest proportion of telephone visits (71.1%). After adjusting for covariates, the overall rates of medication orders (46.6% vs. 44.5%), imaging orders (17.3% vs. 14.9%), lab orders (19.5% vs. 17.2%), and antibiotic orders (7.5% vs. 5.2%) were higher during video visits as compared to telephone visits (p < 0.05). The largest difference within diagnosis groups was for Skin & Soft Tissue conditions, where the rate of medication orders was 9.1% higher than during video visits than telephone visits (45.5% vs. 36.5%, p < 0.05).

Conclusions: We observed statistically significant differences in clinician orders by visit type during telemedicine encounters for common primary care conditions. Our findings suggest that, for certain conditions, visual information conveyed during video visits may promote clinical work-up and treatment.

Keywords: Clinical orders; Primary care; Telemedicine; Visit type.

Fig. 1

Patient cohort and study design

Fig. 2

Proportions of telephone and video visits, by diagnosis group. Notes: Percentages shown are within group proportions. Abbreviations: HTN, Hypertension; DM, Diabetes; GI, Gastrointestinal; MSK, Musculoskeletal

Fig. 3

Predicted probabilities of clinician orders for common primary care diagnoses, by visit type. Notes: Percentages shown are predicted probabilities, adjusted for all patient and encounter covariates. Error bars show 95% confidence intervals, and asterisks (*) represent two-sided p < 0.05. Abbreviations: HTN, Hypertension; DM, Diabetes; GI, Gastrointestinal; MSK, Musculoskeletal