Telemedicine during COVID-19: who gets left behind?

OBJECTIVES: COVID-19 resulted in rapid implementation of telemedicine. It is critical to explore patient factors that drive telemedicine use to recognize any disparities in healthcare access that may emerge with remote healthcare. Our objective was to compare the characteristics and subsequent health care utilization of gynecologic oncology patients who had a scheduled telemedicine visit from 3/23/20-5/18/20. METHODS: We conducted a retrospective study of women whose scheduled in-person office visits were converted to telemedicine. Video visits were encouraged and telephone visits were the default option. We compared patient factors between women who cancelled versus those who completed their telemedicine visits. Subsequent health care utilization included in-person office appointments, emergency department (ED) or operative care was compared between those who cancelled versus completed telemedicine visits. We also evaluated patient differences in completion of video vs telephone visits. Cancellation rates were compared to a historical control of in-person gynecologic oncology office visits (3/23- 5/18/19). Multivariate logistic regression analyses were performed to assess odds of completing a telehealth appointment. RESULTS: Of the 185 women with scheduled telemedicine visits, 158 (85%) completed and 27 (15%) cancelled. Telemedicine had significantly higher cancellations compared to office visits a year ago, 15% vs. 3% (p<0.001). Women who completed telemedicine visits tended to be older than those who cancelled (p=0.008). No other patient characteristics differed between those who completed versus cancelled telemedicine visits, and no differences were observed in subsequent health care utilization between groups. A total of 30% of patients had not activated their electronic patient portals prior to the study period. 124 (75%) of women had telephone visits. Being employed and a new, rather than established patient, was associated with video visits, p=0.009 and 0.003, respectively. In univariate analyses, older patients (OR 0.96; 95% CI 0.93, 0.99) and Medicare insurance (OR 0.21, 95% CI 0.05, 0.90) were less likely to cancel a visit. No patient characteristics were associated with visit cancellation in multivariate analyses. [Figure: see text] CONCLUSIONS: Telemedicine implementation faces ongoing challenges. High cancellation rates and patient's reluctance to use video or activate their electronic patient portals may limit a durable telemedicine platform. No patient characteristics were associated with the successful completion of telemedicine visits, and completion of telemedicine visits was not associated with subsequent healthcare utilization. More in-depth studies of patient engagement with telemedicine are needed to optimize healthcare delivery beyond the COVID-19 pandemic.