Transfer of Patients With Acute Renal Conditions From Rural Emergency Departments After Telenephrology Implementation

Rural hospitals frequently transfer patients because on-site nephrology coverage is limited. We evaluated whether implementation of telemedicine nephrology was associated with changes in nephrology transfer patterns across three rural hospital EDs within a large non-profit health system.

We conducted a multicenter, retrospective interrupted time-series study using monthly operational counts abstracted from quality-assurance dashboards (January 2018–July 2024). Tele-nephrology was implemented at one site in August 2019 and at two sites in October 2022. The primary outcome was the monthly number of all nephrology transfers. We estimated immediate level and post-implementation slope changes at site-specific go-live dates using segmented Negative Binomial regression with calendar-month fixed effects and site effects; robust standard errors accounted for serial correlation. As a complementary summary, we compared the 12 months pre-deployment with the 12 months post-deployment for each site and overall, reporting incidence rate ratios (IRRs) with 95% confidence intervals (CIs).

We evaluated 652 total nephrology transfer encounters across the prespecified 12-month windows (36 site-months pre and 36 site-months post). Aggregated across sites, total transfers numbered 344 pre-deployment and 308 post-deployment, corresponding to 9.6 vs 8.6 transfers per site-month and an overall IRR 0.90 (95% CI 0.77–1.04). By site (total transfers), counts changed from 37 to 12 at the earliest adopter (IRR 0.32; 95% CI 0.17–0.62), 180 to 182 at one later adopter (IRR 1.01; 95% CI 0.82–1.24), and 127 to 114 at the second later adopter (IRR 0.90; 95% CI 0.70–1.16). Negative Binomial segmented models produced inferences consistent with these pre/post summaries, indicating a large immediate reduction at the earliest adopter and no clear first-year change at the later adopters.

Implementation of telemedicine nephrology was associated with a marked decrease in total transfers at the earliest adopter and no detectable change at the two later adopters during their first post-deployment year. Aggregated across sites, total transfers were ~10% lower after deployment, reflecting a trend toward fewer transfers that did not reach statistical significance.