Effect of telemedicine support for intraoperative anaesthesia care on postoperative outcomes: the TECTONICS randomised clinical trial

Abstract

Background: Telemedicine may help improve care quality and patient outcomes. Telemedicine for intraoperative decision support has not been rigorously studied.

Methods: This was a single-centre randomised clinical trial of unselected adult surgical patients. Patients were randomised to receive usual care or decision support from a telemedicine service, which provided real-time recommendations to intraoperative anaesthesia clinicians based on case reviews and physiological alerts. ORs were randomised 1:1. The co-primary outcomes were 30-day all-cause mortality, respiratory failure, acute kidney injury, and delirium in the intensive care unit, analysed by intention to treat.

Results: Between July 1, 2019, and January 31, 2023, a total of 35,302 patients were randomised to receive telemedicine support, with 36,625 receiving usual care. Telemedicine clinicians provided review in 11,812/35,302 cases, with alerts delivered to 2044/35,302 patients. Telemedicine support had no effect on any of the co-primary outcomes. Within 30 days, 630/35,302 (1.8%) patients randomised to telemedicine died within 30 days, compared with 649/36,625 (1.8%) receiving usual care (relative risk [RR]1.01, 95% confidence interval [CI] 0.87-1.16, P=0.98). Telemedicine support did not alter postoperative respiratory failure [telemedicine 1071/33,996 (3.2%) vs usual care 1130/35,236 (3.2%), RR 0.98, 95% CI 0.88-1.09, P=0.98], acute kidney injury [telemedicine 2316/33 251 (7.0%) vs usual care 2432/34,441 (7.1%); RR 0.99, 95% CI 0.92-1.06, P=0.98], or delirium [telemedicine 1264/3873 (32.6%) vs usual care 1298/4044 (32.1%), RR 1.02, 95% CI 0.94-1.10, P=0.98].

Conclusions: In this large randomised clinical trial, intraoperative telemedicine decision support using real-time alerts and case reviews had no impact on adverse postoperative outcomes.

Clinical trial registration: NCT03923699.

Keywords: acute kidney injury; decision support; delirium; machine learning; postoperative mortality; randomised trial; respiratory failure; telemedicine.

Fig 1

CONSORT flow diagram. ACT hours = 06:15 to 16:00. ‘No flowsheet data’ means no data returned for a specified medical record number. ACT, Anesthesiology Control Tower; AKI, acute kidney injury; CAM-ICU, confusion assessment method for the ICU.

Fig 2

Primary outcome rates over time. Bars indicate pointwise 95% confidence intervals from a linear GEE model using the same clustering as the primary analysis and HC1 adjustment. Data aggregated over 3-month intervals. GEE, generalised estimating equation.