Objectives: To review the evidence on harms associated with patient-clinician real-time encounters using video-based telehealth and determine the effectiveness of any related patient safety practices (PSPs). PSPs are interventions, strategies, or approaches intended to prevent or mitigate unintended consequences of healthcare delivery and improve patient safety. This review provides information that clinicians and health system leaders need to determine how to minimize harms from increasing real-time use of telehealth.
Methods: We followed rapid review processes of the Agency for Healthcare Research and Quality Evidence-based Practice Center Program. We searched PubMed, EMBASE, and Cochrane to identify eligible studies published from 2012 to 2022, supplemented by a search for unpublished evaluations and white papers. Outcomes of interest included: adverse events (any harm to patients due to medical care), other specified harms (i.e., preventable hospitalizations, inappropriate treatment, missed or delayed diagnoses, duplication of services, privacy breaches), and implementation factors for any PSPs.
Findings: Our search retrieved 7,155 citations, of which 23 studies (including 6 randomized controlled trials [RCTs]) were eligible for review. Fourteen studies reported on adverse events or unintended effects of telehealth; these studies were conducted in diverse settings, with four studies in behavioral health, two each in rehabilitation, transplant, and Parkinson’s care, and one each in postoperative, termination of pregnancy, community health, and hospital-at-home settings. Adverse events such as death, reoperation, infection, or major complications were infrequent in both telehealth and usual care groups, making it difficult to find statistically significant differences. One RCT found telehealth resulted in fewer medication errors than standard care. Thirteen studies examined preventable hospitalizations or emergency department (ED) visits and reported mixed findings; six of these studies were in postoperative care and two were in urological care. Of the 6 RCTs, 3 showed no difference in risk of hospitalization or ED visits for telehealth compared to usual care, and 3 showed reduced risk for patients receiving telehealth. We found no studies on the effectiveness of PSPs in reducing harms associated with real-time telehealth.
Conclusions: Studies have evaluated the frequency and severity of harms associated with real-time video-based telehealth encounters between clinicians and patients, examining a variety of patient safety measures. Telehealth was not inferior to usual care in terms of hospitalizations or ED visits. No studies evaluated a specific PSP. More research is needed to improve understanding of harms associated with real-time use of telehealth and how to prevent or mitigate those harms.