Study inclusion and exclusion criteria.

Background
Virtual consultations are being increasingly incorporated into routine primary care, as they offer better time and geographical flexibility for patients while also being cost-effective for both patients and service providers. At the same time, concerns have been raised about the extent to which virtual care is safe for patients. As of now, there is no validated methodology for evaluating the safety nuances and implications of virtual care. This study aims to identify patient safety indicators that could be used to evaluate the safety of virtual consultations in primary care.
Methods
A literature search will be performed in Ovid MEDLINE/PubMed, Embase, and Cochrane Library for relevant articles published over the last 10 years (2014–2024). The systematic review will include randomized and non-randomized controlled trials and observational studies with adult populations that compare synchronous patient-provider virtual consultations (telephone or video) or multicomponent interventions involving synchronous remote consultations with face-to-face consultations. The outcome of interest will be patient safety indicators extracted from the studies. The quality of randomized controlled trials will be assessed with the Cochrane Risk of Bias Tool, and the Newcastle-Ottawa Scale will be used to analyze risk of bias in observational studies.
Discussion
Considering the growing adoption of virtual medical care worldwide, a robust and comprehensive evaluation of its safety and quality is now a system-wide priority. Therefore, one of the primary strengths of this proposed systematic review is its focus on a topic of great importance and timeliness, specifically addressing the existing knowledge gap in this area. By publishing this protocol, we demonstrate the transparency and reliability of our research strategy and aim to minimize the risk of selection bias. Potential limitations include the heterogeneity of measures and outcomes, as well as a lower-than-expected number of studies in subgroup analyses, which may negatively influence the statistical significance in data synthesis.
Trial registration
PROSPERO registration number:CRD42023464878.