OP53 Teledermatology — Case Study Of An Evidence Synthesis Of A Complex Intervention Using Real-World Data

Introduction Advances in digital technologies are changing how users interact with health care. While technologies such as telehealth can be trialed in a randomized controlled trial (RCT), evidence may come predominantly from real-world data. Synthesizing this type of data is challenging, particularly for complex interventions. We present the case study of an evidence synthesis of teledermatology for triaging referrals from primary care. Methods A systematic review was undertaken of the clinical efficacy, effectiveness, impact on service utilization, and safety of teledermatology-supported triage of primary care referrals, compared with traditional face-to-face dermatology consultations. Searching (through to June 2024), screening and extraction were undertaken as outlined in our protocol (CRD42024608084). All comparative study designs were considered, including chart reviews and single-center experiences. Risk of bias was assessed using the revised Cochrane risk-of-bias tool for randomized trials and the ROBINS-I tool for all other designs. Meta-analysis was undertaken, where appropriate, and an evidence and gap map developed. Results Ninety-one studies were identified: 10 RCTs, two quasi-RCTs, 56 prospective non-RCTs, and 23 retrospective non-RCTs. Study samples ranged between 15 and 106,500 participants, with most completed in a single center. There was significant heterogeneity in study populations, indications, interventions, and outcome definitions and measurement. Tables, graphs, and plots (including forest plots) were used to produce a coherent narrative, assisting the decision-maker in understanding the complex evidence base. Overall findings and strength of evidence were summarized using an evidence and gap map, enabling the decision-maker to consider the balance of outcomes for teledermatology-supported triage, including clinical non-inferiority, and efficiency in management. Conclusions As policymakers must make decisions based on the best evidence available at a point in time, it is important to capture all relevant information, including real-world data. This case study shows that real-world data can be used to inform assessments of complex interventions. Traditional evidence-synthesis presentation methods can be adapted to facilitate the inclusion of these data.