1579. Evaluation of MATCH: an Electronic Individual Patient-Focused Management System Aimed at Preventing Cytomegalovirus Disease Following Solid Organ Transplantation

Background Cytomegalovirus (CMV) infection is common among solid-organ transplant (SOT) recipients and may cause CMV disease, if not promptly treated. Strategies to prevent CMV disease include chemoprophylaxis and pre-emptive monitoring and treatment of emerging subclinical infection. To optimize the implementation of these strategies as part of routine care, we developed and implemented a proactive and patient-tailored CMV management system for SOT patients (the MATCH program) in our center. Two key performance characteristics of success of MATCH are diagnosing CMV at low levels and avoiding CMV disease at diagnosis; these characteristics are assessed here before (2007–2010), during (2011–2012) and after (2013–2015) the implementation of the MATCH program. Methods In MATCH, SOT recipients follow a personalized, yet standardized, plan for monitoring, prophylaxis and pre-emptive therapy depending on underlying risk for CMV infection. The plan is composed in accordance with the recipient’s a priori risk as to CMV IgG serostatus and is continually updated during the post-transplant course according to patient′s current situation. Each individual patient plan is produced and implemented by a rule-based artificial intelligence (AI) platform, harvesting relevant real-time data from electronic medical records. Via predefined algorithms, plans and revisions are created and alerts are generated in case of missed planned monitoring for or molecular detection of CMV infection. Prior to its implementation, prevention of CMV disease was left at the discretion of the individual physician. Results A total of 603, 357, and 531 patients received an SOT before, during and after implementing MATCH, resp., of whom 88 (14.6%), 56 (15.7%) and 119 (22.4%) developed CMV infection within the first year of transplantation (Table 1). Among those with CMV infection, the % with high viral load decreased as did the % with CMV disease at the time of diagnosis of CMV infection during and after the implementation of MATCH relative to before (Figure 1). Conclusion The implementation of a rule-based AI platform guiding routine prevention of CMV disease among SOT recipients was associated with improved CMV-specific outcome, indicating its ability to identify the CMV infection sooner after onset and before causing disease. Disclosures All authors: No reported disclosures.