Background: Background: Patients transferred between hospitals undergo a high-risk transition of care, where communication is asynchronous, information is commonly lost, and mortality is disproportionately high. Prior studies have shown that standardizing communication through checklists can improve outcomes, and conversely lost documentation has been associated with higher mortality. The lack of interoperability of electronic health records (EHR) is a critical barrier to high quality hand-off in this high-risk transition of car. Improving the flow of information such as regional participation of health information exchanges (HIE) has the potential to improve the cost and outcomes of this vulnerable population, but this remains to be tested in practice. Other methods of interoperability, such as Epic Care-Everywhere, require informed consent to be signed prior to accessing information, which is challenging when a patient is at another facility and possibly medically unstable. The impact of interoperability has yet to be tested in prospective interventional trials.
Methods: We conducted a prospective randomized stepped wedge interventional trial to leverage the high regional prevalence of epic care everywhere of non-emergent transfers to a representative tertiary care center. All inpatient services including all surgical subspecialties and critical care units were randomized into one of 4 blocks which determined order of implementation. A multi-disciplinary intervention including (1) Epic Care-everywhere informed consent was signed (2) images were requested to be electronically “pushed” (3) an encounter was generated, and (4) time was allowed to review records prior to verbal hand off and structured with a note. The intervention was staggered in 3 month periods to each of the four blocks. The primary outcomes was inpatient mortality after transfer, secondary measures included length of stay, return transfer, escalation of care after transfer, and provider satisfaction. Differences-in-differences (DiD) analysis was used comparing outcomes against a 1 year prospectively collected baseline using differences in differences analysis.
Results: Following intervention we compared the outcomes of 1110 patients following intervention against 2221 controls. 92% of transfers had information made available prior to hand off. We observed no significant improvement in inpatient mortality (p = 0.53), nor length of stay (p = 0.58). However, we observed a significant reduction in rate of escalation of care within 24 hours (1.8% to 1.2%, DiD coef, p=0.012). There were a higher rate of transfers directly to the ICU following intervention (p=0.001). Before and after the intervention providers remained positive regarding the potential for HIE to improve inter-hospital transfers; however, there was a decline in the belief that access was practical during routine care particularly among services that lacked 24-hour in-house call.
Conclusions: In this single center interventional study, we find that integrating EHR interoperability into the hand off process did not improve mortality or length of stay, but did reduce early escalation of care. Many providers found a more involved transfer process disruptive to normal workflow and patient care. We conclude that interoperability efforts are a necessary step toward improved communication, however not sufficient in the absence of dedicated staff allowing sufficient time to review records and images during the transfer process.