Robotic process automation versus manual screening to identify patients at risk for perioperative myocardial infarction/Injury: a prospective, single-blinded, paired reader-controlled trial

Abstract

Abstract Background Perioperative myocardial infarction/injury (PMI) can be missed entirely without scheduling pre- and postoperative troponin levels. Currently, manual screening (MS) to identify high-risk patients before surgery is resource-intensive and not implemented in most hospitals despite guideline recommendations. Purpose Robotic Process Automation-Screening (RPAS) is a promising intervention to reduce workload and improve screening accuracy. However, its efficacy in a relevant clinical setting remains uncertain. Methods In this prospective, single-blinded, paired reader-controlled trial, patients undergoing surgery at our hospital in Switzerland were consecutively enrolled. MS and RPAS were carried out simultaneously to identify patients at risk for PMI according to predefined screening criteria (Figure 1). Healthcare professionals performing MS were blinded to RPAS results. Discrepant identification was reviewed by an independent clinician blinded to the origin of the identification (MS or RPAS), generating a gold standard list of paired reader-controlled patients to investigate the primary diagnostic endpoint: relative true positive fraction (rTPF). Results Overall, 656 patients were included in the study, of whom 76 were at risk for PMI according to the gold standard list. RPAS was superior for patients at risk identification compared with MS (74 [97.4%] vs 62 [81.6%] true positive identifications, rTPF 1.19 [95% confidence interval: 1.08–1.32], P=0.004). The number needed to screen (NNS) among patients at risk to identify one additional true positive using RPAS was 6.34. RPAS and MS showed a sensitivity of 0.97 (0.91-0.99) vs 0.82 (0.71-0.89), and specificity of 0.98 (0.97-0.99) vs 1.0 (0.99-1.00), respectively (Figure 2). Conclusions RPAS was superior to standard-of-care MS in identifying patients at risk for PMI. Our findings support the implementation of RPAS in the study setting, with further evaluation needed to assess its effectiveness in clinical practice.Figure 1 Figure 2

Ähnliche Arbeiten