Impact of an In-Hospital Endocarditis Team and a State-Wide Endocarditis Network on Perioperative Outcomes

Background: Infective endocarditis (IE) requires multidisciplinary management. We established an endocarditis team within our hospital in 2011 and a state-wide endocarditis network with referring hospitals in 2015. We aimed to investigate their impact on perioperative outcomes. Methods: We retrospectively analyzed data from patients operated on for IE in our center between 01/2007 and 03/2018. To investigate the impact of the endocarditis network on referral latency and pre-operative complications we divided patients into two eras: before (n = 409) and after (n = 221) 01/2015. To investigate the impact of the endocarditis team on post-operative outcomes we conducted multivariate binary logistic regression analyses for the whole population. Kaplan–Meier estimates of 5-year survival were reported. Results: In the second era, after establishing the endocarditis network, the median time from symptoms to referral was halved (7 days (interquartile range: 2–19) vs. 15 days (interquartile range: 6–35)), and pre-operative endocarditis-related complications were reduced, i.e., stroke (14% vs. 27%, p < 0.001), heart failure (45% vs. 69%, p < 0.001), cardiac abscesses (24% vs. 34%, p = 0.018), and acute requirement of hemodialysis (8% vs. 14%, p = 0.026). In both eras, a lack of recommendations from the endocarditis team was an independent predictor for in-hospital mortality (adjusted odds ratio: 2.12, 95% CI: 1.27–3.53, p = 0.004) and post-operative stroke (adjusted odds ratio: 2.23, 95% CI: 1.12–4.39, p = 0.02), and was associated with worse 5-year survival (59% vs. 40%, log-rank < 0.001). Conclusion: The establishment of an endocarditis network led to the earlier referral of patients with fewer pre-operative endocarditis-related complications. Adhering to endocarditis team recommendations was an independent predictor for lower post-operative stroke and in-hospital mortality, and was associated with better 5-year survival.