Internal Medicine Resident University Of Saskatchewan Saskatoon, Saskatchewan, Canada
Background: In Saskatoon and across Canada, injection drug use-associated infective endocarditis (DUA-IE) admissions are rising. At the Royal University Hospital in Saskatoon, a retrospective cohort of patients admitted with IE utilized ICU 38.7% of the time. We sought to determine if ICU utilization changed at our center after implementation of our Multi-disciplinary clinical ENDOcarditis pathway (MENDO) in September 2023.
METHODS AND RESULTS: This pre-intervention/post-intervention quality improvement cohort study compared outcomes for both a retrospective and a prospective (MENDO) cohort of all IE patients (both DUA-IE and non-DUA-IE) admitted to RUH over an 18-month period. Admitted patients (≥17 yrs. old) admitted to RUH diagnosed with IE by either Cardiology or Infectious Diseases specialists could be referred. The team consisted of a clinical physician lead, care coordinator, peer support worker, addictions specialist, and multiple specialists/surgeons. The pathway promoted early treatment of substance use withdrawal, standardization of care, unbiased specialist consultation, and psychosocial support. The MENDO team aimed to increase local cardiac intervention rate by 25% and improve in-hospital survival rates compared to the retrospective cohort. ICU utilization was also compared given that at face value, increasing surgical intervention rate was predicted to increase ICU utilization.
During the prospective period, 104 patients with IE were admitted to our local hospital. Ten patients ultimately did not have IE leaving 94 patients available for analysis. Our retrospective (n=75; 1-Jan-2022 to 31-May-2023) and MENDO (n=94; 25-Sep-2023 to 30-Nov-2024, cutoff date for interim analysis) cohorts were compared. The in-hospital mortality after MENDO implementation was 24.3% compared to 25.1% in the retrospective cohort (p=ns). Cardiovascular intervention rate increased by 58% (12 vs. 19) after the intervention, and this was significant in our IDU population (p=0.04). Correspondingly, overall post-operative cardiovascular ICU utilization by patient days increased by 146% (104 days vs. 46 days (p < 0.0002)). Despite this, overall ICU utilization measured by patient days in the ICU was 261 in the retrospective cohort compared to 247 in the MENDO cohort (two-tailed p-value, 0.03).
Conclusion: Although the MENDO pathway is improving surgical access and increasing post-surgical ICU bed utilization, this has translated into an overall reduction in ICU use and re-allocation of ICU bed resources. This demonstrates more appropriate use of ICU resources using proactive standardization of DUA-IE care through multi-disciplinary discussion and intensive psychosocial support at our local hospital.
