Our study is among the largest to assess 2016 CCS guideline uptake at a quaternary care hospital and capture 30-day postoperative cardiac events. B-type natriuretic peptide adherence at MUHC was 52.4%, with lower rates for troponin and ECG. Although we observed an increase in adherence over the study period, this was modest and only statistically significant for BNP. While we observed a small decrease in MINS rates between study years, there were no significant differences in MI, mortality, CHF events, or cardiac arrest. Notably, the increased use of biomarkers did not translate to a higher incidence of MINS, suggesting that expanded biomarker use may not imply increased detection. While physician follow-up increased, data on changes in cardiac prescriptions or imaging requests were not captured and therefore a reduction in MINS cannot be inferred by increased testing or follow-up.

The observed adherence rates are comparable to those of other Canadian centres (Table 5),19,20,21 however, existing variability is likely due to multiple factors, namely: 1) differences in institutional practices and infrastructure at various points along the perioperative trajectory, 2) the availability of biomarker testing (e.g., rural sites), 3) surgical/cardiovascular risk of the patient cohort, and 4) variability in physician decision-making when ordering biomarkers.

Although strong evidence exists regarding the prognostic capability of preoperative BNP, there have been no studies to date in patients with elevated BNP to recommend management that improves perioperative cardiovascular outcomes or impacts on health care costs. This has prompted recent guidelines (AHA 2024 and ESC 2022) to issue more conservative recommendations regarding its use in the preoperative setting to further stratify patients at high risk.10,25 Similarly, while postoperative troponin elevation has clear prognostic implications with respect to increased short- and long-term mortality, optimal MINS management remains uncertain and current recommended therapies (e.g., aspirin and statins) are based largely on limited perioperative observational data and inferences from nonsurgical patients with cardiovascular disease showing reductions in short- and long-term mortality.26,27,28 While one randomized trial showed that dabigatran administered to patients post-MINS reduces rates of major vascular complications, it has rarely been incorporated into practice.29 Several retrospective studies have also demonstrated that patients with MINS who received early cardiology referral experienced significant reductions in 30-day mortality.19,30,31,32 Our study clearly demonstrates that elevated cardiac biomarkers are a trigger for cardiology and internal medicine follow-up and are associated with increased incidence of postoperative MACE. The first step towards improving cardiac outcomes in populations at high risk is developing a system that enables their prompt identification followed by longitudinal surveillance. MINS detection therefore offers an opportunity for risk stratification and medical optimization, which can improve outcomes; nevertheless, careful clinician evaluation of patient risk, benefits, and downstream cost implications should be central in a comprehensive perioperative evaluation. Any incentivization or funding towards creating standardized national/provincial perioperative protocols incorporating biomarkers would need to be preceded by robust randomized data showing improved patient outcomes and cost effectiveness.

Changes in institutional practices spanning the start of the preoperative consult to discharge from the surgical ward have been shown to influence biomarker adherence.20 When critically analyzing our perioperative pipeline, we identified optimization of electronic health systems to automate biomarker orders and identify patients at high risk as most likely to improve adherence. In addition, multidisciplinary team re-education on prognostic implications of elevated biomarkers as well as handling of preprinted order sheets at surgical admission can improve successful guideline implementation. While adherence rates between Canadian sites are similar, the BNP adherence rate at MUHC was markedly higher than in a study conducted in Alberta (52.4% vs 6.8%).19 This discrepancy is likely owing to the study period largely predating the release of the CCS guidelines and including multiple rural sites, most of which did not have access to BNP.

Consistent with prior studies, patients with elevated BNP and/or troponin experienced higher incidences of 30-day MINS, MI, mortality, CHF, and cardiac arrest. Compared with a large 2014 meta-analysis, our cohort had lower reported rates of death or nonfatal MI in patients with elevated BNP (7.8% vs 21.8%).33 This discrepancy is likely explained by individual studies overreporting MI since this outcome was inconsistently defined as well as the inclusion of urgent/emergent surgeries. Regarding postoperative troponins, this study’s reported incidence of 30-day mortality in patients with elevated troponin compared with normal troponin (7.8% vs 0.6%) is similar to the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study (9.1% vs 1.1%), the largest prospective international cohort study in the field.23 Though our cohort’s patient mean age was higher and included a significantly higher proportion of patients undergoing thoracic surgeries than VISION, lack of data on patient preoperative comorbidities limits a direct comparison of baseline risk.

Guideline adherence varied by specialty, with the highest adherence seen in vascular surgery, likely reflecting elevated surgical and patient risk. We would hypothesize the low adherence in gynecological surgery to be from perceived lower risk despite nearly all surgeries carrying an intermediate risk of postoperative MACE and 73% of patients having an RCRI ≥ 1. Lower than expected adherence in thoracic surgery may be owing to video-assisted thoracoscopic surgery being perceived as less invasive despite carrying a high risk of postoperative MACE. Relatively poor adherence in this surgical population (comprising 27.5% of total patients) was likely a significant contributor to the low reported rates of MINS and MI. Overall, patients undergoing higher-risk surgeries (across all subspecialties) were more likely to have BNP/troponin measured, suggesting that inherent surgical risk likely influenced biomarker ordering decisions.

This study has several limitations that should be considered. First, chart review was limited to the MUHC’s electronic medical records which may underestimate the number of 30-day events in patients presenting at external centres. Second, findings from a single quaternary centre may not generalize to other settings, particularly rural sites. Third, long-term outcomes and physician follow-up beyond thirty days were not captured. Fourth, the retrospective study design limited insight into clinician decision-making (e.g., decision to forego BNP testing despite indication as it may not have been felt to change outcomes). Lastly, surveillance bias likely contributed to higher MINS/MI detection in patients with an elevated BNP; nevertheless, this was mitigated by measuring clinical outcomes that did not rely on troponins. Despite its limitations, this study’s strengths include a relatively large surgical cohort with a study period beginning 2 years after guideline release, allowing adequate time for in-hospital implementation. High-risk surgery was classified objectively on the basis of ESC criteria rather than subjective physician assessment. Lastly, alternative causes of troponin elevation were explored, allowing distinction between MINS and type 2 MI.

Regarding future directions, more studies exploring the impact of increased testing (biomarker and/or cardiac imaging) and physician follow-up on long-term outcomes and health care expenditure are needed. Research examining adherence pre- and post-ESC/AHA guideline revisions as well as survey studies documenting physician attitudes towards biomarker prescriptions in risk stratification are also necessary. Randomized trials are required to help fill the gap on optimal MINS management and further justify expanded BNP/troponin surveillance. Lastly, qualitative studies exploring challenges faced by perioperative departments in guideline implementation as well as comparative studies across different Canadian institutions could help identify best practices from high-adherence centres.

Overall, about half of the patients undergoing noncardiac surgery in our cohort underwent preoperative BNP screening as recommended by CCS guidelines. As postoperative cardiac ischemia is common and associated with increased 30-day morbidity and mortality, greater efforts must be deployed to identify and screen patients at high risk who can benefit from physician follow-up and cardiac optimization. In addition, more studies exploring the impact of increased testing and physician follow-up on long-term outcomes and health care expenditure are needed.