A Prospective International Multicentre Cohort Study of Intraoperative Heart Rate and Systolic Blood Pressure and Myocardial Injury After Noncardiac Surgery: Results of the VISION Study

Abstract

Background: The association between intraoperative cardiovascular changes and perioperative myocardial injury has chiefly focused on hypotension during noncardiac surgery. However, the relative influence of blood pressure and heart rate (HR) remains unclear. We investigated both individual and codependent relationships among intraoperative HR, systolic blood pressure (SBP), and myocardial injury after noncardiac surgery (MINS).

Methods: Secondary analysis of the Vascular Events in Noncardiac Surgery Cohort Evaluation (VISION) study, a prospective international cohort study of noncardiac surgical patients. Multivariable logistic regression analysis tested for associations between intraoperative HR and/or SBP and MINS, defined by an elevated serum troponin T adjudicated as due to an ischemic etiology, within 30 days after surgery. Predefined thresholds for intraoperative HR and SBP were: maximum HR >100 beats or minimum HR <55 beats per minute (bpm); maximum SBP >160 mm Hg or minimum SBP <100 mm Hg. Secondary outcomes were myocardial infarction and mortality within 30 days after surgery.

Results: After excluding missing data, 1197 of 15,109 patients (7.9%) sustained MINS, 454 of 16,031 (2.8%) sustained myocardial infarction, and 315 of 16,061 patients (2.0%) died within 30 days after surgery. Maximum intraoperative HR >100 bpm was associated with MINS (odds ratio [OR], 1.27 [1.07-1.50]; P < .01), myocardial infarction (OR, 1.34 [1.05-1.70]; P = .02), and mortality (OR, 2.65 [2.06-3.41]; P < .01). Minimum SBP <100 mm Hg was associated with MINS (OR, 1.21 [1.05-1.39]; P = .01) and mortality (OR, 1.81 [1.39-2.37]; P < .01), but not myocardial infarction (OR, 1.21 [0.98-1.49]; P = .07). Maximum SBP >160 mm Hg was associated with MINS (OR, 1.16 [1.01-1.34]; P = .04) and myocardial infarction (OR, 1.34 [1.09-1.64]; P = .01) but, paradoxically, reduced mortality (OR, 0.76 [0.58-0.99]; P = .04). Minimum HR <55 bpm was associated with reduced MINS (OR, 0.70 [0.59-0.82]; P < .01), myocardial infarction (OR, 0.75 [0.58-0.97]; P = .03), and mortality (OR, 0.58 [0.41-0.81]; P < .01). Minimum SBP <100 mm Hg with maximum HR >100 bpm was more strongly associated with MINS (OR, 1.42 [1.15-1.76]; P < .01) compared with minimum SBP <100 mm Hg alone (OR, 1.20 [1.03-1.40]; P = .02).

Conclusions: Intraoperative tachycardia and hypotension are associated with MINS. Further interventional research targeting HR/blood pressure is needed to define the optimum strategy to reduce MINS.

Figure 1.

Patient flow diagram showing the number of cases included and excluded from each analysis.

Figure 2.

Forest plot summarizing multivariable logistic regression models for highest and lowest intraoperative heart rate (HR) and systolic blood pressure (SBP). Dependent variables are myocardial injury after noncardiac surgery (MINS), myocardial infarction, and mortality within 30 d after surgery. Highest intraoperative HR was dichotomized according to a threshold of >100 beats per minute (bpm) with HR ≤100 bpm as the reference category. Lowest intraoperative HR was dichotomized according to the threshold of <55 bpm with HR ≥55 bpm as the reference category. Highest intraoperative SBP was dichotomized according to a threshold of >160 mm Hg with SBP ≤160 mm Hg as the reference category. Lowest intraoperative SBP was dichotomized according to the threshold of <100 mm Hg with SBP ≥100 mm Hg as the reference category. The x-axis shows odds ratios and the error bars show 95% confidence intervals. Full multivariable models are presented in Supplemental Digital Content, Tables 1–4, http://links.lww.com/AA/C70.

Figure 3.

Forest plot summarizing multivariable logistic regression models for the duration of high/low intraoperative heart rate (HR) and systolic blood pressure (SBP). The dependent variable was myocardial injury after noncardiac surgery (MINS) within 30 d after surgery. There were 4 separate regression models for duration of intraoperative HR >100 beats per minute (bpm), intraoperative HR <55 bpm intraoperative SBP >160 mm Hg, and intraoperative SBP <100 mm Hg. For each model, duration was stratified into 4 approximately equal quartiles. The reference categories were patients with “normal” HR or SBP, for example in the analysis of duration of HR >100 bpm, the reference group was patients with HR ≤100 bpm. The x-axis shows odds ratios and the error bars show 95% confidence intervals. The full multivariable regression models are presented in Supplemental Digital Content, Tables 3, 4, 7, and 8, http://links.lww.com/AA/C70.

Figure 4.

Forest plot summarizing multivariable logistic regression models for combinations of highest/lowest intraoperative systolic blood pressure (SBP) and heart rate (HR). The dependent variable was myocardial injury after noncardiac surgery (MINS) within 30 d after surgery. The sample was categorized according to highest intraoperative SBP >160 mm Hg, lowest intraoperative SBP <100 mm Hg, highest intraoperative HR >100 beats per minute (bpm), and lowest intraoperative HR <55 bpm. For highest SBP and HR, the reference group was SBP ≤160 and HR ≤100; for lowest SBP and highest HR, the reference group was SBP ≥100 and HR ≤100; for highest SBP and lowest HR, the reference group was SBP ≤160 and HR ≥55; and for lowest SBP and lowest HR, the reference group was SBP ≥100 and HR ≥55. The x-axis shows odds ratios and the error bars show 95% confidence intervals. The results presented are summaries of adjusted analyses (as per the primary analysis). Full multivariable models are presented in Supplemental Digital Content, Tables 9–12, http://links.lww.com/AA/C70.