Methods

This observational cohort study was designed for the risk prediction of perioperative and postoperative events after CABG. Data collection and variables definition have been described in detail elsewhere.

Briefly, we included all patients referred to our department for CABG from September 1998 to August 2002 in a prospective longitudinal cohort study.

Clinical recording began approximately 1 month before surgery in case of nonurgent CABG, during the anesthetist outpatient visit, and was continued during hospitalization. Clinical features and patients’ medical histories were recorded as follows: gender, age, and history of coronary artery disease, cardiac surgery, stroke or transient ischemic attack. The following cardiovascular risk factors were noted: smoking, defined by active smoking or discontinuation <2 years before CABG; diabetes mellitus, if fasting glucose was >7.0 mmol/L or in case of ongoing oral antidiabetic and/or insulin therapy; hypertension, when elevated blood pressure was reported in the medical chart and/or in case of ongoing antihypertensive therapy; dyslipidemia, defined either by fasting blood cholesterol level on admission >240 mg/dl, low-density lipoprotein cholesterol level >160 mg/dl, or the use of lipid-lowering medication before surgery. Overweight was defined by a body mass index >30 kg/m ² . Peripheral arterial disease was defined by any history of clinical disease, or by an abnormal ankle brachial index (<0.85 or >1.50). Symptomatic heart failure was diagnosed clinically according to the New York Heart Association classification. HR was measured on admission and during the first postoperative outpatient visit (baseline measurement) on an electrocardiogram recorded in the supine position after a 5-minute rest. In case of sinus rhythm or pacing, HR was measured on 3 cardiac cycles. In patients with arrhythmia, it was measured over a 6-second period, to minimize the risk for HR overestimation to 2 beats/min.

The left ventricular ejection fraction was determined by preoperative ventriculography and considered altered when <0.40. Coronary artery stenosis was considered significant when >70% diameter reduction or >50% for the left main coronary artery.

Ongoing therapies at first postoperative visit were recorded, especially HR-lowering drugs (β blockers or nondihydropyridine calcium channel blockers), as well as statins, angiotensin-converting enzyme inhibitors, and angiotensin antagonists.

For every patient, we collected information on the number of bypasses and anastomoses and the use of an on-pump or off-pump technique. Revascularization was considered complete if all diseased coronary arteries were bypassed. Urgent CABG was defined by symptomatic left main coronary artery stenosis, refractory angina pectoris, or an unstable hemodynamic condition.

All patients were systematically scheduled for a first postoperative visit between the second and third months after surgery. For each patient, electrocardiography was performed for postoperative rest HR determination. Thereafter, annual visits were scheduled. Patients who did not meet their appointments were contacted by telephone. In case of death or patients’ not returning calls, family physicians were contacted for information. Hospital records were screened for patients with suspected cardiovascular events and in case of in-hospital death.

The primary outcome in our study was all-cause mortality. The secondary outcome was composite, combining any of the following events, whichever occurring first: death, secondary coronary revascularization, nonfatal acute coronary syndromes, nonfatal stroke or transient ischemic attack, and vascular surgery.

Continuous variables are reported as mean ± SD and categorical variables as numbers and percentages. All data collected before surgery, including HR, were included in a univariate analysis model. Univariate analysis used Kaplan-Meier survival curves with log-rank tests. Independent predictors of survival were calculated using a Cox regression model, using a backward stepwise procedure. A p value <0.05 was considered to indicate statistical significance. Analyses were performed using StatView version 5.0 (SAS Institute Inc., Cary, North Carolina).

Results

Among the 1,022 patients enrolled, 32 died during the first month after CABG and were thus excluded from this analysis. Thirty-five additional patients were excluded because of missing data regarding preoperative (n = 5) and postoperative (n = 30) HR. Among the remaining 955 patients, 161 underwent concomitant valvular and/or vascular surgery and were thus excluded from statistical analysis. The baseline characteristics of the remaining 794 patients are listed in Table 1 . Mean preoperative and 2-month postoperative ambulatory baseline HRs at rest were 64.2 ± 13.0 and 67.4 ± 12.6 beats/min, respectively. The overall distribution of HR measured during the postoperative visit is displayed in Figure 1 . The preoperative and postoperative HRs were found to be correlated ( Figure 2 ), although important disparities were apparent in many cases.

Overall distribution of postoperative HR at rest in the study cohort and with respect to the presence or the absence of β blockers (BB).

Correlation between preoperative and postoperative HR at rest. adm = admission; fu = follow-up.

During the mean follow-up period of 3.2 ± 1.3 years, 40 patients (5.0%) died, 15 (1.9%) of cardiovascular causes. Acute coronary syndromes occurred in 27 patients (3.4%), and 24 (3.0%) patients underwent redo coronary revascularization. Fourteen patients (1.8%) experienced strokes or transient ischemic attacks, and 34 patients (4.3%) required additional vascular surgery. Altogether, the secondary outcome occurred in 98 patients (12.3%). Patients with postoperative rest HRs ≥90 beats/min were at greater risk for death ( Figure 3 ) and at greater risk for the occurrence of the secondary outcome ( Figure 3 ), starting at 75 beats/min. The association between postoperative HR and mortality is listed in Table 2 . Compared with postoperative HR <60 beats/min, HR >90 beats/min was associated with increased risk for mortality after multiple adjustments. However, when fully adjusted to confounding factors, this association did not remain significant.

Kaplan-Meier survival curves with number of subjects at risk by HR categories, with respect to the primary (A) and secondary composite (B) end points (p values unadjusted).

Stay updated, free articles. Join our Telegram channel

Join