Systolic and Diastolic Blood Pressure Variability as Risk Factors for Adverse Events After Coronary Artery Bypass Grafting

Abstract

This study assesses whether blood pressure variability is associated with adverse outcomes after coronary artery bypass grafting surgery.

Although hypertension increases surgical risk, the contribution of chronic outpatient blood pressure variability (BPV) to surgical risk is unclear. Blood pressure variability correlates with altered arterial wall stiffness and atherosclerotic plaque volume and has been associated with subsequent hospitalization, stroke, renal failure, and death.^1,2,3 Patients undergoing coronary artery bypass grafting (CABG) procedures exhibit abnormalities in vascular compliance and endothelial function that predispose them to adverse cardiovascular outcomes.^4,5 We hypothesized that long-term, visit-to-visit BPV before surgery is a previously unrecognized risk factor for adverse outcomes after CABG surgery.

Methods

We reviewed 21 813 blood pressure recordings from 1334 consecutive patients undergoing isolated CABG surgery between July 2012 and July 2017 at 1 institution. Data were prospectively collected and retrospectively analyzed. Isolated CABG procedures, rather than all patients undergoing cardiac surgery, were studied, because risk factors for isolated CABG surgery are well known and because this approach ensured a more homogenous patient population. We defined BPV as the SD in systolic or diastolic blood pressures between patient encounters, requiring at least 10 preoperative blood pressure measurements up to 3 years before the date of surgery, using a methodology that we and others have previously used for nonsurgical outpatients.^1,2,3

The Sanford and University of North Dakota institutional review boards approved this study. The requirement for informed consent was waived.

Outcomes of death, renal failure, stroke, and combined major morbidity and mortality were predefined with Society of Thoracic Surgeons (STS) definitions. Patients’ predefined operative risk for mortality was calculated using the STS Risk Score, and the influence of BPV on outcomes was assessed with and without risk adjustment using the STS Risk Score. The association between preoperative BPV and outcomes after CABG surgery was explored using rank-order analysis and analysis of covariance for testing the equality of slopes.

Results

A total of 405 patients met inclusion criteria. The Table details preoperative patient characteristics. The Figure shows the association between the rank and the value of the BPV as a way of identifying a place where the BPV associations suddenly change by sharply increasing. For systolic SD, this is at an SD rank of approximately 17 and for diastolic SD, approximately 18. This suggests but does not define a cutoff value for the BPV that affects patient outcomes.

Table. Demographics, Outcomes, and Blood Pressure Variability.

Preoperative Factors	No. (%)
Total No.	405 (100)
Sex
Male	309 (76.3)
Female	96 (23.7)
Race/ethnicity
White	382 (94.3)
Other	23 (5.7)
Age, mean (SD) [range]	67.9 (9.7) [30-88]
BMI, mean (SD) [range]	31.0 (5.8) [19.0-52.2]
<25	53 (13.1)
25-29.9	126 (31.1)
30-34.9	131 (32.3)
>35	95 (23.5)
Tobacco use	88 (21.7)
Hypertension	344 (84.9)
Diabetes mellitus	185 (45.7)
End-stage renal disease	6 (1.5)
Operative mortality	9 (2.2)
Outcomes
Permanent stroke	4 (1.0)
Renal failure	8 (2.0)
Major morbidity and operative mortality, mean (SD) [range]	23 (5.7)
Per STS score
Operative mortality	2.0 (2.6) [0.2-26.0]
Permanent stroke	1.0 (0.7) [0.2-4.4]
Renal failure	4.3 (5.1) [0.4-42.6]
Morbidity and mortality	15.2 (10.5) [4.1-83.4]
Blood pressure variability, mean (SD) [range]
Blood pressure readings, No.	23.8 (15.6) [10-87]
Systolic SD	13.5 (4.6) [3.6-32.2]
Diastolic SD	8.0 (2.5) [2.3-24.8]
Systolic change	12.2 (3.9) [3.6-30.5]
Diastolic change	7.3 (2.3) [1.9-18.2]

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); STS, Society of Thoracic Surgeons.

Figure. Blood Pressure Variability.

A, Slope of blue line: systolic SD = 8.351 + 0.459 × rank; slope of orange line: systolic SD = −26.83 + 2.410 × rank; P = .01; difference in slopes: F_1,19 = 67.81; P < .001. B, Slope of blue line: diastolic SD = 4.200 + 0.362 × rank; slope of orange line: diastolic SD = −13.35 + 1.313 × rank; P = .01; difference in slopes: F_1,19 = 51.79; P < .001. Dots indicate individual measurements; dotted blue lines indicate the point at which morbidity and mortality risks increase.

Greater diastolic and systolic BPV correlated with increased event rates of combined major morbidity and mortality. After risk adjustment for the STS Risk Score, diastolic BPV remained independently associated with combined major morbidity and mortality (odds ratio [OR], 1.15 [95% CI, 1.01-1.32]; P = .04). Risk-adjusted systolic BPV was associated with renal failure after CABG surgery (OR, 1.14 [95% CI, 1.02-1.28]; P = .02), and a nonsignificant value was obtained for the association of systolic BPV and combined morbidity and mortality (OR, 1.07 [95% CI, 0.99-1.166]; P = .09).

Discussion

Preoperative diastolic BPV was found to be an independent risk factor for predefined combined morbidity and mortality before and after risk adjustment. Additionally, the significant change in the slope of the rank-order analysis indicates that higher levels of both systolic and diastolic BPV are significantly more likely to be associated with adverse outcomes, suggesting a threshold level of variability that may be important (Figure). Systolic BPV was also associated with renal failure after CABG surgery, which is common and associated with significant long-term adverse sequelae. Larger data sets will be needed to confirm the association of systolic variability with patient outcomes and more precisely define the threshold or proportionality of BPV that puts patients at risk.

The mechanism by which BPV affects outcomes is unclear. Long-term BPV increases atherosclerotic plaque vulnerability.¹ While this study addressed only patients undergoing CABG surgery, patients with BPV and coronary artery disease undergoing other procedures might also be at increased risk for perioperative complications. In addition, BPV may be a marker for medication noncompliance,⁶ which may affect surgical outcomes, as nonadherence to secondary prevention is known to adversely affect outcomes. Future studies requiring larger patient populations may address the role of this and other potential confounders. Diastolic and systolic BPV are calculated variables easily flagged within an electronic medical record. Recognition of high BPV may permit additional risk stratification that is clinically useful and easily quantified for patients undergoing CABG procedures.