Variability in Surgeons’ Perioperative Practices May Influence the Incidence of Low-Output Failure After Coronary Artery Bypass Grafting Surgery
Background—Postoperative low-output failure (LOF) is an important contributor to morbidity and mortality after coronary artery bypass grafting surgery. We sought to understand which pre- and intra-operative factors contribute to postoperative LOF and to what degree the surgeon may influence rates of LOF.
Methods and Results—We identified 11 838 patients undergoing nonemergent, isolated coronary artery bypass grafting surgery using cardiopulmonary bypass by 32 surgeons at 8 centers in northern New England from 2001 to 2009. Our cohort included patients with preoperative ejection fractions >40%. Patients with preoperative intraaortic balloon pumps were excluded. LOF was defined as the need for ≥2 inotropes at 48 hours, an intra- or post-operative intraaortic balloon pumps, or return to cardiopulmonary bypass (for hemodynamic reasons). Case volume varied across the 32 surgeons (limits, 80–766; median, 344). The overall rate of LOF was 4.3% (return to cardiopulmonary bypass, 2.6%; intraaortic balloon pumps, 1.0%; inotrope usage, 0.8%; combination, 1.0%). The predicted risk of LOF did not differ across surgeons, P=0.79, and the observed rates varied from 1.1% to 10.2%, P<0.001. Patients operated by low-rate surgeons had shorter clamp and bypass times, antegrade cardioplegia, longer maximum intervals between cardioplegia doses, lower cardioplegia volume per anastomosis or minute of ischemic time, and less hot-shot use. Patients operated on by higher LOF surgeons had higher rates of postoperative acute kidney injury.
Conclusions—Rates of LOF significantly varied across surgeons and could not be explained solely by patient case mix, suggesting that variability in perioperative practices influences risk of LOF.
Coronary artery bypass grafting (CABG) surgery is one of the most commonly performed and intensely studied cardiac surgical procedures in the United States. Nonetheless, patients remain at risk for considerable morbidity, including injuries to the patient’s myocardium. Low-output failure (LOF) is one of the most significant adverse sequelae of isolated CABG surgery. Previous work has revealed that postoperative left ventricular dysfunction is a significant contributor to mortality, especially among patients who present to surgery with preserved left ventricular function.1 Although several traditional preoperative predictors of LOF have been reported, surgical factors may also contribute to postoperative left ventricular dysfunction, including temperature management, separation from cardiopulmonary bypass (CPB), CPB time, ischemic time, postoperative ischemia, ischemic reperfusion injury, or myocardial infarction.
During CPB, the heart is usually arrested with interruption of myocardial perfusion (Figure 1). As a myocardial protection technique, cardioplegia is commonly used to reduce myocardial injury during this period of ischemic arrest. Myocardial protection has been the focus of a large number of animal and clinical research studies; nevertheless, considerable variability exists across surgeons in their myocardial protection technique, including the type, timing, volume, route, and temperature of cardioplegia solution. Although each variant has theoretical benefits, there is no consensus in the literature regarding the optimal strategy for cardiac protection. As a result, a surgeon’s practice is often influenced by his/her training, local practice patterns, and differing interpretations of the literature. It is possible that variations in myocardial protection strategies may result in inadequate myocardial preservation leading to myocardial stunning, ischemic injury, or reperfusion injury, increasing a patient’s risk of developing LOF postoperatively.
WHAT IS KNOWN
Postoperative low-output failure is an important contributor to morbidity and mortality after coronary artery bypass grafting surgery.
Postoperative left ventricular dysfunction is a also an important contributor to mortality after coronary artery bypass grafting, especially among patients with normal preoperative left ventricular function.
Surgical factors can contribute to postoperative left ventricular dysfunction
WHAT THE STUDY ADDS
Predicted risk of low-output failure did not differ among 32 surgeons (P=0.79), in northern New England, whereas the observed rates varied from 1.1% to 10.2%, P<0.001.
Rates of low-output failure are driven not by patient case mix but rather by perioperative surgical practices, including approaches to cardioplegia.
We sought to identify whether any apparent differences in rates of LOF among patients operated on by 32 cardiac surgeons within northern New England were attributed to patient and disease characteristics or aspects of surgical practice. We conducted this study among a contemporary, multicenter series of 11 838 patients undergoing nonemergent, isolated CABG surgery at 8 medical centers in northern New England from 2001 through 2009.
The Northern New England Cardiovascular Disease Study Group is a voluntary research consortium composed of clinicians, research scientists, and hospital administrators, representing all 8 medical centers in Maine, Vermont, and New Hampshire where cardiac surgery is performed. Since 1987, the Northern New England Cardiovascular Disease Study Group has maintained a prospective registry of all patients undergoing cardiac surgery in the region. The group fosters continuous improvement in the quality of care for patients with cardiovascular disease in the region by studying processes of care, evaluating clinical outcomes data, and providing timely and accurate feedback of data to clinicians.
For this study, patients who had emergent surgery, preoperative ejection fractions (EFs) <40%, preoperative intraaortic balloon pumps, patients who underwent concomitant procedures, as well as patients undergoing off-pump CABG surgery were excluded. To minimize the effect of a surgeon’s operative caseload on rates of LOF, we excluded surgeons who did not perform at least 80 procedures during the time period. Our final cohort consisted of 11 838 patients undergoing isolated CABG surgery at 8 medical centers in northern New England from 2001 through 2009.
Previous publications by the Northern New England Cardiovascular Disease Study Group have discussed in detail our data collection methodology and definitions.2 In short, we prospectively collected the following preoperative variables: (1) demographics: age, sex, body mass index, body surface area; (2) comorbidities: diabetes mellitus, vascular disease, chronic obstructive pulmonary disease, and renal insufficiency (dialysis or creatinine ≥2); (3) cardiac anatomy and function: left main stenosis ≥90%, ≥3 vessel disease, preoperative EF, and number of diseased vessels); and (4) cardiac history: prior myocardial infarction, congestive heart failure, prior CABG surgery.3 Several intraoperative variables were collected, including patient parameters (hematocrit, lowest core temperature); surgical technique (CPB and cross-clamp duration, number of anastomoses, use of a side biting clamp, use of arterial grafts); and myocardial protective strategies (route and type of cardioplegia, maximum interval between cardioplegia dosage, induction and maintenance cardioplegia temperature). Cardiothoracic surgeons assessed patient acuity (elective, urgent, emergent) using definitions described previously.2
LOF was defined as the need for ≥1 of the following: (1) ≥2 inotropes at 48 hours after surgery; (2) use of an intra- or post-operative intraaortic balloon pump; and (3) return to CPB for hemodynamic reasons for ≥10 minutes. Given that surgeons may choose to return patients to CPB for technical reasons or for control of hemorrhage, only those who returned for hemodynamic instability that was not secondary to hemorrhage were considered to have LOF.
For analysis, procedures were divided into approximate terciles based on the surgeon’s observed LOF rate (<2.1%, 2.1%–5.9%, ≥5.9%). Each stratum included all patients operated on by surgeons represented within each stratum, irrespective of the patient’s ventricular function.
Standard statistical methods were used to compare the characteristics of patients having LOF to those free from LOF, including analysis of variance or Kruskal-Wallis tests for continuous data and χ2 tests for categorical data. For ordered categories, we report the nonparametric test of trend (P for trend). The expected rates of LOF for each surgeon were calculated using multivariable logistic regression, adjusting for the following characteristics: age, sex, EF, number of diseased vessels, left main disease, priority at the time of surgery, prior CABG surgery, vascular disease, diabetes mellitus, renal failure or elevated creatinine, chronic obstructive pulmonary disease, and body mass index. We accounted for clustering of surgeons using a random effects approach for the reporting of clinical outcomes.
Our multivariable logistic regression model (with the addition of surgeon) was used to estimate the relative importance of both preoperative characteristics and surgeon in predicting the occurrence of LOF.
All analyses were performed using the STATA 11.0 program (Stata Corporation, College Station, TX).4
Protection of Human Subjects
Institutional review board approval was obtained at each participating medical center. Seven of our 8 member centers’ Institutional review boards have designated the Northern New England Cardiovascular Disease Study Group as a Quality Improvement Registry, and therefore patient consent was not required. Written patient consent was obtained for the 1 remaining center. The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.
Surgical volume among 32 surgeons varied from 80 to 766 cases (median 344) over the time period of the study. The overall rate of LOF was 4.3% (≥2 inotropes at 48 hours, 0.8%; balloon pump only, 1.0%; return to CPB only, 2.6%; combination, 1.0%). Although the predicted risk of LOF did not differ across surgeons, P=0.79, the observed rates varied from 1.1% to 10.2%, P<0.001 (Figure 2).
Baseline characteristics of the patients across surgeon LOF groups are shown in Table 1. Only relatively small absolute differences in the patient and disease characteristics were observed. Patient and disease characteristics accounted for 42% of the variation in LOF, whereas surgeon accounted for 58%.
Patients operated on by surgeons with a high rate of LOF (≥5.9%) were more likely to have left main stenosis ≥90%, 3-vessel disease, peripheral vascular disease, a recent myocardial infarction, and urgent versus elective surgery. They were less likely to undergo a secondary cardiac procedure, have a lower EF, or have congestive heart failure. Although statistically significant, the absolute differences were small.
Intraoperative care practices across surgeon LOF groups are shown in Table 2. Similar to Table 1, only relatively small absolute differences were noted in most instances. Ischemic times, as measured by pump time and clamp time (including clamp time per anastomosis), were shorter in surgeons with low rates of LOF. Patients operated on by low LOF rate surgeons had more use of a single clamp technique, higher average number of distal anastomoses per diseased vessel, and greater percentage of patients with lower core temperature. Surgeons with low rates of LOF increased their utilization of side-biting clamps among patients with >3 anastomoses. Surgeons with high rates of LOF did not change their proximal techniques based on the number of required anastomoses.
Myocardial Preservation Practices
Table 3 shows the myocardial preservation practices across surgeon LOF groups. Patients operated on by surgeons with a low rate of LOF were more likely to receive solely antegrade cardioplegia, a crystalloid-based solution (most accounted for by 1 surgeon in the low-rate group). Surgeons with low rates of LOF gave less cardioplegia volume per clamp time or per anastomosis and were less likely to use a hot shot, whether with cold or tepid/warm maintenance temperatures. In addition, surgeons with a low rate of LOF had longer maximum intervals between cardioplegia delivery, especially those lasting >25 minutes in duration.
Table 4 shows the risk of in-hospital outcomes across surgeon LOF groups. Although there was no statistical difference in most of the reported (mortality, stroke, renal failure or insufficiency, and atrial fibrillation) clinical outcomes across surgeon groups, absolute rates increased across the groups. Rates of acute kidney injury were statistically higher among high LOF surgeons, P=0.01.
The present report describes a regional investigation into the development of LOF after isolated on-pump CABG surgery. Although preoperative factors certainly help to explain some of the variation in rates of LOF, most of the variation is attributed to surgeon (58%). Given this predominating influence, along with the opportunity to modify those factors found to be associated with higher LOF rates, we sought to determine intraoperative practices that may be associated with or explain variability in rates of postoperative LOF. Although the predicted rate of LOF, which averages 4%, did not differ across surgeons, the observed rate varied greatly, with a limit between 1.1% and 10.2% across surgeons. Differences in comorbid conditions were small across surgical strata, suggesting that selection or referral bias does not explain differences in observed LOF rates.
We noted important differences in the intraoperative practices across the 3 surgical strata. For instance, patients operated on by low LOF surgeons had decreased clamp and pump time and cooler core body temperatures. Surgeons with low rates of LOF more frequently used antegrade perfusion and maintained their patients and cardioplegia solution at cooler temperatures. Surgeons with low rates of LOF used less volume of cardioplegia per anastomosis and clamp time and were less likely to use a hot shot. Although the absolute rates of reported clinical outcomes increased across strata of LOF surgeons, the only significant trend occurred among acute kidney injury.
Of interest, patients in the lowest LOF strata were more likely to have a higher preinduction heart rate. In a prior study, we observed that higher preinduction heart rates were associated with higher rates of mortality after CABG surgery in northern New England.5 Aboyans et al6 confirmed this finding. Although the current results may at first glance appear to contradict these prior observations, unlike our present report, these prior studies included patients who had a preoperative EF <40%. In addition, the primary outcome of the current study is LOF, whereas the outcome of interest in the other studies was mortality. Further research should explore the relationship between heart rate and LOF.
Historically, improvements in myocardial protection have certainly contributed to the decline in morbidity after cardiac surgery. In cardiac surgery’s infancy, speed was the mantra whereby surgeons were taught to minimize ischemic times to protect the myocardium. Subsequently, cardioplegia practices were developed as an adjunctive treatment. Although CPB or ischemic times are convenient metrics for estimating the difficulty of a given case or a patient’s risk of intraoperative injury, such a measure likely is neither sensitive nor specific. For instance, such a measure does not provide information regarding how well different coronary territories were protected during the construction of the distal anastomoses and do not reveal the technical difficulties encountered by challenging distal anatomy. Accordingly, our present findings suggest that further investigation is necessary to improve our understanding of how to minimize LOF, given the complexity of the relationship between myocardial protection and perioperative morbidity.
The goal of cardioplegia is to provide adequate and uniform distribution of cardioplegia and preserve the myocardium during the planned ischemic period (Figure 1). Several cardioplegia practices have developed to reduce myocardial injury, including alternative types, temperatures, dosages, and routes of cardioplegia delivered (online-only Data Supplement). Although some trial evidence exists, these findings lack generalizability because of the: (1) inadequate detail concerning a surgeon’s myocardial preservation practice, (2) focus on singular aspects of myocardial preservation, (3) relatively small sample sizes, and (4) inherent nature of designing and executing randomized trials (restrictive entrance criteria into trials result in the reduced generalizability of findings). As a result, uncertainty exists as to which cardioplegia strategies offer the most effective myocardial preservation for a given patient, especially as it relates to the extent and location of their coronary disease.
Most studies on the topic of LOF have focused on using preoperative patient and disease characteristics to predict the occurrence of postoperative LOF. For instance, Rao et al7 investigated the predictors of LOF among 4558 patients undergoing isolated CABG surgery between 1990 and 2003 at The Toronto Hospital. Rao found that 9% of patients had LOF during the index admission. Patients with LOF were older, had more diseased vessels, greater burden of comorbid conditions, longer pump time and ischemic duration, and risk of adverse sequelae. The following factors were found to increase a patient’s risk of LOF: lower preoperative EF, repeat operations, surgical acuity, female sex, diabetes mellitus, older age, left main coronary artery disease, recent myocardial infarction, and triple vessel disease.
Although patient-related factors undoubtedly influence the development of LOF, we hypothesized that variability in myocardial protection practices also impact rates of postoperative LOF. Of particular importance is the finding that although the predicted risk of LOF did not significantly vary across surgeons, the observed rate of LOF varied >8-fold. Furthermore, surgeons with low rates of LOF were more likely to operate on patients with worse left ventricular function, as well as patients presenting with congestive heart failure. Patients operated on by surgeons with a high rate of LOF (≥5.9%) were more likely to have left main stenosis ≥90%, 3-vessel disease, peripheral vascular disease, higher EFs, and urgent surgery. These findings suggest that although surgeons in northern New England have marked differences in rates of LOF after isolated CABG surgery, we cannot rule out some influence of both patient selection and intraoperative practices. We found differences in the intraoperative approach. For instance, patients operated by low LOF rate surgeons were more likely to have antegrade cardioplegia, longer maximum intervals between cardioplegia doses, lower volume of cardioplegia per minute of ischemic time or anastomosis, and less hot-shot use. Further work should focus on elucidating the mechanisms underlying these practices. For instance, is the speed in which one conducts the anastomosis a reflection of the skill of the surgeon or the order of difficulty of performing the distal anastomosis? Is it more important than the type of cardioplegia used? In this study, we found that low LOF rate surgeons have the shortest ischemic duration per anastomoses (low, 16 minutes/anastomoses; medium, 18; high: 17), while paradoxically, a greater percentage of their patients had over 25 minutes in between cardioplegia dosages (low, 24%; medium, 10; high, 13).
We recognize some limitations to our present study. First, in this regional observational cohort study, we used logistic regression to adjust for potentially confounding factors. Although we cannot rule out the influence of unmeasured confounding, we accounted for traditional factors known at the time of surgery, including patient demographics and extent of comorbid disease. As such, we can observe associations but cannot suggest or prove cause and effect. There may be unmeasured difference across these strata that we have not considered, including nontechnical skills such as communication, leadership, and situational awareness.8 Second, we have used 3 surrogates for LOF. Of the 3 surrogates, the use of an intraaortic balloon pumps and return to CPB may reflect a surgeon’s reluctance to use inotropic support, a strategy that increases myocardial work. Some surgeons prefer mechanical support to pharmacological support for this reason. On the other hand, the presence of ≥2 inotropes at 48 hours is likely an apt surrogate for postoperative LOF. Nonetheless, our definition is similar to other series that have relied on the use of inotropic support and balloon pumps to define LOF in the setting of cardiac surgery.
We found an 8-fold variability in observed rates of LOF across surgeons practicing within northern New England. Observed rates could not be completely explained by patient and disease characteristics. Differences in rates of LOF were also not explained by center-level characteristics. Surgeons having high, relative to low, rates of LOF have different cardioplegia practices and had significantly higher rates of acute kidney injury, although not other traditional clinical outcomes. Additional studies are warranted to assist in our efforts aimed at identifying which sets of factors are most predictive of lower rates of LOF, including technical (the type, timing, temperature, and route of cardioplegia, technical speed, quality of the distal anastomoses, surgeon’s preferences for mechanical support in the operating room and immediately postoperative) and nontechnical components (situational awareness, communication, and teamwork). Such studies will provide direction to investigators who wish to reduce morbidity from CABG surgery through decreasing the incidence of LOF.
Sources of Funding
Dr Likosky was supported by a grant from the Agency for Healthcare Research and Quality (1K02HS015663-01A1). This work was partially funded by the Northern New England Cardiovascular Disease Study Group.
The online-only Data Supplement is available at http://circoutcomes.ahajournals.org/lookup/suppl/doi:10.1161/CIRCOUTCOMES.112.967091/-/DC1.
- Received November 23, 2011.
- Accepted June 21, 2012.
- © 2012 American Heart Association, Inc.
- O’Connor GT,
- Birkmeyer JD,
- Dacey LJ,
- Quinton HB,
- Marrin CA,
- Birkmeyer NJ,
- Morton JR,
- Leavitt BJ,
- Maloney CT,
- Hernandez F,
- Clough RA,
- Nugent WC,
- Olmstead EM,
- Charlesworth DC,
- Plume SK
- O’Connor GT,
- Plume SK,
- Olmstead EM,
- Coffin LH,
- Morton JR,
- Maloney CT,
- Nowicki ER,
- Tryzelaar JF,
- Hernandez F,
- Adrian L
- 3.↵Coronary artery surgery study (CASS): A randomized trial of coronary artery bypass surgery. Survival data. Circulation. 1983;68:939–950.
- 4.↵Stat. Stata Statistical Software: Release 11.0. College Station, TX: Stata Corporation.
- Fillinger MP,
- Surgenor SD,
- Hartman GS,
- Clark C,
- Dodds TM,
- Rassias AJ,
- Paganelli WC,
- Marshall P,
- Johnson D,
- Kelly D,
- Galatis D,
- Olmstead EM,
- Ross CS,
- O’Connor GT
- Aboyans V,
- Frank M,
- Nubret K,
- Lacroix P,
- Laskar M