Abstract

Objective

To study the effect of blood glucose (BG) control with insulin in preventing hyperglycemia during and after coronary artery bypass grafting (CABG) surgery in nondiabetic patients.

Methods

In a randomized clinical trial, 120 nondiabetic patients who underwent elective CABG surgery were enrolled for study of whether the control of hyperglycemia was a need in such a surgery in a teaching heart hospital. The patients were randomly divided into study (n = 60) and control (n = 60) groups. In the study group, insulin was infused to maintain BG level between 110 mg/dL and 126 mg/dL (a modified insulin therapy protocol, and in the control group, the patients were excepted). Insulin therapy was limited to intraoperative period. BG levels during surgery and up to 48 hours after surgery and early postoperative complications were compared between the study and control groups.

Results

One hundred seventeen patients completed the study (59 patients in study group and 58 in control group). Peak intraoperative BG level in the study group was 126.4 ± 17.9 mg/dL and in the control group was 137.3 ± 17.6 mg/dL (p = 0.024). The frequencies of severe hyperglycemia (BG ≥ 180 mg/dL) were 6 of 59 (10.1%) in the study group and 19 of 58 (32.7%) in the control group during operation (p = 0.002). Peak postoperative BG level in the study group was 194.8 ± 41.2 mg/dL and was 199.8 ± 43.2 mg/dL in the control group (p = 0.571). There was no hypoglycemic event in either group. The frequencies of early postoperative complications were 10 of 59 (16.9%) in the study group and 19 of 58 (32.7%) in the control group (p = 0.047).

Conclusions

Hyperglycemia (BG ≥ 126 mg/dL) is common in nondiabetic patients undergoing CABG surgery. A modified insulin therapy to maintain BG level between 110 mg/dL and 126 mg/dL may be acceptable for avoiding hypoglycemia and keeping intraoperative BG levels in acceptable range in nondiabetics.

Keywords

blood glucose; coronary artery bypass; hyperglycemia; insulin;

---

1. Introduction

Hyperglycemia is an independent risk factor for perioperative mortality and morbidity in patients undergoing cardiac surgery.1, 2, 3, 4 Perioperative hyperglycemia is associated with increased costs and resource utilization in patients undergoing coronary artery bypass grafting (CABG) surgery with or without diabetes.⁵ Although glucose intolerance may be seen in a fraction of nondiabetic patients, yet in nondiabetic patients without glucose intolerance, severe hyperglycemia has been observed during stressful processes, such as in myocardial infarction and major surgeries.6, 7

Severe hyperglycemia is common during cardiac surgery,1, 2, 3, 8 even with no exemption in nondiabetic patients.⁹ However, most studies have been performed in diabetic patients undergoing CABG surgery. Recent evidences showed no significant reduction in overall mortality in hospitalized patients who underwent intensive insulin therapy; hence, some benefits in reducing neurological and infectious complications with tight glycemic control were offset by increased risk of hypoglycemia.10, 11, 12, 13 Thus, every study on glycemic control must be considered to have the risk of hypoglycemia especially in nondiabetic patients.

The purpose of this study was to investigate the effect of a modified blood glucose (BG) control protocol with insulin in preventing hyperglycemia during CABG surgery (primary objective) and the incidence of early postoperative complications (secondary objective) in nondiabetic patients.

2. Methods

After institutional ethics committee approval and obtaining informed written consent from all the patients, we conducted a double-blinded randomized clinical trial in nondiabetic patients. One hundred twenty nondiabetic patients of American Society of Anesthesiologists status Class II or III who underwent elective CABG surgery from December 2008 to October 2009 in Madani Heart Hospital, a referral teaching center in northwest of Iran, were included in the study.

Patients with American Society of Anesthesiologists status Class IV, those who received insulin or oral hypoglycemic agents before surgery and those in whom inotropic drugs or intra-aortic balloon pump (IABP) were used, were excluded from the study. Also, patients with considerable intraoperative blood loss, hyperkalemia that required insulin and glucose infusion for treatment, and who had seen cardiac arrest before were excluded from the study.

The study sample size determination was based on the results of Gandhi et al⁴ that the frequency of postoperative adverse events (excluding death) was 37%, and considering α = 0.05 and power = 0.8 and by using an online power/sample size calculator to compare two proportions (URL: http://www.stat.ubc.ca/∼rollin/stats/ssize/b2.html), we obtained a sample size of 57 patients in each group. We rounded this number to 60 for each group. Then, patients were randomly divided into two groups: a study group (n = 60) in which insulin was infused if hyperglycemia (BG ≥ 126 mg/dL) happened and a control group (n = 60) in which only BG level was monitored every 30 minutes (Fig. 1). Randomization process was done using “Graphpad” software available from URL http://www.graphpad.com.¹⁴ To assure that randomization concealment was done, the process of randomization was performed by a person not from the research team, and the randomization list was kept as a secret until the end of the study.

The anesthetic technique was the same in both groups. Anesthesia was induced with benzodiazepine (midazolam, 0.05–0.1 mg/kg); opioid (fentanyl, 25–40 μg/kg, or sufentanil, 2.5–4 μg/kg); and muscle relaxant (cisatracurium, 0.2 mg/kg). Anesthesia was maintained with midazolam (1–2 μg/kg/min), fentanyl (1–2 μg/kg/h), and cisatracurium (1–3 μg/kg/min). In this study, the patients underwent “CABG with cardiopulmonary bypass (on-pump)” or “without cardiopulmonary bypass (off-pump),” and the proportion is shown in Table 1. The ratio of “on-pump/off-pump” in study group was 28/31 and in control group was 23/25. There was no significant difference between the two groups regarding the type (technique) of CABG (p = 0.564).

After induction of anesthesia, arterial BG level was measured with glucometer (ACCU-CHEK Active, Ireland). In the control group, no intervention was done unless the BG level exceeded 200 mg/dL (treated by bolus insulin). The intervention (insulin therapy) was limited to intraoperative period. In the study group, insulin was infused to maintain BG level between 110 mg/dL and 126 mg/dL (a modified insulin therapy protocol). During the postoperative period, insulin bolus was administered if BG level exceeded 200 mg/dL [as a part of the intensive care unit (ICU) management in both groups not related to the study protocol].

Because of concerns about hypoglycemia occurrence, we decided to do more conservatively and change the target BG level from tight control (80–110 mg/dL) to semitight control (110–126 mg/dL). The measurement was performed every 30 minutes intraoperatively until the closure of the sternum and thereafter every 2 hours up to 48 hours postoperatively. In the study group, insulin was used according to insulin infusion protocol in Appendix 1 (modified from Cammu et al).¹⁵

Of the 120 patients who enrolled in the randomized study, three were lost from the study: one patient in the study group because of excessive intraoperative blood loss and two patients in the control group because of severe hemodynamic instability, usage of high-dose inotropes, and IABP.

During ICU stay, all patients were visited daily by cardiac anesthesiologists who collected data, including postoperative variables and significant respiratory, nephrological, and neurological complications. The postoperative chest radiographies were examined daily for evidence of pulmonary complications, such as plural effusion, atelectasis, and pneumothorax. Cardiovascular status included cardiac rhythm and the need of inotropic drugs or IABP to maintain hemodynamic stability. Significant arrhythmias were those that caused hemodynamic instability and required treatment. The necessity of using inotropic drug was defined as the need of a moderate to high dose of one of the inotropic drugs for more than 1 hour. Bleeding was defined as greater than or equal to 200 cc/h or more than 1000 cc in 24 hours. Tamponade was diagnosed by clinical examination and Transthoracic Echocardiography and confirmed by re-exploration. Postoperative renal insufficiency was considered as the creatinine concentration exceeding 2 mg/dL or exceeding the preoperative value for more than 1 mg/dL. Stroke was described as a neurological abnormality that was confirmed by a neurologist. Infection was confirmed by positive culture of blood, sputum, urine, plural or mediastinal fluid, and incisional discharge. Additional data collected included the length of ICU stay and in-hospital mortality.

Data analysis was performed using the SPSS for Windows program package version 16.0 (SPSS Inc., Chicago, IL, USA). Categorical variables were compared with Chi-square or Fischer’s exact test as appropriate. Continuous variables were analyzed with independent-sample t test. Mann-Whitney test was used as a nonparametric test. Repeated-measures analysis of variance was done in both groups to evaluate whether the variations of BG level during or after operation were statistically significant or not. All tests were two tailed, and a level of significance was set at p ≤ 0.05.

3. Results

The background characteristics and intraoperative parameters of patients were comparable in both groups (Table 1). Intraoperative and postoperative variation of BG levels in both groups are illustrated in Fig. 2, Fig. 3. As shown in Fig. 2, BG level increased from induction of anesthesia up to end of surgery, which was more prominent in the control group. This increasing pattern continued up to 12 hours postoperatively and then faded in both groups. There was no hypoglycemic event in either group. Of the 117 patients who completed the study, 55 (47.0%) experienced intraoperative BG level greater than or equal to 126 mg/dL (29 of 59 patients in study group and 26 of 58 patients in control group). The mean peak BG level was 126.4 ± 17.9 mg/dL in study group and 137.3 ± 17.6 mg/dL in control group (p = 0.024). In this study, we defined severe hyperglycemia as BG level greater than or equal to 180 mg/dL. The frequencies of severe intraoperative hyperglycemia were 4 of 59 (6.8%) in study group and 19 of 58 (32.7%) in control group (p = 0.002). The total insulin used during operation in 29 of 59 patients in the study group was 11.69 ± 6.20 units.

In postoperative period, patients in both groups experienced higher BG levels (Fig. 3). Postoperative peak BG was not statistically different in the two groups (Table 2). Most of the patients in the study group (51 of 59; 86.4%) and nearly all of the patients in the control group (56 of 58; 96.5%) had at least one episode of BG level greater than or equal to 180 mg/dL in the postoperative period (p = 0.371).

Thirty patients (25.6%) experienced at least one complication in early postoperative period (in ICU). The frequency of postoperative complications was 10 of 59 (16.9%) in study group as opposed to 19 of 58 (32.7%) in control group (p = 0.047). More than one complication was seen in some patients. In the study group, six patients experienced cardiac complications (new-onset atrial fibrillation in five patients and myocardial infarction in one); three patients had prolonged (>18 hours) mechanical ventilation in ICU. Three patients experienced delirium, and two patients were returned to the operating room because of excessive bleeding. In control group, five patients experienced cardiac complications (new-onset atrial fibrillation in four patients and myocardial infarction in one), two patients experienced delirium, and one patient experienced cerebrovascular accident; in three patients, the creatinine level rose to more than twofold of the preoperative value, two patients had deep sternal infection, and three patients were returned to the operating room because of excessive bleeding (Table 2). One patient in each group died during postoperative ICU stay because of cardiogenic shock.

4. Discussion

In the present study, we found that hyperglycemia is common during and after CABG so that considering a target BG level of 126 mg/dL, 47.0% of nondiabetic patients undergoing CABG surgery require insulin therapy. Insulin therapy to contain BG levels to 110–126 mg/dL is successful and safe in keeping intraoperative BG levels in acceptable range especially after the first 90 minutes of the surgery (Fig. 2). Considering the high BG levels during postoperative period, we are of the opinion that intraoperative glycemic control is not effective in preventing postoperative hyperglycemia.

The BG level in both groups raised from induction of anesthesia until the end of the surgery, and this rise continued up to 12 hours postoperatively. This could be mainly because of stress response to surgery, which continued throughout the postoperative period and was exacerbated by awakening from anesthesia and cessation of insulin infusion in this period (these factors explain more pronounced hyperglycemia during postoperative period). In addition, dextrose 5% infusion for on-pump CABG patients and dextrose 3.3% and sodium chloride 0.3% solution infusion for off-pump CABG patients were surgeons’ routines in the first postoperative 12 hours, which predisposed patients to more severe hyperglycemia during this period (Fig. 3).

It has been shown that high levels of glucose irrespective of diabetic status could lead to the formation of abnormal proteins and toxic effects through nonenzymatic glycosylation.¹⁶ Hyperglycemia has been shown to worsen neurological injury after focal and global cerebral ischemia, probably because of anaerobic glycolysis-induced conversion of glucose to lactate, which causes intracellular acidosis and impairment of cellular metabolism.¹

Intraoperative hyperglycemia is an independent risk factor for complications (e.g. deep sternal infection, septicemia, urinary tract infection, stroke, delirium, coma ≥24 hours, prolonged mechanical ventilation, pneumonia, acute or worsening renal failure, heart block, cardiac arrest, and atrial fibrillation), inclusive of death, after cardiac surgery.⁴ It must be noted that the mentioned complications are direct results of hyperglycemia not related to acute or chronic lack of insulin and can be seen in nondiabetic patients.5, 17, 18

Azarfarin and Alizadeh Asl⁹ reported that the prevalence of at least one episode of severe hyperglycemia (BG ≥ 180 mg/dL) was 54.6% in nondiabetic patients during and 24 hours after CABG surgery. Perioperative hyperglycemia is associated with increased resource utilization in patients undergoing CABG with and without diabetes.⁵ Maintaining normoglycemia during surgery would improve the outcomes.⁴ Some studies showed that intensive insulin therapy to maintain normal levels of BG (80–110 mg/dL) during intensive care improves the survival and reduces morbidity in critically ill patients after sustaining complicated, high-risk, or extensive surgery or trauma.6, 7 However, other investigations showed no significant reduction in overall mortality in hospitalized patients who underwent intensive insulin therapy, and in some reports, mortality was increased.10, 11, 12, 13 In these studies, the most serious concern was hypoglycemia, and ongoing research must primarily recommend the avoidance of hypoglycemia by revising the BG level control limits and more frequent measurements of BG level.

There are limited published studies about the effects of glycemic control on the outcome in nondiabetic patients during cardiac surgery, but there are some reports that address this issue in diabetic and nondiabetic patients as a single group, for example, Gandhi et al⁴ showed that the relationship between intraoperative glucose concentrations and adverse events was continuous, with each 20-mg/dL increase in glucose concentration greater than 100 mg/dL being associated with a 34% increase in the likelihood of experiencing eventfulness. Estrada et al⁵ reported that higher perioperative glucose was not associated with higher mortality and infection rate among patients with or without diabetes, but among patients with diabetes, each 50-mg/dL glucose increase was associated with an increase in postoperative stay by 0.99 days, hospitalization charges by $4320, and hospitalization cost by $2870. Similarly, among patients without diabetes, each 50-mg/dL glucose increase was associated with an increase in postoperative days by 0.58 days and hospitalization charges by $1552.⁵

Arabi et al¹⁹ reported that intensive insulin therapy was not associated with improved survival among patients in medical or surgical ICU and was associated with increased occurrence of hypoglycemia. Also, Chaney et al¹ reported that attempting to maintain normoglycemia during cardiopulmonary bypass with insulin therapy was associated with postoperative hypoglycemia. Nevertheless, considering a conservative glucose control protocol in this study, we did not find any hypoglycemic event in our patients, and this is one of the strengths of this trial.

Gandhi et al²⁰ found that tight intraoperative glucose control compared with conventional treatment does not reduce perioperative death or morbidity in cardiac surgery patients (in both diabetics and nondiabetics). Also, they observed more death and stroke in intensive treatment group; however, by considering the comparable incidences of intraoperative and postoperative hypoglycemia in both the study groups, the higher rates of death and stroke may be not related to insulin therapy (may be related to, e.g. older patients’ age or cardiovascular dysfunction). In our study, overall postoperative complications were lower in the study group. However, it does not seem that a single intervention, such as glycemic control, causes such a significant difference in the incidence of postoperative complications. Therefore, more studies with larger number of cases are needed to confirm this relationship.

5. Conclusion

Hyperglycemia in perioperative period is quite common in nondiabetic patients undergoing CABG surgery. Considering a target BG level of 126 mg/dL, nearly half of the nondiabetic patients undergoing CABG surgery may benefit from a modified intraoperative insulin therapy with multiple measurements of BG level in avoiding hypoglycemia.

5.1. Limitations

Because of concerns of occurrence of hypoglycemia during insulin therapy in nondiabetic patients, we changed tight-control protocol target from “70–110 mg/dL” to more conservative protocol target “110–126 mg/dL.”

Because the need of close observation for glycemic control in postoperative period was impossible for us, we discontinued insulin therapy at the end of surgery. Considering higher prevalence and intensity of hyperglycemia in early postoperative period in nondiabetic patients, we strongly suggest extending glycemic control to postoperative period in these patients.

Appendix 1. Institutional continuous intravenous insulin infusion protocol

Before induction of anesthesia, fasting blood glucose (BG) level is obtained. After induction of anesthesia, an insulin infusion (1 U/mL in normal saline) is initiated, as shown in Table A1. All intravenous fluids are free of glucose.

The target BG level is 110–126 mg/dL. Frequency of BG testing is as follows: every 30 minutes in the operating room and, thereafter, in the ICU, every 2 hours (or every hour if >50-mg/dL differences). Insulin titration according to BG level is performed as follows:

1. When BG level is less than 60 mg/dL, stop insulin and give 30% dextrose in water (6 g). Recheck BG level in 15 minutes.

2. When BG level is 80–110 mg/dL, stop insulin. Recheck BG level in 15 minutes.

3. When BG level is 110–126 mg/dL and greater than 50 mg/dL lower than that of the last test, decrease rate by 2 U/h. If BG level is 50–20 mg/dL lower than that of last test, decrease rate by 1 U/h. If BG level is less than 20 mg/dL lower or higher than that of last test, maintain rate.

4. When BG level is 126–160 mg/dL and greater than 50 mg/dL lower than that of last test, decrease rate by 1 U/h. If BG level is 50–20 mg/dL lower than that of last test, maintain the same rate. If BG level is less than 20 mg/dL lower or higher than that of last test, increase rate by 1 U/h.

5. When BG level is 160–200 mg/dL and greater than 50 mg/dL lower than that of last test, decrease rate by 1 U/h. If BG level is 50–20 mg/dL lower than that of last test, maintain the same rate. If BG level is less than 20 mg/dL lower or higher than that of last test, increase rate by 1 U/h.

6. When BG level is greater than 200 mg/dL and has not decreased, then increase rate by 2 U/h. If BG level is less than 50 mg/dL lower than that of last test, increase rate by 1 U/h. If BG level is 100–50 mg/dL lower than that of last test, maintain the same rate. If BG level is greater than 100 mg/dL lower than that of last test, decrease rate by 1 U/h.