The association between hemoglobin A1C values and deep sternal wound infections in diabetes patients undergoing cardiac surgery

Fohl, Alexander L.a; Butler, Simona O.b,c; Patil, Preethi V.d; Zrull, Christina A.d; Kling-Colson, Sued; Dubois, Elizabethd; Sweeney, Jennifer L.e; Haft, Jonathan W.f; Gianchandani, Roma Y.d

Cardiovascular Endocrinology:
doi: 10.1097/XCE.0b013e32835eca16
Original articles

Background: Deep sternal wound infections (DSWI) after cardiac surgery are a major cause of morbidity and mortality, especially in patients with diabetes mellitus (DM). Although data of postoperative blood glucose (BG) control on surgical outcomes is well established, the impact of a high A1C on morbidity and mortality is still unclear.

Objective: The purpose of this study was to evaluate the association between preoperative glucose control, as measured by A1C with postoperative outcomes especially DSWI, in cardiac surgery patients with known DM whose postoperative BG was controlled to a goal of 100–140 mg/dl.

Methods: This is a single-center, retrospective observational study of DM patients who were stratified according to their preoperative A1C: good glycemic control (A1C <7.0%), moderate glycemic control (A1C 7.0–8.5%), and poor glycemic control (A1C >8.5%). Postoperative glycemic management was standardized. Cox regression model was used to determine whether A1C was an independent risk factor of DSWI.

Results: In 861 diabetes patients with similar postoperative BG control after cardiac surgery, the total incidence of DSWI was 2.8%. Six hundred and sixteen qualified and were stratified by A1C. DSWI rates were 2.3% in good glycemic control, 4.3% in moderate glycemic control, and 8.1% in poor glycemic control groups. After multivariate adjustment, a higher A1C was associated with an increased incidence of DSWI (hazard ratio=1.38, P=0.009).

Conclusion: In cardiac surgery patients with DM, despite standardized control of immediate postoperative hyperglycemia, a high preoperative A1C was associated with an increased incidence of DSWI.

Author Information

aIndiana University Health Ball Memorial Hospital, Muncie, Indiana

bDepartment of Pharmacy Services, University of Michigan Hospitals and Health Centers

cUniversity of Michigan College of Pharmacy

dDepartment of Internal Medicine, Division of Metabolism, Endocrinology and Diabetes

Departments of eInfection Control and Epidemiology

fCardiac Surgery, University of Michigan, Ann Arbor, Michigan, USA

Correspondence to Roma Y. Gianchandani, MD, Department of Internal Medicine, Division of Metabolism, Endocrinology and Diabetes, University of Michigan, Domino Farms Lobby G, Ste 1500, 24 Frank Lloyd Wright Drive, PO Box 482, Ann Arbor, MI 48106-0482, USA Tel: +1 734 647 5400; fax: +1 734 936 9240; e-mail:

Received July 9, 2012

Accepted December 19, 2012

Article Outline
Back to Top | Article Outline


Glucose is readily taken up by erythrocytes where it glycates the lysine side chain and the amino terminals of hemoglobin. The extent of glycosylated hemoglobin is directly proportional and a surrogate to the patient’s serum blood glucose (BG) concentration. Hemoglobin A1C (A1C) is used to diagnose and monitor glycemic control in patients with diabetes mellitus (DM). It is a strong prognostic indicator for microvascular complications and to a much lesser degree for macrovascular complications of DM. The American Diabetes Association (ADA) recommends an A1C goal of less than 7% to reduce long-term complications of DM, and suggest modifying this target in patients with comorbidities 1–3.

Poor BG control is also associated with an increased risk of infections. Observational and prospective data have shown an independent association between postoperative BG control and infections, morbidity, mortality, and length of stay with improvement in these outcomes demonstrated when BG is managed to appropriate goals after surgery 4–19. In patients undergoing cardiac surgery, these data have evolved over the last decade. Observational studies from the Portland group aimed to keep BG under 200 mg/dl through postoperative day (POD) 2 to prospective randomized tight control in the Leaven trial and now to a modified goal of 140–180 mg/dl in and after the NICE-SUGAR trial. Although BG goals have been evolving, it is clear that postoperative BG control is standard of care in this population and cannot be ignored.

After cardiac surgery, the incidence of deep sternal wound infections (DSWI) is between 1 and 6% 14,15,20–23; however, its related mortality is reported to be as high as 35% 16,17,20,24. Although the correlation between postoperative hyperglycemia and DSWI, hospital length of stay, morbidity and mortality is well established 4–19 the impact of preoperative BG levels on these outcomes is less clear. The question still remains whether patients with poor BG control preoperatively are at a higher risk of infections or morbidity even after BG is managed to appropriate goals postoperatively. In 1997, Gordon et al. 25 briefly described the association of a high A1C and an increased risk of nosocomial infections in coronary artery bypass grafting (CABG) patients. Several other studies have analyzed this relationship but have not included postoperative glucose control into the study design. If this association holds true, it would make an important case for interventions to improve BG control before an elective procedure.

Back to Top | Article Outline


The purpose of this study was to evaluate glucose control before surgery as measured by A1C levels in cardiac surgery patients with DM, and determine its association with postoperative complications, specifically DSWI, after management of BG to appropriate goals in the immediate postsurgery period. The secondary endpoints included 30- and 90-day mortality.

Back to Top | Article Outline

Experimental design and methods


DSWI and Organ/Space Surgical Site Infections (SSI) were identified according to the criteria determined by the Centers for Disease Control and Prevention (CDC). The infection had to occur within 30 days after the operative procedure if no implant was left in place, involved deep soft tissues (for DSWI) or a surgical site manipulated excluding the skin incision, fascia, or muscle layers and had at least one of the following: (a) an organism isolated from culture of mediastinal tissue or fluid, (b) evidence of DSWI seen during operation, or (c) one of the following: either chest pain, sternal instability, or fever in combination with either purulent discharge from the mediastinum or an organism isolated from a blood or mediastinal drainage culture 14,26,27. This information was gathered and maintained by the infection control group at the University of Michigan Hospitals and Health Centers. Our process to follow DSWIs is stringent. Patients report back to surgery if they have wound discharge, pain, etc. and their readmission is picked up by the Infection Control group of the hospital which follows each and every case and receives results of all positive sternal wound cultures. If a patient goes to another hospital with an infection we are contacted and the patient is transferred. Once a patient is identified as a suspected infection, Infection Control Practitioners review the patient’s on-line medical chart to determine whether the patient meets criteria for a DSWI according to CDC definitions.

Time to event was defined as the days from the surgical procedure to the day the DSWI was detected. The preoperative period was defined as time before surgery, intraoperative defined as the period during the surgical procedure or anesthesia, and postoperative defined as the period after anesthesia and surgery. The perioperative period included the preoperative, intraoperative, and postoperative periods.

This study was approved by the University of Michigan’s Institutional Review Board, and the requirement of written informed patient consent was waived by the Institutional Review Board. This was a single-center, retrospective observational cohort of adult cardiac surgery patients with type 1 or type 2 DM admitted to the University of Michigan Hospitals and Health Centers from 1 January 2005 through 31 December 2009. The cardiac surgery procedures primarily included CABG, valve repairs/replacements, and aortic aneurysm repairs performed through a chest incision.

Only patients who had a documented preoperative A1C within 3 months before surgery were included in the study. A1C was necessary because it is the standard test to measure long-term glucose control in comparison to a fasting BG which measures a single point in glucose metabolism. As A1C is affected by blood transfusions and severe anemia, a presurgical level was a key inclusion criterion. Patients with blood hemoglobin concentration less than 9 mg/dl were excluded to limit the confounding effect of anemia on A1C. Procedures which did not require a chest incision as well as left ventricular assist device (LVAD) surgeries, and heart transplantations were excluded. LVADs are implanted foreign objects and therefore can be the nidus of an infection. Heart transplant patients are immunosuppressed and the steroid treatment exacerbates hyperglycemia. The consort diagram Fig. 1, provides a schematic of the inclusion and exclusion criteria.

Hemoglobin A1C either from an outside blood report or institutionally checked was acceptable. The institutional laboratory used the Tosoh Bioscience glycohemoglobin analyzer HLC-723 G7 (Tosoh Bioscience Inc., San Francisco, California, USA) and testing was performed by high-performance liquid chromatography. During the study period, adherence to standard infection control practice core process measures were actively reviewed and pursued, as were infection control practices to reduce SSI. These included prophylactic antibiotic received within 1 h before surgical incision, antibiotic selection for surgical patients, antibiotics discontinued within 48 h after surgery end time, controlled 6 a.m. postoperative serum glucose on POD 1 and 2, and appropriate hair removal. Standardized protocols for bathing, staphylococcus decolonization, intraoperative glucose control, and antibiotic prophylaxis were used. For patients having isolated coronary artery bypass procedures, cefazolin is our drug of choice for nonpenicillin allergic patients and vancomycin for penicillin allergic patients. Until 2009, for patients having valve or aortic procedures vancomycin alone was used but after 2009, vancomycin plus cefuroxime was used. All prophylactic antibiotics were given for 48 h. The chest was closed by faculty surgeons or fellows depending on the nature of the case and the availability of the various trainees.

All patients with DM undergoing cardiac surgery during this time period were followed postoperatively by a Hospital Intensive Insulin Program. This team oversees the intravenous insulin drip protocol and then transitions and adjusts subcutaneous insulin until discharge. In the cardiothoracic ICU, insulin drips are initiated for BG over 140 mg/dl checked twice an hour apart or a single measurement over 200 mg/dl. Insulin was titrated along the ICU protocol to maintain BG concentrations between 100 and 140 mg/day. This standardized management should reduce the confounding effect of postoperative BG levels on the preoperative glycemic analysis. In addition, surgical follow-up was stringent with postoperative phone calls and a surgical appointment scheduled 4–6 weeks after discharge. Using the anesthesia end date as day 0, the 6 a.m. BG values on POD 1 and 2, which are a component of the Surgical Care Improvement Project measures, are followed by and were included in the analysis of mean postoperative glucose values on those days.

All patients included in the analysis were stratified based on A1C cutoffs into the following groups: good glycemic control (GGC) if their A1C was less than 7.0%, moderate glycemic control (MGC) if their A1C was between and included 7.0 and 8.5%, and poor glycemic control (PGC) if their A1C was greater than 8.5%. The level of 7% for GGC was based on the 2011 recommendations of the ADA for A1C goal 2. This recommendation is derived from the Diabetes, Control and Complications trial data to prevent microvascular complications 28 and on recent ACCORD and ADVANCE 29,30 trial data which caution against hypoglycemia with tight glucose control. An A1C greater than 8.5% was defined as poor control because it corresponds to an estimated average glucose of 200 mg/dl 3, the level of BG at which immune function and ability to fight infections are compromised.

Baseline demographic, operative, BG, and complications data were collected prospectively by Hospital Intensive Insulin Program and the cardiac surgery teams. An average daily BG value was calculated for day of surgery (DOS) and POD 1 and 2. Data extracted from these sources and the electronic medical record included the diagnosis and treatment of DM, demographics, surgical procedure, baseline medical history, cross-clamp time (used as a surrogate for length of surgery), surgical length of stay, and postoperative complications 30- and 90-day mortality. Laboratory values included preoperative A1C and intraoperative and postoperative BG levels (Tables 1 and 2).

Back to Top | Article Outline
Statistical analysis

Descriptive statistics were used to report baseline characteristics and information regarding hospitalization for the entire cohort. Continuous measures are summarized by mean and SD and discrete measures by number with the characteristics (N) and percent. When the measure was continuous, group means were compared by a two sample t-test or a one-way analysis of variance (if more than two groups): when discrete a χ2 test was used. The probability of DSWI was plotted against time for three strata of A1C using Kaplan–Meier curves. Cox proportional hazards model was used for multivariate analysis. Although A1C was organized into risk groups for descriptive purposes, it was analyzed as a continuous predictor in the Cox regression model. On the basis of the univariate association with DSWI, potential confounders were entered into Cox proportional hazard regression models along with A1C. To adjust for perioperative glucose control, two different mean glucose values were computed: mean glucose value for the DOS and the mean glucose values for POD 1 through 2. The a-priori α of 0.05 was considered statistically significant. Data variables were analyzed using SPSS (version 17; SPSS, Inc., Chicago, Illinois, USA) and SAS (9.1; SAS Institute, Cary, North Carolina, USA) statistical software.

In the multivariate analysis, because of the small sample size, potential confounders were adjusted for in separate models. The covariates included were age, surgery type, mean BG on DOS, combined mean BG on POD 1 and 2, preoperative insulin use, time of surgery and peripheral vascular disease.

Back to Top | Article Outline


Between 2005 and 2009, 861 DM patients underwent a cardiac surgical procedure at the University of Michigan Hospitals. Of these, 109 did not meet inclusion criteria for type of procedure (56 LVADs, 42 heart transplants, excluded because of a foreign device or steroids), 10 were endovascular procedures without a chest incision and one patient received postoperative steroids. Consequently, 752 patients were considered for inclusion. An additional 129 were missing a documented preoperative A1C leaving 623 patients. Seven of these were excluded because of missing documentation of postoperative BG monitoring. Ultimately, 616 patients met all the adequate criteria and to be included in the analysis (Fig. 1).

Table 1 shows the baseline characteristics. The mean age was 64±10 years with a mean BMI of 31±8 kg/m2. Fifty percent were smokers and 80% had hypertension and hypercholesterolemia. These patients were stratified according to preoperative A1C as:

GGC group – 296 patients.

MGC group – 210 patients.

PGC group – 110 patients.

Those with a higher preoperative A1C were significantly younger (P<0.0001), had higher BG concentrations (P<0.0001) on the DOS, POD 1, and 2, were more likely to be on insulin before surgery (P<0.001) and to have peripheral vascular disease (P=0.04).

Of the total of 861 DM patients, 24 patients (2.8%) were diagnosed with a DSWI. Of the 616 patients who qualified for the study and were stratified into A1C groups, the cumulative percent of DSWI in the PGC group was 8.2% compared with 4.3% in the MGC and 2.0% in the GGC group (Figs 2 and 3). Mean BG on DOS in those with DSWI and without DSWI was 159.9 and 149.9 mg/dl, respectively, and on POD 1 through 2 was 137.9 and 140 mg/dl and not statistically significant (P=0.41) (Table 2).

Six hundred and thirteen (99.5%) and 605 (98.2%) patients had postoperative glucose values less than 200 mg/dl on POD 1 and 2, respectively, and this was not significant between the three groups either on POD 1 (P=0.06) or POD 2 (P=0.18). To analyze whether DSWI rates for patients in the PGC group were a function of postoperative glucose control, models were fitted to the data from the subgroup where A1C was greater than 8.5% that included mean BG values from POD 1 and 2, both on the original scale and after a log transformation. The tests of fit of these variables were not significant. Higher A1C percentage was associated with an increased incidence of DSWI (hazard ratio=1.38, P=0.009) for each unit increase in A1C (Fig. 4).

Back to Top | Article Outline


In this study of patients with known DM, a higher preoperative A1C level was associated with a proportional increase in DSWI rates (a 38% increase in DSWI for each unit increase in A1C). Because the confounding effects of postoperative BG were well controlled, the preoperative BG association to outcomes is especially valid and is this study’s major strength. In the last decade, several trials have well characterized the association between postoperative BG levels and poor outcomes 12,31,32; therefore, the findings of this study add yet another dimension to glucose control and surgical complications.

Furnary et al. 19 found that increasing BG levels after surgery were directly associated with increasing rates of DSWI and mortality 5,19. Their data recommend continuous insulin infusions up to POD 3 to effectively improve these outcomes. The results of the Leuven surgical trial, the NICE-SUGAR, and other prospective trials reiterate postsurgical BG control, although to different goals. Recent trials have therefore shifted their focus to investigate the relationship between presurgical glycemic control and postsurgical outcomes. These trials however have not stringently controlled for postoperative glucose. Latham et al. 33 prospectively evaluated BG control before and after surgery in 1044 cardiac surgery patients with and without diabetes. He found postoperative BG levels to be an independent risk factor for SSI in all patients although DM patients had higher rates. Unlike our study, their mean preoperative A1C was not statistically different in those with or without SSI (8.4 vs. 7.8%). Despite this lack of A1C correlation with infection, Latham’s group 33 recommended a screening preoperative A1C to identify patients most likely to develop postoperative hyperglycemia. In another cardiac surgery cohort evaluated by Medhi et al. 34, an A1C greater than 7% predicted a prolonged length of stay after surgery, which was used as a surrogate marker for morbidity. Unlike Medhi, this study did not detect any difference in the length of stay between patients with or without DSWI (P=0.73) and between the A1C stratifications (P=0.18). Alserius et al. 35 found that a preoperative A1C greater than 6% increased the risk of DSWI and mortality up to 3 years after bypass surgery. The practicality of achieving an A1C less than 6% in DM patients is associated with significant hypoglycemia. This approach offers no advantage and can have worse outcomes as seen in recent outpatient trials 29,30. The current ADA recommendations 2 reflect these data and suggest a target A1C less than 7% with minimal hypoglycemia.

Very high A1Cs of greater than 8.6% were found by Halkos et al. 31 to quadruple mortality and also increase the risk of postoperative complications (DSWI, cerebrovascular accidents, and renal failure). We found no association between A1C and mortality (Table 2) and our DSWI rates were similar to other published cardiac surgery trials 14,15,20–23. In a recent study by Knapik et al. 36, 39% of patients undergoing CABG had an A1C greater than 7%. An A1C between 7 and 8% was associated with perioperative myocardial infarction, but not infectious outcomes. Unfortunately for the analysis, their hospital protocol automatically rescheduled patients with A1Cs higher than 8%, corresponding to an estimated average glucose of 183 mg/dl 3, at which level infection risk becomes a significant concern and we had highest rates of DSWIs.

Insulin resistance during surgery is evaluated in cardiac, colorectal, and gall bladder procedures 36–39. The association between intraoperative insulin resistance, preoperative glycemic control, and adverse events after cardiac surgery was investigated by Sato et al. 37 who found insulin resistance during surgery rather than a known diagnosis of DM, increased the risk of complications. In several studies, a high BMI was associated with DSWI but in our study this relationship did not attain statistical significance probably because of the smaller study size which was a subset of the DM population.

As with other retrospective analyses, our study is limited by multiple factors. A large number of patients did not meet inclusion criteria because of missing preoperative A1C data which could result in selection bias. As A1C testing occurred before surgery, the impact of this exclusion should be minimal. The 616 patients who were included could have preoperative medical care and comorbidities potentially confounding the association of A1C and DSWI; therefore, an expansive list of baseline demographic characteristics were analyzed. There also were a relatively small number of DSWIs which could weaken the statistical power to estimate risks after multivariable correction for confounders. For the same reason, all confounders could not be adjusted for in the same model.

Clinical limitations were a change in the antimicrobial prophylaxis administered following cardiothoracic surgeries from vancomycin to the addition of cefuroxime for postoperative prophylaxis. Although preoperative antibiotic prophylaxis or antibiotic washes administered before and during sternotomy were not evaluated, the practices were stable during the study period. The severity of disease before surgery was not recorded, and was assumed to be similar for the entire population.

The strengths of this study are the use of a well-collected and prospective data for all cardiac surgery procedures. To overcome the statistical limitations, we also evaluated baseline characteristics of the large excluded group that did not have a preoperative A1C recorded. Statistically significant differences were lower BMIs, less hypertension and chronic lung diseases, less use of insulin, and better BG control on POD 1 and 2, all characteristics suggesting a lower risk group. The major strength of this study was that postoperative BG management was stringent and used standardized insulin protocols; therefore, no statistically significant differences between BG control were found between patients with and without DSWI (P=0.07). DSWI cases were identified and classified with the assistance of a hospital infection control team using the CDC criteria which complement those defined by The Society of Thoracic Surgeons 40.

The results of this study suggest that DM patients at highest risk of infections can be identified by preoperative A1C levels. Interventions to improve BG before surgery may impact outcomes, even though postoperative BG is controlled to appropriate goals. Therefore optimizing A1C in the available time before cardiac surgery may help improve patient outcomes.

Back to Top | Article Outline


In patients with DM undergoing cardiac surgery, higher A1C levels demonstrated an increased incidence in DSWI. The association of preoperative BG control and postoperative complications in this, and other surgical cohorts needs to be evaluated by prospective clinical trials.

Back to Top | Article Outline

This work utilized the Biostatistics Core of the Michigan Diabetes Research and Training Center funded by DK020572 from the National Institute of Diabetes and Digestive and Kidney Diseases.

Back to Top | Article Outline
Conflicts of interest

Dr Roma Y. Gianchandani is on the Speaker’s Bureau of Sanofi. For the remaining authors there are no conflicts of interest.

Back to Top | Article Outline


1. Lu CM, Nicoll D, McPhee SJ, Pignone M ‘Common laboratory tests’ (chapter). Pocket guide to diagnostic tests, 5e. Available at: [Accessed 5 September 2011]
2. . Standards of medical care in diabetes – 2011. Diabetes Care. 2011;34(Suppl 1):S11–S61
3. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine R. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008;31:1–6
4. Lazar HL, McDonnell M, Chipkin SR, Furnary AP, Engelman RM, Sadhu AR, et al. The Society of Thoracic Surgeons practice guideline series: blood glucose management during adult cardiac surgery. Ann Thorac Surg. 2009;87:663–669
5. Furnary AP, Gao G, Grunkemeier GL, Wu Y, Zerr KJ, Bookin SO, et al. Continuous insulin infusion reduces mortality in patients with diabetes undergoing coronary artery bypass grafting. J Thoracic Cardiovasc Surg. 2003;125:1007–1021
6. Finney S, Zekveld C, Elia A, Evans T. Glucose control and mortality in critically ill patients. JAMA. 2003;290:2041–2047
7. Krinsley J. Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proc. 2003;78:1471–1478
8. Krinsley J. Effect of an intensive glucose management protocol on the mortality of critically ill adult patients. Mayo Clin Proc. 2004;79:992–1000
9. Clement S, Braithwaite SS, Magee MF, Ahmann A, Smith EP, Schafer RG, Hirsch IB. Management of diabetes and hyperglycemia in hospitals. Diabetes Care. 2004;27:553–591
10. Fish L, Weaver T, Moore A, Steel L. Value of postoperative blood glucose in predicting complications and length of stay after coronary artery bypass grafting. Am J Cardiol. 2003;92:74–76
11. Anderson R, Brismar K, Barr G, Ivert T. Effects of cardiopulmonary bypass on glucose homeostasis after coronary artery bypass surgery. Eur J Cardiothorac Surg. 2005;28:425–430
12. Zerr K, Furnary A, Grunkemeier G, Bookin S, Kanhere V, Starr A. Glucose control lowers the risk of wound infection in diabetics after open heart operations. Ann Thorac Surg. 1997;63:356–361
13. Hruska L, Smith J, Hendy M, Fritz V, McAdams S. Continuous insulin infusion reduces infectious complications in diabetics following coronary surgery. J Card Surg. 2005;20:403–407
14. Sjogren J, Malmsjo M, Gustafsson R, Ingemansson R. Poststernotomy mediastinitis: a review of conventional surgical treatments, vacuum-assisted closure therapy and presentation of the Lund university hospital mediastinitis algorithm. Eur J Cardiothorac Surg. 2006;30:898–905
15. Kramer R, Groom R, Weldner D, Gallant P, Heyl B, Knapp R, Arnold A. Glycemic control and reduction of deep sternal wound infection rates. Arch Surg. 2008;143:451–456
16. Borger MA, Rao V, Weisel RD, Ivanov J, Cohen G, Scully HE, David TE. Deep sternal wound infection: risk factors and outcomes. Ann Thorac Surg. 1998;65:1050–1056
17. Loop FD, Lytle BW, Cosgrove DM, Mahfood S, McHenry MC, Goormastic M, et al. Sternal wound complications after isolated coronary artery bypass grafting: early and late mortality, morbidity, and cost of care. Ann Thorac Surg. 1990;49:179–186
18. Crabtree T, Codd J, Fraser V, Bailey M, Olsen M, Damiano R. Multivariate analysis of risk factors for deep sternal infection after coronary artery bypass grafting at a tertiary care medical center. Semin Thorac Cardiovasc Surg. 2004;16:53–61
19. Furnary A, Zerr K, Grunkemeier G, Starr A. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg. 1999;67:352–360
20. Lu J, Grayson A, Jha P, Srinivasan A, Fabri B. Risk factors for sternal wound infection and mid-term survival following coronary artery bypass surgery. Eur J Cardiothorac Surg. 2003;23:943–949
21. Ridderstolpe L, Gill H, Granfeldt H, Ahlfeldt H, Rutberg H. Superficial and deep sternal wound complications: incidence, risk factors and mortality. Eur J Cardiothorac Surg. 2001;20:1168–1175
22. Stahle E, Tammelin A, Bergstrom R, Hambreus A, Nystrom S, Hansson H Eur J Cardiothorac Surg. 1997;11:1146–1153
23. Douville EC, Asaph JW, Dworkin RJ, Handy JR Jr, Canepa CS, Grunkemeier GL, Wu Y. Sternal preservation: a better way to treat most sternal wound complications after cardiac surgery. Ann Thorac Surg. 2005;78:1659–1664
24. Grossi EA, Culliford AT, Krieger KH, Kloth D, Press R, Baumann FG, Spencer FC. A survey of 77 major infectious complications of median sternotomy: a review of 7949 consecutive operative procedures. Ann Thorac Surg. 1985;40:214–223
25. Gordon S, Serkey J, Barr C, Cosgrove D, Potts W Infection Control and Hospital Epidemiology, Vol. 18, No. 5, Part 2. The Seventh Annual Meeting of SHEA, The Society for Hospital Epidemiology of America, April 27-29, 1997 (May, 1997), pp. P1-P64
26. Horan T, Andrus M, Dudeck M. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36:309–332
27. Garner J, Jarvis W, Emori T, Horan T, Hughes J. CDC definitions for nosocomial infections. Am J Infect Control. 1988;16:128–140
28. . The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329:977–986
29. . Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008;358:2545–2559
30. . Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358:2560–2572
31. Halkos ME, Puskas JD, Lattouf OM, Kilgo P, Kerendi F, Song HK, et al. Elevated preoperative hemoglobin A1C level is predictive of adverse events after coronary artery bypass surgery. J Thorac Cardiovasc Surg. 2008;136:631–640
32. McAlister FA, Man J, Bistritz L, Amad H, Tandon P. Diabetes and coronary artery bypass surgery: an examination of perioperative glycemic control and outcomes. Diabetes Care. 2003;26:1518–1524
33. Latham R, Lancaster A, Covington J, Pirolo J, Thomas C. The association of diabetes and glucose control with surgical-site infections among cardiothoracic surgery patients. Infect Control Hosp Epidemiol. 2001;22:607–612
34. Medhi M, Marshall MC Jr, Burke HB, Hasan R, Nayak D, Reed G, et al. HbA1c predicts length of stay in patients admitted for coronary artery bypass surgery. Heart Dis. 2001;3:77–79
35. Alserius T, Anderson R, Hammar N, Nordqvist T, Ivert T. Elevated glycated hemoglobin is a major risk marker in coronary artery bypass surgery. Scand Cardiovasc J. 2008;42:392–398
36. Knapik P, Ciesla D, Filipiak K, Knapick M, Zembala M. Prevalence and clinical significance of elevated preoperative glycosylated hemoglobin in diabetic patients scheduled for coronary artery surgery. Eur J Cardiothorac Surg. 2011;39:484–489
37. Sato H, Carvalho G, Sato T, Lattermann R, Matsukawa T, Schricker T. The association of preoperative glycemic control, intraoperative insulin sensitivity, and outcomes after cardiac surgery. J Clin Endocrinol Metab. 2010;95:4338–4344
38. Gustafsson U, Thorell A, Scoop M, Ljungqvist O, Nygren J. Hemoglobin A1c as a predictor of postoperative hyperglycemia and complications after major colorectal surgery. Br J Surg. 2009;96:1358–1364
39. Thorell A, Nygren J, Essén P, Gutniak M, Loftenius A, Andersson B, Ljungqvist O. The metabolic response to cholecystectomy: insulin resistance after open compared with laparoscopic operation. Eur J Surg. 1996;162:187–191
40. Quality performance measures. 2011. Available at: [Accessed 5 September 2011]

diabetes mellitus; outcomes; preoperative care

© 2013Wolters Kluwer Health Lippincott Williams Wilkins