Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Risk Factors for Surgical Site Infections after Colorectal Resection in Diabetic Patients Rishabh Sehgal, MB, BCh, Arthur Berg, PhD, Rafael Figueroa, MD, Lisa S Poritz, MD, Kevin J McKenna, MD, David B Stewart, MD, Walter A Koltun, MD, FACS Surgical site infections (SSIs) are a well known complication of gastrointestinal surgery and associated with an increased morbidity, mortality and overall cost. Diabetes mellitus (DM) is a risk factor for SSI. However, there is no clear consensus as to which other risk factors play a significant role. The goal of this study was to identify risk factors associated with SSI in patients with DM undergoing colorectal resection. STUDY DESIGN: A retrospective review was conducted of DM patients who underwent colorectal resection from June 2000 to June 2009 at Milton S Hershey Medical Center, Division of Colorectal Surgery. Individual measures were analyzed using chi-square, t-test, and Mann-Whitney U tests, and statistical significance was confirmed using a multiple logistical regression model. RESULTS: There were 183 DM patients included in the study, 28 (15%) of whom developed SSI. Glucose levels were significantly higher in the SSI group for each time interval, 0 to 6 hours (211 mg/dL, p ⫽ 0.03), 0 to 48 hours (176 mg/dL, p ⫽ 0.001), and 48 to 96 hours (167 mg/dL, p ⫽ 0.012) postoperatively. Other measures significantly associated with SSI included the use of drains (p ⫽ 0.05) and the use of prophylactic antibiotics for more than 24 hours (p ⫽ 0.02). Body mass index and stoma creation approached statistical significance (p ⫽ 0.08, 0.07, respectively). The type of hypoglycemic regimen, immunosuppression, and emergency surgery were not associated with an increased rate of SSI. CONCLUSIONS: Higher than normal glucose control at all postoperative time intervals was associated with SSI. The majority of glucose levels were below the American Diabetes Association recommended level of 200 mg/dL, but patients still developed SSI. Type of perioperative glucose control did not affect the incidence of SSI. These data suggest that DM patients undergoing colectomy should have glucose tightly controlled, avoid placement of drains, and receive antibiotics for less than 24 hours. (J Am Coll Surg 2011;212:29–34. © 2011 by the American College of Surgeons) BACKGROUND: Surgical site infections (SSI) pose an ongoing problem in surgical practice. They lead to patient dissatisfaction, delayed wound healing, increased risk of incisional hernias, extended length of hospital stay (LOS), and an increased risk of other nosocomial complications. Furthermore, there is an increased expense associated with SSI. The CDC estimates an additional cost of $2,734 to $26,019 to the American health care system per SSI, with an annual figure estimated between $130 million and $845 million.1 Therefore, reducing the incidence of SSIs is important in improving both the individual patient’s health and reducing the overall cost of health care. Patients undergoing colorectal resection are at an increased risk for developing SSI due to the bacterial load of the colon. The incidence of SSIs after colorectal resection ranges from 3% to 30%.2 A wide variety of risk factors thought to play a role in the pathogenesis of SSIs have been studied. Obesity, steroid use, anemia, bowel preparation, hypothermia, nutritional status, and blood transfusion are some of the factors considered to be associated with development of an SSI. The Center for Medicare and Medicaid Services (CMS) has implemented protocols for prophylactic antibiotic choice, timing of administration, and duration of use aimed at reducing SSI, but as Disclosure Information: Nothing to disclose. Presented at the American College of Surgeons 95th Clinical Congress, Chicago, IL, October 2009. Received July 15, 2010; Revised September 7, 2010; Accepted September 10, 2010. From the Division of Colon and Rectal Surgery, Department of Surgery (Sehgal, Berg, Figueroa, Poritz, McKenna, Stewart, Koltun), and Department of Public Health Sciences (Berg), Penn State Milton S Hershey Medical Center, Hershey, PA. Correspondence address: Rishabh Sehgal, MB, BCh, Division of Colon and Rectal Surgery, Penn State Milton S Hershey Medical Center, PO Box 850, Hershey, PA 17033. email: [email protected] or [email protected] © 2011 by the American College of Surgeons Published by Elsevier Inc. 29 ISSN 1072-7515/11/$36.00 doi:10.1016/j.jamcollsurg.2010.09.011 30 Sehgal et al Colorectal Resection and Surgical Site Infection J Am Coll Surg Table 1. Criteria for Defining a Surgical Site Infection Abbreviations and Acronyms BMI DM LOS SSI ⫽ ⫽ ⫽ ⫽ body mass index diabetes mellitus length of stay surgical site infection of yet it is unclear if these interventions have had a significant impact. Diabetes mellitus (DM) has been shown to play a significant predisposing role for SSIs in many surgical disciplines such as orthopaedic, cardiac, and general surgery.3-6 Poor glycemic control, especially in the immediate postoperative period, has been associated with an increased incidence of SSI.6 The American Diabetes Association (ADA) has recommended that perioperative blood glucose levels be kept below 200 mg/dL to minimize SSI.7 However, in this higher risk group of DM patients, it is unclear what other factors play a role in SSI occurrence. For example, whether severity of DM defined by the need for insulin versus oral agents, or whether prolonged use of prophylactic antibiotics might alter SSI incidence is unknown. The purpose of this study was to investigate other potential risk factors in the development of SSIs in this high risk diabetic group undergoing colorectal resection. Superficial incisional surgical site infection Infection occurs within 30 days after the operation and infection involves only skin or subcutaneous tissue of the incision and at least one of the following: 1. Purulent drainage, with or without laboratory confirmation, from the superficial incision. 2. Organisms isolated from an aseptically obtained culture of fluid or tissue from the superficial incision. 3. At least one of the following signs or symptoms of infection: pain or tenderness, localized swelling, redness, or heat and superficial incision is deliberately opened by surgeon, unless incision is culture-negative. 4. Diagnosis of superficial incisional surgical site infection by the surgeon or attending physician. All patients with DM who underwent colorectal resection (segmental colectomy, total proctocolectomy, anterior resection, or abdominal perineal resection) by the Division of Colon and Rectal Surgery at the Milton S Hershey Medical Center, Penn State College of Medicine, between June 2000 and June 2009, were identified. Patients who underwent a simple colostomy closure without segmental resection and patients who did not have a primary closure of their wound were excluded. Data were collected from patient hospital records, clinic notes, operative records, and electronic medical records. The study was approved by the Milton S Hershey Medical Center Institutional Review Board. did or did not develop an SSI according to the definition set by the CDC (Table 1). Postoperative glucose levels were divided into 3 time intervals. The first postoperative glucose (within 6 hours of operation) was a single value. The second and third time values were a mean of all the glucose readings taken between 0 and 48 hours (including the first glucose) and 48 and 96 hours, respectively. The 0- to 6-hour time period was in part chosen because intraoperative blood sugars were infrequently measured, so this 0- to 6-hour period was effectively the first glucose after operation, so it was the best available measure of intraoperative glucose control. The 0- to 48-hour measurement is thought to be an indicator of early postoperative glucose control and by definition, included the first glucose value as well. Operative factors analyzed included creation of a stoma, approach (open, laparoscopic, including hand assisted, or laparoscopic converted to open), mode of skin closure (staples vs subcutaneous closure), use of intra-abdominal or pelvic drains, preoperative bowel preparation, prophylactic antibiotic use, preoperative steroid or immunosuppression use, evidence of intra-abdominal infection at operation, use of antibiotics longer than 24 hours, requirement for blood products, estimated blood loss, core intraoperative body temperature, and operative time. Data collection Statistical analyses Demographic data collected included age at surgery, sex, diagnosis, procedure performed, length of stay (LOS), readmission for SSI or any other reason, the presence of a preoperative stoma (ileostomy or colostomy), surgical wound classification, American Society of Anesthesiology (ASA) score (I to IV), preoperative hypoglycemic regimen (diet controlled, oral hypoglycemics, or insulin), and body mass index (BMI). Patients were subdivided into those who The R statistical software system (version 2.9.2, http:// www.r-project.org/) was used to perform univariate and multivariate statistical analyses. The Fisher’s exact test, 2-sample t-test, and Mann-Whitney U test were used to test the association of individual predictors with SSI. For the multivariate analysis, logistic regression was performed. Variable selection for consideration in the multivariate logistic regression was based on statistical significance or METHODS Subjects Vol. 212, No. 1, January 2011 Sehgal et al Table 2. Patient Characteristics Not Significantly Associated with Surgical Site Infection on Univariate Analysis Variable Age, y*, (mean ⫾ SE) Sex, n (%) Female Male Diagnosis, n (%) Colorectal cancer Diverticulitis Inflammatory bowel disease Other Preoperative stoma, n (%) Yes No Preoperative hypoglycemic regime, n (%) Diet only Oral hypoglycemics Insulin SSI (n ⴝ 28, 15%) No SSI p (n ⴝ 155, 85%) Value 63.6 ⫾ 12 65.6 ⫾ 12.8 12 (43) 16 (57) 64 (41) 91 (59) 10 (36) 6 (21) 70 (45) 33 (21) 9 (32) 3 (11) 25 (16) 27 (18) 0.15 0.67 0.29 Table 3. Operative Measures Not Statistically Associated with Surgical Site Infection on Univariate Analysis 0.14 6 (4) 149 (96) 0.38 4 (14) 14 (50) 10 (36) 31 was done as an elective or emergent colectomy, the patient’s physical status (ASA grade), and wound classification did not correlate with SSI. The approach to the colectomy (open or laparoscopic) and the mode of skin closure also had no influence on development of SSI (Table 3). Glucose values at all 3 time points were significantly higher in the patients who developed an SSI than those who did not (Table 4, Fig. 1). The largest difference in glucose values between patients with or without SSI was Variable 3 (10) 25 (90) Colorectal Resection and Surgical Site Infection 24 (15) 94 (60) 37 (25) *Statistical analysis performed using Mann–Whitney test. Statistical analysis for remaining variables performed using Fisher’s exact test. SSI, surgical site infection. marginal statistical significance in the univariate analysis while also considering the medical relevance of the variables as potential predictors. RESULTS A total of 191 diabetic patients were identified who underwent colorectal resection from June 2000 to June 2009 at the Milton S Hershey Medical Center by the Division of Colon and Rectal Surgery. Eight patients, none of whom had a wound infection, had no glucose readings recorded in the chart and were therefore excluded. Out of the evaluable 183 patients, 28 (15%) patients developed an SSI. Patient characteristics not significantly associated with SSI on univariate analysis are listed in Table 2 and included age, sex, diagnosis (colorectal cancer, diverticulitis, inflammatory bowel disease, or miscellaneous) and the presence of a preoperative stoma. Interestingly, the type of home diabetic regimen the patient used, ie, diet controlled versus oral hypoglycemics versus insulin, did not have any relationship to the development of an SSI. Operative measures that were not significantly associated with an SSI included the use of perioperative steroids, LOS, the use of blood products, estimated blood loss, average core body temperature, and the use of preoperative bowel preparation. Furthermore, whether the procedure SSI (n ⴝ 28) No SSI (n ⴝ 155) Perioperative steroids, % Yes 14 13 No 86 87 Length of stay, d*, (mean ⫾ SD) 8.03 ⫾ 3.75 7.58 ⫾ 4.93 Blood transfusion, % Yes 27 19 No 73 81 Estimated blood loss, mL*, (mean ⫾ SD) 388 ⫾ 445 290 ⫾ 353 Core body temperature, oC*, (mean ⫾ SD) 36.15 ⫾ 0.76 35.96 ⫾ 1.31 Length of operation, min*, (mean ⫾ SD) 161 ⫾ 81 164 ⫾ 83 Bowel preparation, % Yes 67 66 No 33 34 Elective, % 81 82 Emergency, % 19 18 ASA grade, % II 32 32 III 61 58 IV 7 10 Wound classification,% II 93 91 IV 7 (III–IV) 9 Side of colectomy, % Left 79 74 Right 21 26 Approach, % Open 89 87 Laparoscopic to open 7 2 Laparoscopic 4 11 Skin closure Clips 78 85 Sutures 10 8 p Value 0.77 0.55 0.28 0.11 0.35 0.77 1 0.43 1 0.72 0.75 0.25 0.33 *Statistical analysis performed using Mann–Whitney test. Statistical analysis for remaining variables performed using Fisher’s exact test. ASA, American Society of Anesthesiologists; SSI, surgical site infection. 32 Sehgal et al Colorectal Resection and Surgical Site Infection J Am Coll Surg Table 4. Glucose Measurements (Normal 70 to 120 mg/dL) Time interval, h 0–6 0 – 48 48 – 96 SSI (n ⴝ 28) No SSI (n ⴝ 155) 211.11 ⫾ 13.62 g/dL 182.39 ⫾ 4.34 g/dL 176.77 ⫾ 7.41 g/dL 160.44 ⫾ 2.64 g/dL 167.85 ⫾ 5.03 g/dL 155.50 ⫾ 2.43 g/dL p Value 0.03 0.001 0.012 Glucose measurements are mean ⫾ SE. Statistical analysis performed using Mann–Whitney test. SSI, surgical site infection. seen with the immediate postoperative glucose. Also, this immediate postoperative glucose was the highest glucose in both the SSI and non-SSI patient groups, with the highest value (211 mg/dL) seen in the SSI group. Measures other than glucose that were significantly associated with SSI on univariate analysis included higher BMI, use of prophylactic antibiotics greater than 24 hours, stoma creation and the use of pelvic or intra-abdominal drains (Table 5). When these significant univariate factors were put into a multiple logistical regression model, glucose levels, the use of drains, and prophylactic antibiotics for more than 24 hours remained significantly associated with SSI; BMI and stoma creation missed absolute statistical significance (Table 6). DISCUSSION The aim of this study was to identify potential risk factors for SSI in diabetic patients undergoing colorectal resection. Our results concur with those from other studies in the vascular, cardiothoracic, and orthopaedic literature that show high perioperative glucose levels to be associated with an increased risk of developing SSI.6 The American Diabetes Association recommends glucose levels be less than 200 mg/dL perioperatively in order to avoid SSI. However, in 0.006 our cohort, most of the glucose levels measured were below 200 mg/dL in the patients who still developed SSIs. By dividing our patients’ glucose measurements into 3 time intervals, we sought to identify if there was a specific time frame during which hyperglycemia may be more critical in the development of an SSI. In both groups of our patients, those with and without SSI, the immediate postoperative blood glucose was the highest of the 3 time points measured, suggesting this may be a key time point for intervention. Other studies in cardiac patients have also shown that elevated intraoperative blood glucose is associated with SSI.8 Our data suggest, however, that the American Diabetes Association recommendation of 200 mg/dL may still be 0.003 Table 6. Significant Factors Associated with Surgical Site Infection on Multivariate Analysis Table 5. Significant Factors Associated with Surgical Site Infection on Univariate Analysis Factors BMI, kg/m2*, (mean ⫾ SD) Prophylactic antibiotics used ⬎24 h, % Yes No Stoma creation, % Yes No Drains, % Yes No SSI (n ⴝ 28) No SSI (n ⴝ 155) p Value 35 ⫾ 8.6 32 ⫾ 16.5 0.03 70 23 45 55 67 33 Measurement 41 59 0.04 60 40 Figure 1. Postoperative glucose values. The following blood glucose values were recorded: the first postoperative glucose (0 to 6 hours), all glucose between 0 and 48 hours and all glucose between 48 and 96 hours. Patients with and without surgical site infection (SSI) were compared. At all time points the mean glucose was significantly higher in the patients with an SSI. Orange bar, SSI; green bar, no SSI. 42 58 *Statistical analysis performed using Mann–Whitney test. Statistical analysis for remaining variables performed using Fisher’s exact test. BMI, body mass index; SSI, surgical site infection. Glucose Drains Antibiotic ⬎24 h Body mass index, kg/m2 Stoma creation p Value 0.02 0.05 0.02 0.08 0.07 Logistic regression using enter method and incorporating all variables with p⬍0.1 on univariate analysis. Vol. 212, No. 1, January 2011 Sehgal et al too high, and lower blood sugars should be sought to minimize SSI, because the majority of the measured glucose levels were below 200 mg/dL in those who developed SSIs. The majority of the studies in the literature that have looked at perioperative glucose management have been done primarily in the critical care and cardiac surgical setting. In 2003, Furnary and colleagues8 published a study that looked at 3,554 diabetic patients undergoing coronary artery bypass grafting. All patients were treated with either intraoperative subcutaneous insulin or with continuous insulin infusion for hyperglycemic control. Observed mortality with continuous insulin infusion was significantly lower than with subcutaneous insulin. Furthermore, better postoperative glucose control was also shown to decrease wound infections.8,9 Data from the critical care and cardiac literature may be able to be extrapolated to other types of surgery, but large studies looking at perioperative glycemic control in patients with diabetes outside of the critical care and cardiac surgical realm are still limited. One might postulate that the more severe the diabetes the higher the risk of SSI. The type of hypoglycemic regimen can be used as a surrogate for disease severity. One can suggest that diet controlled patients might have the least severe diabetes, those on oral hypoglycemic medications have intermediate DM severity, and those with insulin dependent diabetes have the most severe category of disease. Using these severity categories, our data did not show a significant difference between these groups in the risk of developing an SSI. Ideally, hemoglobin A1c levels could be used to assess longer term glycemic control. However due to the retrospective nature of this study these values were unavailable. It appears, however, that absolute perioperative glucose control is more important than severity of DM as reflected by hypoglycemic regimen. It can be argued that elevated glucose may be an early sign of postoperative infection as opposed to a causative factor. Certainly it is possible that during our 48- to 96hour time point, patients who would ultimately get an SSI may have already started to develop such an infection that could have been reflected in increased blood sugars. However, the highest blood glucose levels were found in the immediate postoperative period within 6 hours of operation, which would be very unlikely to have been caused by a wound infection at such an early time point. Apart from glucose control, other measures that were significantly associated with development of SSIs were the use of pelvic or intra-abdominal drains and the use of prophylactic broad-spectrum antibiotics for longer than 24 hours. Prolonged use of both of these measures has been shown to be associated with SSI in other patient populations.10,11 Both prolonged antibiotic use and placement of Colorectal Resection and Surgical Site Infection 33 drains may have occurred because there were intraoperative findings that made the operating surgeon concerned about subsequent infection, suggesting that it was the intraoperative finding, as opposed to the drain or antibiotics themselves, that increased the SSI risk. However, in the SSI group of patients with drains, only 1 of 18 of those drains was placed in an abscess cavity. Therefore, the drains in our patients were primarily placed for reasons other than obvious findings of infection at the time of operation. The overuse of drains may increase the risk of SSI due to retrograde contamination along the drain, creating a pathway for subcutaneous soft tissue contamination. All of our patients received appropriate anerobic and gram-negative antibiotic coverage on induction of anesthesia. Using a single dose of preoperative antibiotic is considered acceptable in operations lasting up to 3 hours and without gross soilage during the operation.10 However, use of antibiotics beyond 24 hours can be indicated in more complicated operative procedures in which infection is found. Increasing the length of operation has also been shown to be a risk factor for SSI.10,12 In theory, prolonging the length of the operation could potentially increase bacterial exposure to the wound and reduce efficacy of the antibiotic within the tissues, suggesting the need for prolonged antibiotics.10 However, in our study, the duration of the procedure was not a significant variable relating to SSI, although this could be due to the somewhat limited sample size. Obesity not only makes the procedure technically more challenging and potentially longer, but there is also an increased amount of fat or dead space in the wound that may lead to an increased risk of developing a wound infection. A study conducted by Itani and associates13 looking at the association of BMI and the type of antibiotic prophylaxis, revealed that in patients undergoing elective colorectal surgery there was a higher incidence of SSI in patients with BMI ⱖ 30 kg/m2 irrespective of the type of prophylactic antibiotic used. Obesity is already a well known risk factor for development of SSI.14,15 In our study, increased BMI was significant on univariate analysis and approached significance on multivariate analysis. This suggests that weight loss in the obese patient before purely elective procedures may improve overall SSI risk in the DM patient. Creation of a stoma was also significantly associated with the development of SSI in this study. These results echo several results from other similar studies.16 Creation of a stoma may increase the risk of developing an SSI for 2 reasons. First, the stoma may have been created because of intraoperative findings that could lead to an increased risk of infection, and second, the potential for enteric contamination of a fresh abdominal wound from the exposed 34 Sehgal et al Colorectal Resection and Surgical Site Infection bowel lumen. Stoma creation should be used judiciously, and bowel opened or matured only after the incision is closed and protected from possible contamination. This study has some limitations. Though suggestive, tighter glucose control was not definitively shown to prevent wound infection rates. Larger multicenter prospective trials will be necessary to prove this, but would be worthwhile in order to obtain data that can then be used in quality control programs. However, based on our findings, the American Diabetes Association recommendations of glucose levels being kept less than 200 mg/dL may still be too high in DM patients undergoing colorectal resection. Tighter control of perioperative blood glucose levels may lead to lower SSI rates. To summarize, in diabetic patients undergoing colorectal surgery, to minimize wound infections, prophylactic antibiotics should be stopped within 24 hours, and the use of drains should be kept to a minimum. Weight loss before purely elective colorectal surgery may also improve SSI risk. Stomas should be used only when necessary and tight perioperative blood glucose controls may improve the risk of developing SSI. 4. 5. 6. 7. 8. 9. 10. 11. Author Contributions Study conception and design: Sehgal, Figueroa, Poritz, Koltun Acquisition of data: Sehgal, Figueroa Analysis and interpretation of data: Berg Drafting of manuscript: Sehgal, Poritz, Koltun Critical revision: Stewart, McKenna, Poritz, Koltun REFERENCES 1. Center of Disease Control and Prevention. Surgical site infections. Available at: http://www.cdc.gov/ncidod/dhqp/FAQ_ SSI.html. Accessed October 10, 2010. 2. Smith RL, Bohl JK, McElearney ST, et al. Wound infection after elective colorectal resection. Ann Surg 2004;239:599–605; discussion 605–607. 3. Chen S, Anderson MV, Cheng WK, Wongworawat MD. Dia- 12. 13. 14. 15. 16. J Am Coll Surg betes associated with increased surgical site infections in spinal arthrodesis. Clin Orthop Relat Res 2009;467:1670–1673. Ogihara H, Takeuchi K, Majima Y. Risk factors of postoperative infection in head and neck surgery. Auris Nasus Larynx 2009; 36:457–460. Trussell J, Gerkin R, Coates B, et al. Impact of a patient care pathway protocol on surgical site infection rates in cardiothoracic surgery patients. Am J Surg 2008;196:883–889. McConnell YJ, Johnson PM, Porter GA. Surgical site infections following colorectal surgery in patients with diabetes: association with postoperative hyperglycemia. J Gastrointest Surg 2009;13:508–515. American Diabetes Association. Standards of medical care in diabetes–2008. Diabetes Care 2008;31:S12–54. Furnary AP, Gao G, Grunkemeier GL, et al. Continuous insulin infusion reduces mortality in patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg 2003; 125:1007–1021. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care1999;22: 1408–1414. Tang R, Chen HH, Wang YL, et al. Risk factors for surgical site infection after elective resection of the colon and rectum: a single-center prospective study of 2,809 consecutive patients. Ann Surg 2001;234:181–189. Drapeau CM, D’Aniello C, Brafa A; Italian Group on Surgical Infections in Plastic Surgery, et al. Surgical site infections in plastic surgery: an Italian multicenter study. J Surg Res 2007; 143:393–397. Cruse PJ, Foord R. The epidemiology of wound infection. A 10-year prospective study of 62,939 wounds. Surg Clin North Am 1980;60:27–40. Itani KM, Jensen EH, Finn TS, et al. Effect of body mass index and ertapenem versus cefotetan prophylaxis on surgical site infection in elective colorectal surgery. Surg Infect (Larchmt) 2008;9:131–137. Lynch RJ, Ranney DN, Shijie C, et al. Obesity, surgical site infection, and outcome following renal transplantation. Ann Surg 2009;250:1014–1020. Mehrabi A, Fonouni H, Wente M, et al. Wound complications following kidney and liver transplantation. Clin Transplant 2006;20:97–110. KonishiT, WatanabeT, Kishimoto J, Nagawa H. Elective colon and rectal surgery differ in risk factors for wound infection: results of prospective surveillance. Ann Surg 2006;244:758–763.