Risk factors for Staphylococcus aureus surgical site infections following breast operations

Transcription

1 University of Iowa Iowa Research Online Theses and Dissertations Spring 2016 Risk factors for Staphylococcus aureus surgical site infections following breast operations Elaina Rose O'Neill University of Iowa Copyright 2016 Elaina O'Neill This thesis is available at Iowa Research Online: Recommended Citation O'Neill, Elaina Rose. "Risk factors for Staphylococcus aureus surgical site infections following breast operations." MS (Master of Science) thesis, University of Iowa, Follow this and additional works at: Part of the Clinical Epidemiology Commons

2 RISK FACTORS FOR STAPHYLOCOCCUS AUREUS SURGICAL SITE INFECTIONS FOLLOWING BREAST OPERATIONS by Elaina Rose O Neill A thesis submitted in partial fulfillment of the requirements for the Master of Science degree in Epidemiology in the Graduate College of The University of Iowa May 2016 Thesis Supervisor: Professor Loreen Herwaldt

3 Graduate College The University of Iowa Iowa City, Iowa CERTIFICATE OF APPROVAL This is to certify that the Master's thesis of MASTER'S THESIS Elaina Rose O Neill has been approved by the Examining Committee for the thesis requirement for the Master of Science degree in Epidemiology at the May 2016 graduation. Thesis Committee: Loreen Herwaldt, Thesis Supervisor Christine Petersen Elizabeth Chrischilles

4 ACKNOWLEDGEMENTS I would like to thank my committee, Dr. Loreen Herwaldt, Dr. Christine Petersen, and Dr. Elizabeth Chrischilles for their guidance and above all patience. I would like to acknowledge Drs. Ingrid Lizarraga, Sonia Sugg, Jerrod Keith, and Margaret Olsen for their input on this project. I would like to thank Jean Pottinger and Barabara Zilles for providing a list of patients who had breast operations and a list of patients with surgical site infections. A big thank you to Kim Williams for reminding me of deadlines, keeping me on track and getting me to graduation. ii

5 ABSTRACT Background. Surgical site infections (SSIs) cause many adverse outcomes for patients including increased length of hospital stay, hospital costs, morbidity, and psychological distresses. Staphylococcus aureus is one of the most common causes of SSIs in the United States. Objective. Identify risk factors for Staphylococcus aureus SSIs following breast operations. Design. Retrospective nested case-control study of SSIs among women undergoing breast operations. Setting. An academic health center. Patients. We studied patients undergoing breast operations at the University of Iowa Hospitals and Clinics from 7/1/2004 through 9/30/2015. Cases were patients who acquired SSIs meeting the National Healthcare Safety Network definition and whose SSIs were caused by S. aureus. We randomly selected two controls for each case from patients who had breast operation during the study period and did not meet the SSI definition. Controls were selected randomly from uninfected patients whose operations occurred during the same month and year as a case. Results. Forty two (1.2%) patients acquired S. aureus SSIs after 3494 breast operations. SSIs were identified a mean of 27.8 days after the breast operations; 54.76% were deep incisional infections. Poisson regression analysis revealed that S. aureus SSIs following breast operations at UIHC have been increasing at a statistically significant rate. Bivariable analysis identified several patient and procedure related risk factors that increased the risk for S. aureus SSIs. Patient-related factors included a diabetes mellitus, iii

6 active skin disease, prior chemotherapy, breast cancer, hypertension, and preoperative hemoglobin. Procedure-related factors included ASA score > 2, a mastectomy followed by immediate reconstruction, sentinel lymph node biopsy (SLN), drain placement, procedure time, and estimated blood loss. A multivariable analysis of patient factors found only breast cancer maintained significance. A similar analysis of procedure factors found that drain placement remained significant. The combined model contained breast cancer, drain placement, and mastectomy followed by immediate reconstruction as significant variables. Conclusions. S. aureus SSIs following breast operations have been increasing at UIHC. Possible remediable risk factors include blood glucose levels, blood pressure, timing of chemotherapy, and drain placement and care. These results will help doctors at UIHC design interventions to prevent S. aureus SSIs following these procedures. iv

7 PUBLIC ABSTRACT This study examined risk factors for infections caused by the staph bacteria following breast operations. Patient factors that increased the infection risk included a diagnosis of diabetes, high blood pressure, chemotherapy before the procedure, and breast cancer diagnosis. Procedural factors that increased the risk of infection included having a mastectomy followed by immediate reconstruction and drain placement. Patients who had infections of their operative sites were compared with patients that did not acquire infections. Patient and procedural risk factors affected a patient s risk of infection. Diabetes, high blood pressure, chemotherapy before the procedure, breast cancer diagnosis, mastectomy with immediate reconstruction, and drain placement were all factors that contributed to the risk of SSI following breast operations at UIHC. v

8 TABLE OF CONTENTS LIST OF TABLES... vii LIST OF FIGURES... viii CHAPTER I... 1 INTRODUCTION... 1 Background... 1 Objective... 4 CHAPTER II... 5 METHODS... 5 Study Design... 5 Statistical Analysis... 5 Population... 6 Definition of Key Variables... 6 CHAPTER III... 8 RESULTS... 8 Surgical Site Infections and Other Wound Complications... 8 Bivariable Analyses... 8 Multivariable Logistic Regression... 9 Rate Analyses CHAPTER IV DISCUSSION Overall SSI and S. aureus Rate Analysis Diabetes and Hypertension Breast Cancer and Prior Chemotherapy Use Mastectomy and Immediate Reconstruction CHAPTER V CONCLUSION APPENDIX REFERENCES vi

9 LIST OF TABLES Table A.1: Data Dictionary Table A.2: ICD-9-CM Procedure Codes for Breast Operations Table A.3: NHSN Criteria for SSI Classification Table A.4 Distribution of Potential Risk Factors for Surgical Site Infection Following Breast Operations at UIHC from 7/1/2004 to 9/15/ Table A.5: Risk of SSI following other Wound Complications after Breast Operations at UIHC between 7/1/2004 and 9/30/ Table A.6: Bivariable Analysis of Patient-related Risk Factors for S. aureus SSI following Breast Operations at UIHC from 7/1/2004 to 9/30/ Table A.7: Bivariable Analysis of Procedure-related Risk Factors for S. aureus SSI following Breast Operations at UIHC from 7/1/2004 to 9/30/ Table A.8: Multivariable Analysis of Patient and Procedure related Factors Associated with S. aureus Surgical Site Infections after Breast Operations at UIHC from 7/1/2004 to 9/30/ Table A.9: Poisson Regression Analysis of the Overall SSI Rate and the S. aureus SSI Rate following Breast Operations at UIHC vii

10 LIST OF FIGURES Figure A.1: Overall SSI Rates during the Study Period Figure A.2: S. aureusssi Rates during the Study Period viii

11 CHAPTER I INTRODUCTION Background In 2011, an estimated 722,000 healthcare-associated infections (HAIs) were identified in acute care hospitals in the United States (U.S.). 1 HAIs cause almost 100,000 deaths per year and cost the U.S. healthcare system upwards of $30 billion per year. 2 SSIs - infections that affect the superficial incision, the deep incision, or the deep tissues of the surgical site 3 - are the most common HAIs, accounting for almost 22% of all HAIs in the U.S. 2 These infections increase the length of hospital stays and healthcare costs. They also increase morbidity, psychological distress, and delay other therapies. Breast operations are defined as excision of a lesion or tissue of the breast including radical mastectomy, modified mastectomy, or quadrant resection, lumpectomy, incisional biopsy, or mammoplasty. 4 These operations are clean procedures, which means the operative wound has no evidence of inflammation and the procedure did not enter the respiratory, alimentary, genital, or urinary tracts. 4 The Centers for Disease Control and Prevention (CDC) reports SSI rates of about 2% following breast operations. 5 However, reports from individual hospitals have indicated that SSI rates after these procedures vary from 1% to 28%. 6 One study conducted at Barnes-Jewish Hospital between 1999 and 2002 found an SSI rate of 5.3% after breast procedures. 6 Patients that contract SSIs incur higher hospital costs and have longer hospital stays, with the median cost of an SSI after a breast operation being $10,750 per patient. 6 Xue et al. conducted a meta-analysis of possible patient-related and procedure-related risk factors for SSI after breast operations. 9 They found that older age (odds ratio [OR] 1.73; 95% confidence interval [CI], 1.41 to 2.12), hypertension (OR 1.69; 95% CI, 1.34 to 2.14), higher body mass index (BMI) (OR 1.96; 95% CI, 1.63 to 2.37), diabetes mellitus (OR 1.88; 95% CI, 1.47 to 2.39) and an American Society of Anesthesiologists (ASA) score of 3 or 4 (i.e., the patient had a severe systemic disease that could possibly threaten her life; OR 2.06; 95% CI, 1.24 to 3.40) increased the risk of SSI. 9 Of note, most previously identified risk factors for SSI after breast operations are not remediable. Procedure-related factors associated with SSIs: SSI rates vary with the type of procedure. Patients undergoing breast conserving operations are less likely to acquire SSIs than patients undergoing radical procedures. 7 Patients having operations to remove cancerous lesions from the 1

12 breast are at higher risk of SSI than women having operations for other reasons, such as breast reduction. 6 A study conducted using the HealthCore Integrated Research Database identified an overall SSI rate of 8.1%. Another study conducted by the same researchers found that incidence of SSI increased significantly with subsequent operations. 9et Patients who underwent bilateral procedures (SSI rate 6.1%) were at increased risk of SSIs compared with those who underwent unilateral procedures (SSI rate 4.7%). Furthermore, patients who had mastectomies and reconstructions during the same procedures (SSI rate 10.1 %) were twice as likely to acquire SSIs as patients who did not have immediate reconstructions (SSI rate 5%). 8 Olsen et al. also found that a suboptimal dose of prophylactic antibiotics (OR 2.5; 95% CI, 1.4 to 4.5) increased the risk of SSI. 7 The timing of reconstruction following mastectomy and the dose of the prophylactic antibiotic could be modified. Patient-related factors associated with SSIs: Several researchers have found that BMI (> 30) and diabetes mellitus are associated with SSI. Olsen et al. found an association with an ASA physical status classification > 2 (OR 2.3; 95% CI, 1.2 to 4.2), a BMI > 30 (OR, 2.3; 95% CI, 1.3 to 4.1), 6,7 and diabetes mellitus (OR 3.0; 95% CI, 1.5 to 6.0) 6 and SSI. 6 Preoperative blood glucose levels were not associated with increased SSI risk, but postoperative blood glucose levels during the first five post-operative days were significantly higher in patients with SSIs compared with control patients. However, only 27% of patients had postoperative glucose testing performed. 7 Angarita et al. identified diabetes mellitus, current smoking status, and active skin disorder as risk factors for SSI 22 while Vilar-Compte et al. identified hypertension as a risk factor for SSI (OR 1.6; 95% CI, 1.3 to 2.1). Olsen et al. found no association between current smoking status (OR 1.6; 95% CI, 0.8 to 3.3) or skin disorders (OR 1.8; 95% CI, 0.8 to 3.8) and SSI. Perioperative blood glucose levels and hypertension could also be addressed by treating patients with insulin and blood pressure lowering medications, respectively, during the peri-operative period. Thus, if additional studies indicate that blood glucose levels or hypertension are associated with SSIs after breast operations, clinical trials assessing glucose and hypertension control would be warranted. S. aureus causes about 20% of all SSIs and the number of SSIs caused by this organism has been increasing. 13 In addition, S. aureus is a common cause of SSI after breast operations. Studies by Vilar-Compte et al. and Omar et al. found that S. aureus caused 19% and 40%, respectively, of SSIs following breast cancer operations. 19,20 Mukhtar et al. found that 51% of 2

13 SSIs following breast operations were caused by gram-positive organisms, the most common of which was S. aureus. In fact, S. aureus caused 48.4% of these gram-positive SSIs. 20 Moreover, the investigators found that the percentage of gram-positive SSIs caused by S. aureus increased from 28% between 1997 and 2002 to 49% between 2003 and These infections are also common among patients undergoing breast operations at the UIHC. A recent review of S. aureus SSIs at the UIHC from 1/1/2013 through 8/31/2015 found that 23 patients acquired S. aureus SSIs after breast operations, which was more than double the number after hip operations (n = 10), the procedure with the next highest number of S. aureus SSIs. Patients that carry high numbers of S. aureus in their nares are 3 to 6 times more likely to acquire S. aureus HAIs than patients that carry low numbers of S. aureus or that do not carry this organism Thus, numerous groups have assessed whether decolonizing carriers decreases the risk of S. aureus SSIs. Bode et al. assessed the effect of mupirocin and chlorhexidine bathing compared with placebo on the rate of S. aureus HAIs, in general, and SSIs, in particular. The cumulative incidence of S. aureus SSI in the intervention arm (3.4%) was significantly lower than in the placebo group (7.7%; relative risk [RR], 0.42; 95% CI, 0.23 to 0.75). 16 Deep S. aureus SSIs were also less frequent in the intervention group (RR 0.21; 95% CI, 0.07 to 0.62). 16 More recently, Schweizer et al. assessed the use of twice daily intranasal mupirocin application and once daily chlorhexidine bathing for up to five days before their operations for patients colonized with methicillin-resistant S. aureus (MRSA) or methicillin-susceptible S. aureus (MSSA) and cefazolin and vancomycin as perioperative prophylaxis for patients who carried MRSA rather than vancomycin alone. Patients who did not carry either MRSA or MSSA were given chlorhexidine baths the night before and the morning of their operations. A logistic regression model that controlled for age, diabetes, Charlson comorbidity index, and MRSA history found that the bundle was associated with a significant reduction in complex S. aureus SSIs (OR 0.6; 95% CI, 0.37 to 0.981). 17 The study also found that monthly rates of complex S. aureus SSIs decreased significantly from 36 to 21 per 10,000 operations (RR 0.58; 95% CI, 0.37 to 0.92). 17 Thus, screening patients for S. aureus carriage before patients undergo surgical procedures and treating carriers can decrease the incidence of SSIs and, thereby, decrease adverse outcomes associated with these infections, such as longer lengths of hospital stay. We did not identify any studies among patients undergoing breast operations that evaluated screening patients for S. aureuscarriage and decolonizing carriers before their operations. 3

14 Patients who acquire SSIs suffer various adverse outcomes including longer hospital stays, increased likelihood of readmission, and increased hospital costs. Despite the frequency with which S. aureus causes SSIs after breast operations, we did not identify any studies that assessed risk factors for SSIs caused by this organism and the study by Mukhtar et al. 20 was the only study that assessed the frequency of these infections over time. Given the frequency of these operations, the frequency of SSIs after these procedures, and the costs and other adverse outcomes associated with these SSIs, we need to identify remediable risk factors so that we can identify interventions that will decrease the risk of SSIs after breast operations. Thus, we designed this study to identify risk factors for S. aureus SSIs after breast operations, outcomes of these infections, and trends in the rates of these infections at the UIHC. Objective My primary objective was to study patients undergoing breast operations between July 1, 2004 and September 30, 2015 and to identify remediable patient and procedural risk factors for S. aureusssi after breast operations at the UIHC. Specific Aim 1) To assess patient factors and surgical factors that increase the risk for S. aureusssis following breast operations and to identify outcomes of these SSIs that occur within 90 days of the breast operations, including postoperative length of stay, readmissions, length of stay during readmissions, reoperations, and clinic visits during which antimicrobial agents were prescribed. I hypothesized that higher BMI, diabetes mellitus, hypertension, and an ASA physical status classification of 2 or greater would be significant risk factors of S. aureus SSIs, and that patients who contracted SSI would have longer lengths of stay and would be more likely to be readmitted and to undergo reoperations. Specific Aim 2) To determine the percentage and rates of SSIs following breast operations that are attributable to S. aureus and assess whether these measures have changed over the study period. I hypothesized that the percentage of SSIs caused by S. aureus and the rates of SSIs caused by this organism have increased over the study period. 4

15 CHAPTER II METHODS Study Design This retrospective nested case-control study assessed risk factors for S. aureus SSI after breast operations in patients undergoing these procedures at the UIHC between July 1, 2004 and September 30, Medical records were reviewed and data on potential risk factors were collected and entered into Microsoft Access. The University of Iowa s Institutional Review Board approved this study. Statistical Analysis Standard descriptive statistics including frequencies, percentages, means, medians, and CIs were calculated to evaluate the distributions of potential risk factors for cases and controls. The chi-square test and the Fisher s exact test were used to compare the results of dichotomous variables for cases and controls. A student s t-test or a Wilcoxon rank- sum test was used to compare the results of continuous variables. Bivariate analyses were done to test the association between each potential risk factor and S. aureus SSIs and the association between S. aureus SSIs and each outcome. To identify factors significantly associated with S. aureus SSIs all variables with a p < 0.15 in the bivariate analyses or with a priori clinical significance into a multivariate logistic regression model. Backward elimination was used to identify factors for three final multivariate models: a patient model, a procedure model, and a combined model. Factors with p, 0.15 were included in the models. The quantitative variables were examined for multicollinearity using variance inflation factors (VIFs). A c-statistic was computed to assess the predictive ability of the final model and a Hosmer-Lemeshow goodness-of-fit test was used to assess the fit of the model. The number of SSIs caused by S. aureus per month was divided by the number of operative procedures per month to get the monthly S. aureus SSI rates. The monthly rates of all SSIs were calculated by dividing the number of SSIs caused by any organism per month by the number of operative procedures per month. The trends in rates of SSIs caused by S. aureus and rates of SSIs caused by any organism over the study period were assessed using a Poisson regression model with a log link and with log-transformed monthly procedure counts as the offset variable. 5

16 Population The study population included all female patients who underwent breast operations at UIHC between July 1, 2004 and September 30, A list of all patients undergoing breast operations at the UIHC during the study period was generated using ICD-9 procedure codes (Table 2) for breast operations. For the time period from July 1, 2004 through June 30, 2013, the lists of patients with SSIs after breast operations and of patients undergoing breast operations was obtained from Epic through Clarity to Hedi. For the time period from July 1, 2013 through September 30, 2015, the list patients with SSIs and the list of patients undergoing breast operations were obtained from Epic through Clarity to TheraDoc. A case was defined as any female patient that underwent a breast operation at the UIHC between July 1, 2004 and September 30, 2015 and who met the National Health and Safety Network (NHSN) definition of SSI and the SSI was caused by S. aureus. A control was defined as any female patient that underwent a breast operation at the UIHC between July 1, 2004 and September 30, 2015 and did not meet the NHSN definition of SSI or of SSI caused by S. aureus. Forty-four patients with S. aureus SSIs after breast operations were identified by the Program of Hospital Epidemiology (PHE) at UIHC during the study period. One potential case was eliminated because the patient was male. Two controls were selected randomly from uninfected patients whose operations occurred during the same month and year as a case, to achieve a balance in temporality. No potential controls met the case definition. Definition of Key Variables SSIs were identified based on criteria established by the NHSN (Table 3). A seroma was defined as: 1) accumulation of fluid in the tissue and/or 2) described as a seroma by the physician. A hematoma was defined as: 1) swelling of clotted blood within the tissues and/or 2) described as a hematoma by the physician. Dehiscence was defined as separation of the wound. A stitch abscess was defined as an abscess (an enclosed collection of liquefied tissue) that developed around stitch or suture. Flap necrosis was defined based on a physician diagnosis in the clinical notes. Other wound complications were defined based on descriptions in physicians notes. Previous breast operations were defined as occurring at the same site or at a different site. The same site was defined as being on the same breast while a different site was defined as being 6

17 on the other breast. Previous bilateral operations were defined as a same site operation. The type of breast operation was defined based on the surgical procedure notes. Preoperative infection was defined as an infection caused by any organism occurring at a site other than the operation site within 30 days before the operation. Postoperative infection at a site other than the surgical incision was defined as an infection caused by any organism within 30 days after the operation. Prophylactic antimicrobial agents were given before and, possibly during, the procedure to reduce the risk of infection. The agents given were: cefazolin, clindamycin, and vancomycin. 7

18 CHAPTER III RESULTS Surgical Site Infections and Other Wound Complications 84 patients (2.4%) acquired SSIs and 42 patients (1.1%) acquired S. aureus SSIs after 3494 breast operations. The average number of days to SSI was Of the cases, 33 (78.6%) had to be readmitted for treatment of the SSI and 20 (47.6%) had to be re-operated on. Reoperations included debridement of wounds, implant revisions, and implant removals. 17 cases received an implant during their procedure and 14 (82.4%) had to have the implant removed due to S. aureus SSI. Of the 42 cases, 23 (54.76%) had deep SSIs and 19 (45.24%) had superficial SSIs. Of the cases, 16 (38.10%) had seromas, 8 (19.05%) had wound dehiscence, 4 (9.52%) had flap necrosis, 2 (4.76%) had hematomas, and 2 (4.76%) had stitch abscesses. The average time to SSI was days. All cases (100%) had cultures obtained from their wounds or from wound drainage and cultures from all cases grew S. aureus. In addition, cultures from five (11.9%) cases grew other organisms: Acinetobacter species (sp.) (n = 1), Enterococcus sp. (n=1), Peptostreptococcus sp. (n=1), Prevotella sp. (n=1), Enterobacter cloacae (n=1), Enterococcus faecalis (n=1), Escherichia coli (n=1), and Serratia marcescens (n=1). Bivariable Analyses Cases and controls did not differ significantly with respect to age, weight, BMI, current cigarette use, past cigarette use, any cigarette use, menopausal status, active skin disorders, and blood pressure (Table A.6). Compared with controls, cases were more likely to have a diagnosis of diabetes mellitus, to receive chemotherapy before their procedures, and to have a diagnosis of breast cancer. Bivariable analyses of surgical factors found that cases and controls were similar with respect to: skin preparation, ASA score, laterality of the operation, and having a reduction or augmentation operation. Compared with controls, cases were more likely to have a SLN biopsy, a mastectomy followed by immediate reconstruction with an implant, have a skin-sparing mastectomy, have drains placed during the procedure, have greater blood loss during their procedure, and have longer operation times (Table A.7). 8

19 Multivariable Logistic Regression Three logistic models were built to further evaluate the association between patient and surgical factors with SSIs. (Table A.8). The patient model initially included all patient-related variables that were significant (p < 0.05) or approached significance (p < 0.1) in the bivariable analyses. Preoperative hemoglobin and hypertension were excluded from the logistic regression analysis because both variables had many missing values. Only previous breast cancer diagnosis (p = ) remained statistically significant. Diabetes mellitus (p = ) and active skin disease (p = ) approached statistical significance. The c statistic of indicated the patient model had fairly good predictive power and it fit the data well (Hosmer-Lemeshow p = ). The variables in the model were also tested for multicollinearity (Table A.8). All of the variables included in the patient model had VIFs between 1 and 5 indicating little to moderate correlation between the variables. The surgical model initially included all surgical variables that were significant (p < 0.05) or approached significance (p < 0.1) in the bivariable analyses. Mastectomy followed by immediate reconstruction, SLN biopsy, and drain placement were included in the final model. Drain placement was statistically significant with a p-value < Mastectomy followed by immediate reconstruction (p = ) approached statistical significance. The final model fit the data well (Hosmer-Lemeshow p = ). The c statistic was 0.767, indicating the model had good predictive power. All of the variables included in the procedure model had VIFs between 1 and 5 indicating little to moderate correlation between the variables. The combined model included the final variables from the patient and surgical models. Variables were also included if prior research identified them as risk factors for SSIs. Breast cancer, mastectomy with immediate reconstruction, and drain placement remained significant (Table A.8). This model had a good fit (Hosmer-Lemeshow p = ) and the c statistic was 0.757, suggesting that the model had good predictive ability. All of the variables included in the combined model had VIFs between 1 and 5 indicating little to moderate correlation between the variables. An association between skin-sparing mastectomy and mastectomy with immediate reconstruction was examined and found to be significant. Associations between prior chemotherapy and breast cancer and SLN biopsy and breast cancer were also examined. Both associations were found to be significant. Drain placement and procedure type was assessed for 9

20 association, but no significant associations were found. Drain placement and node biopsy was also assessed for association, and no significant associations were identified. Rate Analyses The overall SSI rates and the rates of S. aureus SSI rates over the study period were examined using Poisson regression analysis (Figures 1 and 2). The overall SSI rate increased over the study period but the increase did not reach statistical significance (p = ). SSIs caused by S. aureus increased significantly over the study period (p = ) by a fold increase each month. 10

21 CHAPTER IV DISCUSSION This study is unique because we evaluated S. aureus SSI rates over time and we evaluated risk factors and outcomes for S. aureus SSIs following breast operations. We used multivariable analysis to assess several possible risk factors for S. aureus SSI after breast operations. We identified several risk factors for SSIs chemotherapy use, timing of reconstructions following mastectomy, drain placement that may be remediable. Overall SSI and S. aureus Rate Analysis Over the study period the overall SSI rate increased, but the increase was not statistically significant. The S. aureus SSI rate also increased, and this increase was statistically significant. We did not find an association between the S. aureus SSI rate and the type of breast procedure or specific surgeons. However, since we did not assess rates of S. aureus SSIs following other operations over the study period, we do not know whether the increased S. aureus SSI rates were unique to breast operations. Kluytmans et al. 14 and other investigators found that nasal S. aureus carriage was a risk factor for S. aureus SSIs. Studies of preoperative decolonization of S. aureus carriers have reduced rates of S. aureus SSI among carriers. Investigators have examined several decolonization strategies; intranasal mupirocin application has been studied most frequently. One randomized clinical trial (RCT) demonstrated that mupirocin significantly decreased the rate of nosocomial infections due to S. aureus, especially among patients with nasal carriage of the organism. 25 Several studies have assessed whether intranasal povidone-iodine (PI) could decrease S. aureus SSI. Phillips et al. 27 found in per protocol analysis that S. aureus deep SSIs developed less often in patients who received PI compared with patients who received mupirocin (p = 0.03). To date, published studies have not assessed whether S. aureus carriage increases the risk of S. aureus SSIs after breast operation or whether decolonizing carriers decreases the risk of these infections after breast procedures. Future studies should assess whether nasal carriage of S. aureus is a risk factor for S. aureus SSI following breast operations. Should nasal carriage of S. aureus be a risk factor, then an intervention study would be warranted. 11

22 Diabetes and Hypertension In the bivariable analyses, diabetes was associated with S. aureus SSIs after breast operations. Several prior studies in other surgical populations found similar associations between postoperative infection and diabetes mellitus. Latham et al. 8 found that hyperglycemia during the immediate postoperative period was significantly associated with SSI among patients undergoing cardiothoracic operations. Hyperglycemia is closely linked with diabetes mellitus. While Olsen et al. 6 did not find diabetes to be associated with SSIs following breast operations, they found that postoperative blood glucose levels within 5 days after the procedures were significantly higher in patients with SSIs compared with uninfected patients. However, only 27% of patients had postoperative glucose testing performed. Our study did not find postoperative glucose levels to be significantly associated with S. aureus SSIs but only 16 (12.6%) of our study population had blood glucose readings recorded in the postoperative period. Though diabetes did not stay in the final combined model, the associations between hyperglycemia and diabetes and between diabetes and SSI indicate that diabetes and perioperative glucose levels merit further investigation because glucose levels could be addressed with preoperative and postoperative insulin therapy. Bivariable analyses identified hypertension as being associated with S. aureus SSIs after breast operations. Vilar-Compte et al. 18 found high blood pressure to be a significant risk factor for SSIs following breast operations. Future studies by the same research team did not find hypertension to be significant. Though this variable did not remain significant in the multivariable analysis, it warrants further study since it is a modifiable risk factor. Breast Cancer and Prior Chemotherapy Use We found that a diagnosis of breast cancer and preoperative chemotherapy were associated with S. aureus SSIs following breast operations. Thirty one (73.81%) cases underwent breast operations for breast cancer and only patients diagnosed with breast cancer received chemotherapy. Although a cancer diagnosis and chemotherapy interacted, the interaction term did not remain significant in the final regression analysis suggesting that both cancer and chemotherapy contribute to the risk of S. aureus SSI after breast operations. Olsen et al. 6 previously identified previous a breast cancer diagnosis as a significant risk factors for any SSI following breast operations. Studies by Broadwater et al. 29 Forouhi et al. 30 did not find a significant difference in wound infections among patients who had preoperative chemotherapy 12

23 and mastectomy compared with mastectomy alone. Further research should assess the relationship between prior chemotherapy and SSI following breast operations. Bivariable analysis identified SLN biopsy as a significant risk factor for S. aureus SSI. SLN remained significant in the overall multivariable logistic regression model. Previous research has not identified SLN as a risk factor for SSI following breast operations. Olsen et al. 6 found that mastectomy with 0 2 lymph nodes removed or mastectomy with 3 or more lymph nodes removed was a significant risk factor for SSI. Our study did not analyze whether the number of lymph nodes removed as a risk factor but a SLN might be associated with the number of lymph nodes removed and, thus, with the risk of S. aureus SSI. Mastectomy and Immediate Reconstruction Mastectomy followed by immediate breast reconstruction (IBR) was significantly associated with S. aureusssis in our study. Several researcher groups previously noted the association between the type of breast operation and any SSI after breast operation. For example Olsen et al. 7 found the SSI incidence in a cohort of younger women to be twice as high following mastectomy with IBR compared with mastectomy alone. Younger women are more inclined than older women to undergo immediate reconstruction given that breasts are considered to be hallmarks of femininity and beauty in many western societies. We also found that skin sparing mastectomies were significantly associated with S. aureusssis. These operations are commonly performed when a patient has decided to undergo IBR since these procedures produce better cosmetic results that do simple mastectomies. 31 Skin sparing mastectomies also confer significant psychosocial benefits and have been successful among women with either early- and late-stage breast cancer However, Bailey et al. 33 also found high SSI rates (up to 55%) following skin sparing mastectomies. Given that skin-sparing mastectomies are most often performed prior to IBR, further studies should be done to evaluate the relationship between these operations, IBR, and SSI, including S. aureus SSI. Drain Placement In our study, the use of drains, but not the number of drains, was significantly associated with S. aureus SSIs following breast operations. A study by Reiffel et al. 38 found that 8.9% of drains placed during various plastic surgeries became colonized with MSSA. However, they did not find an association between S. aureus colonization and SSI. Degim et al. 39 conducted a trial on drain care and found that SSIs occurred within 30 days after the operation significantly more 13

24 often in the control group than in the antisepsis intervention group. The analysis was limited by a small number of SSIs, but patients with greater bacterial colonization of their drain fluid or drain tubing had SSIs more frequently than did patients with less bacterial colonization or with negative cultures. Further studies should be done to assess drain use and drain colonization among patients following breast operations. 14

25 CHAPTER V CONCLUSION Our study used Poisson regression to analyze time-series data (SSI rates by month) and assessed many potential risk factors for S. aureus SSIs after breast operations. We demonstrated that S. aureus SSI rates after breast operations have been increasing, and that the increase reached statistical significance. We identified diabetes mellitus, prior chemotherapy, timing of reconstruction after mastectomy, and drain placement as possibly modifiable risk factors for S. aureus SSI after breast operations. Our results will help doctors at UIHC design interventions to prevent S. aureus SSIs following these procedures. This study also suggested several issues that should be addressed in future studies. We think this study should be repeated prospectively in a larger population to confirm our findings. Such a study could also assess the role of preoperative S. aureus nasal carriage as a risk factor for these SSIs. We also think that future studies should assess the association between preoperative and postoperative blood glucose levels, prior chemotherapy, including adjuvant chemotherapy and neoadjuvant chemotherapy, SLN biopsy, and drain placement with S. aureus SSIs. Subsequently, intervention trials could address the most significant risk factors as a means of decreasing the risk of S. aureus SSIs following breast operations. Patients with S. aureus SSIs were often readmitted or had reoperations, likely increasing the costs attributable to SSI following breast operations. Thus, interventions that prevent these SSIs after breast procedures could improve patient outcomes and decrease healthcare costs. 15

26 APPENDIX 16

27 Table A.1: Data Dictionary Variable Name Definition Levels ProcDate Date of breast operation Continuous DaysBOP Number of days admitted to the hospital Continuous before operation LOS Number of days in hospital after Continuous operation Age Age on day of operation Continuous Ht Height on day of operation Continuous Wt Weight on day of operation Continuous BMI BMI on day of operation Continuous CurrentCig Patient is a smoker at time of operation Yes, no, unknown PastCig Patient is a former smoker at time of Yes, no, unknown operation Quit Time, in months, when patient quit 0 3 months, 4 6 months, 7 9 months, months, > 12 months, Unknown Diabetes Patient had diabetes mellitus at time of Yes, No, Unknown operation MenopausalStatus Patient menopause status at time of Pre-menopausal, Menopausal, Post-Menopausal operation SkinDis Patient had an active skin disorder or Yes, No, Unknown rash at time of operation Insurance Patient insurance at time of operation Iowa Care, Medicare, Medicare with Supplemental Insurance, Medicaid, Private Insurance, No Insurance SysBP Systolic blood pressure at time of Continuous operation DiaBP Diastolic blood pressure at time of Continuous operation Hypoglycemic Patient taking oral hypoglycemic agents within 30 days of operation Yes, no, unknown 17

28 Table A.1 continued Insulin Patient taking insulin within 30 days of Yes, no, unknown operation Steroids Patient taking oral or parenteral steroids Yes, no, unknown within 30 days of operation Chemo Patient received chemotherapy within 4 Yes, no, unknown to 6 weeks of operation Immuno Patient received immunosuppressive Yes, no, unknown agents within 30 days of operation PriorAbx Patient received antimicrobial agents Yes, no, unknown within 30 days of operation PrevOps Patient had previous breast operations Yes, no, unknown SubOps Patient had subsequent breast operation Yes, no, unknown within 30 days PreOpRad Patient received radiation therapy within Yes, no, unknown 30 days prior to operation PostOpRad Patient received radiation therapy within Yes, no, unknown 60 days after operation PreInfect Patient had infection at other site within Yes, no, unknown 30 days prior to operation PostInfect Patient had infection at other site within Yes, no, unknown 30 days after operation OpReason Reason for operation Cancer, augmentation, reduction, reconstruction OpType Type of operation Mastectomy, mastectomy and implant, implant, lumpectomy, flap, reduction only, mastopexy MastType Type of mastectomy, if applicable Simple/Total, modified radical, radical, nipple-sparing, skin-sparing, unknown ImpType Type of implant, if applicable Tissue expander, permanent implant ADM If the patient received an implant, was an Yes, no, unknown acellular dermal matrix used LumpType Patient received wire or needle Yes, no, unknown localization FlapType Type of flap, if applicable Free, pedicled, unknown 18

29 Table A.1 continued SNL Patient received a SLN biopsy Yes, no, unknown ALND Patient received an axillary lymph node Yes, no, unknown dissection Region Area that the breast operation was done Unilateral Bilateral on ASAScore 1, 2, 3, 4, 5, Unknown WoundClass Clean, Clean-Contaminated, Contaminated, Dirty SkinPrep The skin antisepsis received by the CHG, PVI patient ProcTime Continuous EBL Estimated blood loss during the Continuous operation RegBlock Patient received a regional block before Yes, no, unknown the operation Prophylaxis Patient received prophylactic antibiotics Yes, no, unknown prior to operation ProphyAbx Antibiotic received by the patient ProphDose Antibiotic dose received by the patient Continuous ProphTime Time the antibiotic dose was received by Continuous the patient IntraOpAbx Doses of prophylactic antibiotics Continuous received by patient during the operation PostOpUse Same antibiotics given to patient after Yes, no, unknown the operation PostOpDoses Doses of the same antibiotics received after the operation Continuous Drain Drains placed before the end of operation Yes, no, unknown NumDrain The number of drains placed Continuous DrainTime1 The time drain was in place Continuous DrainTime2 The time drain was in place Continuous DrainTime3 The time drain was in place Continuous DrainTime4 The time drain was in place Continuous 19

30 Table A.1 Continued DrainTime5 The time drain was in place Continuous PreopHGB Preoperative hemoglobin level Continuous PostopHGB Postoperative hemoglobin level Continuous PreopGlu Preoperative glucose level Continuous PostopGlu Postoperative glucose level Continuous PreopWBC Preoperative total white blood cell count Continuous PostopWBC Postoperative total white blood cell Continuous count Status Status of the patient Case, Control SSIDate Recorded date of SSI Continuous InfectType Type of SSI Superficial, deep, organ/space Readmit Patient was readmitted due to SSI Yes, no, unknown ORReturn Patient returned to operating room to Yes, no, unknown treat SSI ImpRemove Implant had to be removed due to SSI Yes, no, unknown MeshRemove Mesh had to be removed due to SSI Yes, no, unknown AbxOutpatient Patient received antibiotics as an Yes, no, unknown outpatient OpenWndClinic Patient wound was opened by doctor at Yes, no, unknown clinic visit Seroma Patient experienced a seroma Yes, no, unknown Hematoma Patient experienced a hematoma Yes, no, unknown Dehiscence Patient experienced dehiscence Yes, no, unknown StitchAbscess Patient experienced a stitch abscess Yes, no, unknown FlapNecrosis Patient experience flap necrosis Yes, no, unknown 20

31 Table A.2: ICD-9-CM Procedure Codes for Breast Operations ICD-9 Code Definition Open biopsy of breast Excision or destruction of breast tissue, not otherwise specified Local excision of lesion of breast; lumpectomy, removal of area of fibrosis from breast. Excludes: biopsy of breast ( ) Resection of quadrant of breast Subtotal mastectomy. Excludes: quadrant resection (85.22) Unilateral reduction mammoplasty; Unilateral: amputative mammoplasty, size reduction mammoplasty Bilateral reduction mammoplasty; amputative mammoplasty, reduction mammoplasty (for gynecomastia) Unilateral subcutaneous mammectomy with synchronous implant. Excludes: that without synchronous implant (85.34) Other unilateral subcutaneous mammectomy; removal of breast tissue with preservation of skin and nipple, subcutaneous mammectomy NOS Bilateral subcutaneous mammectomy with synchronous implant. Excludes: that without synchronous implant (85.36) Other bilateral subcutaneous mammectomy Unilateral simple mastectomy; Mastectomy: NOS, complete Bilateral simple mastectomy; Bilateral complete mastectomy Unilateral extended simple mastectomy; Extended simple mastectomy NOS, modified radical mastectomy, simple mastectomy with excision of regional lymph nodes Bilateral extended simple mastectomy Unilateral radical mastectomy; excision of breast, pectoral muscles, and regional lymph nodes [axillary, clavicular, supraclavicular], radical mastectomy NOS Bilateral radical mastectomy Unilateral extended radical mastectomy; Excision of breast, muscles, and lymph nodes [ axillary, clavicular, supraclavicular, internal mammary, and mediastinal], extended radical mastectomy NOS Bilateral extended radical mastectomy Augmentation mammoplasty, not otherwise specified Unilateral breast implant Bilateral breast implant; Breast implant NOS Fat graft to breast Mastopexy Total reconstruction of the breast, not otherwise specified; perforator flap, free Latissimus dorsi myocutaneous flap Transverse rectus abdominis myocutaneous (TRAM) flap, pedicled. Excludes: transver rectus abdominis myocutaneous (TRAM) flap, free (85.73) Transverse rectus abdominis myocutaneous (TRAM) flap, free. Excludes: transverse rectus abdominis myocutaneous (TRAM) flap, pedicled (85.72) 21

32 Table A.2 continued Deep inferior epigastric artery perforator (DIEP) flap, free Superficial inferior epigastric artery (SIEA) flap, free Gluteal artery perforator (GAP) flap, free Other total reconstruction of breast. Excludes: deep inferior epigastric artery perforator (DIEP) flap, free (85.74), gluteal artery perforator flap, free (85.76) lastissimus dorsi myocutaneous flap (85.71), perforator flap, free (85.70), superficial inferior epigastric artery (SIEA) flap, free (85.75), total reconstruction of breast, not otherwise specified (85.70), transverse rectus abdominis myocutaneous (TRAM) flap, free (85.73), transverse rectus abdominis myocutaneous (TRAM) flap, pedicled (85.72) Revision of implant of breast Removal of implant of breast Insertion of breast tissue expander; insertion (soft tissue) of tissue expander (one or more) under muscle or platysma to develop skin flaps for donor use Removal of breast tissue expander 22

33 Table A.3: NHSN Criteria for SSI Classification Superficial Incisional SSI Infection occurs within 30 days after the operative procedure AND involves only skin and subcutaneous tissue of the incision AND must meet one of the following criteria: a. purulent drainage from the superficial incision. b. organisms isolated from an aseptically obtained culture of fluid or tissue from the superficial incision. c. 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, and is culture-positive or not cultured. A culture-negative finding does not meet this criterion. d. diagnosis of superficial incisional SSI by the surgeon or attending physician. Deep Incisional SSI Infection occurs within 30 days after the operative procedure if no implant is left in place or within one year if implant is in place and the infection appears to be related to the operative procedure AND involves deep soft tissues of the incision AND must meet one of the following criteria: a. purulent drainage from the deep incision but not from the organ/space component of the surgical site. b. a deep incision spontaneously dehisces or is deliberately opened by a surgeon and is culture-positive or not cultured when the patient has at least one of the following signs or symptoms: fever (>38 C), or localized pain or tenderness. A culture-negative finding does not meet this criterion. c. An abscess or other evidence of infection involving the deep incision is found on direct examination, during reoperation, or by histopathologic or radiologic examination. d. Diagnosis of a deep incisional SSI by a surgeon or attending physician. Organ/Space SSI Infection occurs within 30 days after the operative procedure if no implant is left in place or within one year if implant is in place and the infection appears to be related to the operative procedure AND involves any part of the body, excluding the skin incision, fascia, or muscle layers, that is opened or manipulated during the operative procedure AND must meet one of the following criteria: a. purulent drainage from a drain that is placed into the organ/space. b. organisms isolated from an aseptically obtained culture or fluid or tissue in the organ/space. c. an abscess or other evidence of infection involving the organ/space that is found on direct examination, during reoperation, or by histopathologic or radiologic examination. d. diagnosis of an organ/space SSI by a surgeon or attending physician. 23

34 Table A.4 Distribution of Potential Risk Factors for Surgical Site Infection Following Breast Operations at UIHC from 7/1/2004 to 9/15/2015 Variable Name SSI (N=42) No SSI (N=84) Age, Mean (range) 54.6 (29 84) 52.1 (20 86) BMI, Mean (range) 29.2 ( ) 30.4 ( ) Diabetes 10 (24.39%) 8 (9.76%) Hypertension 13 (59.09%) 12 (35.29%) Any Previous Cancer Diagnosis 34 (80.95%) 45 (53.57%) ASA Score > 2 34 (80.95%) 55 (65.48%) Indication for Procedure Cancer 31 (73.81%) 41 (48.81%) Augmentation 1 (2.38%) 6 (7.14%) Reconstruction 5 (11.90%) 7 (8.33%) Reduction 3 (7.14%) 8 (9.52%) Type of Procedure Mastectomy 10 (23.81%) 11 (13.10%) Mastectomy and Implant 13 (30.95%) 5 (5.95%) Implant 2 (4.76%) 15 (17.86%) Lumpectomy 5 (11.90%) 32 (38.10%) 24