Abstract Objective To determine the attributable costs associated with surgical site infection (SSI) following breast surgery. Design and Setting Cost analysis of a retrospective cohort in a tertiary care university hospital. Patients All persons who underwent breast surgery other than breast-conserving surgery from July 1, 1999, through June 30, 2002. Main Outcome Measures Surgical site infection and hospital costs. Costs included all those incurred in the original surgical admission and any readmission(s) within 1 year of surgery, inflation adjusted to US dollars in 2004. Results Surgical site infection was identified in 50 women during the original surgical admission or at readmission to the hospital within 1 year of surgery (N = 949). The incidence of SSI was 12.4% following mastectomy with immediate implant reconstruction, 6.2% following mastectomy with immediate reconstruction using a transverse rectus abdominis myocutaneous flap, 4.4% following mastectomy only, and 1.1% following breast reduction surgery. Of the SSI cases, 96.0% were identified at readmission to the hospital. Patients with SSI had crude median costs of $16 882 compared with $6123 for uninfected patients. After adjusting for the type of surgical procedure(s), breast cancer stage, and other variables associated with significantly increased costs using feasible generalized least squares, the attributable cost of SSI after breast surgery was $4091 (95% confidence interval, $2839-$5533). Conclusions Surgical site infection after breast cancer surgical procedures was more common than expected for clean surgery and more common than SSI after non–cancer-related breast surgical procedures. Knowledge of the attributable costs of SSI in this patient population can be used to justify infection control interventions to reduce the risk of infection. The overall surgical site infection (SSI) rate following mastectomy reported to the Centers for Disease Control and Prevention National Nosocomial Infections Surveillance System from 1992 through 2004 was 1.98%.1 In contrast, SSI rates published in individual reports in the literature range from 1% to 28%. Surgical site infection rates have been reported to be lowest following breast-conserving surgery, ranging from 1.5% to 10.8%.2-8 The incidence of SSI following mastectomy ranges from 2.8% to 25% in the surgical and infection control literature,4-7,9-18 although only 2 studies4,5 report an incidence of less than 5%. The SSI rate following breast cancer reconstructive surgery also seems to be relatively high (range, 6.3%-28%), based on a few reports in the literature.10,19-22 Thus, the SSI rates following breast surgery seem to be much greater than the nationally reported incidence of 2.0% and much higher than what is expected for clean surgical procedures. Given the state of fiscal constraints within the US health care system, it is important to calculate the cost-effectiveness of infection control interventions to justify their use from an economic perspective. Cost-effectiveness analyses require accurate estimates for the attributable costs of hospital-acquired infections, which are lacking for SSI. To our knowledge, there is only one report in the literature describing the attributable costs of SSI following breast surgery. Reilly et al23 estimated the costs associated with SSI after a variety of different surgical procedures, including breast surgery. The costs included in this study were those associated with additional hospital length of stay, outpatient and home health visits, and outpatient antibiotics. Using the total costs attributable to infection presented in their publication, the mean attributable costs of SSI following breast surgery equaled £359 (approximately $574 in the United States).23 It is not clear what types of breast surgery were included in the analysis (mastectomy, breast-conserving surgery, or cosmetic surgical procedures) or what proportion of the observed SSI were diagnosed and treated solely in the outpatient setting. Infections managed exclusively in an outpatient setting will presumably be associated with lower costs than infections requiring rehospitalization. In addition, British health care delivery and use are different from those in the United States, so the costs attributable to SSI after breast surgery determined in this study may not be comparable to the costs of SSI in the United States. Despite the publication of a number of studies describing costs associated with SSI, the magnitude of economic costs attributable to these infections is still unclear. Costs associated with SSI will likely vary depending on the type of surgical procedure. Surgical site infections following cardiothoracic or neurosurgical procedures, with the potential for severe morbidity and mortality, would presumably have the largest attributable costs, while infection following surgery not involving major organ spaces would be expected to have lower costs. This is consistent with the few adjusted estimates of SSI costs that have been reported.23-30 We studied the hospital-associated costs of SSI following mastectomy and breast reconstructive surgery to determine more precisely the attributable costs of SSI following these procedures. Methods Study population Procedures eligible for inclusion in this study included all mastectomy and breast reconstructive surgical procedures performed at Barnes-Jewish Hospital (BJH) from July 1, 1999, through June 30, 2002. Barnes-Jewish Hospital is a 1251-bed tertiary care hospital affiliated with Washington University School of Medicine. International Classification of Diseases, Ninth Revision (ICD-9) procedure codes were used to identify eligible surgical admissions and outpatient surgical procedures, including breast reduction, augmentation, mastectomy, breast reconstruction with transverse rectus abdominis myocutaneous (TRAM) or other muscle flap, and insertion of a breast implant (ICD-9 procedure codes 85.31-85.48, 85.7, 85.85, and 85.95). Admissions for breast-conserving surgery (excisional biopsy, lumpectomy, or partial mastectomy) and mastopexy only were also excluded, because inpatient readmission for therapy of SSI after these procedures is rare. Using these criteria, 949 admissions with breast surgery were included in the original cohort. Approval for this study was obtained from the Washington University School of Medicine Human Studies Committee. Identification of ssi case patients Surgical site infections following breast surgery were identified by electronic surveillance using data from the BJH Medical Informatics database. Indicators suggestive of SSI included ICD-9 diagnosis codes consistent with wound infection (codes 682.2, 682.3, 998.5, 998.51, 998.59, and 996.69) or breast surgery complication (codes 611.0, 996.79, 998.3, and 998.83) and/or positive microbiology wound culture results during the original surgical admission or any readmission (inpatient or outpatient surgical) within 1 year of surgery. All potential wound infections identified by electronic surveillance were verified by review of the medical records. Centers for Disease Control and Prevention–National Nosocomial Infections Surveillance System criteria were used to define SSI. Surgical site infection required at least one of the following: purulent drainage from the incision, organisms isolated from an aseptically obtained culture, peri-incisional erythema or warmth and the incision was opened by a physician, abscess found on radiologic or histopathologic examination, or a diagnosis of SSI by the surgeon. Only infections diagnosed during the original surgical admission or at readmission to the hospital or to outpatient surgery were included; infections diagnosed and managed solely in the outpatient setting were excluded, because by definition outpatient infections would not have associated hospital costs. Variable creation and definitions Data regarding variables potentially associated with increased costs were abstracted for the cohort from the BJH Medical Informatics database. The data available for the entire cohort included ICD-9 discharge diagnosis and procedure codes, pharmacy and microbiology data, and demographic information. Preexisting comorbidities, carcinoma in situ, malignancy, and history of malignancy were defined using the ICD-9 diagnosis codes assigned during the original surgical admission and the coding criteria of Deyo et al.31 Comorbidities selected for analysis included those generally associated with increased risk of SSI (eg, diabetes mellitus) and comorbidities that may be more frequent in the population of patients with SSI and would most likely be associated with increased costs (eg, serious heart disease). In addition, frequencies were determined for all ICD-9 diagnosis codes assigned to the surgical cohort; all comorbidities present in more than 1% of the cohort were included in the statistical analyses. The comorbidities analyzed included diabetes mellitus, serious heart disease (myocardial infarction or congestive heart failure), renal failure, chronic obstructive pulmonary disease, and hypertension. Diagnosis of metastatic disease was defined using ICD-9 diagnosis codes assigned during the original surgical admission and any readmission to the hospital within 1 year of surgery. Central venous catheterization was defined using ICD-9 procedure codes (code 38.93 or 86.07) for the original surgical admission and any hospital readmission within 1 year of surgery. Surgical procedures performed during the original surgical admission were defined using ICD-9 procedure codes. All possible logic checks were performed for variables created from ICD-9 diagnosis and procedure codes, and discrepant results were verified using the electronic medical record. The number of ICD-9 codes assigned during the surgical admission was used as an indicator of severity of illness rather than the Charlson comorbidity score (adapted for use with administrative data), to determine the costs associated with specific comorbidities in this population. The Charlson comorbidity score is a summary score in which weights are assigned to various comorbidities associated with mortality; because our intent was to develop a precise determination of attributable costs, we used the individual comorbidities rather than a summary score in the statistical models to calculate costs.32 In addition, the Charlson comorbidity score includes diagnosis of any malignancy in the scoring scheme. Because we wanted to specifically control for costs associated with tumor stage in the patients diagnosed as having in situ or invasive breast cancer in our statistical models, this was an additional rationale for not using an aggregated comorbidity score in our models to adjust for severity of illness. Pharmacy data available from the BJH Medical Informatics database were also used to define some underlying disease variables, including insulin use in the original surgical admission or within 1 year of surgery, insulin and oral hypoglycemic agents as markers of diabetes mellitus, and digoxin use as a marker of serious heart disease. These variables were compared with the comorbidity variables previously described, to verify the diagnostic variables and to assess the effect of variables created by combination of the diagnostic and pharmacy information. Cost data Cost data for the cohort from the perspective of the hospital, including the hospital costs of the original surgical admission and hospital costs for all readmissions within 1 year of surgery, were obtained from the BJH cost accounting database (Trendstar; McKesson Corp, Alpharetta, Georgia). Eligible hospital readmissions included inpatient, same day, or ambulatory surgery and outpatient surgery admissions. All readmission hospital costs within 1 year of surgery were included in the analyses, to hopefully capture all costs involved in the initial diagnosis and therapy of infection. Because SSI in some patients may require multiple admissions to the hospital for antibiotic or surgical treatment, we included readmission costs for all members of the cohort for a 1-year follow-up to capture these costs. The cost accounting database uses a set-down allocation method to calculate costs, which included indirect variable, direct variable, and fixed costs. All patient charge codes received during hospitalizations were recorded, and the departmental cost for each charge code calculated was based on each department's actual cost components multiplied by the charges for that code, divided by total departmental charges. Costs were summed across each department to provide total hospital costs for each admission and were inflation adjusted to 2004 US dollars using the medical care component of the Consumer Price Index (accessed at http://www.bls.gov). Only hospital-associated costs were included in the analysis; physician costs were not included. Statistical analyses The χ2 test was used to compare categorical variables, and the Mann-Whitney test was used for continuous variables. Multivariate analyses to determine costs attributable to SSI were performed by ordinary least-squares and feasible generalized least-squares (FGLS) regression, using natural log-transformed costs as the dependent variable, because of the highly skewed distribution of total costs. These methods can be used to estimate the costs associated with variables of interest, taking into account variation in the occurrence of other factors significantly associated with expenditures. The coefficients obtained in the FGLS model represent the mean change in the variable of interest per unit increase in X, when the other variables in the model are held constant. In this study, the change in X represents the mean difference in the natural logarithm of costs between individuals with infection and those without infection, holding all other predictors of costs constant. All independent variables were checked for colinearity. The models were checked for functional form misspecification using the Ramsey regression specification error test and for heteroscedasticity using the Breusch-Pagan test.33 The total mortality within 1 year of surgery in the cohort was 1.5% (13/878); because it was so low, no attempt was made to adjust the FGLS models for mortality. Data for admissions in which the standardized residuals for the final model were large (>3 SDs from the mean of 0) were all checked for accuracy. Because the FGLS model used the natural logarithm of costs as the dependent variable, an intermediate regression was performed to predict costs given that the dependent variable in the model was ln(costs).33 Propensity score methods were also used to determine the association between SSI and costs.34 A nonparsimonious logistic regression model was used to obtain predicted values for SSI. Patients with SSI were then matched 1:1 with control subjects based on the propensity score. Total costs were compared between the matched pairs by the Wilcoxon signed rank test, and the mean and median differences in total costs were calculated. Statistical analyses were performed using commercially available software programs (SPSS, version 12.0 [SPSS Inc, Chicago, Illinois] and Stata, version 9.2 [Stata Corp, College Station, Texas]). Results Ssi rates in the cohort A cohort of 949 admissions in which ICD-9 procedure codes for mastectomy or breast reconstruction were assigned was established for the 2-year period of the study. Most surgical procedures (90.6%) resulted in at least an overnight hospitalization, with only 4 patients who underwent mastectomy (0.7%) discharged on the day of surgery. A total of 7.6% (14/185) of reduction-only surgical procedures, 50.8% (33/65) of delayed implant placement in patients with cancer, and 62.5% (15/24) of cosmetic augmentations were performed in outpatient surgery. Surgical site infection was identified in 50 patients within 1 year of surgery in this cohort (SSI rate, 5.3%). Thirteen infections (4.4%) followed mastectomy only, and 22 were breast implant associated, requiring surgical removal of the implant in 18. Surgical site infection in 45 patients involved the breast incision(s), 7 involved the abdominal incision in patients who had undergone TRAM reconstruction, and 1 involved the back incision in a woman who had undergone latissimus dorsi reconstruction. Three patients had SSI following TRAM reconstruction involving the breast and abdominal incisions. Surgical site infection rates were lowest following non–cancer-related surgical procedures. Rates of SSI according to the type of surgery performed are shown in Table 1. For the outcome and cost analyses, any readmissions to the hospital within 1 year of surgery, including readmissions in which eligible breast surgical procedures were performed, were included in the readmission costs associated with the original surgical admission. A total of 61 patients had 1 or more follow-up breast surgical procedures performed within 1 year of the original surgery that were included in readmission costs, reducing the number of breast surgical admissions in the outcome cohort to 888. Ten patients had surgical admissions separated by more than 1 year; these were considered separate events. Forty patients (4.5%) included in the final breast surgical admission cohort were male; 33 males underwent subcutaneous mastectomies or reduction mammaplasty because of gynecomastia, while 7 underwent unilateral mastectomies for surgical removal of a breast malignancy. Characteristics of patients with ssi and outcomes Characteristics identified in the surgical cohort and their associations with subsequent SSI are shown in Table 2. All of the patients with SSI following breast surgery were female, and there was no association between SSI and race, age, and malignant or metastatic breast cancer. Patients with SSI were significantly more likely to have an underlying diagnosis of diabetes mellitus, to be obese (defined by a body mass index [calculated as weight in kilograms divided by height in meters squared] > 30), to have undergone mastectomy or a surgical procedure involving placement of a breast implant, and to have a central venous catheter placed within 1 year of surgery (excluding catheters placed to administer antibiotics due to SSI). Patients with SSI had significantly more ICD-9 diagnosis codes assigned during the breast surgical admission. Surgical site infection was significantly less likely in patients who had undergone only reduction surgery, compared with patients who had other procedures performed. Information on antibiotic prophylaxis and drain use and management for individual patients was not available, because those data were not maintained electronically. During the study period, the standard of care at our institution was to give all patients undergoing the included breast surgical procedures 1 g of cefazolin as prophylaxis within 1 hour before incision. In a separate analysis of a case-control subset at our institution that overlapped the period of this study, 88% of patients were given cefazolin prophylactically, with the remainder receiving vancomycin, a combination of ampicillin and sulbactam, clindamycin, or other antibiotics. Only 2% of patients in the case-control study were not given prophylactic antibiotics (M.A.O.; M. Lefta, BS; J.R.D.; K.E.B.; R. Aft, MD; R. Matthews, MPH; J.M.; and V.J.F.; unpublished data, 2006). Drains were placed routinely in the remaining breast tissue or mastectomy bed in patients who underwent mastectomy, flap reconstruction, or reduction, but were not used routinely in patients undergoing cosmetic augmentation or delayed reconstruction with an implant. Patients undergoing simple mastectomy typically had a single drain placed in the mastectomy bed, and patients undergoing modified radical mastectomy had 2 drains, 1 in the mastectomy bed and 1 in the axilla. Patients undergoing reconstruction with a TRAM flap had 2 drains placed in the abdomen and 1 in the reconstructed breast, and patients undergoing latissimus dorsi flap reconstruction had 1 drain placed in the back and 1 in the reconstructed breast. Patients were discharged from the hospital with the drain(s) in place, and the drain(s) were generally removed between 7 and 21 days after surgery when the drainage was less than 1 ounce per 24 hours. The median time from surgery until diagnosis of SSI was 25 days (range, 2-315 days; mean [SD], 46.6 [62.0] days). Of the 50 patients with SSI, 41 (82.0%) had their infections diagnosed within 60 days of surgery. Only 2 patients were diagnosed as having an SSI during the original surgical admission; almost three-quarters of infected patients (72.0%) were treated during a single readmission to the hospital (Table 3). Thirty patients with SSI (60.0%) underwent subsequent surgery because of infection: 18 involved breast implant removal, while 12 involved debridement or incision and drainage of the wound in the operating room. Surgical site infections in the remaining 20 patients were treated with intravenous antibiotics with or without bedside drainage or debridement (Table 3). Costs associated with ssi Complete cost data were not available for 4 patients, including 1 with an SSI (following mastectomy plus immediate placement of an implant), resulting in a population of 884 surgical admissions available for analysis. The comparisons of the total costs and hospital length of stay for the surgical admission plus all readmissions within 1 year of surgery for the cohort are shown in Table 4. Patients with SSI had significantly higher hospital costs associated with surgery and during the 1-year period after surgery compared with uninfected patients, and they had a significantly longer total length of hospital stay (P < .001 for both, Mann-Whitney test). The crude median costs attributable to SSI totaled approximately $10 750, but, as can be seen in Table 4, the range of costs for the infected and uninfected control patients was broad. The crude median length of stay attributable to SSI equaled 4.3 days, although the range of length of stay was also broad for SSI case patients and uninfected control patients (Table 4). Feasible generalized least squares was used to calculate the attributable costs of SSI in the year after breast surgery. The natural logarithm of the total costs was used as the dependent variable in the regression model, because of the highly skewed nature of total costs. The operative variables associated with significantly increased costs included the type of donor flap used to reconstruct breast tissue (free or nonmicrovascular TRAM or latissimus dorsi flap), placement of an implant immediately following mastectomy, and bilateral surgery (Table 5). The diagnosis of malignant disease was associated with significantly increased costs, and metastatic disease was associated with even greater costs (indicated by the progressively larger values for the β coefficient). Correspondingly, placement of a central venous catheter, used as a proxy for chemotherapy, was associated with increased costs. The only comorbidities associated with significantly increased costs in this population were serious heart disease (acute or old myocardial infarction, congestive heart failure, or digoxin administration in the hospital in the year after surgery) and renal failure. A diagnosis of male gynecomastia and reduction surgery in females was associated with significantly decreased costs. Controlling for the other variables in the model, costs for nonwhite persons were 7.9% higher in this breast surgical cohort compared with costs incurred by whites (Table 5). Excluding noncancer surgical procedures, diagnoses of breast cancer with lymph node metastases and metastatic breast cancer within 1 year of surgery were more common in nonwhite compared with white persons (31.7% vs 25.9% for lymph node metastases [P = .13] and 6.5% vs 2.5% for metastatic breast cancer [P = .01]), suggesting that nonwhites present at a more advanced stage of cancer than do whites. The attributable cost of SSI in this cohort, adjusted for the variables previously described and others listed in Table 5, was $4091 (95% confidence interval, $2839-$5533) (in 2004 US dollars). The final model presented in Table 5 had an adjusted coefficient of determination (R2) of 0.52, indicating that approximately 52% of the variation in actual costs was explained by the model. In addition to the FGLS model, total costs were also compared between SSI cases and controls matched based on the propensity scores. A nonparsimonious logistic regression model was developed with SSI as the dependent variable. Uninfected control patients were matched to SSI case patients based on the propensity score, with 2 SSI case patients unable to be matched. The difference in total costs between the matched pairs was significant (P < .001, Wilcoxon signed rank test). The mean difference in total costs between the matched SSI and control pairs was $5700, with a median difference in costs of $3492. Comment Knowledge of the outcomes of breast surgery, including SSI and their associated costs, is necessary to understand the impact of these infections in a population of surgical patients. Our study, using a retrospective cohort of breast surgical patients treated at a university-affiliated tertiary care hospital, shows that SSI after breast cancer surgery is more common than would be expected for clean surgery and more common than SSI rates after similarly extensive breast surgical procedures in patients without cancer. Given the relatively high incidence of SSI after these procedures, it is important to understand their associated costs, to develop prevention strategies and make informed decisions concerning the potential cost-effectiveness of interventions designed to reduce the risk of SSI. For this reason, it is important to quantify the costs attributable to infection as precisely as possible, to determine the true magnitude of hospital-associated costs and, ultimately, societal costs because of SSI. In this study, we used FGLS to calculate attributable hospital costs of SSI after breast surgery, controlling for differences in surgical procedures, comorbidities, stage of breast cancer, and other factors associated with increased costs in this surgical population. We chose this method because it allows for the analysis of data from an entire cohort of surgical patients, in contrast to matched-pair studies in which case patients with SSI are matched to control patients without infection after surgery, resulting in analysis of only a selected subset of the control population. The inability to match some infected patients with unusual combinations of comorbidities or severe underlying illness with comparable control patients is an inherent disadvantage of the matched-pair method. The ordinary least-squares method using log-transformed costs as the dependent variable has recently been shown to accurately predict costs following coronary artery bypass graft surgery in a validation sample.35 Most published reports36,37 addressing the costs of SSI have used a matched case-control study design, although this study design is known to result in potentially inaccurate cost estimates, because of selection bias. The estimated costs of SSI determined in studies in which some form of adjustment for patient acuity, and other factors associated with increased risk of SSI, were performed, either by matching or other statistical methods, range from approximately $3400 to $17 700. The adjusted costs of SSI have been lower following general or mixed surgical procedures ($3382-$5038)26,30 and highest following orthopedic or cardiothoracic surgery ($17 708 for orthopedic SSI and $12 419-$14 211 following coronary artery bypass surgery).27-29 These cost estimates are not necessarily comparable, even for the same type of surgery, because of differing statistical methods to determine attributable costs, different duration of follow-up after surgery, and the inclusion of different sources of cost data in the various studies. In the present study, we calculated crude median costs attributable to SSI after breast surgical procedures equal to $10 759. After accounting for the variety of operative methods, diagnoses, and comorbidities, the adjusted attributable costs of SSI equaled $4091. We found similar results using propensity scores and calculation of median differences in costs between matched SSI cases and controls with similar propensity to develop an infection. Our cost estimates included costs incurred during the original surgical admission and all hospital readmissions within 1 year of surgery for all patients in the cohort. We included all readmission costs during a 1-year follow-up for several reasons. First, the standard definition of deep SSI used by the Centers for Disease Control and Prevention–National Nosocomial Infections Surveillance System includes infections with onset within 1 year of surgery in patients with foreign bodies implanted during surgery. Of the patients in our cohort, 21.0% had a breast implant placed during surgery and, thus, we needed to extend the period of observation to include infections associated with implants. Second, therapy due to SSI required multiple readmissions to the hospital in some patients, so we needed to include all of these readmissions to capture all costs associated with therapy due to SSI. We included readmission costs for control patients to model the costs directly attributable to SSI. For example, patients with SSI may also have had complications of chemotherapy during a readmission hospitalization, so not all of the costs incurred during that readmission were directly attributable to SSI. By including all readmission costs for case patients and controls, we could use the ordinary least-squares method to model the costs attributable to each of the variables in our model. The only underlying comorbidities associated with increased costs in this patient population were serious heart disease and renal failure, most likely due to the occurrence of breast cancer in primarily elderly women. Nonwhite race was also associated with significantly increased costs in our ordinary least-squares model, possibly because of later stage at diagnosis of breast cancer in African American women compared with white women.38 This is supported by the finding that more nonwhite persons in our cohort were diagnosed as having metastatic breast cancer within 1 year of surgery than were white persons, consistent with more advanced disease at diagnosis in nonwhite persons. Our results are consistent with the hypothesis that SSI following surgery involving major organ spaces are associated with greater costs than SSI following less extensive surgical procedures. Our calculated attributable cost of $4091 is much less than the attributable costs of SSI reported after cardiothoracic surgery24,27,28 and orthopedic surgery,29 but much higher, however, than the value reported by Reilly et al23 (approximately $574) in breast surgical patients. The higher attributable cost associated with SSI that we calculated may be primarily because of the inclusion of different types of surgical procedures in our study compared with that of Reilly and colleagues. We excluded breast-conserving surgery from our cohort because of the rarity of inpatient readmission for therapy due to SSI in this population at our institution (M.A.O., J.R.D., and V.J.F., unpublished data, 2006). Reilly et al included costs incurred in the hospital and outpatient setting and, if their population included women undergoing breast-conserving surgery, they may have been able to capture the outpatient costs associated with SSI following this procedure, resulting in the much lower value reported compared with our result. The attributable cost of SSI we calculated, $4091 per patient, includes only hospital-associated costs and, thus, certainly represents an underestimate of the true societal costs of serious SSI in this surgical population. Our study did not include physician costs or costs due to SSI associated with excess clinic visits, outpatient surgical procedures performed in the general or plastic surgery clinics, outpatient antibiotic use, or home health care and, thus, the total attributable costs we calculated represent, at best, the minimum costs associated with serious SSI resulting in readmission to the hospital or outpatient surgery. Ultimately, the goal of studying the incidence and costs associated with SSI is to develop and implement interventions to reduce the occurrence of this adverse surgical event. Potential interventions to reduce the incidence of SSI in this patient population include strategies to optimize the timing and dosage of prophylactic antibiotics administered before the surgical incision, glucose control in diabetic patients, promotion of meticulous hand hygiene, and strategies to promote timely removal of drains, among others. Interventions to reduce the incidence of SSI following breast cancer surgical procedures are essential to reduce not only morbidity in these patient populations but also costs to the individuals and to society. Correspondence: Margaret A. Olsen, PhD, MPH, Division of Infectious Diseases, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8051, St Louis, MO 63110 (email@example.com). Accepted for Publication: September 4, 2006. Author Contributions:Study concept and design: Olsen, Brandt, Dietz, and Fraser. Acquisition of data: Olsen, Chu-Ongsakul, Mayfield, and Fraser. Analysis and interpretation of data: Olsen, Chu-Ongsakul, Dietz, and Fraser. Drafting of the manuscript: Olsen and Fraser. Critical revision of the manuscript for important intellectual content: Chu-Ongsakul, Brandt, Dietz, Mayfield, and Fraser. Statistical analysis: Olsen. Obtained funding: Fraser.Administrative, technical, and material support: Chu-Ongsakul, Mayfield, and Fraser. Study supervision: Brandt, Dietz, and Fraser. Financial Disclosure: None reported. Funding/Support: This study was supported by Epicenter Prevention Program Cooperative Agreement VR8/CCU715087 from the Centers for Disease Control and Prevention (Dr Fraser). Additional Contributions: Christopher S. Hollenbeak, PhD, and Donald Nichols, PhD, provided insight and comments concerning the creation of the ordinary least-squares and generalized least-squares cost models; and Barton Hamilton, PhD, provided suggestions concerning the use of propensity score methods. References 1. National Nosocomial Infections Surveillance System, National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32 (8) 470- 485PubMedGoogle ScholarCrossref 2. Seltzer MH Partial mastectomy and limited axillary dissection performed as same day surgical procedure in the treatment of breast cancer. Int Surg 1995;80 (1) 79- 81PubMedGoogle Scholar 3. Keidan RDHoffman JPWeese JL et al. Delayed breast abscesses after lumpectomy and radiation therapy. Am Surg 1990;56 (7) 440- 444PubMedGoogle Scholar 4. Tejler GAspegren K Complications and hospital stay after surgery for breast cancer: a prospective study of 385 patients. Br J Surg 1985;72 (7) 542- 544PubMedGoogle ScholarCrossref 5. Platt RZucker JRZaleznik DF et al. Prophylaxis against wound infection following herniorrhaphy or breast surgery. J Infect Dis 1992;166 (3) 556- 560PubMedGoogle ScholarCrossref 6. Rotstein CFerguson RCummings KMPiedmonte MRLucey JBanish A Determinants of clean surgical wound infections for breast procedures at an oncology center. Infect Control Hosp Epidemiol 1992;13 (4) 207- 214PubMedGoogle ScholarCrossref 7. Tran CLLanger SBroderick-Villa GDiFronzo LA Does reoperation predispose to postoperative wound infection in women undergoing operation for breast cancer? Am Surg 2003;69 (10) 852- 856PubMedGoogle Scholar 8. Bertin MLCrowe JGordon SM Determinants of surgical site infection after breast surgery. Am J Infect Control 1998;26 (1) 61- 65PubMedGoogle ScholarCrossref 9. Chen JGutkin ZBawnik J Postoperative infections in breast surgery. J Hosp Infect 1991;17 (1) 61- 65PubMedGoogle ScholarCrossref 10. Vinton ALTraverso LWZehring RD Immediate breast reconstruction following mastectomy is as safe as mastectomy alone. Arch Surg 1990;125 (10) 1303- 1308PubMedGoogle ScholarCrossref 11. Morimoto KKinoshita H Once-daily use of ofloxacin for prophylaxis in breast cancer surgery. Chemotherapy 1998;44 (2) 135- 141PubMedGoogle ScholarCrossref 12. Broadwater JREdwards MJKuglen CHortobagyi GNAmes FCBalch CM Mastectomy following preoperative chemotherapy: strict operative criteria control operative morbidity. Ann Surg 1991;213 (2) 126- 129PubMedGoogle ScholarCrossref 13. Beatty JDRobinson GVZaia JA et al. A prospective analysis of nosocomial wound infection after mastectomy. Arch Surg 1983;118 (12) 1421- 1424PubMedGoogle ScholarCrossref 14. Lipshy KANeifeld JPBoyle RM et al. Complications of mastectomy and their relationship to biopsy technique. Ann Surg Oncol 1996;3 (3) 290- 294PubMedGoogle ScholarCrossref 15. Nieto ALozano MMoro MTKeller JCarralafuente C Determinants of wound infections after surgery for breast cancer. Zentralbl Gynakol 2002;124 (8-9) 429- 433PubMedGoogle ScholarCrossref 16. Hayes JABryan RM Wound healing following mastectomy. Aust N Z J Surg 1984;54 (1) 25- 27PubMedGoogle ScholarCrossref 17. Badr el Din ACoibion MGuenier C et al. Local postoperative morbidity following pre-operative irradiation in locally advanced breast cancer. Eur J Surg Oncol 1989;15 (6) 486- 489PubMedGoogle Scholar 18. Say CCDonegan W A biostatistical evaluation of complications from mastectomy. Surg Gynecol Obstet 1974;138 (3) 370- 376PubMedGoogle Scholar 19. Alderman AKWilkins EGKim HMLowery JC Complications in postmastectomy breast reconstruction: two-year results of the Michigan Breast Reconstruction Outcome Study. Plast Reconstr Surg 2002;109 (7) 2265- 2274PubMedGoogle ScholarCrossref 20. Vandeweyer EDeraemaecker RNogaret JMHertens D Immediate breast reconstruction with implants and adjuvant chemotherapy: a good option? Acta Chir Belg 2003;103 (1) 98- 101PubMedGoogle Scholar 21. Van Geel ANContant CMWai RTSchmitz PIEggermont AMMenke-Pluijmers MM Mastectomy by inverted drip incision and immediate reconstruction: data from 510 cases. Ann Surg Oncol 2003;10 (4) 389- 395PubMedGoogle ScholarCrossref 22. Nahabedian MYTsangaris TMomen BManson PN Infectious complications following breast reconstruction with expanders and implants. Plast Reconstr Surg 2003;112 (2) 467- 476PubMedGoogle ScholarCrossref 23. Reilly JTwaddle SMcIntosh JKean L An economic analysis of surgical wound infection. J Hosp Infect 2001;49 (4) 245- 249PubMedGoogle ScholarCrossref 24. Boyce JMPotter-Bynoe GDziobek L Hospital reimbursement patterns among patients with surgical wound infections following open heart surgery. Infect Control Hosp Epidemiol 1990;11 (2) 89- 93PubMedGoogle ScholarCrossref 25. Zoutman DMcDonald SVethanayagan D Total and attributable costs of surgical-wound infections at a Canadian tertiary-care center. Infect Control Hosp Epidemiol 1998;19 (4) 254- 259PubMedGoogle ScholarCrossref 26. Kirkland KBBriggs JPTrivette SLWilkinson WESexton DJ The impact of surgical-site infections in the 1990s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol 1999;20 (11) 725- 730PubMedGoogle ScholarCrossref 27. Jenney AWJHarrington GARusso PLSpelman DW Cost of surgical site infections following coronary artery bypass surgery. ANZ J Surg 2001;71 (11) 662- 664PubMedGoogle ScholarCrossref 28. Hollenbeak CSMurphy DDunagan WCFraser VJ Nonrandom selection and the attributable cost of surgical-site infections. Infect Control Hosp Epidemiol 2002;23 (4) 177- 182PubMedGoogle ScholarCrossref 29. Whitehouse JDFriedman DKirkland KSRichardson WJSexton DJ The impact of surgical-site infections following orthopedic surgery at a community hospital and a university hospital: adverse quality of life, excess length of stay, and extra cost. Infect Control Hosp Epidemiol 2002;23 (4) 183- 189PubMedGoogle ScholarCrossref 30. Perencevich ENSands KECosgrove SEGuadagnoli EMeara EPlatt R Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis 2003;9 (2) 196- 203PubMedGoogle ScholarCrossref 31. Deyo RACherkin DCCiol MA Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45 (6) 613- 619PubMedGoogle ScholarCrossref 32. Charlson MEPompei PAles KMacKenzie CR A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40 (5) 373- 383PubMedGoogle ScholarCrossref 33. Wooldridge JM Introductory Econometrics: A Modern Approach. Mason, OH Thomson Learning2003; 34. Rubin DB Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127 (8, pt 2) 757- 763PubMedGoogle ScholarCrossref 35. Austin PCGhali WATu JV A comparison of several regression models for analysing cost of CABG surgery. Stat Med 2003;22 (17) 2799- 2815PubMedGoogle ScholarCrossref 36. Smith HL Matching with multiple controls to estimate treatment effects in observational studies. Sociol Methodol 1997;27325- 353Google ScholarCrossref 37. Ming KRosenbaum PR Substantial gains in bias reduction from matching with a variable number of controls. Biometrics 2000;56 (1) 118- 124PubMedGoogle ScholarCrossref 38. Ghafoor AJemal AWard ECokkinides VSmith RThun M Trends in breast cancer by race and ethnicity. CA Cancer J Clin 2003;53 (6) 342- 355PubMedGoogle ScholarCrossref
Archives of Surgery – American Medical Association
Published: Jan 1, 2008
Keywords: cancer,surgical procedures, operative,surgical wound infection,breast surgery,breast,reconstructive surgical procedures,breast reconstruction with tram flap,mastectomy,breast cancer,breast conserving surgery,infections,reduction mammaplasty,patient readmission,risk reduction
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera