Breast disease and breast cancer management form a major part of healthcare delivery, constituting one of the most frequent elective surgeries performed globally  . Uncomplicated surgical outcomes are important in optimising functional, cosmetic and oncological outcomes. Surgical site occurrence (SSO) including wound infection, wound dehiscence and deep infection, with or without implant loss pose septic and oncological risks to patients    . There is a spectrum from minor wound infection to implant loss with increasing costs for the health care system   . Reducing SSO will benefit patient’s physical and psychological outcomes, facilitate adjuvant treatment, and optimise long term cosmetic and oncological outcomes   .
Wound infection and adverse wound events are multifactorial    . Identifying the relative importance of the contributing factors is challenging. A bundle approach to wound care has been shown to facilitate better outcomes    . There are few reports of the use of bundled approaches to reducing SSO in breast surgery  . The National Mastectomy Audit suggests that current rates of SSO are unacceptable  .
The aim of this meta-analysis is to identify key risk factors and interventions that may alter the incidence of SSO in patients undergoing breast surgery.
2.1. Search Strategy and Study Eligibility
An ethically approved meta-analysis of the literature was undertaken to incorporate articles relating to breast wound care, breast wound infection, breast surgical site adverse outcomes, infected related implant loss, and return to the operating theatre. Existing research optimising wound care in surgery was reviewed to determine current strategies to improve wound outcomes. Key risk factors and interventions for SSO were identified in three keys phases of care, pre-, intra- and post-operative periods.
The methods of analysis and inclusion criteria were specified in advance to avoid selection bias and documented in a protocol which was registered and published with the International Prospective Register of Systematic Reviews (PROSPERO) (ID 42016039883). This meta-analysis adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement  and Cochrane Handbook for Systematic Reviews of Interventions  .
A systematic review and meta-analysis of all published English articles was conducted using PubMed, Scopus and Cochrane Library electronic databases from 2010 to 2017. Medical Subject Headings used terms which included risk factors (“risk factor*”), surgical site infections (“surgical site infection*”, “wound infection*”), breast surgery (“breast surg*”), and interventions (“intervention*”). The following search strategies were used in our meta-analysis: (“risk factor*” AND “breast surg*” AND “infection*”), (“intervention*” AND “breast surg*” AND “infection*”), (“risk factor*” OR “intervention*” AND “breast surg*” AND “surgical site infection*”), and (“risk factor*” OR “intervention*” AND “breast surg*” AND “wound infection*”). Studies that were case studies or meta-analysis, not related to breast surgery, did not report key outcomes, or where data was inadequate for interpretation via meta-analysis, or duplicate studies were excluded.
Eligibility assessment was performed independently in a blinded standardised manner by two reviewers (SV and MG). Disagreements between reviewers were resolved by discussion between the two review authors and if no agreement could be reached, it was planned a third reviewer (AJ) would decide.
2.2. Data Extraction and Quality Assessment
We developed a standardised data extraction sheet and one reviewer (SV) extracted the following data from included studies and the second reviewer (MG) checked the extracted data. Discrepancies were resolved by discussion and consultation with another reviewer (AJ). Two reviewers (SV and MG) independently assessed each published study for the quality of study design by using the Methodological Index for Nonrandomised Studies (MINORS) score whereby the global ideal score is 16 for non-comparative studies and 24 for comparative studies  . Articles scoring > 10 for non-comparative studies and >15 for comparative studies, using MINORS criteria were included in the final analysis (Table S1). Risk of bias across studies was not assessed as there were too few included studies per outcome.
Information was extracted from each included study on: 1) Characteristics of participants 2) Inclusion and exclusion criteria 3) Risk factor or type of intervention 4) Well-reported outcome measurements (including a clear report of surgical site infections or breast wound infections).
2.3. Data Synthesis and Analysis
The odds ratio (OR) and risk ratio (RR) of surgical site infections (SSI) were the primary measure of risk factors and intervention effect respectively. The meta-analyses were performed by computing the OR or RR using Mantel-Haenszel method and random-effects model to combine variables of interest. OR or RR and 95% Confidence Intervals (CI) for each risk factor and intervention were calculated. No additional analyses were done. The analysis was performed by using Review Manager Version 5  .
The Centres for Disease Control and Prevention (CDC) definitions   for surgical site infection were used. These are classified into superficial, deep or organ/space relating to implant/infection. The ASEPSIS (Additional treatment, Serous discharge, Erythema, Purulent exudate, Separation of deep tissues, Isolation of bacteria, Stay duration as inpatient > 14 days) scoring system was also used to quantify surgical site infections  .
WHO classification of nutritional status according to Body Mass Index (BMI)  was used to categorise underweight (BMI < 18.5), normal weight (BMI 18.5 - 24.9), pre-obesity (BMI 25.0 - 29.9), obesity class 1 (BMI 30.0 - 34.9), obesity class 2 (BMI 35.0 - 39.9), and obesity class 3 (BMI > 40). Patients were considered to have diabetes mellitus only if they were taking oral hypoglycaemic agents and/or on insulin. Patients were considered smokers if they were currently smokers or had smoked cigarettes in the year before admission for surgery. The tumour-node-metastasis (TNM) classification  for breast cancer was used to determine breast cancer staging. The American Society of Anesthesiology (ASA) Physical Status classification  was defined as follows: 1) normal healthy patient; 2) mild systemic disease; 3) patient with severe systemic disease; 4) severe systemic disease with constant threat to life; and 5) moribund patient not expected to survive without surgery. Neoadjuvant chemotherapy treatment was defined as administration of chemotherapeutic agents for cancer within 30 to 90 days prior to surgery        . Neoadjuvant radiotherapy included patients who had treatment within 90 days before surgery. Breast reoperations were defined as re-excision or mastectomy within 180 days of initial breast surgery.
To determine the level of risk, a number of classification systems were used in this study. These included the surgical wound classification   and the National Nosocomial Infection Surveillance (NNIS) Risk Index  .
This meta-analysis reviewed 1606 articles for risk factors and interventions of SSI in breast surgery. 64 studies were found to be suitable after eligibility analysis and 49 studies were included for quantitative analysis of this meta-analysis (Figure 1). Characteristics of the studies included in this meta-analysis are listed in Table S2. Significant and insignificant factors affecting breast SSO that are not included in the quantitative meta-analysis are listed in Table 1 and Table 2 respectively.
A number of statistical significant causative factors and interventions for SSI in breast surgery in the key phases of care were identified; seven in pre-operative, three in intra-operative and four in post-operative.
Table 1. Significant factors for breast SSO not included in the quantitative meta-analysis.
Table 2. Insignificant factors for breast SSO not included in the quantitative meta-analysis.
Figure 1. Prisma flow diagram.
3.1. Pre-Operative Phase
Significant pre-operative risk factors (Figure 2 and Figure 3) for developing SSI in breast wounds are class 3 obesity versus non-obese (OR = 2.90, CI = 2.10 - 3.99, p < 0.001), diabetes mellitus (OR = 2.52, CI = 1.78 - 3.59, p < 0.001), smoking (OR = 2.39, CI = 1.57 - 3.63, p < 0.001), American Society of Anesthesiologists (ASA) Physical Status classification ≥III (OR = 2.37, CI = 1.51 - 3.74, p < 0.001), class 2 obesity versus non-obese (OR = 1.70, CI = 1.07 - 2.70, p = 0.03), class 1 obesity when compared to non-obese and normal BMI respectively (OR = 1.61, CI = 1.40 - 1.86, p < 0.001; OR = 1.93, CI = 1.59 - 2.35, p < 0.001), overweight versus normal BMI (OR = 1.66, CI = 1.15 - 2.40, p = 0.007), and hypertension (OR = 1.63, CI = 1.39 - 1.90, p < 0.001). Overall, being overweight or obese versus normal BMI and being obese versus non-obese was significant for increasing the incidence of SSI (OR = 1.70, CI = 1.36 - 2.13, p < 0.001; OR = 1.84, CI = 1.52 - 2.24, p < 0.001 respectively). Interventions shown to be statistically significant in reducing surgical site infections in breast surgery (Figure 6) is antibiotics prophylaxis (RR = 0.58, CI = 0.36 - 0.95, p = 0.03).
Insignificant pre-operative risk factors are neoadjuvant radiotherapy (OR = 1.26, CI = 0.55 - 2.89, p = 0.58), neoadjuvant chemotherapy (OR = 0.96, CI =
Figure 2. (a) Obese vs non obese; (b) Overweight/obese vs normal.
Figure 3. (a) Diabetes; (b) Smoking; (c) ASA; (d) Hypertension.
0.84 - 1.10, p = 0.55), age ≥ 50 years old (OR = 1.26, CI = 0.97 - 1.64, p = 0.09) and steroids use (OR = 1.04, CI = 0.81 - 1.32, p = 0.78). Hair removal (RR = 1.26, CI = 0.46 - 3.44, p = 0.66) was not shown to be statistically significant in reducing breast SSI.
3.2. Intra-Operative Phase
Significant intra-operative risk factors (Figure 4) are surgical wound classifications 3 or 4 (OR = 6.16, CI = 2.52 - 15.02, p < 0.001), the use of surgical drains (OR = 2.80, CI = 1.06 - 7.38, p = 0.04), and axillary lymph node dissection (ALND) (OR = 1.46, CI = 1.18 - 1.80, p < 0.001).
Figure 4. (a) Surgical wound class; (b) Surgical drains; (c) ALND.
Insignificant intra-operative risk factors are inpatient admission (OR = 3.59, CI = 0.18 - 72.11, p = 0.40), operative time > 2 hours (OR = 2.87, CI = 0.32 - 25.47, p = 0.34), immediate breast reconstruction (IBR) versus mastectomy only (OR = 2.66, CI = 0.72 - 9.83, p = 0.14), IBR versus delayed reconstruction (OR = 1.39, CI = 0.73 - 2.64, p = 0.32), a cellular dermal matrix (ADM) use (OR = 1.32, CI = 0.22 - 8.06, p = 0.06), breast cancer stage II - IV versus breast cancer stage 0 - I (OR = 1.24, CI = 0.54 - 2.88, p = 0.61), breast cancer versus prophylactic stage (OR = 1.10, CI = 0.82 - 1.47, p = 0.53), and sentinel lymph node biopsy (SLNB) (OR = 0.46, CI = 0.06 - 3.57, p = 0.46).
3.3. Post-Operative Phase
Adjuvant radiotherapy (OR = 1.77, CI = 1.26 - 2.50, p = 0.001) and re-operated patients (OR = 1.65, CI = 1.01 - 2.70, p = 0.05) are significant post-operative risk factors (Figure 5). Interventions shown to be statistically significant in reducing surgical site infections in breast surgery (Figure 6) are post-operative antibiotics (RR = 0.57, CI = 0.33 - 0.98, p = 0.04), and drain antisepsis care (RR = 0.15, CI = 0.03 - 0.82, p = 0.03).
Adjuvant chemotherapy (OR = 1.98, CI = 0.97 - 4.06, p = 0.06) was found to be an insignificant postoperative risk factor for breast SSI. Duration of post-operative antibiotics ≥24 hours versus <24 hours (RR = 0.75, CI = 0.51 - 1.10, p = 0.14) and the administration of antibiotics until drain removal versus antibiotics for 24 hours (RR = 1.00, CI = 0.56 - 1.80, p = 0.99) were not shown to be statistically significant in reducing SSI in breast surgery.
Figure 5. (a) Adjuvant radiotherapy; (b) Reoperation.
Figure 6. Antibiotics vs none.
This meta-analysis identifies a number of peri-operative factors associated with adverse wound outcomes. Given the volume of breast surgery, both benign and malignant, reduction of adverse wound outcomes is vital; particularly SSO. The importance of preventing SSO is critical in patients having breast surgery (mastectomy) with reconstruction using alloplastic (implant) material and anticipated to have adjuvant chemotherapy and/or radiation treatment. In patients having SSO, their outcome can be compromised as the timing to proceed with necessary chemotherapy or radiation could be delayed, with potential reduction on survival.
Defining and collecting surgical site infection data is somewhat problematic. There have been many reviews of the nomenclature of wound complications. Terms such as surgical site occurrence (SSO) were introduced in 2010  . Current definitions of SSO are subject to debate in many areas of surgery. The standardized definition of an SSI, developed by the Centres for Disease Control and Prevention (CDC), is an infection occurring in part of the body where surgery took place, including superficial, deep, and organ space infection  . It has been suggested that some SSIs are not relevant and in an effort to add more transparency, the term “surgical site occurrences requiring procedural interventions” (SSOPI) has been introduced recently  . Another term that has been used is “surgical site event”  .
Post-operative breast surgery infections even when delayed, or initially thought to be indolent, can be devastating with implant loss or delays in adjuvant treatment. Recently it has been suggested that SSIs following breast cancer surgery decrease oncological survival   , while others do not support this   .
Disregarding considerations regarding the reliability of definitions and surveillance, it is clear that understanding risk factors for infection is crucial to preventing SSO and optimising care. The cost of SSI after breast surgery was reported at $4091 by one study  . The care bundle is not a novel concept, but it is integral for the provision of a team-based approach to patient care. The risk factors identified by this study were grouped into pre-operative, operative and post-operative, which may aid in a tailored approach to intervention.
Patient factors such as obesity and the degree of obesity, smoking, diabetes, recent surgery, and anaesthetic risk will significantly increase the SSI risk. The Breast Care team needs to consider these factors while initially tailoring an optimal surgical strategy and even consider modifying the use of implant based reconstruction in these high-risk patients.
Neo-adjuvant chemotherapy is increasingly used and is not associated with an increased SSI risk.
This meta-analysis has identified significant risk factors for developing breast related adverse surgical site infections. Planned strategy to mitigate against these should be incorporated into Breast Surgery Care Bundles. Current SSO levels in breast surgery are unacceptably high and need to be addressed. Incorporating wound bundles as key reportable performance indicators or as a mandatory field in oncologist registries may encourage their wider adoption. This may help address concerns expressed about the incidence of wound infection mastectomy  .
M. S., S. V., and A. J. contributed to the idea conception. S. V., M. G., and A. J. did the literature search and analysis. S. V., M. S., R. D. and M. V. did the co-writing and editing of the article. S. V., M. G., and A. J. did the statistical analyses and interpretations. M. S., R. D., M.V., S. V., and A. J. did the approval of final article submission.
This study was ethically approved by the Galway University Hospitals Research Ethics Committee.
This study was supported by funding from the School of Medicine, National University of Ireland, Galway; Breast Development Fund Letterkenny; and Donegal Clinical Research Academy (DCRA).
Table S1. MINORS grading of studies included in the meta-analysis.
Table S2. Characteristics of studies included in the meta-analysis.
ADM: Acellular Dermal Matrix,
ALND: Axillary Lymph Node Dissection,
AMP: Antimicrobial Prophylaxis,
ASA: American Society of Anesthesiologists Physical Classification System,
BCS: Breast Conserving Surgery,
BMI: Body Mass Index,
BR: Breast Reconstruction,
CDC: Centers for Disease Control and Prevention Guideline for Prevention of Surgical Site Infections,
CI: Confidence Interval,
DCIS: Ductal Carcinoma in Situ,
DR: Delayed Reconstruction,
ECOG: Eastern Cooperative Oncology Group performance status,
IBBR: Implant-Based Breast Reconstruction,
IBR: Immediate Breast Reconstruction,
IORT: Intraoperative Radiotherapy,
IR: Immediate Reconstruction,
ITEBR: Immediate Tissue Expander-Based Breast Reconstruction,
MINORS: Methodological Index for Nonrandomised Studies,
MLD: Manual Lymphatic Drainage,
MRM: Modified Radical Mastectomy,
NNIS: National Nosocomial Infection Surveillance,
NPWT: Negative Pressure Wound Therapy,
OR: Odds Ratio,
RCT: Randomised Controlled Trial,
PI: Permanent Implant,
RR: Risk Ratio,
SLNB: Sentinel Lymph Node Biopsy,
SR: Subsequent Reconstruction,
SSI: Surgical Site Infection,
SSO: Surgical Site Occurrence,
TE: Tissue Expander,
TEBR: Tissue Expander-Based Reconstruction,
TM: Total Mastectomy,
TNM: Tumour Node Metastasis classification,
WLE: Wide Local Excision.