Introduction to Surgical-Site Infections

Surgical-site infections (SSIs) are an increasingly important postoperative problem known to complicate 2% to 5% of all surgeries in the United States, which translates to an estimated 500,000 postoperative SSIs.^[1] Clinicians and providers are facing ever-increasing pressure to reduce the incidence and the burden of SSIs. At the same time, the epidemiology and the management paradigm of SSIs are steadily changing. SSIs are an active area of research, with data constantly emerging on new treatment and management strategies aimed at reducing incidence and prevalence. The latest data from the combined meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) and Infectious Diseases Society of America (IDSA) are put into perspective in this report, focusing on clinical questions central to preventing infection and improving infection control.

Update on Risk Factors for SSI

Factors including type of surgery, age, comorbid conditions, and patient-specific variables increase the risk for SSI. Outcomes of patients who underwent colorectal surgery were analyzed, including 307 cases of inflammatory bowel disease and 273 cases of colon cancer.^[2] The overall rate of SSI found was 12.6%, higher than rates typically reported by the Centers for Disease Control and Prevention (CDC).^[3] Compared with patients undergoing bowel surgery for colon cancer, patients with inflammatory bowel disease were not at higher risk for overall SSI. Patients with this disease were, however, at higher risk for incisional SSI (odds ratio [OR], 2.57; 95% confidence interval [CI], 1.32-4.97), but the definitions used for subtypes of SSIs were unclear.^[4]

Independent risk factors for postoperative meningitis after craniotomy were identified by analyzing data from 456 consecutive craniotomies performed at a hospital in Greece,^[5] including 38 cases of meningitis. Compared with uninfected patients, patients with postoperative meningitis were more likely to have a surgical drain in place, have undergone a revision surgery, or have another infection (most often ventilator-associated pneumonia).

Key Clinical Question: Is functional status a predictor of SSI in the elderly?

A subanalysis of a previously published study^[6] was conducted to examine risk factors for SSI due to methicillin-resistant Staphylococcus aureus (MRSA) in elderly patients treated in 7 hospitals.^[7] Elderly infected patients were compared with 2 sets of controls: patients with SSI due to methicillin-sensitive S. aureus and uninfected patients. Compared with both sets of controls, elderly patients who required assistance with 3 or more activities of daily living were at higher risk for SSI due to MRSA. Therefore, functional status may predict risk for SSI in the elderly.

Key Clinical Question: Is the presence of a urinary catheter before surgery a risk factor for SSI?

A retrospective, case-control study was conducted at a community hospital to determine specific risk factors for gram-negative SSIs.^[8] The researchers identified 101 patients with gram-negative SSI during a 10-year period and matched these cases by surgeon and procedure. Patients with an antibiotic allergy were found to be less likely to receive appropriately timed antimicrobial prophylaxis than were controls (P = .03). Independent predictors of gram-negative SSI included previously identified variables such as diabetes mellitus and American Society of Anesthesiologists (ASA) score. In addition, the presence of a urinary catheter before surgery was a strong predictor of gram-negative SSI (OR, 3.74; 95% CI, 1.11-12.62). Although the independent predictors identified could simply be predictors of overall SSI, the finding of urinary catheter as a specific risk factor for gram-negative SSI is plausible and requires further study.

Key Clinical Question: Are pens used to mark the patient's skin in the operating room a vector for transmission of multidrug-resistant organisms?

Researchers contaminated 2 different types of pens (Sharpie [Sanford Corporation, Oak Brook, Illinois] and Securline [Thermo Fisher Scientific, Inc., Waltham, Massachusetts]) with a standardized inoculum of MRSA, vancomycin-resistant enterococci (VRE), Escherichia coli, or Pseudomonas species. The pens were subsequently swiped across blood agar plates at 8 different time points from 0 minutes to 1 week.^[9] All contaminated markers led to bacterial growth initially, but only 1 Sharpie pen exhibited further growth of VRE at 24 hours. Securline markers led to growth of all organisms up to 4 hours and growth of E coli and Pseudomonas species up to 24 hours after contamination with inocula. The researchers concluded that Sharpie-brand pens led to no increased risk for transmission in the operating room but that Securline pens should be used only once.

New SSI Surveillance Data

Key Clinical Question: Can SSI surveillance be implemented to identify improvements in patient outcomes?

A Web-based reporting and analysis program to assist with SSI surveillance in patients undergoing hip and knee arthroplasty and gastrectomy was piloted for 13 months in several hospitals in South Korea.^[10] Data were collected on demographic, clinical, and operative variables, and patients with SSI were compared with patients without SSI to identify opportunities for improvement. The researchers concluded that this program could be expanded for use throughout Korea, much like the National Healthcare Safety Network (NHSN) system is used in the United States.

Key Clinical Question: What is the correlation of the NNIS risk index score with SSI?

Canadian researchers replicated the CDC’s NNIS/NHSN system by pooling data on 469,349 elderly surgical patients from multiple administrative databases. These included province-wide hospital discharge databases, physician claims databases, and anesthesiologist billing databases.^[11] They were able to successfully abstract duration of surgery and ASA score for each patient. Assuming all patients underwent clean or clean-contaminated procedures, the researchers calculated an NNIS risk index score (0, 1, and 2) for all procedures. Analysis showed that the risk for SSI increased incrementally with each NNIS risk index score (0 = 2% risk, 1 = 4% risk, and 2 = 9% risk).

Key Clinical Question: Does feedback on surgeon performance have an impact on rate of SSI?

In addition to standard surveillance based on CDC definitions, French researchers monitored best practices based on national guidelines and provided immediate feedback of SSI rates to surgeons.^[12] The investigators concluded that the following factors led to lower rates of SSI: use of active surveillance, rapid feedback to surgeons, and monitoring for compliance with national guidelines. However, it is unclear whether individual interventions led to a decrease or whether all interventions were synergistic.

MRSA-Related SSI Outcomes

Key Clinical Question: Are outcomes of MRSA-related SSIs different?

A multicenter study of 8302 patients with SSI at 97 US hospitals was conducted from 2003 to 2007.^[13] Results show that the proportion of MRSA-related SSIs increased from 16% to 21% during the study period. The researchers compared outcomes of MRSA-related SSIs with outcomes of SSIs due to other pathogens. Outcomes were worse in patients with MRSA-related SSI. In addition, adjusted length of stay was increased by 1 day (P < .001) and cost was increased by $1023 (P < .001).

Two retrospective, multicenter, case-control studies also examined outcomes in patients with MRSA-related SSIs.^[14,15] The first study compared 150 patients who had MRSA-related SSI with 231 uninfected surgical patients matched by hospital, type of procedure, and year of procedure.^[14] The second study compared the same set of patients who had MRSA-related SSI with128 patients who had methicillin-susceptible S aureus (MSSA)-related SSI.^[15] Compared with uninfected patients, those with MRSA-related SSI were:

• Hospitalized for 16 additional days;
• 3 times more likely to be discharged to a care facility;
• 7 times more likely to die; and
• 30 times more likely to be readmitted to the hospital after discharge.

For all measures, the P values were < .001.

Compared with patients who had MSSA-related SSI, those with MRSA-related SSI were:

• Hospitalized for 6 additional days;
• 2 times more likely to be discharged to a facility; and
• 2.6 times more likely to die.

These results demonstrate differences in patients with MRSA vs MRSA-related SSIs and uninfected patients, which affect outcomes and mortality.

Preventing SSIs With Antibiotic Prophylaxis

Key Clinical Question: Does the timing of initiation of antibiotic prophylaxis matter?

The Centers for Medicare & Medicaid Services created Surgical Infection Prevention (SIP) in 2002 in an effort to decrease the incidence of morbidity and mortality associated with postoperative SSI.^[16] SIP includes 3 key performance measures related to antibiotic prophylaxis:

• Initiation of intravenous antimicrobial prophylaxis within 1 hour before the initial incision (2 hours are allowed for the administration of vancomycin and fluoroquinolones);
• Use of an antimicrobial prophylactic agent consistent with published guidelines; and
• Discontinuation of the prophylactic antimicrobial agent within 24 hours after surgery end time.

Many US hospitals and healthcare systems around the world have adopted the SIP guidelines. However, recent findings show that adherence to these performance measures is low,^[17-19] particularly the recommendation that antibiotic prophylaxis be initiated within 1 hour before the initial incision.^[19]

A retrospective, multicenter study examined compliance to guidelines for optimal surgical prophylaxis.^[20] A total of 2441 cases of 3 common surgical procedures (total hip replacement, colectomy, and hysterectomy) from 26 hospitals in France were analyzed. Compliance with recommendations for surgical prophylaxis was approximately 90% for total hip replacement surgery but was less for colectomy and hysterectomy. Antibiotics were initiated after the incision in 23% of colectomy and 22% of hysterectomy procedures. Furthermore, antibiotics that were not recommended in the guidelines were administered in more than one third of the hysterectomy procedures (36%). A multicenter Italian study of compliance with preoperative antibiotic prophylaxis showed similar results.^[21] Ninety-nine (20%) of the 494 evaluable surgical patients who underwent general, orthopedic, vascular, neurosurgical, and gynecologic surgeries did not receive the first dose of the indicated antibiotic within 1 hour of the initial surgical incision. Furthermore, 21% of patients received the wrong antibiotic or an inadequate dose.

Improving Compliance Through Workflow Redesign

Key Clinical Question: What does it take to improve compliance with practice guidelines?

Although the principles of surgical antimicrobial prophylaxis are well established, adherence to guideline recommendation is suboptimal. A top-down approach to restructuring workflow in the perioperative setting is necessary, and a clinically initiated program is needed to educate and give feedback to healthcare providers to improve adherence to surgical prophylaxis guidelines. The following is an example of a quality improvement initiative that designated roles and accountability for specific healthcare workers, made systematic changes to the workflow of antibiotic prophylaxis, and provided education and feedback to overcome barriers to complying with guidelines.

Staff at the Tufts New England Medical Center in Boston convened a committee consisting of an infectious disease physician, an anesthesiologist, a cardiac surgeon, and epidemiology, operating room nursing, and quality improvement staff. This oversight committee overhauled the existing system of antibiotic prescribing and administration in the operating room. Anesthesiologists were given the responsibility for appropriate antibiotic selection and were given pocket cards that detailed the appropriate antibiotic choices by surgical procedures. The rate of compliance with SIP guidelines increased to 93% within 1 year.^[22]

Institutional and policy changes included the following:

• Administering antibiotics only in the operating room to avoid premature administration;
• Documenting time of antibiotic administration and any deviations from recommended guidelines on printed, preformatted operative notes;
• Verification by nurses of items on a checklist that linked the administration of antibiotics to the “Time Out” so that the surgical procedure could not begin until antibiotics had been given; and
• Periodic clinician education on postoperative orders to reduce antibiotic use beyond 24 hours after surgery.

The researchers concluded that the improvements were due to systematic changes in the hospital and the operating room, institution of a written protocol, and education of staff.

Preoperative Use of HMG-CoA Reductase Inhibitors

Key Clinical Question: Does statin therapy decrease the risk of SSI?

Activities of 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase inhibitors (statins) have been shown to have anti-inflammatory and even antimicrobial effects in vitro.^[23,24] The mechanisms by which statins mediate these protective effects have been recently elucidated. Statins reduce endothelial production and release of proinflammatory cytokines such as interleukin-6 and tumor necrosis factor alpha.^[25] In addition, statins inhibit bacterial entry into cells by inhibiting acting-mediated endocytosis.^[26] In a study of 469,349 elderly patients undergoing elective surgery in from 1992 to 2006, investigators tested preoperative therapy with HMG CoA reductase inhibitors and associated risk for SSI.^[11] A total of 68,387 (15%) patients received preoperative statin therapy; 53,565 of these cases were matched to 53,565 patients who underwent similar elective procedures but did not receive statin therapy. After adjustment for demographic characteristics, health resource utilization, medications, and comorbid conditions, statins were not shown to be associated with an increased or decreased risk for SSI (OR, 1.00; 95% CI, 0.94-1.06).

Treatment of Prosthesis-Related and Sternal Wound Infections

Key Clinical Question: What is the latest data on difficult-to-treat infections?

A retrospective case series examined 26 patients with documented acute hardware infection following prosthesis insertion who received daptomycin-based therapy.^[27] Sixteen (61%) patients had an infected hip or knee joint prosthesis and were infected with MRSA or MSSA. The median dose of daptomycin was 6 mg/kg, and the median duration of treatment was 42 days (range, 14-83 days). All patients had baseline and posttreatment creatine phosphokinase levels drawn, and the study endpoint was the proportion of hardware retention at 12 months. Twelve-month follow-up showed no daptomycin-related adverse events. Twenty-two (85%) patients retained their prostheses. Of note, however, 18 (82%) patients with retained hardware at 12 months received suppressive antibiotic therapy when daptomycin treatment ended.

A prospective case series of linezolid plus rifampicin as salvage therapy in 49 patients with prosthetic joint infection was analyzed.^[28] Many of the patients who received linezolid and rifampicin also received and did not respond to therapy with teicoplanin. The mean duration of treatment with linezolid plus rifampicin was 3 months. The proportion of patients undergoing surgical washout and debridement was not provided. By using clinical criteria of cure, the proportion of patients who successfully retained their prostheses in the linezolid plus rifampicin treatment group was 69.3% after 1 year of follow-up. Two patients developed thrombocytopenia and 30 developed mild anemia, but these conditions were not dose-limiting.

Rifampicin is attracting increasing interest as a potential adjunctive therapy for difficult-to-treat gram-positive infections. A large10-year, retrospective, case-cohort study investigated patients with deep sternal wound infection following cardiothoracic surgery. Coagulase-negative staphylococci and S aureus accounted for nearly 83% of all sternal wound infections.^[29] Notably, Enterobacteriaceae organisms were implicated in 11% of cases. All patients were treated with antibiotics, and 95% were also treated with surgical debridement. The median duration of antibiotic treatment was 48 days (interquartile range [IQR], 41-73 days). According to clinical criteria, treatment failed in 26patients; 9 of these patients died of sepsis (median time to death, 11 days [IQR, 10-27] after diagnosis). The 1-year rate of cure of sternal wound infections was 78.3%. In the multivariate analysis, after adjustment for age, microorganism, and treatment duration, an antibiotic regimen containing rifampicin was strongly associated with lower risk for treatment failure (hazard ratio [HR], 0.3; 95% CI, 0.1-0.7; P = .009). Outcomes did not differ between treatment durations of 6 weeks and 3 months (HR, 0.7; 95% CI, 0.2-2.2; P =.511).

References