Evidence Links Nicotine Use to Surgical Site Infections

Every day in acute care hospitals in the United States, at least 1 patient in 25 experiences a health care-related infection.¹   Among these, 21% are surgical site infections (SSI), mainly related to Staphylococcus aureus, Escherichia coli, Klebsiella, or Enterococcae species.¹ Similar SSI prevalence (26%) and infecting organism patterns were reported in a consecutive cohort of 250 patients undergoing major surgery in Tanzanian surgical wards² where logistic regression identified high odds ratios (OR) for significant risk factors predicting an SSI. If the sole skin preparation for surgery was iodine, the likelihood of an SSI was multiplied by 17.6. If a wound drain was used, an SSI was 15.3 times as likely. This was confirmed by a meta-analysis of 12 randomized controlled trials of laparascopic cholecystectomy procedures.³ Cigarette smoking multiplied the odds of having an SSI by 9.6. This was more than the effect of having an existing premorbid illness in this Tanzanian sample (OR: 6.1) or of having an operation lasting at least 3 hours (OR: 3.2). Wound care professionals can correct some of these risk factors with more appropriate skin preparation or wound drainage but, as responsible prospective patients, we owe it to ourselves and our loved ones to take charge of our own habits to reduce the SSI risk factors we can change. Tobacco use is a significant risk factor for SSI in a wide range of major surgical procedures,² neurosurgery,⁴ or dermatologic surgery,⁵ plus other complications including wound dehiscence, flap or graft necrosis, or prolonged healing time.⁵ Two more references reviewed in this Evidence Corner add to the compelling evidence linking nicotine use to SSI.^6,7

Laura Bolton, PhD
Adjunct Associate Professor
Department of Surgery, Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ

  Reference: Rao S, Stolle EC, Sher S, Lin CW, Momen B, Nahabedian MY. A multiple logistic regression analysis of complications following microsurgical breast reconstruction. Gland Surg. 2014;3(4):226-231.

  Rationale. Microvascular free flap breast reconstruction (MBR) has improved outcomes compared to traditional breast reconstruction in high-risk patients, but few studies have evaluated a consecutive series of high-risk patients undergoing all types of MBR.

  Objective. Conduct a 5-year retrospective patient record review of a cohort of all high-risk subjects undergoing MBR to better inform surgeons of the impact of specific risk factors on outcomes of this surgery.

  Methods. Records of all patients who underwent MBR by the senior author from July 2005 through July 2010 were analyzed for risk factors (ie, age, body mass index [BMI], smoking history), medical history (ie, diabetes, cancer, hypertension, coronary artery disease, American Society of Anesthesiologists (ASA) Score, connective tissue disease), adjunct therapies (eg, radiation, chemotherapy) type of flap including arteries and perforators involved, and whether the timing of the reconstruction was immediate or delayed. Outcomes investigated included all complications as well as leech use and partial or total flap loss.

  Results. Among 352 consecutive patients undergoing 490 MBR procedures during the 5-year study period, the only significant predictors of wound infection were active smoking (OR 4.3; P = 0.0081) and higher values of BMI (P < 0.00001). Active smoking also increased the likelihood of a patient developing a seroma (OR 36; P < 0.0001) or pneumonia (OR 17.1; P < 0.0001).

  Authors’ Conclusions. This study corroborates prior findings that active tobacco use or obesity are significant risk factors for wound infection and other complications of MBR. Patients choosing to undergo MBR should be informed of the complication profiles of these 2 risk factors.

  Reference: van Walraven C, Musselman Reilly. The Surgical Site Infection Risk Score (SSIRS): a model to predict the risk of surgical site infections. PLOS/One. 2013;8(6):e67167. DOI: 10.1371/journal.pone.0067167.

  Rationale. Surgical site infections (SSI) occur in 2%-5% of the 30 million operations performed in the United States annually, increasing health care costs, patient pain, and readmissions. Accurately quantifying SSI risk would help inform decisions for SSI preventive strategies and facilitate risk-adjusted comparisons among facilities and health care providers.

  Objective. The authors used data from the American College of Surgery National Surgical Quality Improvement Program (ACS-NSQIP) to derive and internally validate a model for predicting whether a patient will develop an SSI within 30 days following surgery.

  Methods. With appropriate ethics board approval, uniformly trained Surgical Clinical Reviewers (SCR) collected standardized de-identified data on subjects at least 18 years of age from participating hospital ACS-NSQIP databases involving a variety of general, vascular, or specialty surgeries stratified across high-volume and low-volume hospitals. Data on all ACS-NSQIP preoperative and intraoperative covariates were considered as possible risk factors as long as there was less than 1% missing data for the variable. The primary outcome was a superficial, deep, or organ space SSI occurring from 2 to 30 days after surgery. Surgical site infection risk was classified within surgery type based on observed records and divided by the expected likelihood of an SSI for surgeries corresponding to the first 3 numbers of the current procedural terminology code (CPT3). Surgeries with CPT3 < 1 were less likely to experience an SSI than those with CPT3 >1. Logistic regression was used to identify covariates predicting an SSI and qualifying them for incorporation into the SSI Risk Score (SSIRS). The C-statistic, with a range from 0 to 1, was used to test if the SSIRS discriminated between patients with an SSI or those without an SSI within 2 to 30 days after surgery. A C-statistic value of 0.5 indicated the SSIRS discrimination was no better than chance alone. The SSIRS was then calibrated by comparing the expected risk of an SSI calculated from the SSIRS model for each patient in a separate validation sample, compared to the actual observed SSI risk for that patient. Finally, performance of the SSIRS model was compared to that of the previously developed National Nosocomial Infection Surveillance (NNIS) Basic SSI Risk Index.

  Results. Of 363,040 qualifying surgeries in the ACS-NSQIP database, 181,894 were used for SSIRS model development and 181,146 were used for model validation. Most patients were middle-aged. About half underwent inpatient elective surgery; one-third, ambulatory surgery; and about one-tenth had emergency surgery. The final SSIRS model had high discriminatory validity (C-statistic = 0.8). Strong independent predictors of an SSI 2-30 days after surgery included CPT3 code, patient variables (ie, smoking, BMI > 35, peripheral vascular disease, metastatic cancer, steroid use, or preoperative sepsis) and procedural variables (ie, inpatient or emergent setting; a contaminated, dirty, or infected operative site; ASA score > 3; use of general instead of local anesthesia; conducting more than 1 procedure during surgery; and operating time either shorter than 30 minutes or longer than 3.5 hours). Surgical site infection prediction was within the 95% confidence interval for the SSIRS on 89.7% of the validation sample and had better discriminatory validation than the NNIS Basic Risk model.

  Authors’ Conclusions. The SSIRS as derived and validated predicts and discriminates SSI risk well across a wide range of surgeries. Further external validation should be pursued, especially incorporating antibiotic prophylaxis as a covariate.

  Both studies reviewed here^6,7 provide compelling evidence that patients acting responsibly could limit their own likelihood of developing an SSI by stopping nicotine use and, if possible, by reducing body weight. These are factors that only patients can control. Equally compelling evidence⁸ reminds us that health care professionals, too, neglect long-recognized basic procedures that prevent infections. These include health care workers’ thorough hand disinfection between patients,⁹ appropriate timing and use of antibiotics,⁹ or using moisture-retentive wound dressings to reduce likelihood of infection in acute or chronic wounds.^10,11 In low-resource settings, health care workers may further reduce the likelihood of an SSI by avoiding sole use of iodine skin cleansing2 or inappropriate percutaneous drain use.^2,3 The less-invasive practice of daily wound probing is an effective alternative to wound drains reported to reduce SSI.¹² It is time for patients to join health care workers by getting back to basics⁸ in stopping the SSI epidemic. If we truly want to reduce SSI, everyone contributing to this epidemic, even patients, should manage controllable risk factors like smoking. We owe it to ourselves, our patients, and their loved ones not to let a single break in the chain of responsibility initiate the distress and expense of an SSI.

This article was not subject to the WOUNDS peer-review process.