After an exhaustive review of the literature for evidence of best practices, the CDC has issued new guidelines1 for the prevention of the most common and costly healthcare-associated infection: surgical site infections (SSIs).

The goal of the CDC’s Healthcare Infection Control Practices Advisory Committee (HICPAC) was to thrash through the thicket of data, controlled trials, and observational studies in an attempt to bring the best evidence to bear on a guideline that had not been updated since 1999. They certainly had some success, but the bar they were bound to was a high one, and many issues fell into the dreaded “no recommendation/unresolved issue” category.

“Sadly, in surgical infections, there are tons of unresolved issues. There just aren’t any studies,” Dale Bratzler, DO, MPH, MACOI, FIDSA, said recently in Portland at the conference of the Association for Professionals in Infection Control and Epidemiology. An internist and chief executive officer of the Oklahoma Foundation for Medical Quality, Bratzler was a member of HICPAC when the SSI guidelines were being revised and was heavily involved in the effort.

“SSIs are the most common and most expensive HAIs reported in hospitals,” he said. “They are extremely costly and have substantial patient morbidity and mortality. The risk factors include operation type, patient host factors, the [individual] surgeon, and other things. No single intervention is going to take care of the problem of surgical site infections.”

Currently, SSIs infect some 160,000 to 300,000 patients annually. In estimates cited in the new CDC SSI guideline, approximately 80 million surgical procedures were performed in the United States at inpatient hospitals and ambulatory hospital-affiliated or freestanding surgical centers in 2006.1

“Two percent to 5% of all operations end up with an SSI, and when that happens the patient is more likely to die, or to get more inpatient and outpatient treatment,” Bratzler said. “There is substantial increase in the length of stay when they are diagnosed, but as you know, more than half of SSIs are not diagnosed until the patient has already gone home, which then results in their readmission.”

With regard to CMS Value-Based Purchasing and other incentivized reimbursement programs, “a single patient with an SSI can impact the payment programs for doctors, hospitals and others through all these various programs,” he said. “It is very substantial and something that you want to prevent.”

While the contributing factor of any given infection to subsequent mortality can be somewhat undefined for some infections, the link to developing an SSI and subsequent death is more firmly established. An estimated 77% of deaths among SSI patients are directly attributable to the infection, Bratzler said. Surgical site infections add 7 to 11 additional postoperative hospital days and cost up to $10 billion annually, he said.

Risk Factors

As mentioned, the risk factors that can contribute to the development of an SSI are legion, but Bratzler said some of them can be broken down under the following four major areas:

  • Host factors: Includes age, morbid obesity, malnutrition, prolonged preoperative stay, cancer, and diabetes.
  • Endogenous flora/microbial factors: Includes nasal/skin carriage, virulence, adherence, and inoculum.
  • Surgical procedures: Includes abdominal site, wound classification, poor hemostasis, drains/foreign bodies, and urgency of the surgery.
  • Surgical team and hospital practice factors: Includes razor shaves, intraoperative contamination, prophylactic antibiotic timing, selection and duration, preoperative screening for resistant organisms and decolonization, surgeon’s skill, and surgical volume.

“There are a lot of risk factors for surgical site infections. Some of those are in your control and some are not,” Bratzler said. “I love the inspirational concept of getting to ‘zero.’ I also think it sometimes drives underreporting and maybe some ‘gaming’ of reporting. I think that we just have to acknowledge that we might not be able to prevent all [SSIs], at least based on the current level of scientific knowledge that we have. That said, it is generally considered that 50% of surgical site infections are probably preventable if we did all of the evidence-based practices."

Bratzler reminded that virtually all surgical wounds become contaminated in the operating room, noting that it is a misconception that the OR is some kind of sterile environment.

“Different operations have different risks,” he said. “So, colorectal surgery or vascular surgery of a lower extremity have a much higher risk of SSIs than total knee replacement. It is largely because the inoculum of the bacteria that are likely to get into the wound.”

In addition, similar operations performed by the same surgeon in different patient populations will have different rates of infection, he noted.

“That’s because there are different patient host factors, but I will tell you that the same operation done exactly the same way by different surgeons will have a different SSI rate,” Bratzler said. “Frankly, we don’t know yet how to measures surgical technique and address that as a risk factor. Clearly, there are studies that show the rate of infections varies by surgeon.”

Unresolved Issues

The aforementioned areas where there is insufficient data to make any recommendation include the following unresolved issues in surgical infection prevention:

  • weight-based antimicrobial dosing;
  • intraoperative antimicrobial irrigation;
  • antimicrobial soaking of prosthetic devices;
  • antimicrobial dressings applied to surgical incision;
  • optimal target for blood glucose control;
  • value of HbA1C for predicting SSI;
  • best strategy for maintaining normothermia;
  • oxygenation in non-endotracheal intubation surgery;
  • best mechanism to deliver postoperative oxygen and the optimal Fi02;
  • optimal timing of preoperative bathing.

For all orthopedic surgery key questions except antimicrobial prophylaxis duration, no randomized, controlled trials were identified and only observational studies reviewed.

“I will tell you that for orthopedic surgery we made the decision, because there was not a single randomized trial, to look at observational studies,” he said. “We didn’t find any evidence that we could hang our hats on there, either, with one exception: Stop the antibiotics at the end of surgery duration. That was the only recommendation we made in orthopedic surgery.”

The new HICPAC SSI recommendations were ranked as follows:

  • Category IA: A strong recommendation supported by high- to moderate-quality evidence suggesting net clinical benefits or harms.
  • Category IB: A strong recommendation supported by low-quality evidence suggesting net clinical benefits or harms or an accepted practice (e.g., aseptic technique) supported by low- to very low-quality evidence.
  • Category IC: A strong recommendation required by state or federal regulation. “We had no category IC recommendations,” Bratzler said.
  • Category II: A weak recommendation supported by any quality evidence suggesting a trade-off between clinical benefits and harms. “This last category is kind of flip of the coin about whether it’s helpful or not,” he said.

SSI Recommendations

The following are some of the major CDC/HICPAC recommendations to prevent SSIs, with comments from Bratzler.

  • In clean and clean-contaminated procedures, do not administer additional prophylactic antimicrobial agent doses after the surgical incision is closed in the operating room, even in the presence of a drain. (Category IA)

“Most of you guys should be cheering because we gave a Category IA recommendation for all operations — stop antibiotics once the incision is closed,” he told APIC attendees. “That is controversial with some [surgical] societies, but we didn’t find a single study that showed any benefit of antibiotics post-incision closure. [The wound] is a dead vascular space with no blood flow. The way you fix those infections is reopen the wound.”

Indeed, the characteristics of a wound carved out of a section of tissue makes it unlikely that antibiotics are going to penetrate to kill the bacteria in the surgical site. Of course, with the current focus on preserving antibiotic efficacy, any move to reduce unnecessary use of antimicrobials is helpful to prevent the rise of multidrug -esistant bacteria.

“Historically, we thought if we just gave a bunch of antibiotics and gave them for a long time, we could prevent surgical infections,” Bratzler said. “It’s important to know that you have this avascular wound — this space — and then we put tight sutures and other things to hold the skin and soft tissues together. You can give antibiotics until the cows come home and you’re not going to prevent these infections once the wound is closed. Nothing is going to get in there, and giving antibiotics is going to have very little impact on infections.”

  • Administer preoperative antimicrobial agents only when indicated based on published clinical practice guidelines, and timed such that a bactericidal concentration of the agents is established in the serum and tissues when the incision is made. (Category IB)
  • Administer the appropriate parenteral prophylactic antimicrobial agents before skin incision in all cesarean section procedures. (Category IA)

“Again, few randomized trials, except for cesarean section, where we gave Category IA based on multiple randomized trials showing reduced risk of SSIs if you give the antibiotic before the incision is made,” he said.

  • The literature search did not identify sufficient randomized, controlled trial evidence to evaluate the benefits and harms of intraoperative redosing of parenteral prophylactic antimicrobial agents for the prevention of SSI. (No recommendation/unresolved issue.)

“There are no randomized trials on these, but there are a lot of observational studies that suggest that this is a good thing to do and shows that it can reduce surgical site infection,” he said. “There are also pharmacological data showing if you don’t give a big enough dose, you’ll never exceed the MIC [minimum inhibitory concentration]. That data is very strong, but again, there were no randomized trials of intraoperative redosing.”

  • Consider the use of triclosan-coated sutures for the prevention of SSI. (Category II)

“A big controversy, and we gave it a weak recommendation,” Bratzler said. “However, there is really good data that if you are using braided, non-absorbable sutures that biofilm develops. There are very good reports of patients that had persistent draining that were ultimately traced to non-absorbable sutures that had biofilm on them and were contaminated. If you are using monofilament [sutures], the risk is less and if you are using absorbables the risk is less. But non-absorbable, braided sutures are clearly a risk.”

  • Implement perioperative glycemic control and use blood glucose target levels less than 200 mg/dL in patients with and without diabetes. (Category IA)

“We recommended this for all operations, not just cardiac surgery. There is good data in a variety of opertions that it reduces the risk of SSIs. There is substantial increased risk for SSIs in patients with diabetes and hyperglycemia.”

  • Maintain perioperative normothermia. (Category IA)

“We gave a [high] recommendation for keeping patients warm in the operating room,” he said. “We couldn’t tell you exactly what temperature to target or any best practices for keeping patients warm. There are a variety of different devices out there.”

  • For patients with normal pulmonary function undergoing general anesthesia with endotracheal intubation, administer increased FiO2 during surgery and after extubation in the immediate postoperative period. To optimize tissue oxygen delivery, maintain perioperative normothermia and adequate volume replacement. (Category IA)

“Why [normal pulmonary function]? Because that is the only group of patients that’s been studied, if they didn’t have ‘normal pulmonary function’ that got excluded from the studies,” Bratzler said. “But there is evidence there that you can reduce SSIs in those patients. Also at the same time if you are keeping those patients warm and have adequate volume replacement.”

A Bath’s a Good Thing

  • Advise patients to shower or bathe (full body) with soap (antimicrobial or nonantimicrobial) or an antiseptic agent on at least the night before the operative day. (Category IB). “We actually did not find randomized trial evidence in surgery that using chlorhexidine [CHG] or anything else was that much better than using soap and water,” he said. “But taking a bath’s a good thing. We don’t have anything against using CHG; we just didn’t find randomized trial evidence. This recommendation did not include the population colonized with MRSA, where there is good evidence that you should use CHG.”
  • Perform intraoperative skin preparation with an alcohol-based antiseptic agent unless contraindicated. (Category IA)

“We recommend you use an alcohol base, so there is povidone iodine-based alcohol, there is CHG-based alcohol. Alcohol kills bacteria very rapidly,” Bratzler said. “In the preoperative phase, we recommended an alcohol-based antiseptic.”

  • Consider intraoperative irrigation of deep or subcutaneous tissues with aqueous iodophor solution for the prevention of SSI. Intraperitoneal lavage with aqueous iodophor solution in contaminated or dirty abdominal procedures is not necessary. (Category II)

“This is a weak recommendation, but perhaps in limited operations, particularly spine surgery, which was shown to be a little bit beneficial but not for abdominal operations,” he said.


  1. Berríos-Torres SI, Umscheid CA, Bratzler DW, et al. For the Healthcare Infection Control Practices Advisory Committee. Centers for Disease Control and Prevention Guideline for the Prevention of Surgical Site Infection, 2017. JAMA Surg 2017;152(8):784-791.