With the fear of antibiotic resistance in mind, the WHO yesterday published new guidelines for carrying out surgical procedures aimed at reducing the incidence of surgical site infections (SSIs), which occur in 11% of procedures in low and middle income countries (20 times higher than the incidence in high-income settings) and affect millions of lives and cost healthcare systems billions of dollars. The WHO tasked a group of 20 experts to come up with key unanswered questions related to surgical infection control and carry out meta-analyses using published studies to assess the likelihood of harm or benefit from different care protocols.

These analyses produced 29 clinical recommendations, which fall into 2 categories: STRONG recommendations that the experts felt had enough good quality data to be confident that the benefits outweighed the risks and would be adaptable across most income settings; and CONDITIONAL recommendations where the data is less robust and implementation should be decided on a more case-by-case basis. You can read the full discussion of these guidelines in two Lancet Infectious Disease reviews published yesterday, one article covering preoperative decisions and the other on intra/post-operative care.

In this post, I am not going to comprehensively discuss all the recommendations (I can already hear sighs of relief), but would like highlight a few guidelines that are interesting examples of context-dependent advice and some new changes that are just generally surprising:

1. Prepping the patients:

• Bathing is standard before surgery to help reduce bacterial burden on the skin, and often in the hospitals, this involves using antiseptic washes. But the data did not particularly show improved SSI outcomes in using these antiseptic compounds compared to plain old soap (echoes of the recent FDA triclosan ban in handsoaps, anyone?), so it's now a CONDITIONAL recommendation to just use plain old soap and water, no doubt easier on lower income settings.

• BUT, the big exception to this rule is if the patient is colonized with S. aureus. Here, the data showed that preoperative nasal treatment with mipirocin with and without body washes containing chlorhexidine gluconate had marked benefit for patients undergoing thoracic or orthopedic surgery. So if you're a carrier, it is STRONGLY recommended to use antiseptics. And while the data is less clear on other types of surgery, given that S. aureus carriage is a well known risk factor across different procedures, this recommendation may translate quite broadly.

• Who remembers the sitcom tableau of a hapless man being shaved in some embarrassing area before surgery? Well, that's a thing of the past. The data did not show a significant difference in SSI outcomes if hair was removed before surgery or not. But interestingly, there was a slightly worse outcome if you were shaved or hair was trimmed using clippers (microabrasions?), so it's STRONGLY recommended to avoid shaving patients. Well, there goes a comedic trope.

2. During/after surgery:

• Patients are ventilated alongside general anesthesia and it is now STRONGLY recommended that they also be given air with high oxygen saturation (80%). While it is not clear how it works (enhanced neturophil activity?), providing increased oxygen during AND continuing after surgery better prevented SSIs. The major drawback for this approach in most clinical settings is that the oxygen (and concomitant safety resources) needed will be hard to swing on limited budgets.

• Remember when antiseptics were generally no better than plain soap? Well, triclosan (and perhaps other antimicrobial) coating on sutures is a CONDITIONAL recommendation. While the data is still coming in, there appears to a benefit of coated sutures over the standard materials. So maybe the dividing line between the utility of antiseptics is if you are using them in internal rather than external settings?

• Finally, it's known that prophylactic antibiotics given to patients before surgery are important, and it's now STRONGLY recommended to give drugs closer to the procedure in order to sustain high systemic concentration of drug at the height of microbial exposure. But, interestingly, the analyses also found that that prolonged antibiotic treatment after surgery had no appreciable benefit in preventing SSIs. So without evidence of benefit and potentially selecting for resistant organisms via extended dosing it is now STRONGLY recommended that you get only a single dose of antibiotics and only before the operation. That means no more leaving leaving the hospital with a bottle of penicillin to polish off for the next week or so, 'just in case'.

These are only a few examples of the many recommendations presented, but what struck me about the guidelines was 1) it is very important (and probably overdue) to dig into the data with the explicit goal of actually harnessing the power of research to change clinical practices, and 2) most of the recommendations were CONDITIONAL, meaning that the amount and quality of data we have was just not enough to get clear answers to these important questions. Add in the mind-boggling number of variables that differ between studies (differences in treatment times, dosing, types of procedures, patient populations), and it's clear we still have a lot of work left to do in the clinical research setting. But retroactive meta-analyses like this can start pointing out the missing pieces and spur future studies, and it gives us comfort that even as we're groping in the dark, we know there's at least something in the shadows to find.