Perioperative Antibiotic Prophylaxis

Timing within 30-60 minutes before incision is the most important factor in whether prophylaxis works. Within that window 30 minutes is optimal, and 60 minutes still gives therapeutic levels. The timeline shows what happens either side of it.

Working with the anaesthetist. Coordinate the timing with anaesthesia. The antibiotic goes in once the patient is in theatre with IV access established, and before skin preparation and draping. Common practice is to give it at induction, about 30 minutes before incision, which lands in the optimal window and avoids a too-early dose in the holding area.

Vancomycin is the exception. It needs a 1-2 hour IV infusion, because rapid infusion causes red man syndrome through histamine release. Start it 120 minutes before incision so that the infusion is complete by the time the knife goes in.

Cefazolin is dosed by weight, and children receive 30 mg/kg IV to a maximum of 2g.

Why weight matters. Volume of distribution increases with body weight, and an obese patient has more tissue for the antibiotic to penetrate. A 2g dose achieves adequate tissue levels in patients of 70-90kg, and 3g is necessary for adequate levels above 120kg.

The evidence. Studies show that 1g gave inadequate tissue levels in many patients, whereas 2g achieves the MIC90 for S. aureus in bone and tissue in 95% of patients. Underdosing leaves subtherapeutic levels and is associated with increased SSI risk, which makes it as bad as the wrong timing.

Other agents. Doses for the alternatives:

• Vancomycin - 15 mg/kg IV (maximum 2g per dose)
• Gentamicin - 5 mg/kg IV as a single dose (maximum 320mg)
• Clindamycin - 900mg IV, the beta-lactam allergy alternative

Therapeutic levels must be maintained until the wound is closed, and redosing matters as much as the first dose. Redose when either:

• The operation exceeds two half-lives of the drug
• Blood loss exceeds 1500mL

Cefazolin. Its half-life is 1.8-2 hours with normal renal function, so it is redosed at 4 hours if surgery is ongoing. The redose is the same as the first dose (2g or 3g by weight), repeated every 4 hours until wound closure.

Blood loss. Beyond 1500mL, redose regardless of the time, because dilution reduces serum and tissue concentrations. The rule applies to every agent except gentamicin, which is never redosed. Estimate the loss intraoperatively from suction, sponges and the field.

In practice. The practical steps:

• Anaesthesia tracks time and blood loss
• Set a timer for 4 hours from the first cefazolin dose
• Tell the surgeon when a redose is given (some surgeons want to know the timing)
• Document each dose in the anaesthetic record

Stop within 24 hours. For most procedures a single preoperative dose is sufficient, including clean orthopaedic surgery such as THA, TKA and arthroscopy. Multiple RCTs show a single dose equals multiple doses for SSI prevention, and additional postoperative doses bring no benefit. IDSA, SCIP and AAOS guidance is to stop within 24 hours.

If it is continued. Some surgeons continue for 24 hours, from tradition or for medicolegal reasons. That is acceptable but no more effective than a single dose, and routine prophylaxis should never exceed 24 hours.

Why more is worse. Beyond 24 hours SSI does not fall further: the effect plateaus, and multiple RCTs show no benefit. What prolonged prophylaxis adds is harm:

• Antimicrobial resistance, selecting resistant S. aureus and Enterobacteriaceae
• C. difficile infection
• Adverse effects: allergy, renal toxicity, drug interactions
• Cost, without benefit

The exceptions. Antibiotics continue beyond 24 hours in three situations:

• Open fractures - until the wound is definitively closed, within a maximum set by the Gustilo type (see Open Fractures)
• Contaminated surgery - gross contamination (faecal, pus) may warrant 24-48 hours, which is treatment, not prophylaxis
• Established infection - a treatment regimen, not prophylaxis

Stewardship. The measures that keep courses short:

• Document the planned duration at surgery (e.g. "cefazolin 2g pre-op, stop at 24h")
• Automatic stop orders at 24 hours as the default
• Antimicrobial stewardship teams review prolonged courses
• Educate surgeons on the evidence against prolonged prophylaxis

The organisms. After total hip and knee arthroplasty and clean spine surgery, S. aureus causes 30-40% of SSI, coagulase-negative staphylococci (S. epidermidis) another 30-40%, and streptococci 10-15%.

Why cefazolin. It is a first-generation cephalosporin, narrow in spectrum and so appropriate for prophylaxis, with excellent tissue penetration that achieves therapeutic levels in bone and soft tissue. Its half-life of about 2 hours, long for a cephalosporin, allows single-dose prophylaxis. Give a single dose, or 24 hours if the surgeon prefers, timed, dosed and redosed as above.

MRSA high-risk patients. Some surgeons add vancomycin to cefazolin. This is controversial, with no strong evidence behind it.

Antibiotic cement. PMMA bone cement loaded with gentamicin or vancomycin gives local, high-concentration antibiotic release. It may reduce infection in high-risk patients (revision, immunosuppressed), and joint registries (e.g. AOANJRR, Swedish/Norwegian) report benefit of antibiotic-loaded cement particularly in revision and high-risk arthroplasty. It provides local levels but no systemic coverage during surgery, so it does not replace systemic prophylaxis: give IV cefazolin as well. Cement is an adjuvant.

Intra-wound vancomycin powder. Separate from cement, vancomycin powder sprinkled into the wound before closure is widely used, especially in spine surgery, and increasingly studied in arthroplasty. Local levels far exceed the MIC for staphylococci while serum levels stay low, with negligible systemic absorption, so it targets the gram-positive organisms that cause most SSI without systemic vancomycin toxicity.

Does the powder work? Large retrospective and observational spine series suggested reduced deep SSI, but higher-quality data are mixed: the prospective FORTIFY-type/randomised evidence has not confirmed a consistent benefit. It remains an adjunct, not standard of care, and supplements, never replaces, correctly timed systemic prophylaxis. The harms to know:

• No gram-negative cover, with a theoretical concern of shifting deep infections toward gram-negative or polymicrobial organisms
• Seroma formation and wound-healing problems
• In spine, a theoretical effect on fusion/pseudarthrosis at high doses

Give intravenous antibiotics as soon as possible, ideally within 1 hour of injury (BOAST Open Fractures 2017; NICE NG37), following the published network antibiotic guideline. The Gustilo-Anderson type then sets both the regimen and how long it runs; the table follows the EAST guideline (Hoff 2011).

Duration. EAST stops prophylaxis 24 hours after wound closure for Type I and II fractures; for Type III it continues for 72 hours after injury or not more than 24 hours after soft-tissue coverage has been achieved. NICE NG37 and BOAST set no antibiotic duration, but require definitive soft-tissue cover within 72 hours of injury if it is not achieved at debridement. Extended prophylaxis in open fractures is evidence-based, but must be balanced against the emergence of resistant organisms and C. difficile risk.

Contamination changes the cover. Add cover for the organisms the environment brings:

• Farm or faecal - add high-dose penicillin for potential clostridial contamination (EAST)
• Aquatic or marine - add doxycycline or a fluoroquinolone (Aeromonas, Vibrio)

Gentamicin. For prophylaxis it is given as a single daily dose (5 mg/kg, maximum 320mg). It is not redosed intraoperatively: it is nephrotoxic, and the single dose provides coverage. Monitor renal function if it continues beyond 48 hours.

Allergy or MRSA. Vancomycin replaces cefazolin for gram-positive cover, with gentamicin for Type III. Avoid fluoroquinolones: EAST found they offer no advantage over cephalosporin/aminoglycoside regimens and may give higher infection rates in Type III fractures.

Who is at risk. MRSA risk factors are:

• Known MRSA colonisation (nasal swab positive)
• Recent hospitalisation (within 6 months)
• Nursing home resident
• Chronic wounds, indwelling catheters
• Previous MRSA infection
• High local MRSA prevalence (greater than 20%)

Vancomycin alone. 15 mg/kg IV (maximum 2g per dose). It covers MRSA and MSSA, but is weaker against MSSA than cefazolin.

Vancomycin plus cefazolin. Dual prophylaxis adds cefazolin 2g IV for its better MSSA and streptococcal cover. There is no strong evidence of benefit over vancomycin alone, and some institutions use it for high-risk patients (obesity, diabetes, revision). For a carrier who tolerates beta-lactams, the decision table below treats vancomycin alone as the wrong answer for exactly that reason: worse MSSA cover.

Decolonisation. Screen high-risk patients with a nasal swab. Preoperative decolonisation reduces MRSA carriage and SSI in colonised patients:

• Nasal mupirocin - 2% ointment twice daily for 5 days before surgery
• Chlorhexidine baths - 2-4% solution daily for 5 days before surgery

Which allergy matters. A true IgE-mediated reaction (anaphylaxis, angioedema, bronchospasm) is an absolute contraindication to cephalosporins. A non-IgE reaction (rash, GI upset) may tolerate them, because cross-reactivity is low. Many reported "penicillin allergies" (70-80%) are not true IgE reactions, and many are childhood rashes.

Take the history first. Clarify the allergy before defaulting to an alternative, and consider preoperative allergy testing in elective cases. After a non-severe reaction to penicillin, cefazolin is generally safe, after discussion with the patient.

Cross-reactivity. Penicillin-cephalosporin cross-reactivity is 1-3% overall, historically over-quoted as 10%; penicillin-carbapenem is about 1%. It is driven by the similarity of the R1 side chain, not the shared beta-lactam ring: penicillins and cephalosporins with a similar R1 side chain (e.g. amoxicillin and cefadroxil/cefprozil) genuinely cross-react, and those with dissimilar side chains do not. Cefazolin's R1 side chain is unique, shared with no penicillin and no other commonly used cephalosporin, so its cross-reactivity with penicillin is negligible (near zero).

Alternatives for true beta-lactam allergy. When cefazolin is contraindicated, the options are:

• Vancomycin 15 mg/kg IV - covers S. aureus (MSSA and MRSA) and S. epidermidis; less effective than cefazolin for MSSA, but acceptable; no gram-negative or anaerobic cover
• Clindamycin 900mg IV - covers S. aureus (MSSA, some MRSA resistance), Streptococcus and anaerobes, with good bone penetration; no gram-negative cover, and resistance is increasing in some S. aureus strains
• Vancomycin 15 mg/kg + gentamicin 5 mg/kg IV - when gram-negative cover is needed (open fracture, contaminated case); broad coverage, but a nephrotoxic combination
• Aztreonam - a monobactam with no cross-reactivity with penicillins; covers gram-negatives only (no Staph cover), so rarely used for prophylaxis, but can be combined with vancomycin for broad coverage