- Design
- updated systematic review and meta-analysis of 25 observational studies, random-effects pooling
- Population
- patients undergoing total knee arthroplasty, with and without periprosthetic joint infection
- Primary outcome
- factors associated with periprosthetic joint infection
- Effect
- BMI 30 or above OR 5.72 (2.65 to 12.36); pulmonary disease OR 5.54; transfusion OR 2.27; drain 24 hours or longer OR 2.23
An updated systematic review pooled 25 observational studies of periprosthetic joint infection after total knee arthroplasty, 24 of them contributing to quantitative synthesis, and reported odds ratios for each risk factor.
The largest associations were a body mass index of 30 or above (OR 5.72, 95% CI 2.65 to 12.36), coexisting pulmonary disease (OR 5.54, 1.93 to 15.96) and a history of hormone therapy (OR 4.88, 2.90 to 8.22). Then postoperative urinary tract infection (OR 3.59, 1.15 to 11.21), an ASA score of 3 or more (OR 2.73, 1.02 to 7.32), rheumatoid arthritis (OR 2.64, 1.38 to 5.02), blood transfusion (OR 2.27, 1.59 to 3.25), drain left in for 24 hours or longer (OR 2.23, 1.50 to 3.32), type 2 diabetes (OR 2.09, 1.45 to 3.01), preoperative anaemia (OR 1.82, 1.67 to 1.99), smoking (OR 1.65, 1.34 to 2.04), intraoperative blood loss above 200 ml (OR 1.51, 1.03 to 2.22) and male sex (OR 1.39, 1.27 to 1.51). Prolonged operative time was also associated. Age was reported narratively because studies classified it too differently to pool.
Sorted by what a surgeon can change, the list is short and useful: anaemia corrected before the list rather than transfused during it, glycaemic control, smoking cessation, drains out inside 24 hours, operative time, and a urinary tract infection treated rather than observed. The rest - obesity, rheumatoid disease, lung disease, ASA class - are for risk stratification and for an honest consent conversation about a knee that may need revising.
- Correct preoperative anaemia in clinic rather than transfusing around the operation
- Remove drains within 24 hours, or do not place one
- Treat a postoperative urinary tract infection promptly rather than watching it
- Optimise glycaemic control and stop smoking before listing, not at admission
- Use the high-odds factors for consent and follow-up intensity, not to refuse surgery
Why it matters
Half of this list can be changed in clinic weeks before the operation, and the other half belongs in the consent conversation.
Don't overread it
All included studies were observational and used differing infection definitions; these are associations, not causes, and mostly unadjusted.
The statistics, in plain English
These are unadjusted pooled odds ratios from observational studies, so they overlap heavily: obesity, diabetes and ASA class travel together, and each odds ratio partly reflects the others. An interval as wide as 1.93 to 15.96 for pulmonary disease means the direction is reliable and the size is not. Transfusion and blood loss are the clearest example of association standing in for something else - the patient who needs blood is different from the patient who does not, before any transfusion happens.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for orthopaedics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free