What actually predicts infection after knee replacement

Design

updated systematic review and meta-analysis of observational studies, nine databases searched to May 2025, Newcastle-Ottawa quality assessment, PROSPERO-registered

Population

25 observational studies of patients undergoing total knee arthroplasty, 24 in the quantitative synthesis

Primary outcome

factors associated with periprosthetic joint infection after total knee arthroplasty

Effect

BMI >=30 OR 5.72 (2.65-12.36); pulmonary disease 5.54 (1.93-15.96); hormone therapy 4.88 (2.90-8.22); postoperative UTI 3.59 (1.15-11.21); drain >=24 h 2.23 (1.50-3.32); anaemia 1.82 (1.67-1.99)

This updated review pooled 25 observational studies, 24 contributing quantitatively, on factors associated with periprosthetic joint infection after total knee arthroplasty, with quality assessed on the Newcastle-Ottawa Scale and the protocol registered on PROSPERO.

The large effects are worth memorising. Obesity with a BMI of 30 or above carried an odds ratio of 5.72 (95% CI 2.65 to 12.36), coexisting pulmonary disease 5.54 (1.93 to 15.96), and a history of hormone therapy 4.88 (2.90 to 8.22). Then a middle tier: postoperative urinary tract infection 3.59 (1.15 to 11.21), ASA 3 or above 2.73 (1.02 to 7.32), rheumatoid arthritis 2.64 (1.38 to 5.02), blood transfusion 2.27 (1.59 to 3.25), a drain left 24 hours or longer 2.23 (1.50 to 3.32), type 2 diabetes 2.09 (1.45 to 3.01). Then smaller but precise: preoperative anaemia 1.82 (1.67 to 1.99), smoking 1.65 (1.34 to 2.04), blood loss over 200 ml 1.51 (1.03 to 2.22), male sex 1.39 (1.27 to 1.51).

Separate the modifiable from the fixed, because that is the only use for a list like this. Anaemia, smoking, glycaemic control, transfusion avoidance and drain duration are things a preoperative pathway can act on; sex, rheumatoid arthritis and ASA class are things a consent conversation should reflect. The authors are explicit that these are associations from heterogeneous observational studies without standardised PJI definitions or consistently adjusted estimates, so the odds ratios overlap and cannot simply be multiplied together into a patient's risk.

• Correct preoperative anaemia — it is common, modifiable, and carried a precise 1.82 odds ratio
• Treat drain duration as a decision: 24 hours or more doubled the odds
• Build smoking cessation and glycaemic optimisation into the preoperative pathway rather than the consent form
• Treat and document postoperative urinary tract infection promptly; it was among the strongest postoperative associations
• Do not add or multiply these odds ratios — they come from different studies with overlapping populations and unadjusted definitions

The statistics, in plain English

The widest intervals belong to the largest odds ratios — pulmonary disease at 1.93 to 15.96 and urinary tract infection at 1.15 to 11.21 — meaning those factors are probably real but their size is barely pinned down, usually because few studies and few events contribute. Preoperative anaemia has the tightest interval (1.67 to 1.99), so it is the most reliably estimated even though its effect is modest. Because most contributing studies reported unadjusted associations, some of these factors are partly measuring each other: obesity, diabetes, ASA class and operative time travel together.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.