Previous arthroscopy was associated with more infection after arthroplasty, but not at every joint

Design

meta-analysis of 31 observational studies, generalised linear mixed model, subgroup by joint

Population

patients undergoing knee, hip or shoulder arthroplasty with or without prior arthroscopy

Primary outcome

postoperative overall and deep infection after arthroplasty

Effect

overall infection OR 1.33 (95% CI 1.14–1.56); deep infection OR 1.38 (1.17–1.63); pooled infection rate 2%

A prior arthroscopy is common history in patients coming to joint replacement, and whether it raises the later infection risk has been argued from single-centre series. This meta-analysis pooled 31 observational studies with a generalised linear mixed model, and stratified by joint.

Prior arthroscopic surgery was associated with higher overall infection after arthroplasty (odds ratio 1.33, 95% CI 1.14 to 1.56) and higher deep infection (1.38, 1.17 to 1.63). The joint-specific pattern is the useful part: previous knee arthroscopy carried raised infection risk, previous shoulder arthroscopy was associated particularly with deep infection, and hip arthroscopy showed no significant association at all. Pooled overall and deep infection rates were both 2% in this population.

What this changes is the pre-operative conversation and the index of suspicion afterwards, not the antibiotic prophylaxis regimen — nothing here tested a different prophylaxis. A patient with previous knee or shoulder arthroscopy should have that history recorded where the operating team will see it, and a low threshold for investigating a painful prosthesis applies more strongly to them.

• Record previous arthroscopy, and which joint, in the pre-operative assessment
• Apply a lower threshold for investigating a painful knee or shoulder prosthesis in these patients
• Do not change prophylaxis on this evidence — no prophylaxis strategy was compared
• Quote the absolute risk when consenting: pooled infection rates were 2%, not 33% higher than something large
• Treat the hip result as genuinely null rather than underpowered-but-probably-real

Why it matters

It turns a vague sense that prior surgery matters into a joint-specific question worth asking before the list.

Don't overread it

These are observational studies pooled together — prior arthroscopy marks a patient at higher risk, it does not prove it causes the infection.

The statistics, in plain English

An odds ratio of 1.33 on a baseline infection rate of 2% translates to a small absolute increase — roughly half a percentage point — which is why the consent conversation should use the absolute figures. The authors report obvious heterogeneity between joints, and that is the finding rather than a nuisance: pooling knee, hip and shoulder into one number would have hidden that the hip result is null. Observational pooling also cannot separate the arthroscopy itself from whatever made the patient need one.

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.