Two-stage revision is the gold standard for managing infected total hip and knee arthroplasties. The aim was to assess the effect of duration between stages on reinfection rate at one year.

A systematic review and meta-analysis was conducted on all studies investigating reinfection rate with documented interval between first and second stages. Total hip (THR) and total knee replacements (TKRs) were included but analysed separately. The effect size of studies was stratified according to sample size then with study quality.

All papers up until November 2015 (including non-English language) were considered. From 3827 papers reviewed, 38 cohorts from 35 studies were included, comprising 23 THR and 15 TKR groups. Average study quality was 5.6/11 (range 3–8). Funnel plots calculated to assess for bias indicated significant asymmetry at lower sample sizes in both groups.

In the TKR group, studies with 0–3 months between stages showed a significantly lower reinfection rate than 3–6 months (9.5% 21/222 vs 20.7% 28/135, p<0.01). A similar trend was seen in the THR group (6.1% vs 10.7%, p<0.05). No difference was observed for either group between 3–6 and 6–9 months.

There is no consensus regarding the appropriate duration between surgeries in two-stage revisions for infection. Studies stratified by sample size and quality indicate an increased reinfection rate past three months. Published guidance is no substitute for clinical decision-making but the conclusions from this study are to recommend against routine delay of more than 3 months between first and second stage revisions for infected THR and TKR.