Wessex Gauvain Society Annual Meeting

Authors:
A Chowdhury, S Islam, T Ranoboldo, K Sheen, K Wilcocks, S Sampalli, A Elmorsy

Hospital:
Salisbury NHS Foundation Trust

Introduction
Intra-articular corticosteroid injections are frequently utilised in patients with glenohumeral arthritis for diagnostic and therapeutic purposes. It is debated whether these injections increase the risk of infection following subsequent shoulder arthroplasty. We evaluated the relationship between prior corticosteroid injection and infection following shoulder arthroplasty.

Methods
A computerised search was performed, using databases Medline, EMBASE and CINAHL, supplemented from Google Scholar and reference lists. Studies with comparative data of post-arthroplasty infection following previous corticosteroid injection versus controls were included. Data were extracted and quality assessment performed by two reviewers. Quantitative analysis was performed where possible.

Results
Six studies fulfilled the inclusion criteria, one prospective and five retrospective cohort designs, with 87,820 patients. On pooled analysis, a previous steroid injection did not increase the risk of infection (relative risk (RR)1.16, 95% CI 0.98-1.36, p=0.08). It did, however, confer an increased risk of revision surgery (RR1.52, 95% CI 1.32-1.75, p<0.001). On subgroup analysis for timing of injection, one given <3 months of arthroplasty conferred a greater risk of infection than no injection (RR 2.30, 95% CI 1.13-4.69, p=0.02) or one given at 3-12 months before arthroplasty (RR3.32, 95% CI 1.43-7.72, p=0.005). An injection at 3-12 months before arthroplasty did not increase the risk of infection over controls (RR0.89, 95% CI 0.63-1.25, p=0.50).

Conclusions
This review has found there to be a time-dependent relationship between prior corticosteroid injection and infection post-arthroplasty, with an increased risk <3 months before arthroplasty. Thus, an interval of at least 3 months is recommended between injection and arthroplasty.