RESUMEN
BACKGROUND: Corticosteroid (CS) can be injected in a blind fashion (landmark-guided) or with ultrasound (US) guidance, and this may contribute to varying clinical results. We conducted this systematic review and meta-analysis to assess the effectiveness of US-guided versus landmark CS injections in the treatment of adult patients with shoulder pain. METHODS: We searched MEDLINE (via PubMed), Scopus, Web of Science, EBSCO, and Cochrane Library for randomized controlled trials (RCTs) comparing US-guided versus landmark CS injection regarding visual analogue scale (VAS), functional scores, disability scores, abduction degree, and side effects. The data were pooled as mean difference (MD), standardized mean difference (SMD), or risk ratios (RRs), with 95% confidence intervals (CIs), using R software (meta package 4.9-0) for windows. Subgroup analysis and leave-one-out analysis were conducted. RESULTS: Eighteen RCTs, with a total of 1010 patients, were included in this meta-analysis. The pooled estimate favored the US-guided over landmark CS injection in terms of the mean change of VAS between 6 weeks and baseline (SMD = - 0.48, 95% CI [- 0.79, - 0.17]), the shoulder functional scores (SMD = 0.35, 95% CI [0.05, 0.65]) and shoulder abduction degree (MD = 8.78, 95% CI [3.11, 14.46]). Whilst no significant difference was found between the compared group regarding the overall shoulder disability scores (SMD = - 0.51, 95% CI (- 1.25, 0.22]) and side effects (RR = 0.45, 95% CI [0.15, 1.34]). None of the eligible study analyzed the cost-effectiveness of the US-guided method compared with the landmark method for CS injection. CONCLUSION: Our analysis showed that US-guided CS injection was effective in the treatment of various shoulder diseases. Further research on the cost-effectiveness of US-guided CS methods is needed.