Three RCTs were included in the review (3,994 patients). Of these, two trials included data on high-risk patients (2,887 patients). Follow-up ranged from two to three months. All three RCTs used blinded assessors; only one trial did not use patient blinding. Random-effects meta-analyses were reported for all results.
The pooled estimate was not statistically significant for any sternal wound infection, although it suggested some benefit for gentamicin impregnated collagen sponge in addition to systemic antibiotics. There was high heterogeneity (Ι²=75%). The results were more consistent (Ι²=0%) for deep sternal wound infections, suggesting some benefit for gentamicin, but this was not statistically significant.
For high-risk patients, there was no statistically significant difference for any sternal wound infection, and there was high heterogeneity (Ι²=91%). However, for deep sternal wound infection, there was a statistically significant benefit found for gentamicin compared with control groups (OR 0.62; 95% CI 0.39 to 0.98; Ι²=0%; two trials).