AimsDelayed postoperative inoculation of orthopaedic implants with persistent wound drainage or bacterial seeding of a haematoma can result in periprosthetic joint infection (PJI). The aim of this in vivo study was to compare the efficacy of vancomycin powder with vancomycin-eluting calcium sulphate beads in preventing PJI due to delayed inoculation.MethodsA mouse model of PJI of the knee was used. Mice were randomized into groups with intervention at the time of surgery (postoperative day (POD) 0): a sterile control (SC; n = 6); infected control (IC; n = 15); systemic vancomycin (SV; n = 9); vancomycin powder (VP; n = 21); and vancomycin bead (VB; n = 19) groups. Delayed inoculation was introduced during an arthrotomy on POD 7 with 1 × 105 colony-forming units (CFUs) of a bioluminescent strain of Staphylococcus aureus. The bacterial burden was monitored using bioluminescence in vivo. All mice were killed on POD 21. Implants and soft-tissue were harvested and sonicated for analysis of the CFUs.ResultsThe mean in vivo bioluminescence in the VB group was significantly lower on POD 8 and POD 10 compared with the other groups. There was a significant 1.3-log10 (95%) and 1.5-log10 (97%) reduction in mean soft-tissue CFUs in the VB group compared with the VP and IC groups (3.6 × 103 vs 7.0 × 104; p = 0.022; 3.6 × 103 vs 1.0 × 105; p = 0.007, respectively) at POD 21. There was a significant 1.6-log10 (98%) reduction in mean implant CFUs in the VB group compared with the IC group (1.3 × 100 vs 4.7 × 101, respectively; p = 0.038). Combined soft-tissue and implant infection was prevented in 10 of 19 mice (53%) in the VB group as opposed to 5 of 21 (24%) in the VP group, 3 of 15 (20%) in the IC group, and 0% in the SV group.ConclusionIn our in vivo mouse model, antibiotic-releasing calcium sulphate beads appeared to outperform vancomycin powder alone in lowering the bacterial burden and preventing soft-tissue and implant infections.Cite this article: Bone Joint J 2024;106-B(6):632–638.