Oral versus intravenous antibiotics for bone and joint infections: the OVIVA non-inferiority RCT
Scarborough M., Li HK., Rombach I., Zambellas R., Walker AS., McNally M., Atkins B., Kümin M., Lipsky BA., Hughes H., Bose D., Warren S., Mack D., Folb J., Moore E., Jenkins N., Hopkins S., Seaton RA., Hemsley C., Sandoe J., Aggarwal I., Ellis S., Sutherland R., Geue C., McMeekin N., Scarborough C., Paul J., Cooke G., Bostock J., Khatamzas E., Wong N., Brent A., Lomas J., Matthews P., Wangrangsimakul T., Gundle R., Rogers M., Taylor A., Thwaites GE., Bejon P.
Background Management of bone and joint infection commonly includes 4–6 weeks of intravenous (IV) antibiotics, but there is little evidence to suggest that oral (PO) therapy results in worse outcomes. Objective To determine whether or not PO antibiotics are non-inferior to IV antibiotics in treating bone and joint infection. Design Parallel-group, randomised (1 : 1), open-label, non-inferiority trial. The non-inferiority margin was 7.5%. Setting Twenty-six NHS hospitals. Participants Adults with a clinical diagnosis of bone, joint or orthopaedic metalware-associated infection who would ordinarily receive at least 6 weeks of antibiotics, and who had received ≤ 7 days of IV therapy from definitive surgery (or start of planned curative treatment in patients managed non-operatively). Interventions Participants were centrally computer-randomised to PO or IV antibiotics to complete the first 6 weeks of therapy. Follow-on PO therapy was permitted in either arm. Main outcome measure The primary outcome was the proportion of participants experiencing treatment failure within 1 year. An associated cost-effectiveness evaluation assessed health resource use and quality-of-life data. Results Out of 1054 participants (527 in each arm), end-point data were available for 1015 (96.30%) participants. Treatment failure was identified in 141 out of 1015 (13.89%) participants: 74 out of 506 (14.62%) and 67 out of 509 (13.16%) of those participants randomised to IV and PO therapy, respectively. In the intention-to-treat analysis, using multiple imputation to include all participants, the imputed risk difference between PO and IV therapy for definitive treatment failure was –1.38% (90% confidence interval –4.94% to 2.19%), thus meeting the non-inferiority criterion. A complete-case analysis, a per-protocol analysis and sensitivity analyses for missing data each confirmed this result. With the exception of IV catheter complications [49/523 (9.37%) in the IV arm vs. 5/523 (0.96%) in the PO arm)], there was no significant difference between the two arms in the incidence of serious adverse events. PO therapy was highly cost-effective, yielding a saving of £2740 per patient without any significant difference in quality-adjusted life-years between the two arms of the trial. Limitations The OVIVA (Oral Versus IntraVenous Antibiotics) trial was an open-label trial, but bias was limited by assessing all potential end points by a blinded adjudication committee. The population was heterogenous, which facilitated generalisability but limited the statistical power of subgroup analyses. Participants were only followed up for 1 year so differences in late recurrence cannot be excluded. Conclusions PO antibiotic therapy is non-inferior to IV therapy when used during the first 6 weeks in the treatment for bone and joint infection, as assessed by definitive treatment failure within 1 year of randomisation. These findings challenge the current standard of care and provide an opportunity to realise significant benefits for patients, antimicrobial stewardship and the health economy. Future work Further work is required to define the optimal total duration of therapy for bone and joint infection in the context of specific surgical interventions. Currently, wide variation in clinical practice suggests significant redundancy that likely contributes to the excess and unnecessary use of antibiotics. Trial registration Current Controlled Trials ISRCTN91566927. Funding This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 23, No. 38. See the NIHR Journals Library website for further project information.