Annals of Internal Medicine | 2019
Initial oral antibiotic therapy was noninferior to IV therapy for treatment failure in orthopedic infection at 1 y
Abstract
Question In patients with orthopedic infection, is oral antibiotic therapy noninferior to IV therapy for the first 6 weeks of therapy? Methods Design Randomized controlled trial (RCT) (Oral versus Intravenous Antibiotics for Bone and Joint Infection [OVIVA]). ISRCTN 91566927. Allocation Concealed.* Blinding Blinded* (outcome adjudicators). Follow-up period 1 year. Setting 26 clinical centers in the UK. Patients 1054 adults >18 years of age (median age 60 y, 64% men) who had acute or chronic bone or joint infection (native joint infection requiring excision arthroplasty, native osteomyelitis of the extraaxial skeleton, orthopedic fixation device infection, prosthetic joint infection, or vertebral osteomyelitis with or without associated diskitis or soft tissue infection) and would normally have been treated with IV antibiotics during the first 6 weeks of therapy. Exclusion criteria included Staphylococcus aureus bacteremia on presentation or in the past month. Intervention Oral (n =527) or IV (n =527) antibiotic therapy beginning within 7 days of definitive surgical intervention or at the start of antibiotic therapy for patients managed nonsurgically. The antibiotics used were assigned by accredited infection specialists before randomization. Outcomes Definite treatment failure (defined by prespecified clinical, microbiologic, or histologic criteria) at 1 year. Secondary outcomes included probable or possible treatment failure, early discontinuation of study treatment, serious adverse events (SAEs), and patient-reported health status (European Quality of Life5 Dimensions [EQ-5D-3L] and Oxford Hip and Knee Scores) at 1 year. The prespecified noninferiority limit was 7.5% (upper limit of the 2-sided 90% CI around the unadjusted absolute risk difference) for the primary outcome (limit was changed from 5% during trial). Patient follow-up 96% (intention-to-treat analysis). Main results Median duration of antibiotic therapy was 71 days in the oral group and 78 days in the IV group (P =0.63). Results for treatment failures, early discontinuation, and SAEs are in the Table. The oral therapy group had better Oxford Knee Scores than did the IV treatment group at 1 year (P =0.04); groups did not differ for EQ-5D-3L scores or Oxford Hip Scores (P0.18). Conclusion In patients with orthopedic infection, initial oral antibiotic therapy was noninferior to IV therapy for reducing treatment failure. Initial oral vs IV antibiotic therapy in patients with orthopedic infections Outcomes Event rates At 1 y Oral IV Absolute risk difference (95% CI) NNT (CI) Definite treatment failure 13% 15% 1.4% (5.6 to 2.9) Not significant Early discontinuation of therapy 13% 19% 6.1% (11 to 1.7) 17 (10 to 59) Serious adverse events 26% 28% 1.5% (6.9 to 3.8) Not significant NNH Probable or possible treatment failure 2.0% 1.2% 0.8% (0.8 to 2.5) Not significant Abbreviations defined in Glossary. Absolute risk difference (ARD), NNT, and CI calculated from event rates in article except ARD for definite treatment failure. Presence of 1 clinical, microbiologic, or histologic criterion. Criterion for noninferiority was met because the upper limit of the 2-sided 90% CI of the unadjusted absolute difference in risk (1.4%, 90% CI 4.9 to 2.2) was <7.5%. Commentary The practice of treating osteomyelitis with parenteral therapy was established by the influential 1970 case series by Waldvogel and colleagues, who argued that osteomyelitis requires surgical debridement and prolonged (4-6 weeks) parenteral antibiotic therapy (1). Patients with osteomyelitis are usually given extended courses of parenteral antibiotics despite limited supporting evidence. A 2013 meta-analysis of 8 RCTs (282 patients) suggested that parenteral therapy has no advantage over oral therapy (2). Li and colleagues conducted a pragmatic noninferiority trial that compared oral with IV antibiotics to treat osteomyelitis; the trial was intentionally inclusive and not limited by pathogen, infection site, or antibiotic regimen. Rates of definite treatment failure in the oral therapy group were noninferior to those in the IV therapy group. Early discontinuation occurred more frequently in the IV group than in the oral group, although the groups did not differ for total duration of antibiotic therapy. The per protocol analysis of patients who received 4 weeks of their assigned treatment found that oral therapy remained noninferior to IV therapy for treatment failure. 60% of the patients in this trial had orthopedic hardware present at enrollment, approximately half of whom had retained hardware after surgery; post hoc subgroup analysis of patients who did and did not retain hardware did not find a difference in treatment failure related to use of oral or IV antibiotics. Oral antibiotics offer convenience and cost-savings and eliminate IV access risks. RCTs are warranted to clarify the specific patients and pathogens for whom oral therapy is appropriate treatment for osteomyelitis.