Internal Medicine Portal: Archived - Journal Club Blog

Journal Club: OVIVA

Oral versus Intravenous Antibiotics for Bone and Joint Infection

Also known as the "OVIVA" trial. Discussed December 2019

Key clinical take-aways from this article: 

• Why we should know it: Following initial therapy with intravenous antimicrobials, 6 weeks of oral therapy was non-inferior to intravenous therapy.
• open-label, parallel group, randomized controlled non-inferiority trial completed in 26 UK centers
• Blinding = unethical to expose participants in oral group to risks associated with prolonged courses of IV administered placebo
• participants normally on 6 weeks of IV antibiotics for osteo; excluded bacteremic, nonbacterial, mild osteo, or infectio without an oral regime 
• Primary endpoint = definite treatment failure within 1 year after randomization; secondary endpoints of treatment failure, early discontiuation of randomly assigned treatment strategy, intravenous catheter complications, c.diff, serious adverse events, resource use, health status, etc. 
• Criticisms:
  • no comparison of antibiotics made.
  • Incidence of serious adverse events very high
  • open label design --> bias
  • Antibiotics regimens not pre-specified in protocol
  • Proof of Concept - picked 2 drugs and monitored (individually monitored)
    • Non-inferiority set-up with 5% anticipated rate of treatment failure, which was later adjusted to 7.5% non-inferiority (original margin considered too restrictive).
      • 5% noninferiority margin based on consensus among wide range of researchers, ID specialists, and orthopedic surgeons, and balanced the potential risks/benefits of oral therapy.  They did not mention how they adjusted (if they went back to this original group or not) within the paper, or if they just decided arbitrarily that it was too restrictive.
        • Remember, when things are adjusted, it's still important to double check that the trial is still powered correctly
          • power calculator - clincalc.com/stats/samplesize.aspx

Statistics: what are appropriate statistical tests?

• Permuted blocks - sequence of blocks with prespecified number of treatment assignments in random order [used to avoid selection bias]
• randomize patient between groups within a set of a study population
• treatment assignments within blocks are determined in random order but desired allocation proportions are achieved exactly within block

Citation

Li HK, et al. "Oral versus Intravenous Antibiotics for Bone and Joint Infection". The New England Journal of Medicine. 2019. 380(5):425-436.

https://www.ncbi.nlm.nih.gov/pubmed/30699315