Study: Longer antibiotic course not associated with increased benefit or harm
Click Here to Manage Email Alerts
Key takeaways:
- Researchers used an instrumental variable analysis to assess antibiotic duration.
- There was no greater risk of harm or benefit associated with longer durations of antibiotic therapy.
Using a novel approach accounting for prescriber preference, researchers found that longer durations of antibiotic therapy were not associated with greater risk for harm or benefit compared with shorter durations.
“We’ve conducted previous studies that show wide variability in physician antibiotic prescribing habits, even after accounting for patient characteristics,” Bradley J. Langford, PharmD, BCIDP, pharmacist specialist of antimicrobial resistance and stewardship at Public Health Ontario, told Healio. “One such prescribing habit is the selection of duration, where there is substantial interphysician variability in the proportion of antibiotic prescriptions that are long in duration.”
Langford noted the growing body of evidence that supports the efficacy and safety of shorter antibiotic courses for infections, and said that "the notion of 'shorter is better' still applies to our clinical decision-making," but that quantifying the risk of harm associated with longer courses in a real-world setting it also important.
To do this, Langford and colleagues conducted a population-based cohort study using administrative health data from Ontario, Canada, ultimately including outpatients aged 66 to 110 years who received a prescription for amoxicillin, cephalexin, and/or ciprofloxacin, with prescriptions being categorized as short (3 to 7 days) or long (8 to 14 days) duration.
According to the study, the primary outcome was a composite of antibiotic-related harms, including adverse reactions, Clostridioides difficile infection and antibiotic resistance.
The authors said they used an instrumental variable analysis to reduce risk of bias. Langford explained that observational studies “pose a risk of bias where unmeasured variables may differ between groups” compared with randomized controlled trials. These variables, he said, can confound the relationship between antibiotic duration and measured harms, leading to spurious associations.
“Using an approach called ‘instrumental variable analysis,’ we can use a variable, the ‘instrument’ that mimics randomization that occurs in randomized trials,” he said. “In this case, since the physician prescribing of duration may be more influenced by habit than by patient characteristics, it offers a pseudo-random assignment from the patient perspective.”
Langford said that the team leveraged this instrumental variable approach to try to reduce the potential risk for bias when evaluating the potential harms and benefits of long vs. short course antibiotic therapy.
Overall, the study showed that there was no difference in medically attended visits for antibiotic harms or benefits when comparing short vs. long durations. The adjusted OR for each of the assessed antibiotics were 0.99 for amoxicillin (95% CI, 0.84-1.15), 1.11 for cephalexin (95% CI, 0.9-1.38) and 0.94 for ciprofloxacin (95% CI, 0.74-1.2).
Langford said these data suggest that this specific study population may be at lower risk of harm and benefits of antibiotic therapy, making the impact of duration “less evident.”
“Alternatively, we may have underestimated true harms as patients would have had to present to health care settings in order to be captured in the follow-up period,” he added. “Another possible explanation is that patients may individualize their own therapy durations by not taking all of their antibiotics when prescribed longer durations.”
In a sensitivity analysis where researchers analyzed the data using a propensity-matched approach, they similarly found no evidence of harms. However, they did find an increase in balancing outcomes among those receiving longer durations. This, Langford said, suggests fewer antibiotic benefits with longer durations.
“Prior to this study, the bulk of literature has shown the importance of using the shortest, most effective duration to treat infections while reducing the risk of harm for patients and the population,” Langford said. “While our study did not find an additional risk of harm with longer courses, it also did not detect any benefits associated with these longer durations.”
Collapse
Read more about
Click Here to Manage Email Alerts