This is a trial addressing a practice to which I’m not opposed, and recommended by some other experts: delayed antibiotic prescribing. Delayed prescribing, providing patients with an antibiotic to fill “just in case they get worse”, has seemingly reasonable fundamental components. An encounter with a patient is a single snapshot in time, frequently without a clear picture of the ongoing clinical course. Early bacterial illness may not be clearly apparent as infection, with subsequent days bringing the diagnosis into clearer focus. In the interests of respecting patients’ time and healthcare resource utilization, why not provide a provisional antibiotic prescription to be filled by the patient if they self-assess clinical worsening?
This specific delayed antibiotic prescribing trial was performed in children presenting with an acute, uncomplicated respiratory infection – which, by their inclusion criteria included acute otitis media, pharyngitis, rhinosinusitis, or acute bronchitis. These authors randomized 436 children to one of three arms: immediate antibiotics, delayed antibiotics, or no antibiotics. Parents whose children were enrolled in the delayed antibiotics arm were instructed to start the antibiotic if their children worsened or did not start feeling better within a few days to weeks, depending on the underlying illness at time of enrollment. The primary outcome for the trial was severity and duration of acute symptoms over the following month.
The good news – the trial was a “success”. Children randomized to the delayed antibiotics strategy reported similar numbers of symptom days as those randomized to the immediate antibiotics arm. Whether it was pharyngitis, otitis media, rhinosinusitis or bronchitis, children remained symptomatic – for 4 to 10 days – regardless of their treatment arm. Delayed antibiotic prescribing, then, was not clinically harmful with respect to the primary illness.
With respect to antibiotic use and adverse effects, within 30 days of enrolment nearly the entire immediate antibiotic arm used their prescriptions. In contrast, only 25% of those randomized to delayed antibiotics did so, along with 12% of those who had no antibiotic prescribed. As to be expected, those with fewer antibiotic exposures suffered fewer gastrointestinal side effects. Unscheduled primary care utilization was uncommon and similar across all treatment groups.
So, “success”!
Except for that other niggling, less-heralded study arm: the no antibiotics arm.
The authors conclude by saying “DAP compared to IAP led to greatly reduced antibiotic use and fewer gastrointestinal adverse effects associated with antibiotic intake.” While this is not untrue, the actual final conclusion might more appropriately be: “No practical advice can be provided regarding the appropriateness of a delayed antibiotic strategy, as this trial best demonstrates the no antibiotic strategy most likely the best choice.”
The entire premise of a delayed strategy is, in the context of clinical uncertainty, there is a substantial likelihood the underlying illness will ultimately require antibiotics for successful resolution. In this trial, the authors have selected a scenario where that is not the case – and, therefore, haven’t produced terribly generalizable information regarding delayed antibiotic prescribing strategies.
The authors have, at least, provided some useful insight into human behavior with respect to delayed antibiotics and the rate at which they are filled. But, mostly, they have best demonstrated, yet again, the vast majority of children evaluated for respiratory illnesses are best treated with supportive care and time, rather than antibiotics.
“Delayed Antibiotic Prescription for Children With Respiratory Infections: A Randomized Trial”
https://pediatrics.aappublications.org/content/early/2021/02/09/peds.2020-1323