Discussion

Similar to prior studies, caregivers were highly concerned about potential complications and pain from untreated or delayed treatment and had misconceptions about the need for antibiotics. ^{Reference Kautz-Freimuth, Redaèlli, Samel, Civello, Altin and Stock8,Reference Hansen, Howlett, Del Mar and Hoffmann9} In contrast, caregivers were less concerned about risks associated with antibiotics. Caregivers expressed strong preferences for immediate antibiotics and most caregivers stated they would immediately fill prescriptions written as delayed. Conversely, caregivers did not have a strong preference for antibiotic duration; most wanted whichever length was recommended by their provider. Parental pressure is often cited as a key driver of antibiotic prescribing, ^{Reference Mangione-Smith, McGlynn, Elliott, Krogstad and Brook7} and education for caregivers on the risks and benefits of antibiotic treatment for AOM is likely needed to reduce misconceptions and curtail prescribing.

In contrast to prior clinical trials, caregivers in a community health center environment had low satisfaction with observation and delayed prescribing and were very likely to immediately fill prescriptions written as delayed. Pragmatic studies that broadly examine antibiotic use and satisfaction with observation or delayed prescribing in typical clinical practice environments are needed. Similarly, we previously reported that providers were less likely to use delayed prescribing or observation than immediate prescribing. ^{Reference Frost, Lou, Keith, Byars and Jenkins3} Further studies are warranted to assure that national approaches to AOM management and antibiotic stewardship align with the needs of caregivers and willingness of providers to follow them.

Comparatively, caregivers did not have strong preferences for duration of therapy. We previously found that providers preferred to prescribe the shortest duration of antibiotics needed. ^{Reference Frost, Lou, Keith, Byars and Jenkins3} Shorter durations of antibiotics have been shown to be equally effective for most children aged ≥2 years with AOM and are associated with reduced adverse drug events. ^{Reference Kozyrskyj, Klassen, Moffatt and Harvey10} Thus, prescribing a shorter duration of antibiotics for children aged ≥2 years is likely to substantially reduce overall antibiotic exposure in a way that is acceptable to caregivers and providers.

The strengths of this evaluation include responses from a diverse population in a community-based health system. This evaluation also has limitations. Because this was a single-center project, the data are not generalizable to other organizations and evaluation of patients from diverse settings is warranted. Selection bias likely influenced the results; respondents were more likely to identify as a race other than White or Black and were more likely to have commercial insurance than nonrespondents. In addition, the response rate was low and caregivers who felt strongly about AOM management may have been more likely to respond to the survey. The low response rate may be attributed to the characteristics of our diverse patient population including technology barriers, recruitment using text messages, concerns about being identified, literacy barriers, mistrust of the medical establishment, or lack of an incentive to respond.

In conclusion, careful evaluations of caregivers’ preferences for management and providers’ willingness to adopt recommendations across diverse practice settings and populations are needed to guide AOM management and antibiotic stewardship recommendations. In particular, studies to better understand antibiotic utilization and satisfaction with observation and delayed prescribing in typical clinical practice would be beneficial. In the interim, optimizing the duration of antibiotics prescribed is likely to be a pragmatic mechanism to reduce antibiotic use in a way that is aligned with caregiver preferences.