Development and Preliminary Evaluation of a Parent-Reported Outcome Instrument for Clinical Trials in Acute Otitis Media

Abstract

Acute otitis media (AOM) is the most common childhood diagnosis, leading to prescription of an antibiotic in the United States. Although antibiotics are used in children with AOM, in part, to shorten the duration of symptoms, no instruments have been developed to track early changes in symptoms from the parent's point of view. The goal of the present study was to develop and evaluate a parent-reported symptom scale for children with AOM (AOM-SOS) for use as an outcome measure in AOM treatment trials. From a pool of 28 potential symptoms, we selected 7 on the basis of parent questionnaire, expert interviews, and review of the literature for inclusion in the AOM-SOS. We administered the AOM-SOS to a primary-care sample of children aged 6-25 months enrolled in a study of nasopharyngeal bacterial colonization. Children were seen for well visits, illness visits, and AOM follow-up visits. At each visit, parents completed the AOM-SOS and their children were examined by trained otoscopists. As part of the evaluation of the AOM-SOS, we examined the association between each item on the questionnaire and the clinical diagnosis of AOM while adjusting for the presence of upper respiratory tract infection. To assess responsiveness, we examined the change in AOM-SOS scores in patients with AOM who were seen for follow-up within 3 weeks of diagnosis. We evaluated 264 children (mean age, 12.5 months at entry) at a total of 642 visits. We diagnosed AOM at 24% of the visits. Each item on the questionnaire was significantly associated with the clinical diagnosis of AOM (P < 0.001 for each), before and after adjusting for the presence or absence of upper respiratory infection. The mean AOM-SOS score at visits when AOM was diagnosed was 3.71, compared with 0.96 at visits when AOM was not diagnosed (P < 0.001). Internal reliability of the scale as measured by Cronbach's alpha was 0.84. AOM-SOS scores in children with AOM who were otoscopically improved decreased by an average of 2.81 points (standardized response mean = 0.73). We have developed a short symptom scale for children with AOM. This study provides preliminary data on the performance of the AOM-SOS in a primary care sample of children.