COSTS OF OTITIS MEDIA IN A MANAGED CARE POPULATION

Abstract

Otitis media has been increasing in the past 15 years, and its treatment remains an area of practice variation and innovation.1 The increasing antimicrobial resistance of Streptococcus pneumoniae, short course antimicrobial therapy, national dissemination of guidelines to promote judicious antimicrobial prescribing and the development of a pneumococcal conjugate vaccine effective in young children all may cause changes in the incidence and cost of otitis media in the near future. There is scant empirical evidence about otitis media costs, especially work loss costs. Such information is needed to gauge the impact of current and future interventions against otitis media. The aim of this study was to describe the medical, nonmedical and work loss costs of otitis media among children younger than 14 years of age in a large population. Methods. To estimate the nonmedical costs of otitis media, we interviewed the parents of 300 children younger than 14 years old, 2 weeks after they had made an otitis-related visit sometime during August to November, 1997. The children were enrolled in the Kaiser Permanente Medical Care Program, a group model, nonprofit health maintenance organization (HMO) providing comprehensive medical care to approximately 2.8 million members in Northern California. The interview, conducted in English or Spanish, obtained information on work missed by adult family members to care for the child from the week before the visit until the interview and on other costs such as days missed from day care, extra babysitting and travel-related expenses. To estimate medical costs we performed a retrospective analysis of Kaiser Permanente Medical Care Program computerized utilization data for 1997 using a random sample of 119 970 children younger than 14 years with continuous enrollment and drug coverage in 1997. (A list of the relevant ICD-9 and CPT-4 codes for otitis-related outpatient visits and hospital services is available on request.) Otitis-related visits were classified as being for a new infection, a relapsed infection, follow-up or other otitis-related reasons. First, a visit was considered to be for a new episode if it had a diagnosis of otitis media, if appropriate antimicrobial therapy was prescribed and if it occurred >21 days after the previous otitis visit at which antimicrobial therapy had been prescribed.2 Second, a relapsed visit was similar to a new visit except it occurred within 21 days of a prior otitis visit at which antimicrobial treatment had been prescribed. The third type of visit was a follow-up, defined as either (1) an otitis visit during which antimicrobials were not prescribed or were prescribed for prophylaxis only or (2) an office visit with a “normal examination” diagnosis and occurred within 42 days of an otitis visit. Finally we identified other related visits with diagnoses such as Eustachian tube dysfunction, mastoiditis or hearing loss as well as visits for the following procedures: tympanometry, myringotomy and audiology tests (audiometry, impedance and auditory brainstem response). All costs were summarized at the episode level. An episode was defined as an initial new visit and all otitis-related hospitalizations and visits until the next new otitis visit. The cost of an episode was the total of the costs of these events and of antimicrobial treatments and prophylaxis. We also categorized episodes into simple (with no relapsed visit) or complex (with at least one relapsed visit). Results. Work loss and other costs to families. The parents of 21 of the 387 (5%) children selected at random for the work loss survey said the child did not have otitis media at the visit identified. Interviews were completed with 300 (82%) of the remaining parents. Among these 67% of mothers and 94% of fathers worked for pay. The median income category was between $20 000 and $40 000 for both working mothers and fathers. In 45% of the families at least one adult member missed work due to the child's otitis media visit. The average number of work hours lost per family was 5.9, costing an average $120. The total of other costs incurred by the families (days missed from prepaid day care, extra babysitting, travel-related expenses, copayments and over-the-counter medications) averaged $13 per family. Medical costs. Of the total medical costs outpatient visits comprised $6.28 million (78%), hospitalizations comprised $1.08 million (14%) and antimicrobial treatment and prophylaxis comprised $0.66 million (8%). (Detailed data on incidence and costs of outpatient visits, hospital services and antimicrobial prescriptions are available on request to the authors.) In this cohort there were 112 192 person years of follow-up, for an average cost of approximately $72 per person year of follow-up. Costs of simple vs. complex episodes.Table 1 summarizes the costs of otitis media episodes, which could include more than one outpatient visit as well any hospitalizations and antimicrobials associated with otitis-related visits. For the average simple episode of otitis media (that had no relapsed visits), work loss cost $114 and medical care cost $131. Complex episodes had disproportionately higher work loss as well as other nonmedical and medical costs compared with simple episodes. For the average complex episode work loss cost $404 and medical care cost $327. The trend of increasing average cost with increasing numbers of complex visits was the result not only of increased pediatric outpatient visit and medication costs but also of increased hospitalization and other outpatient visit (e.g. otorhinolaryngology) costs.TABLE 1: Average costs of simple and complex* episodes of otitis media in children <14 years of age, Northern California Kaiser Permanente, 1997Discussion. The key findings of this study were that the nonmedical costs of otitis media are at least equal to its medical costs and that complex otitis media episodes have disproportionately high costs compared with simple episodes. Our study is unique in describing the work loss and other nonmedical as well as medical costs of the full spectrum of otitis media episodes ranging from simple to complex. National estimates of otitis media costs have been made, but most previous estimates have been generated from utilization statistics and cost assumptions rather than empirical observation.1, 3, 4 In addition new clinical interventions may have differential effects on simple vs. complex otitis media. Analyses of national datasets have not contained the level of detail on visit patterns and antibiotic use that would be most useful for monitoring these differential effects on utilization and costs. Two previous studies reporting the nonmedical costs of otitis were limited to first otitis media visits5 or very small samples.6 Those studies had widely disparate findings. In our study both the medical and work loss costs for simple otitis media episodes were more than double those observed in the study of Kaplan et al.5 Generalizability should be considered whenever utilization or costs are studied in a single population. This HMO setting had a relatively cohesive clinical practice group, which published regional guidelines for otitis media treatment and follow-up in 1995 and 1997. Compared with a recent study in a New England health plan,7 we observed a lower rate of myringotomy with ventilatory tube placement and less use of cephalosporins for simple otitis media. The large patient population of the HMO in this study is diverse in race/ethnicity and socioeconomic status, although families with very low and very high incomes are underrepresented.8 The average 5.9 h of parent work loss we observed most likely represents the high end of the possible range because most families in the study population had two parents working for pay. The retrospective nature of our utilization analysis imposes limits on its interpretation. For example our algorithm for classifying visits looked at antibiotic use but did not attempt to differentiate between acute otitis media and middle ear effusion based on visit coding. The computerized analysis may have slightly overestimated the incidence of otitis media visits; ˜5% of parents in the interview sample said that their children did not have otitis media at the visit identified. In addition the chart review conducted for validation found that our algorithm for classifying visits missed antibiotic prescriptions for some patients. Thus the analysis probably overestimated the proportion of visits that were for follow-up and underestimated the proportions for new and relapsed otitis media. However, this limitation would not affect the total incidence of visits for otitis media or our findings on nonmedical costs. Acknowledgments. This study was supported by a grant from Wyeth-Lederle Vaccines and Pediatrics.