Table 2.
Two-Year Primary and Secondary Outcomes, According to Treatment Assignment.*
| Outcome Measure or Child Characteristic | Tympanostomy-Tube (N = 129) | Medical-Management (N = 121) | All Children (N = 250) | Estimated Between-Group Difference (95% CI); P Value† |
|---|---|---|---|---|
| Two-year occurrence of episodes of acute otitis media | ||||
| No. of episodes; no. of child-yr‡ | 384; 259.5 | 378; 242.6 | 762; 502.1 | |
| Yr 1 and 2 combined — rate per child-yr§ | 1.48±0.08 | 1.56±0.08 | 1.52±0.08 | 0.97 (0.84 to 1.12); P = 0.66 |
| Yr 1 | 1.94±0.12 | 2.20±0.14 | 2.07±0.13 | |
| Yr 2 | 1.06±0.09 | 0.97±0.09 | 1.01±0.09 | |
| Frequency distribution of episodes of acute otitis media, yr 1 and 2 combined¶ | ||||
| No. of episodes — no. of children/total no. (%) | ||||
| 0 | 17/108 (16) | 12/100 (12) | 29/208 (14) | |
| 1 or 2 | 41/108 (38) | 41/100 (41) | 82/208 (39) | |
| 3 or 4 | 24/108 (22) | 29/100 (29) | 53/208 (25) | |
| ≥5 | 26/108 (24) | 18/100 (18) | 44/208 (21) | |
| Range — no. of episodes | 0 to 13 | 0 to 16 | 0 to 16 | |
| Estimated severity of episodes of acute otitis media — no. of episodes/total no.(%)‖ | ||||
| Probably nonsevere | 180/336 (54) | 168/333 (50) | 348/669 (52) | |
| Probably severe | 156/336 (46) | 165/333 (50) | 321/669 (48) | 0.91 (0.76 to 1.09) |
| Treatment failure — no. of children/total no. (%)** | ||||
| Had failure | 56/124 (45) | 74/120 (62) | 130/244 (53) | 0.73 (0.58 to 0.92) |
| Did not have failure | 68/124 (55) | 46/120 (38) | 114/244 (47) | |
| Total days with otitis-related symptoms or signs — no. of days per yr (range) | ||||
| Tube otorrhea | 7.96±1.10 (0 to 81) | 2.83±0.78 (0 to 76) | 5.44±0.70 (0 to 81) | 5.21 (2.60 to 7.82) |
| Other symptoms of acute otitis media | 2.00±0.29 (0 to 17) | 8.33±0.59 (0 to 35) | 5.11±0.38 (0 to 35) | −6.32 (−7.55 to −5.10) |
| Total days of antimicrobial treatment — no. of days per yr (range) | 8.76±0.94 (0 to 119) | 12.92±0.90 (0 to 56) | 10.80±0.67 (0 to 119) | −4.50 (−6.82 to −2.18) |
| Child and parent QOL assessments | ||||
| Score on the Otitis Media—6 Survey†† | 1.50±0.03 | 1.55±0.03 | 1.52±0.02 | −0.05 (−0.13 to 0.02) |
| Score on the Otitis Media—6 Survey — children’s overall QOL‡‡ | 8.45±0.07 | 8.37±0.07 | 8.42±0.05 | 0.06 (−0.13 to 0.24) |
| Score on the Caregiver Impact Questionnaire§§ | 10.82±0.53 | 10.93±0.55 | 10.87±0.38 | −0.04 (−1.55 to 1.47) |
| Score on the Caregiver Impact Questionnaire — caregivers’ overall QOL‡‡ | 8.55±0.06 | 8.50±0.06 | 8.53±0.04 | 0.03 (−0.14 to 0.20) |
| Parental satisfaction with treatment assignment¶¶ | 4.64±0.10 | 4.43±0.13 | 4.54±0.08 | 0.25 (−0.06 to 0.56) |
| Use of medical resources other than trial visits — no. of parent reports/no. of parental questionnaires (%)‖‖ | 738/1635 (45) | 672/1628 (41) | 1410/3263 (43) | 1.07 (0.98 to 1.18) |
| Use of nonmedical resources — no. of reported occurrences/no. of parental questionnaires (%)*** | ||||
| Missed work owing to child’s illness | 286/1635 (17) | 256/1628 (16) | 542/3263 (17) | 1.11 (0.88 to 1.41) |
| Special child care arrangements owing to child’s illness | 231/1635 (14) | 195/1628 (12) | 426/3263 (13) | 1.15 (0.89 to 1.48) |
Plus—minus values are means ±SE. Three additional secondary outcome measures — the occurrence of otorrhea, protocol-defined diarrhea, and medication-related diaper dermatitis — were also considered to be adverse events. CI denotes confidence interval, and QOL quality of life.
Estimates for rates and percentages are based on risk ratios; estimates for continuous outcomes are based on least-squares means. Estimates are adjusted for trial site, age at enrollment, and exposure or nonexposure to other children, unless noted otherwise. A P value is provided only for the primary outcome. There was no adjustment for multiple comparisons across the secondary outcomes; results are reported with point estimates and 95% confidence intervals. The confidence intervals are also not adjusted for multiple comparisons and should not be used to infer definitive treatment effects.
Values for the number of episodes are rounded to the nearest whole number. The measure “child-years” comprises both full-year and fractional-year experiences, the latter a consequence of withdrawal or loss to follow-up of some children. For each child with incomplete 2-year follow-up, we imputed the total number of episodes of acute otitis media using multivariate imputation by chained equations with 50 imputations. For the tympanostomy-tube group. 231.1 of the 259.5 child-years (89%) that are listed represented actual experience; corresponding values for the medical-management group were 222.6 of 242.6 child-years (92%). Values for the remaining child-years were imputed.
This was the primary outcome measure. The values shown reflect multivariate imputation. Before imputation, the values were 1.45±0.08 in the tympanostomy-tube group and 1.50±0.08 in the medical-management group.
Data were limited to children with at least 23 months of follow-up.
The current American Academy of Pediatrics clinical practice guideline regarding the management of acute otitis media refers to children with “severe signs or symptoms” as those with “moderate or severe otalgia or otalgia for at least 48 hours or temperature 39°C (102.2°F) or higher.”2 In an effort to simulate that definition, we used scores on two items of the five-item Acute Otitis Media Severity of Symptom (AOM-SOS) scale, version 4.0,13 in which parents are asked to rate symptoms, as compared with their child’s usual state, as “none,” “a little,” or “a lot,” with corresponding scores of 0. 1, and 2. We categorized episodes of acute otitis media as “probably severe” if the parent described the child as having had moderate or severe otalgia (“a lot” of ear tugging: i.e., a score of 2), a temperature 39°C or higher, or an AOM-SOS scale score of more than 6 (range, 0 to 10, with higher scores indicating greater severity of symptoms) on day 1.2,15
Protocol criteria for determination of treatment failure are described in the final version of the protocol. This analysis concerns 244 children with follow-up of any duration; 6 children were not evaluated beyond the enrollment visit. Of the children who had treatment failure, the failure was attributed to the occurrence of frequent episodes of acute otitis media in all 56 children in the tympanostomy-tube group and in 55 of the 74 children in the medical-management group. Of these 55 children, 35 underwent tympanostomy-tube placement and 20 did not. The mean time from randomization to tympanostomy-tube placement in these 35 children was 5.17 months. In the remaining 19 children, the episode-frequency criterion for treatment failure was not met, but treatment failure was instead ascribed to the children’s receipt of tympanostomy-tube placement at parental request. The mean time from randomization to tympanostomy-tube placement in these 19 children was 2.51 months. Seven of these 19 children had had no episodes of acute otitis media after randomization; were these 7 children to be excluded from the analysis, the risk of treatment failure would still be lower in the tympanostomy-tube group than in the medical-management group.
The Otitis Media—6 Survey was scored with the use of an ordinal response scale from 1 (no problem) to 7 (greatest problem).17
Children’s and parents’ overall quality of life was scored with the use of an ordinal response scale from 0 (worst quality of life) to 10 (best quality of life).
Scoring on the Caregiver Impact Questionnaire was expanded to a continuous response scale from 0 (no effect on caregiver) to 100 (greatest effect).18
Parental satisfaction with the child’s treatment assignment was assessed on a five-point scale, with higher numbers indicating greater satisfaction.
Shown is the number of parent reports that the child had seen any health care provider since the last trial visit. Eleven children in the tympanostomy-tube group and two in the medical-management group underwent a second tube-placement procedure at intervals ranging from 191 to 553 days after the first procedure. Eight children in the tympanostomy-tube group and two children in the medical-management group underwent adenoidectomy with or without tonsillectomy, procedures not contemplated in the trial protocol.
Shown is the number of instances in which parents reported missing work or the need for special child care arrangements for at least 1 day.