Effect of Antimicrobial Treatment on the Resolution of Middle-Ear Effusion After Acute Otitis Media

Effect of Antimicrobial Treatment on the Resolution of Middle-Ear Effusion After Acute Otitis Media Abstract Background Acute otitis media (AOM) induces middle-ear effusion (MEE), which affects hearing. The effect of antimicrobial treatment on the resolution of MEE is controversial, and the factors that affect resolution are unknown. Methods We studied the effect of antimicrobial treatment on the time to the resolution of MEE as a secondary objective in our randomized double-blind placebo-controlled trial. Children aged 6 to 35 months with stringently diagnosed AOM were allocated to receive amoxicillin-clavulanate (161 patients) or placebo (158 patients) for 7 days and closely followed for 3 months. This study was registered at ClinicalTrials.gov (identifier NCT00299455). Results The median times to resolution of MEE were 20 days (95% confidence interval [CI], 16–24 days) and 29 days (95% CI, 26–32 days) in the amoxicillin-clavulanate and placebo groups, respectively (P = .10). The resolution of MEE was confirmed in 138 (86%) of 161 and 132 (84%) of 158 patients in the amoxicillin-clavulanate and placebo groups, respectively (P = .59). In multivariable analysis, the resolution of MEE was prolonged most significantly by at least 1 recurrence of AOM during follow-up. MEE resolved in 65 (65%) of 100 patients with a recurrence of AOM during follow-up and in 205 (94%) of 219 of those without a recurrence (P < .001) (median times to resolution, 67 vs 15 days, respectively; P < .001). Conclusions Immediate antimicrobial treatment of AOM does not significantly affect the resolution of MEE in young children. Subsequent recurrences of AOM are a major reason for the persistence of MEE. Acute otitis media (AOM) is among the most common reasons for outpatient sick visits and is the most common reason for antimicrobial treatment in young children [1]. The diagnosis is based on acute symptoms, acute inflammatory signs of the tympanic membrane, and middle-ear effusion (MEE). MEE decreases the flexibility of the tympanic membrane and the vibration of ossicles and thus impairs the conductive component of hearing [2]. Therefore, persistent MEE is the major indication for tympanostomy tube placements, which cost nearly $1.5 billion annually in the United States [3, 4]. Thus, it is important to know whether antimicrobial treatment of AOM affects the resolution of MEE. Most previous randomized double-blind placebo-controlled studies on the treatment of AOM have found that antimicrobial treatment does not affect the resolution of MEE [5–9]. Those studies, however, had relatively long follow-up intervals, which might have hindered the observance of a disappearance and reappearance of MEE [10]. Therefore, they did not report the time to resolution of MEE but, instead, its prevalence. Recently, Tapiainen et al [11] found that antimicrobial treatment significantly accelerates the resolution of MEE. They performed a close 2-month follow-up that included daily parental tympanometry; however, a majority of their study patients were older than 24 months and, thus, beyond the age to be at high risk for the persistence of MEE. We previously reported the results of the primary outcome of our randomized double-blind placebo-controlled study on the treatment of AOM in young children. Amoxicillin-clavulanate accelerated improvement of the children’s overall condition and the resolution of individual symptoms such as fever [12]. We closely followed the resolution of MEE for 3 months. Here, we report the results of our prespecified secondary objective, which was to determine the effect of antimicrobial treatment on the time to resolution of MEE and which factors affect the resolution of MEE, including factors during follow-up as a novel aspect. METHODS Patients and Diagnostic Criteria Children aged 6 to 35 months were eligible for our diagnostic screening when their parents suspected AOM on the basis of the child’s acute symptoms. A list of the exclusion criteria, along with their reasons, is provided in Supplementary Methods. The inclusion criterion was AOM as diagnosed by a study physician. The diagnosis of AOM required the following 3 criteria. First, the child’s MEE had to be detected by means of pneumatic otoscopic examination that revealed at least 2 of the following tympanic-membrane findings: bulging position, decreased or absent mobility, abnormal color or opacity not caused by scarring, and/or air–fluid interfaces. Second, at least 1 of the following acute inflammatory signs in the tympanic membrane had to be present: distinct erythematous patches or streaks and/or increased vascularity over a full, bulging, or yellow tympanic membrane. Third, the child had to have acute symptoms of AOM. A parent of each child provided written informed consent. The protocol was approved by the ethics committee of the Hospital District of Southwest Finland. Study Design and Follow-Up This randomized double-blind placebo-controlled study was initiated by the investigators and was conducted independently of any commercial entities. Our primary objective was to study the efficacy of antimicrobial therapy on treatment failure, and the secondary objective, reported here, was to study the efficacy of antimicrobial treatment for the resolution of MEE. Our hypothesis was that amoxicillin-clavulanate would accelerate the resolution of MEE. Another objective was to study which factors characterized at study entry and which factors documented during follow-up affected the resolution of MEE. Eligible patients were assigned randomly to receive amoxicillin-clavulanate (40/5.7 mg/kg, divided into 2 daily doses) or placebo for 7 days. The placebo was similar to the active treatment in appearance and taste. The description of the study drugs, the randomization procedure, the procedure for concealment of study assignments, and details of the rescue treatment and other open antimicrobial treatments are provided in Supplementary Methods. After the enrollment visit (study day 1), we followed the patients to determine when/if their MEE resolved and the development of recurrences of AOM by closely scheduled visits on study days 3, 8, 15, 30 (45, if MEE had not resolved earlier), and 60. Thereafter, the patients who had persistent MEE were followed up every other week until study day 90 (±10 days) or until referral to tympanostomy tube placement. Apart from the scheduled visits, we arranged additional sick visits for any patient on any day of the week if the parents were concerned about their child’s condition or if they suspected AOM. At each visit, the study physician interviewed the parents about the symptoms of their child, performed a clinical examination (including thorough pneumatic otoscopic and tympanometry examinations), and acquired a nasopharyngeal sample for the detection of respiratory viruses and bacteria (see details in Supplementary Methods). If AOM was diagnosed, the first episode was treated primarily with amoxicillin-clavulanate, and subsequent episodes were treated with amoxicillin. This study was registered at ClinicalTrials.gov (identifier NCT00299455). Outcome The outcome was the time to resolution of MEE. The time was from the day of random assignment to the day on which a study physician confirmed the resolution of MEE, which was based on a pneumatic otoscopic assessment; both tympanic membranes had to have good mobility, complete or partial transparency, and no visibly remnant effusion. Statistics The sample size had been calculated for the primary outcome (see Supplementary Methods), the results of which were published previously [12]. No separate sample-size calculations had been made for the secondary outcome reported here. All analyses were made in the intention-to-treat population. The medians and means of the time-to-outcome data were analyzed with the Kaplan-Meier method and compared with results of the log-rank test. Cox regression was used to study the effect of treatment allocation and of predefined factors on the resolution of MEE. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution . Proportions of patients were compared with the χ2 or Fisher exact test, as appropriate. All P values are 2-sided. The analyses were conducted with SPSS 22 (IBM SPSS Statistics, IBM Corporation, Armonk, NY) by Dr Ruohola. Details of the statistics and definitions for the factors included in the analyses are provided in Supplementary Methods. RESULTS Study Population The intention-to-treat population included 319 patients, 161 in the amoxicillin-clavulanate group and 158 in the placebo group, 53 (34%) of whom received delayed amoxicillin-clavulanate treatment as rescue therapy (see Supplementary Figure 1). The characteristics of the treatment groups were balanced (Table 1). Table 1. Baseline Characteristics of the Intention-to-Treat Population Characteristic  Amoxicillin-Clavulanate Group (N = 161)  Placebo Group (N = 158)  Demographics   Age (mean [range]) (mo)  16 (6–35)  16 (6–35)   Age (n [%])    6–11 mo  57 (35)  59 (37)    12–23 mo  78 (48)  65 (41)    24–35 mo  26 (16)  34 (22)   Male sex (n [%])  92 (57)  90 (57)  Otitis media risk factors (n [%])   Family history of otitis media  90 (56)  98 (62)   Sibling(s) in the household  89 (55)  93 (59)   Day care attendance  87 (54)  86 (54)   Parental smoking  57 (36)  48 (30)   Current use of pacifier  80 (50)  85 (54)  Otitis media history   Previous episodes of AOM (mean [range]) (n)  2 (0–10)  2 (0–10)   Previous episodes of AOM (n [%])    0  43 (27)  51 (32)    1–3  88 (55)  79 (50)    ≥4  30 (19)  28 (18)   Time from previous episode of AOM (mean [range]) (mo)a  3 (0–22)  3 (0–15)   Time from previous episode of AOM (n [%])a    ≤1 mo  36 (22)  31 (20)    >1–3 mo  42 (26)  45 (29)    >3 mo  40 (25)  31 (20)  Ear examination findings at entry (n [%])   Unilateral AOMb  99 (62)  89 (57)   Severe bulging of tympanic membrane  45 (28)  39 (25)   Peaked tympanogramc  32 (20)  25 (16)  Preceding symptoms (n [%])   Fever ≥38°C  64 (40)  46 (29)   Parentally reported ear pain  123 (76)  126 (80)   Decreased activity  76 (47)  59 (37)   Respiratory symptoms  156 (97)  156 (99)  Characteristic  Amoxicillin-Clavulanate Group (N = 161)  Placebo Group (N = 158)  Demographics   Age (mean [range]) (mo)  16 (6–35)  16 (6–35)   Age (n [%])    6–11 mo  57 (35)  59 (37)    12–23 mo  78 (48)  65 (41)    24–35 mo  26 (16)  34 (22)   Male sex (n [%])  92 (57)  90 (57)  Otitis media risk factors (n [%])   Family history of otitis media  90 (56)  98 (62)   Sibling(s) in the household  89 (55)  93 (59)   Day care attendance  87 (54)  86 (54)   Parental smoking  57 (36)  48 (30)   Current use of pacifier  80 (50)  85 (54)  Otitis media history   Previous episodes of AOM (mean [range]) (n)  2 (0–10)  2 (0–10)   Previous episodes of AOM (n [%])    0  43 (27)  51 (32)    1–3  88 (55)  79 (50)    ≥4  30 (19)  28 (18)   Time from previous episode of AOM (mean [range]) (mo)a  3 (0–22)  3 (0–15)   Time from previous episode of AOM (n [%])a    ≤1 mo  36 (22)  31 (20)    >1–3 mo  42 (26)  45 (29)    >3 mo  40 (25)  31 (20)  Ear examination findings at entry (n [%])   Unilateral AOMb  99 (62)  89 (57)   Severe bulging of tympanic membrane  45 (28)  39 (25)   Peaked tympanogramc  32 (20)  25 (16)  Preceding symptoms (n [%])   Fever ≥38°C  64 (40)  46 (29)   Parentally reported ear pain  123 (76)  126 (80)   Decreased activity  76 (47)  59 (37)   Respiratory symptoms  156 (97)  156 (99)  Abbreviation: AOM, acute otitis media. aAmong those patients who had had a previous episode of AOM. bUnilateral AOM was defined as the diagnosis of AOM in only 1 ear, with the other ear showing no otoscopic signs of pathology or only middle-ear effusion without acute inflammatory signs. Data are missing for 2 patients in the amoxicillin-clavulanate group and 2 patients in the placebo group, because an adequate view of the contralateral tympanic membrane was not reached because of thick cerumen. cA peaked tympanogram was defined as a tympanogram with a height (static acoustic admittance) of ≥0.2 millimhos (mmho) and width ≤300 dekapascals (daPa). The selected tympanogram was obtained from the ear with the most severe otoscopic findings. View Large Effects of Amoxicillin-Clavulanate on MEE Resolution The median time to resolution of MEE was 20 days (95% confidence interval [CI], 16–24 days) in patients who were allocated to the amoxicillin-clavulanate group and 29 days (95% CI, 26–32 days) in patients who were allocated to the placebo group (P = .10) (Figure 1A). The complete resolution of MEE was confirmed during the 3-month follow-up in 138 (86%) of the 161 patients in the amoxicillin-clavulanate group and in 132 (84%) of the 158 patients in the placebo group (P = .59). Additional details of the effects of amoxicillin-clavulanate are provided in Table 2. Table 2. Effect of Amoxicillin-Clavulanate Treatment on the Resolution, Persistence, and Prevalence of MEE in Young Children With Stringently Diagnosed AOM Effect  Amoxicillin-Clavulanate Group  Placebo Group  P  Time to resolution of MEE (median [95% CI]) (days)a   All patients  20 (16–24)  29 (26–32)  .10   Age < 24 mo  22 (13–31)  29 (26–32)  .20   Age ≥ 24 mo  16 (15–17)  30 (17–43)  .12  Time to resolution of MEE (mean [95% CI]) (days)a   All patients  35 (30–40)  41 (36–46)  .10   Age < 24 mo  36 (31–42)  42 (37–48)  .20   Age ≥ 24 mo  26 (17–35)  37 (27–47)  .12  Persistence of MEE (n/N [%])a   By 2 wk  80/159 (50)  100/156 (64)  .013   By 3 wk  68/158 (43)  90/155 (58)  .008   By 4 wk  51/157 (33)  60/155 (39)  .25   By 6 wk  39/157 (25)  49/154 (32)  .17   By 8 wk  25/155 (16)  35/150 (23)  .11   By 12 wk  17/155 (11)  18/150 (12)  .77  Prevalence of MEE (n/N [%])b   At 2 wkc  85/157 (54)  107/151 (71)  .002   At 4 wkd  80/146 (55)  79/143 (55)  .94   At 8 wke  58/134 (43)  68/131 (52)  .16  Effect  Amoxicillin-Clavulanate Group  Placebo Group  P  Time to resolution of MEE (median [95% CI]) (days)a   All patients  20 (16–24)  29 (26–32)  .10   Age < 24 mo  22 (13–31)  29 (26–32)  .20   Age ≥ 24 mo  16 (15–17)  30 (17–43)  .12  Time to resolution of MEE (mean [95% CI]) (days)a   All patients  35 (30–40)  41 (36–46)  .10   Age < 24 mo  36 (31–42)  42 (37–48)  .20   Age ≥ 24 mo  26 (17–35)  37 (27–47)  .12  Persistence of MEE (n/N [%])a   By 2 wk  80/159 (50)  100/156 (64)  .013   By 3 wk  68/158 (43)  90/155 (58)  .008   By 4 wk  51/157 (33)  60/155 (39)  .25   By 6 wk  39/157 (25)  49/154 (32)  .17   By 8 wk  25/155 (16)  35/150 (23)  .11   By 12 wk  17/155 (11)  18/150 (12)  .77  Prevalence of MEE (n/N [%])b   At 2 wkc  85/157 (54)  107/151 (71)  .002   At 4 wkd  80/146 (55)  79/143 (55)  .94   At 8 wke  58/134 (43)  68/131 (52)  .16  Abbreviations: AOM, acute otitis media; CI, confidence interval; MEE, middle-ear effusion. aFigures are based on Kaplan-Meier analyses, P values for the medians and means are based on log-rank tests, and P values for the proportions are based on the χ2 or Fisher exact test, as appropriate. bIncludes those patients who attended study clinic visit in the time window. Figures are based on cross-tabulations, and P values are based on the χ2 or Fisher exact test, as appropriate. cAt study days 11–19. dAt study days 24–36. eAt study days 50–70. View Large Figure 1. View largeDownload slide Kaplan-Meier curves for the time to the resolution of middle-ear effusion (MEE) after acute otitis media (AOM). (A) Comparison between treatment allocation groups. (B) Comparison between patients with and those without a recurrence of AOM during follow-up before the resolution of MEE. P values were determined by the log-rank test. Figure 1. View largeDownload slide Kaplan-Meier curves for the time to the resolution of middle-ear effusion (MEE) after acute otitis media (AOM). (A) Comparison between treatment allocation groups. (B) Comparison between patients with and those without a recurrence of AOM during follow-up before the resolution of MEE. P values were determined by the log-rank test. Factors That Affect MEE Resolution Details of the univariable analyses are presented in Supplementary Table 1. The factors that tended (P < .15) to shorten the time to resolution of MEE during the entire 3-month follow-up period were amoxicillin-clavulanate, unilaterality of AOM at entry, and a peaked tympanogram at entry. In contrast, age less than 24 months tended (P < .15) to prolong the time to resolution of MEE, as did the following factors, characterized during follow-up: a recurrence of AOM any time after the 7-day study-drug period and before the resolution of MEE, open antimicrobial treatment, laboratory-confirmed respiratory viral infection, and nasopharyngeal bacterial colonization. The factors that remained significant in the multivariable analyses are shown in Figure 2. During the entire 3-month follow-up period, a recurrence of AOM, respiratory viral infection, and nasopharyngeal bacterial colonization prolonged the resolution of MEE. Of the factors characterized at entry, a peaked tympanogram indicated a favorable prognosis. Age less than 24 months indicated an unfavorable prognosis; the hazard ratio for the time to MEE resolution was 0.74 (95% CI, 0.55–1.01; P = .06). However, age was excluded from the final step of the model, because its P value exceeded the predefined level of significance. Figure 2. View largeDownload slide Forest plot of the factors that had an independent effect on the resolution of middle-ear effusion (MEE). Hazard ratios and 95% confidence intervals (CIs) for the resolution of MEE are based on a multivariable Cox regression model for the entire 3-month follow-up period and on separate models for each of the 3 follow-up months. Each of the 4 models included those patients in whom MEE had not yet resolved. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution. aTotal number was 319, but the model excluded 6 patients because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. bTotal number was 319, but the model excluded 10 patients, 4 because of missing data on unilateral acute otitis media and 6 because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. cThe model did not include 2 of the factors (ie, respiratory viral infection and nasopharyngeal bacterial colonization) because sampling was infrequent during the third month of follow-up. Figure 2. View largeDownload slide Forest plot of the factors that had an independent effect on the resolution of middle-ear effusion (MEE). Hazard ratios and 95% confidence intervals (CIs) for the resolution of MEE are based on a multivariable Cox regression model for the entire 3-month follow-up period and on separate models for each of the 3 follow-up months. Each of the 4 models included those patients in whom MEE had not yet resolved. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution. aTotal number was 319, but the model excluded 6 patients because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. bTotal number was 319, but the model excluded 10 patients, 4 because of missing data on unilateral acute otitis media and 6 because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. cThe model did not include 2 of the factors (ie, respiratory viral infection and nasopharyngeal bacterial colonization) because sampling was infrequent during the third month of follow-up. During the first month of the 3-month follow-up period, the independent factors that prolonged the resolution of MEE were a recurrence of AOM, respiratory viral infection, and nasopharyngeal bacterial colonization. Favorable prognostic factors were unilaterality of AOM and a peaked tympanogram at entry. During the second and third months of the 3-month follow-up period, a recurrence of AOM independently prolonged the resolution of MEE, whereas a peaked tympanogram was a favorable prognostic factor. Unfavorable prognostic factors were age less than 24 months and a sibling(s) in the household, both only during the third month of the 3-month follow-up period. Effect of a Recurrence of AOM During Follow-Up A recurrence of AOM at any time after the 7-day study-drug period significantly prolonged the resolution of MEE (Figure 1B). As Figure 3 shows, the median time to MEE resolution was longest in 100 patients who had at least 1 recurrence (67 days [95% CI, 49–85 days]) and, in contrast, shortest in 219 patients who did not have a recurrence of AOM (15 days [95% CI, 14–16]) (P < .001). The resolution of MEE was documented in 65 (65%) of 100 patients with a recurrence and 205 (94%) of 219 patients without a recurrence (P < .001). According to posthoc subgroup analyses, the disadvantageous effect of a recurrence of AOM on MEE resolution was not modified by amoxicillin-clavulanate treatment or any of the factors characterized at entry or documented during follow-up (see Supplementary Figure 2). Figure 3. View largeDownload slide Median times to resolution of middle-ear effusion. Medians (95% confidence intervals [CIs]) are based on the Kaplan-Meier method, and P values are based on the log-rank test. aData for 4 patients were missing. bData for 6 patients were missing. Figure 3. View largeDownload slide Median times to resolution of middle-ear effusion. Medians (95% confidence intervals [CIs]) are based on the Kaplan-Meier method, and P values are based on the log-rank test. aData for 4 patients were missing. bData for 6 patients were missing. Association of the Symptomatic Relief With the Resolution of MEE The hazard ratios for being asymptomatic by day 8 and day 16 for the time to the MEE resolution were 0.95 (95% CI, 0.74–1.23; P = .70) and 1.23 (95% CI, 0.96–1.61; P = .09), respectively. DISCUSSION In this study, treatment of AOM with amoxicillin-clavulanate in young children did not shorten the time to MEE resolution significantly; instead, new episodes of AOM during follow-up remarkably prolonged MEE. Our finding that treatment with amoxicillin-clavulanate shortly accelerated the resolution of MEE but did not significantly shorten its overall duration is, at first sight, inconsistent with results from a recent study by Tapiainen et al [11], because they concluded that amoxicillin-clavulanate effectively reduces the duration of MEE. However, closer observation revealed that this inconsistency can be explained by the age differences in the study populations. Our study population was at high risk of MEE persistence and AOM recurrence because the mean age was 1.3 years, whereas the population of the Tapiainen et al [11] study was at low risk because the mean age was 4.4 years. In these 2 studies among children younger than 24 months, the mean MEE durations in the amoxicillin-clavulanate and placebo groups were very similar; both studies had an approximately 1-week difference in MEE durations between the treatment groups. Hoberman et al [13] had short follow-up intervals and a study design and population also otherwise comparable to ours. They observed a significant difference between treatment groups in the persistence of MEE at their final follow-up visit 3 weeks after entry. Similarly, Kaleida et al [14] detected a significant difference 2 weeks after study entry. We also observed a statistically significant treatment effect at 2 and 3 weeks, which indicates that antimicrobial treatment accelerated the resolution of MEE. However, our results suggest that this accelerating effect was short lasting, because at 4 weeks and thereafter, no significant treatment effect was observed. The overall difference in the median duration of MEE between our study groups was 9 days, which was statistically insignificant. Other previous placebo-controlled trials did not find any differences between their antimicrobial-treatment and placebo groups [5–9]. However, these studies reported the prevalences of MEE instead of the persistence of MEE. Prevalence figures incorporate persistent MEE and newly developed episodes, and, as results of our analyses show, prevalence figures poorly reflect the true persistence of MEE [10]. Altogether, it seems that in older children, the favorable accelerating effect of antimicrobial treatment on MEE resolution might be sustained. In contrast, at an otitis-prone age, the effect of antimicrobial treatment on the resolution of MEE seems to be modest and short lasting and, thus, clinically insignificant. A recurrence of AOM during follow-up significantly altered the course of MEE resolution. When a recurrence of AOM was diagnosed before the resolution of effusion, only 65% of the children were free of effusion during the 3-month follow-up period, and the resolution time was almost 2 months longer than that in children without a recurrence before MEE was resolved. This finding is novel but not unexpected, because explanatory mechanisms are available. A recurrence of AOM is accompanied by respiratory viral infection, which deteriorates the function of the Eustachian tube [15]. A recurrence also reinduces inflammation and, thus, effusion in the middle ear. The formation of bacterial biofilms, common in recurrent AOM [16], might be an explanation, although recurrences or a recent episode of AOM before study entry did not affect the resolution of MEE. Current guidelines advise differentiating between recurrent episodes of AOM and persistence of MEE (or, in other words, chronic otitis media with effusion) because the persistence of MEE can cause hearing impairment and, thus, is the primary indication for tympanostomy tube placement [3]. Our results indicate that in young children who do not have craniofacial anomalies or known immunodeficiencies, recurrences of AOM and the persistence of MEE cannot be differentiated from each other because recurrences of AOM seem to be a major reason for the persistence of MEE after the initial episode of AOM. The prerequisites for AOM development are respiratory viral infection and nasopharyngeal bacterial colonization, most commonly by Streptococcus pneumoniae, Haemophilus influenzae, and/or Moraxella catarrhalis [17, 18]. It was an unexpected finding that these microbes independently prolonged the resolution of MEE. The effect of respiratory viral infection is probably mediated by inflammation of the respiratory epithelium, which deteriorates the function of the Eustachian tubes [15]. The effect of bacterial nasopharyngeal colonization might be caused by interaction between bacteria and viruses. We recently found that M catarrhalis is independently associated with respiratory symptoms, especially rhinitis, in young children with respiratory viral infection [19]. In our current study, the risk factors for otitis media, such as day care attendance and parental smoking, did not have remarkable effects on MEE resolution. However, in the prevention of MEE persistence, the risk factors should also be recognized. If the risk factors are reduced, the acquisition of respiratory viruses and colonizing bacteria might be reduced, and therefore, the risk of AOM recurrences might be reduced also, which all might prevent the persistence of MEE. According to our results and those of Renko et al [20], a favorable prognostic factor for the resolution of MEE seems to be unilaterality of AOM. However, in our study, a peaked tympanogram was the most favorable prognostic factor. This finding is logical, because a peaked tympanogram is an indicator of small volume of MEE [21, 22]; in these instances, the bulging position of the tympanic membrane is probably a result of positive pressure in the middle ear. The drawback to using a peaked tympanogram as a prognostic factor is that it is a rare finding when stringent criteria are used to diagnose AOM. The major strengths of our study are the nonselective study population at an otitis-prone age, the stringent diagnostic criteria, and the tight follow-up schedule. To our knowledge, our study is the first randomized double-blind placebo-controlled trial in which all patients were closely followed until MEE resolution or referral for tympanostomy tube placement; only 5% of the patients were lost to follow-up. The major methodologic limitation was the subjectivity of pneumatic otoscopy even when added with video recording. However, visibly low levels of MEE are not necessarily obtained by tympanocentesis, and tympanometry often produces a peaked curve that suggests the resolution of MEE. It is impossible currently to determine which volume of MEE is clinically relevant; therefore, we aimed to observe even small volumes of MEE and, thus, to be assured of total MEE resolution, which can be criticized. It could be questioned if a higher dose of amoxicillin-clavulanate might have been effective, although the dose (40 mg/kg per day) we chose had excellent antimicrobial coverage in our study population. A study by Hoberman et al [13] was much like ours, but they chose a dose of 90 mg/kg per day. At their 3-week final control visit, the proportions of patients who had persistent MEE were actually somewhat higher, and the difference between the study groups was smaller than that in our study. Thus, it does not seem that the dose is a major reason why amoxicillin-clavulanate had no overall effect. Our study can also be criticized for the lack of sample-size calculation for the reported secondary outcome. However, although with a larger sample size the difference between treatment-allocation groups would have been statistically significant, we would regard the observed difference to be clinically insignificant. In conclusion, our study reveals that in young children with AOM, immediate antimicrobial treatment (such as with amoxicillin-clavulanate) that has appropriate coverage does not significantly shorten the time to MEE resolution. The major reason for inadequate resolution and, thus, for the persistence of MEE is recurrence of AOM before MEE resolution. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online. Notes Acknowledgments. We thank all the families who participated in this study and the staff at the health center of Turku city primary care for their commitment to the study. We acknowledge Raakel Luoto, MD, PhD, and Elina Lahti, MD, PhD, for their contribution to data collection; Sari Rajamäki, RN, Maarit Rosenblad, RN, and Kaisa Erkkilä, RN, for their assistance at the study clinic; Pentti Huovinen, MD, PhD, Jari Jalava, PhD, and Laura Lindholm, MSc, for bacterial analyses; Matti Waris, PhD, and Raija Vainionpää, PhD, for viral analyses; Tero Vahlberg, MSc, for the allocation procedure; and Eliisa Löyttyniemi, MSc, and Maria Ruohola, BS, for data handling. Financial support. This work was supported by the Fellowship Award of the European Society for Paediatric Infectious Diseases (to A. R.) and by grants from the Foundation for Paediatric Research, the Research Funds from Specified Government Transfers, the Jenny and Antti Wihuri Foundation, the Paulo Foundation, the Maud Kuistila Memorial Foundation, the Emil Aaltonen Foundation, the Finnish Cultural Foundation: Varsinais-Suomi Regional Fund, the Turku University Hospital Research Foundation, the Finnish Medical Foundation, and the Turku University Foundation. Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Schilder AG Chonmaitree T Cripps AWet al.  . Otitis media. Nat Rev Dis Primers  2016; 2: 60– 3. Google Scholar CrossRef Search ADS   2. Marsh RR Baranak CC Potsic WP. Hearing loss and visco-elasticity of middle ear fluid. Int J Pediatr Otorhinolaryngol  1985; 9: 115– 20. Google Scholar CrossRef Search ADS PubMed  3. Rosenfeld RM Schwartz SR Pynnonen MAet al.  . Clinical practice guideline: tympanostomy tubes in children. Otolaryngol Head Neck Surg  2013; 149: S1– 35. Google Scholar CrossRef Search ADS PubMed  4. Mui S Rasgon BM Hilsinger RLet al.  . Tympanostomy tubes for otitis media: quality-of-life improvement for children and parents. Ear Nose Throat J  2005; 84: 418, 20–2, 24. Google Scholar PubMed  5. Mygind N Meistrup-Larsen KI Thomsen Jet al.  . Penicillin in acute otitis media: a double-blind placebo-controlled trial. Clin Otolaryngol Allied Sci  1981; 6: 5– 13. Google Scholar CrossRef Search ADS PubMed  6. Burke P Bain J Robinson D Dunleavey J. Acute red ear in children: controlled trial of non-antibiotic treatment in general practice. BMJ  1991; 303: 558– 62. Google Scholar CrossRef Search ADS PubMed  7. Claessen JQ Appelman CL Touw-Otten FWet al.  . Persistence of middle ear dysfunction after recurrent acute otitis media. Clin Otolaryngol Allied Sci  1994; 19: 35– 40. Google Scholar CrossRef Search ADS PubMed  8. Damoiseaux RA van Balen FA Hoes AWet al.  . Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children aged under 2 years. BMJ  2000; 320: 350– 4. Google Scholar CrossRef Search ADS PubMed  9. Le Saux N Gaboury I Baird Met al.  . A randomized, double-blind, placebo-controlled noninferiority trial of amoxicillin for clinically diagnosed acute otitis media in children 6 months to 5 years of age. CMAJ  2005; 172: 335– 41. Google Scholar CrossRef Search ADS PubMed  10. Mandel EM Doyle WJ Winther B Alper CM. The incidence, prevalence and burden of OM in unselected children aged 1–8 years followed by weekly otoscopy through the “common cold” season. Int J Pediatr Otorhinolaryngol  2008; 72: 491– 9. Google Scholar CrossRef Search ADS PubMed  11. Tapiainen T Kujala T Renko Met al.  . Effect of antimicrobial treatment of acute otitis media on the daily disappearance of middle ear effusion: a placebo-controlled trial. JAMA Pediatr  2014; 168: 635– 41. Google Scholar CrossRef Search ADS PubMed  12. Tähtinen PA Laine MK Huovinen Pet al.  . A placebo-controlled trial of antimicrobial treatment for acute otitis media. N Engl J Med  2011; 364: 116– 26. Google Scholar CrossRef Search ADS PubMed  13. Hoberman A Paradise JL Rockette HEet al.  . Treatment of acute otitis media in children under 2 years of age. N Engl J Med  2011; 364: 105– 15. Google Scholar CrossRef Search ADS PubMed  14. Kaleida PH Casselbrant ML Rockette HEet al.  . Amoxicillin or myringotomy or both for acute otitis media: results of a randomized clinical trial. Pediatrics  1991; 87: 466– 74. Google Scholar PubMed  15. Bakaletz LO. Immunopathogenesis of polymicrobial otitis media. J Leukoc Biol  2010; 87: 213– 22. Google Scholar CrossRef Search ADS PubMed  16. Dagan R Pelton S Bakaletz L Cohen R. Prevention of early episodes of otitis media by pneumococcal vaccines might reduce progression to complex disease. Lancet Infect Dis  2016; 16: 480– 92. Google Scholar CrossRef Search ADS PubMed  17. Chonmaitree T Revai K Grady JJet al.  . Viral upper respiratory tract infection and otitis media complication in young children. Clin Infect Dis  2008; 46: 815– 23. Google Scholar CrossRef Search ADS PubMed  18. Bogaert D De Groot R Hermans PW. Streptococcus pneumoniae colonisation: the key to pneumococcal disease. Lancet Infect Dis  2004; 4: 144– 54. Google Scholar CrossRef Search ADS PubMed  19. Uitti JM Tähtinen PA Laine MKet al.  . Role of nasopharyngeal bacteria and respiratory viruses in acute symptoms of young children. Pediatr Infect Dis J  2015; 34: 1056– 62. Google Scholar CrossRef Search ADS PubMed  20. Renko M Kontiokari T Jounio-Ervasti Ket al.  . Disappearance of middle ear effusion in acute otitis media monitored daily with tympanometry. Acta Paediatr  2006; 95: 359– 63. Google Scholar CrossRef Search ADS PubMed  21. Koivunen P Alho OP Uhari Met al.  . Minitympanometry in detecting middle ear fluid. J Pediatr  1997; 131: 419– 22. Google Scholar CrossRef Search ADS PubMed  22. Helenius KK Laine MK Tähtinen PAet al.  . Tympanometry in discrimination of otoscopic diagnoses in young ambulatory children. Pediatr Infect Dis J  2012; 31: 1003– 6. Google Scholar PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of The Journal of the Pediatric Infectious Diseases Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of the Pediatric Infectious Diseases Society Oxford University Press

Effect of Antimicrobial Treatment on the Resolution of Middle-Ear Effusion After Acute Otitis Media

Loading next page...
 
/lp/ou_press/effect-of-antimicrobial-treatment-on-the-resolution-of-middle-ear-fl6eCv6GIJ
Publisher
Pediatric Infectious Diseases Society
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Journal of the Pediatric Infectious Diseases Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
2048-7193
eISSN
2048-7207
D.O.I.
10.1093/jpids/pix008
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Acute otitis media (AOM) induces middle-ear effusion (MEE), which affects hearing. The effect of antimicrobial treatment on the resolution of MEE is controversial, and the factors that affect resolution are unknown. Methods We studied the effect of antimicrobial treatment on the time to the resolution of MEE as a secondary objective in our randomized double-blind placebo-controlled trial. Children aged 6 to 35 months with stringently diagnosed AOM were allocated to receive amoxicillin-clavulanate (161 patients) or placebo (158 patients) for 7 days and closely followed for 3 months. This study was registered at ClinicalTrials.gov (identifier NCT00299455). Results The median times to resolution of MEE were 20 days (95% confidence interval [CI], 16–24 days) and 29 days (95% CI, 26–32 days) in the amoxicillin-clavulanate and placebo groups, respectively (P = .10). The resolution of MEE was confirmed in 138 (86%) of 161 and 132 (84%) of 158 patients in the amoxicillin-clavulanate and placebo groups, respectively (P = .59). In multivariable analysis, the resolution of MEE was prolonged most significantly by at least 1 recurrence of AOM during follow-up. MEE resolved in 65 (65%) of 100 patients with a recurrence of AOM during follow-up and in 205 (94%) of 219 of those without a recurrence (P < .001) (median times to resolution, 67 vs 15 days, respectively; P < .001). Conclusions Immediate antimicrobial treatment of AOM does not significantly affect the resolution of MEE in young children. Subsequent recurrences of AOM are a major reason for the persistence of MEE. Acute otitis media (AOM) is among the most common reasons for outpatient sick visits and is the most common reason for antimicrobial treatment in young children [1]. The diagnosis is based on acute symptoms, acute inflammatory signs of the tympanic membrane, and middle-ear effusion (MEE). MEE decreases the flexibility of the tympanic membrane and the vibration of ossicles and thus impairs the conductive component of hearing [2]. Therefore, persistent MEE is the major indication for tympanostomy tube placements, which cost nearly $1.5 billion annually in the United States [3, 4]. Thus, it is important to know whether antimicrobial treatment of AOM affects the resolution of MEE. Most previous randomized double-blind placebo-controlled studies on the treatment of AOM have found that antimicrobial treatment does not affect the resolution of MEE [5–9]. Those studies, however, had relatively long follow-up intervals, which might have hindered the observance of a disappearance and reappearance of MEE [10]. Therefore, they did not report the time to resolution of MEE but, instead, its prevalence. Recently, Tapiainen et al [11] found that antimicrobial treatment significantly accelerates the resolution of MEE. They performed a close 2-month follow-up that included daily parental tympanometry; however, a majority of their study patients were older than 24 months and, thus, beyond the age to be at high risk for the persistence of MEE. We previously reported the results of the primary outcome of our randomized double-blind placebo-controlled study on the treatment of AOM in young children. Amoxicillin-clavulanate accelerated improvement of the children’s overall condition and the resolution of individual symptoms such as fever [12]. We closely followed the resolution of MEE for 3 months. Here, we report the results of our prespecified secondary objective, which was to determine the effect of antimicrobial treatment on the time to resolution of MEE and which factors affect the resolution of MEE, including factors during follow-up as a novel aspect. METHODS Patients and Diagnostic Criteria Children aged 6 to 35 months were eligible for our diagnostic screening when their parents suspected AOM on the basis of the child’s acute symptoms. A list of the exclusion criteria, along with their reasons, is provided in Supplementary Methods. The inclusion criterion was AOM as diagnosed by a study physician. The diagnosis of AOM required the following 3 criteria. First, the child’s MEE had to be detected by means of pneumatic otoscopic examination that revealed at least 2 of the following tympanic-membrane findings: bulging position, decreased or absent mobility, abnormal color or opacity not caused by scarring, and/or air–fluid interfaces. Second, at least 1 of the following acute inflammatory signs in the tympanic membrane had to be present: distinct erythematous patches or streaks and/or increased vascularity over a full, bulging, or yellow tympanic membrane. Third, the child had to have acute symptoms of AOM. A parent of each child provided written informed consent. The protocol was approved by the ethics committee of the Hospital District of Southwest Finland. Study Design and Follow-Up This randomized double-blind placebo-controlled study was initiated by the investigators and was conducted independently of any commercial entities. Our primary objective was to study the efficacy of antimicrobial therapy on treatment failure, and the secondary objective, reported here, was to study the efficacy of antimicrobial treatment for the resolution of MEE. Our hypothesis was that amoxicillin-clavulanate would accelerate the resolution of MEE. Another objective was to study which factors characterized at study entry and which factors documented during follow-up affected the resolution of MEE. Eligible patients were assigned randomly to receive amoxicillin-clavulanate (40/5.7 mg/kg, divided into 2 daily doses) or placebo for 7 days. The placebo was similar to the active treatment in appearance and taste. The description of the study drugs, the randomization procedure, the procedure for concealment of study assignments, and details of the rescue treatment and other open antimicrobial treatments are provided in Supplementary Methods. After the enrollment visit (study day 1), we followed the patients to determine when/if their MEE resolved and the development of recurrences of AOM by closely scheduled visits on study days 3, 8, 15, 30 (45, if MEE had not resolved earlier), and 60. Thereafter, the patients who had persistent MEE were followed up every other week until study day 90 (±10 days) or until referral to tympanostomy tube placement. Apart from the scheduled visits, we arranged additional sick visits for any patient on any day of the week if the parents were concerned about their child’s condition or if they suspected AOM. At each visit, the study physician interviewed the parents about the symptoms of their child, performed a clinical examination (including thorough pneumatic otoscopic and tympanometry examinations), and acquired a nasopharyngeal sample for the detection of respiratory viruses and bacteria (see details in Supplementary Methods). If AOM was diagnosed, the first episode was treated primarily with amoxicillin-clavulanate, and subsequent episodes were treated with amoxicillin. This study was registered at ClinicalTrials.gov (identifier NCT00299455). Outcome The outcome was the time to resolution of MEE. The time was from the day of random assignment to the day on which a study physician confirmed the resolution of MEE, which was based on a pneumatic otoscopic assessment; both tympanic membranes had to have good mobility, complete or partial transparency, and no visibly remnant effusion. Statistics The sample size had been calculated for the primary outcome (see Supplementary Methods), the results of which were published previously [12]. No separate sample-size calculations had been made for the secondary outcome reported here. All analyses were made in the intention-to-treat population. The medians and means of the time-to-outcome data were analyzed with the Kaplan-Meier method and compared with results of the log-rank test. Cox regression was used to study the effect of treatment allocation and of predefined factors on the resolution of MEE. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution . Proportions of patients were compared with the χ2 or Fisher exact test, as appropriate. All P values are 2-sided. The analyses were conducted with SPSS 22 (IBM SPSS Statistics, IBM Corporation, Armonk, NY) by Dr Ruohola. Details of the statistics and definitions for the factors included in the analyses are provided in Supplementary Methods. RESULTS Study Population The intention-to-treat population included 319 patients, 161 in the amoxicillin-clavulanate group and 158 in the placebo group, 53 (34%) of whom received delayed amoxicillin-clavulanate treatment as rescue therapy (see Supplementary Figure 1). The characteristics of the treatment groups were balanced (Table 1). Table 1. Baseline Characteristics of the Intention-to-Treat Population Characteristic  Amoxicillin-Clavulanate Group (N = 161)  Placebo Group (N = 158)  Demographics   Age (mean [range]) (mo)  16 (6–35)  16 (6–35)   Age (n [%])    6–11 mo  57 (35)  59 (37)    12–23 mo  78 (48)  65 (41)    24–35 mo  26 (16)  34 (22)   Male sex (n [%])  92 (57)  90 (57)  Otitis media risk factors (n [%])   Family history of otitis media  90 (56)  98 (62)   Sibling(s) in the household  89 (55)  93 (59)   Day care attendance  87 (54)  86 (54)   Parental smoking  57 (36)  48 (30)   Current use of pacifier  80 (50)  85 (54)  Otitis media history   Previous episodes of AOM (mean [range]) (n)  2 (0–10)  2 (0–10)   Previous episodes of AOM (n [%])    0  43 (27)  51 (32)    1–3  88 (55)  79 (50)    ≥4  30 (19)  28 (18)   Time from previous episode of AOM (mean [range]) (mo)a  3 (0–22)  3 (0–15)   Time from previous episode of AOM (n [%])a    ≤1 mo  36 (22)  31 (20)    >1–3 mo  42 (26)  45 (29)    >3 mo  40 (25)  31 (20)  Ear examination findings at entry (n [%])   Unilateral AOMb  99 (62)  89 (57)   Severe bulging of tympanic membrane  45 (28)  39 (25)   Peaked tympanogramc  32 (20)  25 (16)  Preceding symptoms (n [%])   Fever ≥38°C  64 (40)  46 (29)   Parentally reported ear pain  123 (76)  126 (80)   Decreased activity  76 (47)  59 (37)   Respiratory symptoms  156 (97)  156 (99)  Characteristic  Amoxicillin-Clavulanate Group (N = 161)  Placebo Group (N = 158)  Demographics   Age (mean [range]) (mo)  16 (6–35)  16 (6–35)   Age (n [%])    6–11 mo  57 (35)  59 (37)    12–23 mo  78 (48)  65 (41)    24–35 mo  26 (16)  34 (22)   Male sex (n [%])  92 (57)  90 (57)  Otitis media risk factors (n [%])   Family history of otitis media  90 (56)  98 (62)   Sibling(s) in the household  89 (55)  93 (59)   Day care attendance  87 (54)  86 (54)   Parental smoking  57 (36)  48 (30)   Current use of pacifier  80 (50)  85 (54)  Otitis media history   Previous episodes of AOM (mean [range]) (n)  2 (0–10)  2 (0–10)   Previous episodes of AOM (n [%])    0  43 (27)  51 (32)    1–3  88 (55)  79 (50)    ≥4  30 (19)  28 (18)   Time from previous episode of AOM (mean [range]) (mo)a  3 (0–22)  3 (0–15)   Time from previous episode of AOM (n [%])a    ≤1 mo  36 (22)  31 (20)    >1–3 mo  42 (26)  45 (29)    >3 mo  40 (25)  31 (20)  Ear examination findings at entry (n [%])   Unilateral AOMb  99 (62)  89 (57)   Severe bulging of tympanic membrane  45 (28)  39 (25)   Peaked tympanogramc  32 (20)  25 (16)  Preceding symptoms (n [%])   Fever ≥38°C  64 (40)  46 (29)   Parentally reported ear pain  123 (76)  126 (80)   Decreased activity  76 (47)  59 (37)   Respiratory symptoms  156 (97)  156 (99)  Abbreviation: AOM, acute otitis media. aAmong those patients who had had a previous episode of AOM. bUnilateral AOM was defined as the diagnosis of AOM in only 1 ear, with the other ear showing no otoscopic signs of pathology or only middle-ear effusion without acute inflammatory signs. Data are missing for 2 patients in the amoxicillin-clavulanate group and 2 patients in the placebo group, because an adequate view of the contralateral tympanic membrane was not reached because of thick cerumen. cA peaked tympanogram was defined as a tympanogram with a height (static acoustic admittance) of ≥0.2 millimhos (mmho) and width ≤300 dekapascals (daPa). The selected tympanogram was obtained from the ear with the most severe otoscopic findings. View Large Effects of Amoxicillin-Clavulanate on MEE Resolution The median time to resolution of MEE was 20 days (95% confidence interval [CI], 16–24 days) in patients who were allocated to the amoxicillin-clavulanate group and 29 days (95% CI, 26–32 days) in patients who were allocated to the placebo group (P = .10) (Figure 1A). The complete resolution of MEE was confirmed during the 3-month follow-up in 138 (86%) of the 161 patients in the amoxicillin-clavulanate group and in 132 (84%) of the 158 patients in the placebo group (P = .59). Additional details of the effects of amoxicillin-clavulanate are provided in Table 2. Table 2. Effect of Amoxicillin-Clavulanate Treatment on the Resolution, Persistence, and Prevalence of MEE in Young Children With Stringently Diagnosed AOM Effect  Amoxicillin-Clavulanate Group  Placebo Group  P  Time to resolution of MEE (median [95% CI]) (days)a   All patients  20 (16–24)  29 (26–32)  .10   Age < 24 mo  22 (13–31)  29 (26–32)  .20   Age ≥ 24 mo  16 (15–17)  30 (17–43)  .12  Time to resolution of MEE (mean [95% CI]) (days)a   All patients  35 (30–40)  41 (36–46)  .10   Age < 24 mo  36 (31–42)  42 (37–48)  .20   Age ≥ 24 mo  26 (17–35)  37 (27–47)  .12  Persistence of MEE (n/N [%])a   By 2 wk  80/159 (50)  100/156 (64)  .013   By 3 wk  68/158 (43)  90/155 (58)  .008   By 4 wk  51/157 (33)  60/155 (39)  .25   By 6 wk  39/157 (25)  49/154 (32)  .17   By 8 wk  25/155 (16)  35/150 (23)  .11   By 12 wk  17/155 (11)  18/150 (12)  .77  Prevalence of MEE (n/N [%])b   At 2 wkc  85/157 (54)  107/151 (71)  .002   At 4 wkd  80/146 (55)  79/143 (55)  .94   At 8 wke  58/134 (43)  68/131 (52)  .16  Effect  Amoxicillin-Clavulanate Group  Placebo Group  P  Time to resolution of MEE (median [95% CI]) (days)a   All patients  20 (16–24)  29 (26–32)  .10   Age < 24 mo  22 (13–31)  29 (26–32)  .20   Age ≥ 24 mo  16 (15–17)  30 (17–43)  .12  Time to resolution of MEE (mean [95% CI]) (days)a   All patients  35 (30–40)  41 (36–46)  .10   Age < 24 mo  36 (31–42)  42 (37–48)  .20   Age ≥ 24 mo  26 (17–35)  37 (27–47)  .12  Persistence of MEE (n/N [%])a   By 2 wk  80/159 (50)  100/156 (64)  .013   By 3 wk  68/158 (43)  90/155 (58)  .008   By 4 wk  51/157 (33)  60/155 (39)  .25   By 6 wk  39/157 (25)  49/154 (32)  .17   By 8 wk  25/155 (16)  35/150 (23)  .11   By 12 wk  17/155 (11)  18/150 (12)  .77  Prevalence of MEE (n/N [%])b   At 2 wkc  85/157 (54)  107/151 (71)  .002   At 4 wkd  80/146 (55)  79/143 (55)  .94   At 8 wke  58/134 (43)  68/131 (52)  .16  Abbreviations: AOM, acute otitis media; CI, confidence interval; MEE, middle-ear effusion. aFigures are based on Kaplan-Meier analyses, P values for the medians and means are based on log-rank tests, and P values for the proportions are based on the χ2 or Fisher exact test, as appropriate. bIncludes those patients who attended study clinic visit in the time window. Figures are based on cross-tabulations, and P values are based on the χ2 or Fisher exact test, as appropriate. cAt study days 11–19. dAt study days 24–36. eAt study days 50–70. View Large Figure 1. View largeDownload slide Kaplan-Meier curves for the time to the resolution of middle-ear effusion (MEE) after acute otitis media (AOM). (A) Comparison between treatment allocation groups. (B) Comparison between patients with and those without a recurrence of AOM during follow-up before the resolution of MEE. P values were determined by the log-rank test. Figure 1. View largeDownload slide Kaplan-Meier curves for the time to the resolution of middle-ear effusion (MEE) after acute otitis media (AOM). (A) Comparison between treatment allocation groups. (B) Comparison between patients with and those without a recurrence of AOM during follow-up before the resolution of MEE. P values were determined by the log-rank test. Factors That Affect MEE Resolution Details of the univariable analyses are presented in Supplementary Table 1. The factors that tended (P < .15) to shorten the time to resolution of MEE during the entire 3-month follow-up period were amoxicillin-clavulanate, unilaterality of AOM at entry, and a peaked tympanogram at entry. In contrast, age less than 24 months tended (P < .15) to prolong the time to resolution of MEE, as did the following factors, characterized during follow-up: a recurrence of AOM any time after the 7-day study-drug period and before the resolution of MEE, open antimicrobial treatment, laboratory-confirmed respiratory viral infection, and nasopharyngeal bacterial colonization. The factors that remained significant in the multivariable analyses are shown in Figure 2. During the entire 3-month follow-up period, a recurrence of AOM, respiratory viral infection, and nasopharyngeal bacterial colonization prolonged the resolution of MEE. Of the factors characterized at entry, a peaked tympanogram indicated a favorable prognosis. Age less than 24 months indicated an unfavorable prognosis; the hazard ratio for the time to MEE resolution was 0.74 (95% CI, 0.55–1.01; P = .06). However, age was excluded from the final step of the model, because its P value exceeded the predefined level of significance. Figure 2. View largeDownload slide Forest plot of the factors that had an independent effect on the resolution of middle-ear effusion (MEE). Hazard ratios and 95% confidence intervals (CIs) for the resolution of MEE are based on a multivariable Cox regression model for the entire 3-month follow-up period and on separate models for each of the 3 follow-up months. Each of the 4 models included those patients in whom MEE had not yet resolved. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution. aTotal number was 319, but the model excluded 6 patients because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. bTotal number was 319, but the model excluded 10 patients, 4 because of missing data on unilateral acute otitis media and 6 because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. cThe model did not include 2 of the factors (ie, respiratory viral infection and nasopharyngeal bacterial colonization) because sampling was infrequent during the third month of follow-up. Figure 2. View largeDownload slide Forest plot of the factors that had an independent effect on the resolution of middle-ear effusion (MEE). Hazard ratios and 95% confidence intervals (CIs) for the resolution of MEE are based on a multivariable Cox regression model for the entire 3-month follow-up period and on separate models for each of the 3 follow-up months. Each of the 4 models included those patients in whom MEE had not yet resolved. In this work, a hazard ratio of <1.0 indicates that the factor prolonged the time to MEE resolution, and a hazard ratio of >1.0 indicates that the factor accelerated the time to MEE resolution. aTotal number was 319, but the model excluded 6 patients because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. bTotal number was 319, but the model excluded 10 patients, 4 because of missing data on unilateral acute otitis media and 6 because of missing data on respiratory viral infection and nasopharyngeal bacterial colonization. cThe model did not include 2 of the factors (ie, respiratory viral infection and nasopharyngeal bacterial colonization) because sampling was infrequent during the third month of follow-up. During the first month of the 3-month follow-up period, the independent factors that prolonged the resolution of MEE were a recurrence of AOM, respiratory viral infection, and nasopharyngeal bacterial colonization. Favorable prognostic factors were unilaterality of AOM and a peaked tympanogram at entry. During the second and third months of the 3-month follow-up period, a recurrence of AOM independently prolonged the resolution of MEE, whereas a peaked tympanogram was a favorable prognostic factor. Unfavorable prognostic factors were age less than 24 months and a sibling(s) in the household, both only during the third month of the 3-month follow-up period. Effect of a Recurrence of AOM During Follow-Up A recurrence of AOM at any time after the 7-day study-drug period significantly prolonged the resolution of MEE (Figure 1B). As Figure 3 shows, the median time to MEE resolution was longest in 100 patients who had at least 1 recurrence (67 days [95% CI, 49–85 days]) and, in contrast, shortest in 219 patients who did not have a recurrence of AOM (15 days [95% CI, 14–16]) (P < .001). The resolution of MEE was documented in 65 (65%) of 100 patients with a recurrence and 205 (94%) of 219 patients without a recurrence (P < .001). According to posthoc subgroup analyses, the disadvantageous effect of a recurrence of AOM on MEE resolution was not modified by amoxicillin-clavulanate treatment or any of the factors characterized at entry or documented during follow-up (see Supplementary Figure 2). Figure 3. View largeDownload slide Median times to resolution of middle-ear effusion. Medians (95% confidence intervals [CIs]) are based on the Kaplan-Meier method, and P values are based on the log-rank test. aData for 4 patients were missing. bData for 6 patients were missing. Figure 3. View largeDownload slide Median times to resolution of middle-ear effusion. Medians (95% confidence intervals [CIs]) are based on the Kaplan-Meier method, and P values are based on the log-rank test. aData for 4 patients were missing. bData for 6 patients were missing. Association of the Symptomatic Relief With the Resolution of MEE The hazard ratios for being asymptomatic by day 8 and day 16 for the time to the MEE resolution were 0.95 (95% CI, 0.74–1.23; P = .70) and 1.23 (95% CI, 0.96–1.61; P = .09), respectively. DISCUSSION In this study, treatment of AOM with amoxicillin-clavulanate in young children did not shorten the time to MEE resolution significantly; instead, new episodes of AOM during follow-up remarkably prolonged MEE. Our finding that treatment with amoxicillin-clavulanate shortly accelerated the resolution of MEE but did not significantly shorten its overall duration is, at first sight, inconsistent with results from a recent study by Tapiainen et al [11], because they concluded that amoxicillin-clavulanate effectively reduces the duration of MEE. However, closer observation revealed that this inconsistency can be explained by the age differences in the study populations. Our study population was at high risk of MEE persistence and AOM recurrence because the mean age was 1.3 years, whereas the population of the Tapiainen et al [11] study was at low risk because the mean age was 4.4 years. In these 2 studies among children younger than 24 months, the mean MEE durations in the amoxicillin-clavulanate and placebo groups were very similar; both studies had an approximately 1-week difference in MEE durations between the treatment groups. Hoberman et al [13] had short follow-up intervals and a study design and population also otherwise comparable to ours. They observed a significant difference between treatment groups in the persistence of MEE at their final follow-up visit 3 weeks after entry. Similarly, Kaleida et al [14] detected a significant difference 2 weeks after study entry. We also observed a statistically significant treatment effect at 2 and 3 weeks, which indicates that antimicrobial treatment accelerated the resolution of MEE. However, our results suggest that this accelerating effect was short lasting, because at 4 weeks and thereafter, no significant treatment effect was observed. The overall difference in the median duration of MEE between our study groups was 9 days, which was statistically insignificant. Other previous placebo-controlled trials did not find any differences between their antimicrobial-treatment and placebo groups [5–9]. However, these studies reported the prevalences of MEE instead of the persistence of MEE. Prevalence figures incorporate persistent MEE and newly developed episodes, and, as results of our analyses show, prevalence figures poorly reflect the true persistence of MEE [10]. Altogether, it seems that in older children, the favorable accelerating effect of antimicrobial treatment on MEE resolution might be sustained. In contrast, at an otitis-prone age, the effect of antimicrobial treatment on the resolution of MEE seems to be modest and short lasting and, thus, clinically insignificant. A recurrence of AOM during follow-up significantly altered the course of MEE resolution. When a recurrence of AOM was diagnosed before the resolution of effusion, only 65% of the children were free of effusion during the 3-month follow-up period, and the resolution time was almost 2 months longer than that in children without a recurrence before MEE was resolved. This finding is novel but not unexpected, because explanatory mechanisms are available. A recurrence of AOM is accompanied by respiratory viral infection, which deteriorates the function of the Eustachian tube [15]. A recurrence also reinduces inflammation and, thus, effusion in the middle ear. The formation of bacterial biofilms, common in recurrent AOM [16], might be an explanation, although recurrences or a recent episode of AOM before study entry did not affect the resolution of MEE. Current guidelines advise differentiating between recurrent episodes of AOM and persistence of MEE (or, in other words, chronic otitis media with effusion) because the persistence of MEE can cause hearing impairment and, thus, is the primary indication for tympanostomy tube placement [3]. Our results indicate that in young children who do not have craniofacial anomalies or known immunodeficiencies, recurrences of AOM and the persistence of MEE cannot be differentiated from each other because recurrences of AOM seem to be a major reason for the persistence of MEE after the initial episode of AOM. The prerequisites for AOM development are respiratory viral infection and nasopharyngeal bacterial colonization, most commonly by Streptococcus pneumoniae, Haemophilus influenzae, and/or Moraxella catarrhalis [17, 18]. It was an unexpected finding that these microbes independently prolonged the resolution of MEE. The effect of respiratory viral infection is probably mediated by inflammation of the respiratory epithelium, which deteriorates the function of the Eustachian tubes [15]. The effect of bacterial nasopharyngeal colonization might be caused by interaction between bacteria and viruses. We recently found that M catarrhalis is independently associated with respiratory symptoms, especially rhinitis, in young children with respiratory viral infection [19]. In our current study, the risk factors for otitis media, such as day care attendance and parental smoking, did not have remarkable effects on MEE resolution. However, in the prevention of MEE persistence, the risk factors should also be recognized. If the risk factors are reduced, the acquisition of respiratory viruses and colonizing bacteria might be reduced, and therefore, the risk of AOM recurrences might be reduced also, which all might prevent the persistence of MEE. According to our results and those of Renko et al [20], a favorable prognostic factor for the resolution of MEE seems to be unilaterality of AOM. However, in our study, a peaked tympanogram was the most favorable prognostic factor. This finding is logical, because a peaked tympanogram is an indicator of small volume of MEE [21, 22]; in these instances, the bulging position of the tympanic membrane is probably a result of positive pressure in the middle ear. The drawback to using a peaked tympanogram as a prognostic factor is that it is a rare finding when stringent criteria are used to diagnose AOM. The major strengths of our study are the nonselective study population at an otitis-prone age, the stringent diagnostic criteria, and the tight follow-up schedule. To our knowledge, our study is the first randomized double-blind placebo-controlled trial in which all patients were closely followed until MEE resolution or referral for tympanostomy tube placement; only 5% of the patients were lost to follow-up. The major methodologic limitation was the subjectivity of pneumatic otoscopy even when added with video recording. However, visibly low levels of MEE are not necessarily obtained by tympanocentesis, and tympanometry often produces a peaked curve that suggests the resolution of MEE. It is impossible currently to determine which volume of MEE is clinically relevant; therefore, we aimed to observe even small volumes of MEE and, thus, to be assured of total MEE resolution, which can be criticized. It could be questioned if a higher dose of amoxicillin-clavulanate might have been effective, although the dose (40 mg/kg per day) we chose had excellent antimicrobial coverage in our study population. A study by Hoberman et al [13] was much like ours, but they chose a dose of 90 mg/kg per day. At their 3-week final control visit, the proportions of patients who had persistent MEE were actually somewhat higher, and the difference between the study groups was smaller than that in our study. Thus, it does not seem that the dose is a major reason why amoxicillin-clavulanate had no overall effect. Our study can also be criticized for the lack of sample-size calculation for the reported secondary outcome. However, although with a larger sample size the difference between treatment-allocation groups would have been statistically significant, we would regard the observed difference to be clinically insignificant. In conclusion, our study reveals that in young children with AOM, immediate antimicrobial treatment (such as with amoxicillin-clavulanate) that has appropriate coverage does not significantly shorten the time to MEE resolution. The major reason for inadequate resolution and, thus, for the persistence of MEE is recurrence of AOM before MEE resolution. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online. Notes Acknowledgments. We thank all the families who participated in this study and the staff at the health center of Turku city primary care for their commitment to the study. We acknowledge Raakel Luoto, MD, PhD, and Elina Lahti, MD, PhD, for their contribution to data collection; Sari Rajamäki, RN, Maarit Rosenblad, RN, and Kaisa Erkkilä, RN, for their assistance at the study clinic; Pentti Huovinen, MD, PhD, Jari Jalava, PhD, and Laura Lindholm, MSc, for bacterial analyses; Matti Waris, PhD, and Raija Vainionpää, PhD, for viral analyses; Tero Vahlberg, MSc, for the allocation procedure; and Eliisa Löyttyniemi, MSc, and Maria Ruohola, BS, for data handling. Financial support. This work was supported by the Fellowship Award of the European Society for Paediatric Infectious Diseases (to A. R.) and by grants from the Foundation for Paediatric Research, the Research Funds from Specified Government Transfers, the Jenny and Antti Wihuri Foundation, the Paulo Foundation, the Maud Kuistila Memorial Foundation, the Emil Aaltonen Foundation, the Finnish Cultural Foundation: Varsinais-Suomi Regional Fund, the Turku University Hospital Research Foundation, the Finnish Medical Foundation, and the Turku University Foundation. Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Schilder AG Chonmaitree T Cripps AWet al.  . Otitis media. Nat Rev Dis Primers  2016; 2: 60– 3. Google Scholar CrossRef Search ADS   2. Marsh RR Baranak CC Potsic WP. Hearing loss and visco-elasticity of middle ear fluid. Int J Pediatr Otorhinolaryngol  1985; 9: 115– 20. Google Scholar CrossRef Search ADS PubMed  3. Rosenfeld RM Schwartz SR Pynnonen MAet al.  . Clinical practice guideline: tympanostomy tubes in children. Otolaryngol Head Neck Surg  2013; 149: S1– 35. Google Scholar CrossRef Search ADS PubMed  4. Mui S Rasgon BM Hilsinger RLet al.  . Tympanostomy tubes for otitis media: quality-of-life improvement for children and parents. Ear Nose Throat J  2005; 84: 418, 20–2, 24. Google Scholar PubMed  5. Mygind N Meistrup-Larsen KI Thomsen Jet al.  . Penicillin in acute otitis media: a double-blind placebo-controlled trial. Clin Otolaryngol Allied Sci  1981; 6: 5– 13. Google Scholar CrossRef Search ADS PubMed  6. Burke P Bain J Robinson D Dunleavey J. Acute red ear in children: controlled trial of non-antibiotic treatment in general practice. BMJ  1991; 303: 558– 62. Google Scholar CrossRef Search ADS PubMed  7. Claessen JQ Appelman CL Touw-Otten FWet al.  . Persistence of middle ear dysfunction after recurrent acute otitis media. Clin Otolaryngol Allied Sci  1994; 19: 35– 40. Google Scholar CrossRef Search ADS PubMed  8. Damoiseaux RA van Balen FA Hoes AWet al.  . Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children aged under 2 years. BMJ  2000; 320: 350– 4. Google Scholar CrossRef Search ADS PubMed  9. Le Saux N Gaboury I Baird Met al.  . A randomized, double-blind, placebo-controlled noninferiority trial of amoxicillin for clinically diagnosed acute otitis media in children 6 months to 5 years of age. CMAJ  2005; 172: 335– 41. Google Scholar CrossRef Search ADS PubMed  10. Mandel EM Doyle WJ Winther B Alper CM. The incidence, prevalence and burden of OM in unselected children aged 1–8 years followed by weekly otoscopy through the “common cold” season. Int J Pediatr Otorhinolaryngol  2008; 72: 491– 9. Google Scholar CrossRef Search ADS PubMed  11. Tapiainen T Kujala T Renko Met al.  . Effect of antimicrobial treatment of acute otitis media on the daily disappearance of middle ear effusion: a placebo-controlled trial. JAMA Pediatr  2014; 168: 635– 41. Google Scholar CrossRef Search ADS PubMed  12. Tähtinen PA Laine MK Huovinen Pet al.  . A placebo-controlled trial of antimicrobial treatment for acute otitis media. N Engl J Med  2011; 364: 116– 26. Google Scholar CrossRef Search ADS PubMed  13. Hoberman A Paradise JL Rockette HEet al.  . Treatment of acute otitis media in children under 2 years of age. N Engl J Med  2011; 364: 105– 15. Google Scholar CrossRef Search ADS PubMed  14. Kaleida PH Casselbrant ML Rockette HEet al.  . Amoxicillin or myringotomy or both for acute otitis media: results of a randomized clinical trial. Pediatrics  1991; 87: 466– 74. Google Scholar PubMed  15. Bakaletz LO. Immunopathogenesis of polymicrobial otitis media. J Leukoc Biol  2010; 87: 213– 22. Google Scholar CrossRef Search ADS PubMed  16. Dagan R Pelton S Bakaletz L Cohen R. Prevention of early episodes of otitis media by pneumococcal vaccines might reduce progression to complex disease. Lancet Infect Dis  2016; 16: 480– 92. Google Scholar CrossRef Search ADS PubMed  17. Chonmaitree T Revai K Grady JJet al.  . Viral upper respiratory tract infection and otitis media complication in young children. Clin Infect Dis  2008; 46: 815– 23. Google Scholar CrossRef Search ADS PubMed  18. Bogaert D De Groot R Hermans PW. Streptococcus pneumoniae colonisation: the key to pneumococcal disease. Lancet Infect Dis  2004; 4: 144– 54. Google Scholar CrossRef Search ADS PubMed  19. Uitti JM Tähtinen PA Laine MKet al.  . Role of nasopharyngeal bacteria and respiratory viruses in acute symptoms of young children. Pediatr Infect Dis J  2015; 34: 1056– 62. Google Scholar CrossRef Search ADS PubMed  20. Renko M Kontiokari T Jounio-Ervasti Ket al.  . Disappearance of middle ear effusion in acute otitis media monitored daily with tympanometry. Acta Paediatr  2006; 95: 359– 63. Google Scholar CrossRef Search ADS PubMed  21. Koivunen P Alho OP Uhari Met al.  . Minitympanometry in detecting middle ear fluid. J Pediatr  1997; 131: 419– 22. Google Scholar CrossRef Search ADS PubMed  22. Helenius KK Laine MK Tähtinen PAet al.  . Tympanometry in discrimination of otoscopic diagnoses in young ambulatory children. Pediatr Infect Dis J  2012; 31: 1003– 6. Google Scholar PubMed  © The Author(s) 2017. Published by Oxford University Press on behalf of The Journal of the Pediatric Infectious Diseases Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

Journal of the Pediatric Infectious Diseases SocietyOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve Freelancer

DeepDyve Pro

Price
FREE
$49/month

$360/year
Save searches from
Google Scholar,
PubMed
Create lists to
organize your research
Export lists, citations
Read DeepDyve articles
Abstract access only
Unlimited access to over
18 million full-text articles
Print
20 pages/month
PDF Discount
20% off