Saskia Hullegie

Topical or oral antibiotics in childhood AOMd: a randomised controlled non-inferiority trial

Discussion

Due to early termination of the trial non-inferiority of antibiotic-corticosteroid eardrops to oral antibiotics could not be determined in children with AOMd. In our small group of 58 children, we found that those assigned to eardrops had lower resolution rates of ear pain and fever at 3 days, longer parent-reported ear discharge, and slightly higher mean ear pain scores over days 1-3 compared to those receiving oral antibiotics, but they received fewer oral antibiotic courses in three months and had less GI upset and rash.

Strengths and limitations
This is the first report on the comparative effectiveness evidence of antibiotic-corticosteroid eardrops versus oral antibiotic treatment in children with AOM presenting with ear discharge. The pragmatic design of the trial and high rate of data-completeness support applicability of its findings to routine daily practice.

Some limitations deserve further attention. Accrual to our trial was affected by a temporary closure due to study medication supply issues. When this was resolved the COVID-19 pandemic complicated trial recruitment and accrual did not recover after the pandemic restrictions were lifted. This phenomenon has affected many trials worldwide. Further, our non-blinded design could potentially have introduced detection bias. However, detection bias is unlikely to have significantly impacted our findings since we compared two active treatments and – based on our parent panel input – parents do not have strong preferences for one over the other treatment. Also, a double dummy design would have hampered the applicability of trial results to everyday practice.

We chose hydrocortisone-bacitracin-colistin eardrops because they are widely used in the Netherlands and France, do not contain a potentially ototoxic aminoglycoside, cover the most important pathogens involved in AOM and have been proven effective in children with ventilation tubes who present with acute ear discharge. They are however not available in many countries. Despite absence of evidence, we believe that any combination of antibiotic-corticosteroid eardrops with a similar antimicrobial profile, like a quinolone-containing eardrops plus dexamethasone, would have yielded comparable results.

Comparison with existing literature
We initiated this trial after establishing superiority of antibiotic-corticosteroid eardrops over oral antibiotics in children with ventilation tubes who present with acute ear discharge. Our current findings in a small sample of children without ventilation tubes who present with AOMd indicate that these findings cannot be extrapolated to this patient population. While there is a patent passage between ear canal and the middle ear in children with ventilation tubes, the spontaneous eardrum perforation in children with AOMd may close too early to allow antibiotic-corticosteroid eardrops to completely resolve the middle ear inflammation.

During the preparation of our trial we collaborated with the UK based team developing the REST (Runny Ear Study, trial registry number ISRCTN287PS92) addressing the same topic. We harmonised design and outcomes to enable future meta-analysis. This trial however was also terminated early due to issues with its electronic health record system platform and no formal statistical analysis was performed on its sample of 22 children.

Implications for research and practice
We were unable to determine non-inferiority of antibiotic-corticosteroid eardrops to oral antibiotics, but our findings in a small group of children, requiring confirmation, suggest that oral antibiotics may be more effective in resolving symptoms and shortening the duration of ear discharge than antibiotic-corticosteroid eardrops in children with AOMd. That must be balanced against the findings that eardrops are associated with reasonable symptom control, fewer total oral antibiotic courses and less systemic side effects in case there is non-inferiority. Since we were unable to demonstrate non-inferiority of antibiotic-corticosteroid eardrops to oral antibiotics in children with AOMd, current guidelines’ recommendation that clinicians can consider oral antibiotics in this group of children are not unreasonable, but must be balanced against the major public health threat of antibiotic resistance.

Supplementary Table 1. Disease specific quality of life assessed with the otitis media-6 questionnaire at baseline, at 2 weeks and 3 months follow up

Difference in change score ± Eardrops vs oral Abx Δ T2-T3 (p): 0.76 (p =0.207), -0.57 (p= 0.234), -0.08 (p = 0.793), -0.01 (p = 0.988), 0.15 (p = 0.764), 0.29 (p = 0.581), 0.11 (p= 0.773), 0.42 (p = 0.864), -0.78 (p = 0.283)

Oral Abx Δ T2-T3 (SD): 2.6 (2.39), 1.7 (1.69), 0.6 (0.97), 2.4 (1.72), 2.2 (1.73), 1.8 (1.92), 1.9 (1.47), 11.2 (8.80), 3.8 (2.23)

Change score * Eardrops Δ T2-T3 (SD): 3.4 (1.66), 1.1 (1.52), 0.5 (1.07), 2.4 (1.64), 2.3 (1.69), 2.1 (1.59), 2.0 (1.22), 11.6 (7.63), 3.0 (2.56)

Difference in change score ± Eardrops vs oral Abx Δ T0-T2 (p): -1.19 (p=0.006), 0.71 (p = 0.127), 0.07 (p = 0.831), -0.63 (p= 0.097), -0.33 (p= 0.385), -0.55 (p = 0.149), -0.36 (p = 0.164), -1.73 (p = 0.246), -0.32 (p = 0.585)

Oral Abx Δ T0-T2 (SD): 1.3 (1.39), -0.7 (1.69), -0.2 (1.43), 1.1 (1.14), 0.8 (1.42), 0.9 (1.54), 0.54 (0.91), 3.3 (5.49), 0.7 (1.92)

Change score * Eardrops Δ T0-T2 (SD): 0.1 (1.68), 0.0 (1.71), -0.2 (0.82), 0.5 (1.6), 0.5 (1.42), 0.4 (1.29), 0.19 (0.98), 1.5 (5.50), 0.4 (2.32)

Month 3 (n=48)
Oral Abx mean (SD): 2.2 (1.44), 2.1 (1.37), 1.6 (0.69), 1.9 (1.03), 1.9 (0.93), 2.1 (1.20), 2.0 (0.99), 11.7 (5.96), 7.8 (2.09)
Eardrops mean (SD): 1.9 (1.29), 2.0 (1.40), 1.7 (1.34), 1.9 (0.97), 1.9 (1.09), 2.1 (1.45), 1.9 (1.11), 11.4 (6.68), 7.8 (2.09)

Oral Abx; oral antibiotics; °1-7: higher scores indicating more of a problem. *Change score between time points, a positive score indicates deterioration. A score < 0.5 indicates trivial change; 0.5 - 0.9 small change, 1.0 - 1.4 moderate change; and > 1.5 large change; ± using independent t-test.

Week 2 (n=56)
Oral Abx mean (SD): 4.6 (1.63), 3.7 (1.8), 2.0 (1.13), 4.3 (1.3), 3.9 (1.3), 3.8 (1.4), 3.7 (1.11), 22.2 (6.67), 5.1 (1.92)
Eardrops mean (SD): 5.1 (1.68), 3.1 (1.90), 2.2 (1.41), 4.3 (1.67), 4.1 (1.62), 4.1 (1.3), 3.9 (1.33), 22.8 (7.54), 4.8 (1.55)

Baseline (n=57)
Oral Abx mean (SD): 5.9 (1.26), 3.0 (2.05), 1.8 (1.20), 5.4 (1.23), 4.7 (1.42), 4.7 (1.37), 4.3 (0.95), 25.5 (5.68), 4.4 (1.89)
Eardrops mean (SD): 5.2 (1.68), 3.2 (1.83), 2.1 (1.23), 4.8 (1.77), 4.6 (1.79), 4.7 (1.42), 4.1 (1.24), 24.3 (7.43), 4.4 (2.21)

Range of scores: 1 to 7, 1 to 7, 1 to 7, 1 to 7, 1 to 7, 1 to 7, 1 to 7, 6 to 42, 0 to 10
Questionnaire items: Physical suffering, Hearing loss, Speech impairment, Emotional distress, Activity limitations, Caregivers concerns, Mean symptom score, Total score, Visual analog score