Adenoidectomy with or without grommets for children with otitis media: an individual patient data meta-analysis

Abstract

Otitis media (OM) is a leading cause of medical consultations, antibiotic prescription and surgery in children. The surgical procedures offered to children with recurrent or persistent OM are insertion of grommets, adenoidectomy or a combination of the two. There is clear National Institute for Health and Care Excellence guidance for the use of grommets in subgroups of children with persistent OM with effusion (OME), but similar guidance is not available for adenoidectomy, either in persistent OME or in recurrent acute OM (AOM). (1) To develop a model to predict the risk of children referred for adenoidectomy having a prolonged duration of their OM. Then, (2a) to evaluate the overall effect of adenoidectomy, with or without grommets, on OM using individual patient data (IPD) and (2b) to identify those subgroups of children who are most likely to benefit from adenoidectomy with or without grommets. A number of electronic databases were searched from their inception including the Cochrane Ear, Nose and Throat Disorders Group Trials Register, the Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, EMBASE, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), metaRegister of Current Controlled Trials (mRCT), ClinicalTrials.gov, International Clinical Trials Registry Platform (ICTRP), ClinicalStudyResults.org and Google. Studies eligible for inclusion in this IPD meta-analysis were randomised controlled trials (RCTs) in children up to 12 years of age diagnosed with recurrent AOM and/or persistent OME in which adenoidectomy (with or without grommets) was compared with non-surgical treatment or grommets alone. The final selection of eligible studies and the quality assessment were carried out according to standard methods and disagreement was resolved by discussion. A total of 503 articles were identified of which 10 trials were included in the meta-analysis; eight of these were at a low risk of bias and two were at moderate risk. The primary outcome was failure at 12 months, defined by a set of persisting symptoms and signs. In the prognostic analysis 56% of those children referred for adenoidectomy (but randomised to the non-surgical group) failed to improve (38% of the children with recurrent AOM and 89% of the children with persistent OME). Children who had adenoidectomy had a greater chance of clinical improvement. The size of that effect is, in general, small but persists for at least 2 years. Two subgroups of children are most likely to benefit from adenoidectomy: first, children aged < 2 years with recurrent AOM - 16% of those who had adenoidectomy failed at 12 months whereas 27% of those who did not have adenoidectomy failed [rate difference (RD) 12%, 95% confidence interval (Cl) 6% to 18%; number needed to treat (NNT) = 9]; second, children aged ≥ 4 years with persistent OME - 51% of those who had adenoidectomy failed at 12 months whereas 70% of those who did not have adenoidectomy failed (RD 19%, 95% Cl 12% to 26%; NNT = 6). No significant benefit of adenoidectomy was found in children aged ≥ 2 years with recurrent AOM and children aged < 4 years with persistent OME. The need to use a composite end point and the limited number of subgroup variables that could be studied are factors that reduce the robustness of these results; however, we do not believe that this reduces the validity of the conclusions. Adenoidectomy is most beneficial in children with persistent OME aged ≥ 4 years. A smaller beneficial effect was found in children with recurrent AOM aged < 2 years. Consideration must be given to the balance between benefits and harms. Future research is required in a number of key areas, including defining the best methods of selecting, developing and administering patient-reported outcome measures to assess the value of treatments for children with persistent OME and recurrent AOM and upper respiratory infections; investigating the clinical effectiveness and cost-effectiveness of hearing aids (air or bone conduction) and the use of interventions to improve classroom acoustics for children with different degrees of persistence and severity of hearing loss associated with OME; and investigating why professionals' and parents'/carers' treatment preferences vary so much both nationally and internationally. We do not understand why adenoidectomy works in different subgroups at different ages, nor its effects in special populations, such as children with Down syndrome. We also need further research on the impact and optimal management of otitis media in these special situations and others, such as in children with a cleft palate or developmental problems. The study is registered as PROSPERO CRD42011001549. The National Institute for Health Research Health Technology Assessment programme.