Evidence Profile: Child at Risk

Aggregate evidence quality: C, observational studies of children at risk; D, expert opinion on the ability of prompt assessment and management to alter outcomes.

Benefits: optimizing conditions for hearing, speech, and language; enabling children with special needs to reach their potential; avoiding limitations on the benefits of educational interventions because of hearing problems from OME.

Harm: cost, time, and specific risks of medications or surgery.

Benefits-harms assessment: exceptional preponderance of benefits over harm based on subcommittee consensus because of circumstances to date precluding randomized trials.

Policy level: recommendation.

4. WATCHFUL WAITING: CLINICIANS SHOULD MANAGE THE CHILD WITH OME WHO IS NOT AT RISK WITH WATCHFUL WAITING FOR 3 MONTHS FROM THE DATE OF EFFUSION ONSET (IF KNOWN) OR DIAGNOSIS (IF ONSET IS UNKNOWN)

This recommendation is based on systematic review of cohort studies and the preponderance of benefit over harm.

This recommendation is based on the self-limited nature of most OME, which has been well documented in cohort studies and in control groups of randomized trials.

The likelihood of spontaneous resolution of OME is determined by the cause and duration of effusion. For example, ~75% to 90% of residual OME after an AOM episode resolves spontaneously by 3 months. Similar outcomes of defined onset during a period of surveillance in a cohort study are observed for OME. Another favorable situation involves improvement (not resolution) of newly detected OME defined as change in tympanogram from type B (flat curve) to non-B (anything other than a flat curve). Approximately 55% of children so defined improve by 3 months, but one third will have OME relapse within the next 3 months. Although a type B tympanogram is an imperfect measure of OME (81% sensitivity and 74% specificity versus myringotomy), it is the most widely reported measure suitable for deriving pooled resolution rates.

Approximately 25% of newly detected OME of unknown prior duration in children 2 to 4 years old resolves by 3 months when resolution is defined as a change in tympanogram from type B to type A/C1 (peak pressure >200 daPa). Resolution rates may be higher for infants and young children in whom the preexisting duration of effusion is generally shorter, and particularly for those observed prospectively in studies or in the course of well-child care. Documented bilateral OME of 3 months' duration or longer resolves spontaneously after 6 to 12 months in ~30% of children primarily 2 years old or older, with only marginal benefits if observed longer.

Any intervention for OME (medical or surgical) other than observation carries some inherent harm. There is little harm associated with a specified period of observation in the child who is not at risk for speech, language, or learning problems. When observing children with OME, clinicians should inform the parent or caregiver that the child may experience reduced hearing until the effusion resolves, especially if it is bilateral. Clinicians may discuss strategies for optimizing the listening and learning environment until the effusion resolves. These strategies include speaking in close proximity to the child, facing the child and speaking clearly, repeating phrases when misunderstood, and providing preferential classroom seating.

The recommendation for a 3-month period of observation is based on a clear preponderance of benefit over harm and is consistent with the original OME guideline intent of avoiding unnecessary surgery. At the discretion of the clinician, this 3-month period of watchful waiting may include interval visits at which OME is monitored by using pneumatic otoscopy, tympanometry, or both. Factors to consider in determining the optimal interval(s) for follow-up include clinical judgment, parental comfort level, unique characteristics of the child and/or his environment, access to a health care system, and hearing levels (HLs) if known.

After documented resolution of OME in all affected ears, additional follow-up is unnecessary.