Management of childhood otitis media

A prospective audit of 5430 initial consultations and 1602 admissions to Scottish hospitals for surgical treatment of otitis media with effusion in childhood in ear, nose and throat departments throughout Scotland has been carried out. The results suggest that Scottish children are assessed by clinicians of appropriate experience when initially seen in outpatients and only a minority (30%) are listed for surgery following their first visit. Not all children have a hearing test at their initial clinic visit. In those cases where surgical treatment is not advised at the first visit, a policy of "watchful waiting" is preferred to medical treatment by most clinicians. Bilateral dry tap rates varied between zero in Forth Valley and 19% in Lanarkshire. Fifty one per cent of operations were carried out by consultants and only 3% by SHOs. "Best practice" for the initial management of childhood ear problems is widespread in Scotland but there is room for improvement. There is a need for review of the availability of paediatric audiology services.

Ear symptoms are among the commonest reason for general practitioner consultation in the childhood population in the UK, and also constitute the majority of referrals to hospitals in the childhood population, mainly for the management of otitis media with effusion (OME, glue ear). Most children have middle ear effusions at some time during childhood but these are transient in the majority.¹

There is, however, a minority in whom effusions persist over months or years causing hearing loss which in turn potentially impairs speech development and educational performance.^2,3 Some of these children also suffer from recurrent otalgia.

No medical treatment has been shown to influence the natural history of OME.⁴ Both antibiotics⁵ and auto-inflation using the "Otovent" device may produce temporary improvement. The use of ventilation tubes leads to improvement in hearing limited to the period that the tubes are in situ and functioning.^6-9 Adenoidectomy is the only treatment which leads to resolution of OME, but it is only effective in a proportion of cases? In the past the prevention of subsequent development of cholesteatoma has been cited as an indication for surgical treatment of OME, but there is no good evidence that the use of ventilation tubes either prevents or causes this disease.¹⁰

This audit of current practice in the management of OME and childhood ear symptoms was designed to establish current practice in Scotland as the first part of the audit loop. As otolaryngologists, we were also aware that our management of such cases has come under some criticism from individuals outside the speciality.

The infrastructure already in place for the Scottish Tonsillectomy Audit was utilised for the project. Data were collected in a prospective manner by asking clinicians and audiologists to complete proformas covering relevant information. Five separate forms were created to cover each different clinical situation: an initial visit form, a follow-up outpatient form, an inpatient form and a post-operative follow-up form, the fifth form being for audiometric data to be entered by the audiologist. This manuscript presents the results from initial visit, audiology and inpatient forms only. New referrals were recruited over a 15 month period between April 1994 and July 1995. Clinicians were encouraged to recruit all cases presenting to them of children aged under 9 years-of-age with ear symptoms and enter data into the appropriate forms. The data were collated centrally and analysed by the Medical Computing Unit, University of Dundee, Ninewells Hospital.

The results were compared with a set of standards for good practice which the Audit Sub-committee felt constituted a consensus amongst otolaryngologists (Table 1).

Assessment prior to surgery should take place over more than one clinic visit unless documentation from the GP or Communitybased hearing assessment unit is very comprehensive. The review period should allow sufficient time for natural resolution of the condition
Initial assessment should include an age appropriate hearing test
Medical therapy should be generally confined to analgesia with antibiotics for definite infection
Initial assessment should be carried out by staff of adequate seniority and experience

During the patient recruitment period (15 months) 5430 initial visit forms were completed and returned. There was significant variation in the numbers returned from different parts of Scotland. This only partly reflects differences in population size in different areas. Figure 1 presents the numbers of forms returned expressed as proportions of the populations of each area of Scotland within the relevant age group. This indicates a low rate of return from some areas such as Glasgow, as compared with other areas, such as Dumfries and Galloway. Assuming that the prevalence of child-hood ear disease is similar throughout Scotland, this indicates different levels of compliance with the audit.

Figure 1:Number of initial visits forms in relation to population under 9 years-of-age, by area

1 Argyll & Clyde 2 Ayrshire & Arran 3 Dumfries & Galloway 4 Fife 5 Forth Valley 6 Grampian 7 Glasgow 8 Highland 9 Lanarkshire 10 Lothian 11Tayside

The delay between the date of a referral letter being sent and the date of the initial consultation varied widely. Some patients were seen as emergencies on the day of referral while others waited up to 14 months. Most patients were seen within 3 months. Sixty-eight per cent of patients were seen by a consultant at their initial visit. Only a very small number (less than 2%) were seen by relatively inexperienced junior staff. There was some variation between areas, to some extent reflecting different patterns of staffing in different ENT units. Most referrals (74%) were from general practitioners, but a significant minority were referred following screening of hearing by community based units.

A completed audiology form was submitted for initial visits in 66% of cases. The variation in the proportion of patients for whom an audiology form was received in different areas are presented in Figure 2. The first column for each area shows the proportion of initial visit forms accompanied by an audiology form for the whole group. To verify the accuracy of this data, the notes of a sample of a total of 1000 children selected from each area in proportion to the population of children under 9 years of age were examined and missing data were retrieved. The second columns in Figure 3 show the proportions of initial visit forms accompanied by an audiology form within this sample.

Figure 2: Proportions of initial visit forms accompanied by an Audiology Form for the whole group and checked

1 Argyll & Clyde 2 Ayrshire & Arran 3 Dumfries & Galloway 4 Fife 5 Forth Valley 6 Grampian 7 Glasgow 8 Highland 9 Lanarkshire 10 Lothian 11Tayside

The most common initial management strategy was reassessment (i.e. no active intervention), while 20% of patients were discharged from the clinic and only 30% were listed for surgery at the first visit. Some of the children for whom review appointment was arranged were prescribed medical treatment, but this was only employed by a small minority. The most popular treatment was decongestant medication which was recommended in 6.5% of cases. Recurrent otalgia was the sole reason for surgery in a minority of children in all areas. This reason was cited most frequently in Lanarkshire and least often in Forth Valley.

Inpatient data were submitted for 1602 admissions. The waiting times for surgery in different areas are shown in Figure 3. One hundred and nine (7%) children did not undergo the planned operation, leaving 1493 who had surgery, either myringotomies with or without ventilation tubes and/or adenoidectomy. In 43% of non-operated cases this was because the OME had resolved, while in 12% effusions were still present but the hearing thresholds were satisfactory. Thirty-one percent were considered to be unfit for surgery at the time of admission. Fifty-five per cent of these patients had outpatient review appoint-ments arranged and 30% were discharged to the care of their GPs. Surgery was rescheduled in 14%. Of those who under-went surgery, 62% had bilateral effusions, 19% had unilateral effusions and 19% had bilateral dry taps. Mucoid effusions were the most common type reported (69%). Bilateral dry tap rates in the different areas are shown in Figure 4.

1 Argyll & Clyde 2 Ayrshire & Arran 3 Dumfries & Galloway 4 Fife 5 Forth Valley 6 Grampian 7 Glasgow 8 Highland 9 Lanarkshire 10 Lothian 11Tayside

Ventilation tubes were inserted in 76% of ears with effusions and in 45% of those which were found to be dry at the time of myringotomy. The most popular type of ventilation tube was the Shah grommet (57%). Of those undergoing surgery 48.8% underwent adenoidectomy in addition to myringotomies with or without grommet insertion and 4.9% underwent tonsillectomy. The numbers of operations carried out by different grades of surgeon throughout Scotland are shown in Figure 5. Variations between areas reflect staff mix in different areas.

Variations in management of OME in different Scottish health Board areas have been reported by Bisset¹¹ who studied Scottish Morbidity Records (SMR1) data. The data presented in this report confirms this to some extent. It should be pointed out that it is difficult to have a consensus concerning the correct management for OME when there is inadequate data available to indicate the optimal strategy. Bisset¹¹ also reported a decline in the numbers of operations for OME between 1990 and 1994 and a trend towards less variation in types and frequency of intervention.

The results of this audit support the contention that most children with ear problems in Scotland are managed thoughtfully and in line with current accepted best practice. It appears that most are seen by consultants or other experienced clinicians. Variations in the proportions seen by consultants in different departments are largely explained by variations in patterns of staffing between areas with a consultant-based service, such as Forth Valley, and those with a consultant led service, such as Tayside.

Surgical treatment of OME was recommended in 30% of children at the first visit, indicating that a policy of ‘watchful waiting’ is widely practised in Scotland. Some may have previously undergone a period of observation in a community-based hearing assessment unit. The availability of such units varies across Scotland. In other cases, the decision may have been based on considerations such as delayed speech development or on information from the general practitioner. Surgery was most often recommended in children deemed to have bilateral OME on otoscopy. However, in 312 cases surgery was advised when unilateral OME was reported on otoscopy. Such decisions may have been based on audiometric and tympanometric findings.

Our findings concerning audiometric testing at the initial visit are of some concern. The evaluation of hearing is an important part of the initial assessment of this type of case and audiometry prior to surgical treatment is essential. Our results indicate substantial variations in the proportion of children for whom an audiometric data form was returned following the initial visit. It may be relevant that the Initial Visit Forms were completed and returned by the clinician who saw the patient, while the Audiology Form was completed and sent in by the audiologist. The compliance of two participants, therefore, was required for a complete data set to be received and the fact that an Audiology Form was not returned does not mean that a test was not done. The sample of 1000 cases in which the case notes were obtained and missing data was extracted, confirmed that there were no hearing test results available for a significant minority of children. Some of these would have been seen in a clinic where audiometric facilities were absent or inadequate, particularly for younger children. Testing hearing in young children is difficult and requires specific training and considerable skill. All clinicians who assess children with ear problems must have access to adequate paediatric audiology facilities.

It is likely that surgery for some of the children with bilateral normal ears (148 cases) was for recurrent otalgia, usually due to acute otitis media. Other children may have been shown to have fluctuating OME, verified by a period of outpatient observation, which was interfering with their progress. Some children with recurrent acute otitis media also have chronic OME, but others have normal ears between episodes. There is some evidence of benefit from ventilation tube placement in children with recurrent acute otitis media but there is doubt concerning their efficacy in influencing the natural history of the problem.

Data were collected for 1493 operations for OME. In addition, 109 children's operations were cancelled. In 43% of cases this was because of spontaneous resolution of effusions. This illustrates the importance of reassessing such cases prior to surgery. Dry tap rates varied significantly and this is an area where practice could be improved. This could be achieved by pre-operative re-assessment of surgical cases and postponement of surgery in those in whom resolution or substantial improvement is evident. This can be conveniently done at a pre-admission clinic with facilities for audiological assessment. Another factor which may influence dry tap rates is the duration of the interval between the decision to operate and the date of surgery. There are variations in waiting times across Scotland and variations in practice exist in relation to reassessment. In Lothian, for example, children are placed on a provisional waiting list and then reviewed in outpatients by the clinician who listed them prior to being given a date for surgery. OME is well recognised as a remitting condition and, in spite of this, may still have significant effects and require active treatment. Although the effusion may have temporarily resolved, surgery may still be indicated in fully assessed cases.

The results show that compliance with the audit, particularly in some areas of the country, was well below expectations, assuming that the prevalence of OME does not differ widely between different parts of Scotland. The actual level of compliance in different areas cannot be quantified and this potentially introduces bias into the results. It is likely that some clinicians returned forms for a higher number of the relevant cases they saw than others and that some did not return any forms. However, there is no particular reason to believe that these individuals manage childhood ear problems differently or that they are likely to be junior doctors rather than senior staff.

It is not possible from this data to determine surgical rates in different areas of Scotland because, although we have data concerning initial and follow-up consultations, it seems likely that the operations reported above comprise only a proportion of those carried out during the period of the audit.

OME is a common and important condition because of its impact on speech development and educational performance. Appropriate management of the condition is particularly important because of the relatively high aggregate costs of treatment to the NHS. While guidelines can be formulated for its management, based on existing data, there are still significant gaps in our knowledge and scope for individualisation of care.

This audit indicates that there is a widespread appreciation of the high quality management of OME among Scottish otolaryngologists, particularly with regard to the decision making process. Decisions about the management are taken by experienced clinicians. Concerns are raised about the availability of high quality audiology services tailored to the requirements for testing young children. Surgery for childhood otitis media with effusion is carried out by clinicians of appropriate experience. Variations in dry tap rates suggest a need for more widespread reassessment of cases immediately pre-operatively.

We would like to thank the members of the Audit Sub-committee of the Scottish Otolaryngological Society, our Audit Co-ordinator, Mrs Anne McLagan, and Miss Dawn Benvie and Dr Wynn Carter of the University Medical Computing Unit, Ninewells Hospital.

Correspondence to: Dr R.P. Mills, Otolaryngology Unit, Lauriston Building, Royal Infirmary, Edinburgh EH3 9E, U.K.

Management of childhood otitis media with effusion by Scottish otolaryngologists