Why it is vital not to take your eye off practice's expenses
In the first of a two-part update on glue ear, Professor Mark Haggard and Mary Gannon detail what can be done to ensure accurate diagnosis and reduce the huge variation in GP referral rates
Our study among GPs1 found large differences in referral patterns for children's ear problems compared with those for nose and throat complaints. Such differences were noticeable even within practices where the catchment and referral destinations were the same.
Similarity between partners has to reflect shared local constraints such as waiting lists for the single specialty involved and shared school of thought among partners, while differences have to reflect individual clinical uncertainty.
Our study showed differences in referral rates could be reduced by GPs using a structured checklist and video about children's ear problems. Clearly there is a need for more training, referral criteria, audit and instrumentation.
Diagnosis and link to ear infections
In otitis media with effusion (OME) the inflamed middle-ear mucosa generate secretions at such a high rate and viscosity that the eustachian tube cannot drain them. These effusions were formerly thought to be largely sterile, but sensitive modern molecular assays such as PCR now show they are not.
The condition is typically self-limiting general resolution rate 50 per cent in three months but once the condition has been established for much longer, the resolution rate becomes slower. This is probably because OME in some children can also be self-maintaining. After an acute ear infection (AOM) bacteria can survive for many months in a special state of lowered activity.
Bacteria forming a biofilm, for example, resist attacks by the immune system and antibiotics. Despite their low activity, these bacteria nevertheless stimulate further immune system activity and hence continuing effusion. This sub-acute state can flare up again into evident inflammation, re-justifying the label 'acute'.
This explains both the intercurrent (super-added) AOM that can occur in cases viewed primarily as OME and also explains how OME can sometimes appear to occur without evidence of immediately preceding respiratory and ear infection, although that is the usual progression.
AOM will often produce a short-term purulent effusion, followed by a serous one, although the latter will usually resolve within a few weeks. When this effusion takes the mucoid and viscous form the name 'glue ear', associated with longer-term persistence, is truly justified. As both a medical and a vernacular term, glue ear became over-generalised to the whole spectrum of common non-acute otitis.
When effusion lasts more than two weeks the child can be legitimately diagnosed as having OME. But, like most cut-offs on a continuum of degree or of time for diagnosis or treatment, this cut-off is relatively arbitrary.
In such a common condition, diagnosis alone cannot reasonably control the consequent action: severity, persistence, and evident impact or factors predisposing to impact need to be considered. In UK practice, those children who achieve an explicit diagnosis at ENT or community audiology will not be marginal cases just meeting such a diagnostic criterion.
They will have seemed to parents and to the GP to have suffered sufficient impact to justify referral, will have pursued the appointment, and then 'tested positive', ie had evidence in the tympanogram of fluid present plus an identifiable if usually mild hearing loss. Explicit diagnosis does not tend to be made by GPs, partly as they lack the relevant equipment2 but of course a referral or invitation to a parent to re-consult, if problems continue, carries a provisional or an implicit diagnosis.
In OME despite the continuing minor inflammation in the middle ear, there is usually no otoscopically perceptible redness of the type seen in AOM. The eardrum appearances are diverse so diagnosis should not be attempted on ordinary otoscopy alone. Pneumatic otoscopy is very useful to establish eardrum mobility or immobility but is uncommon in UK primary care.
General clinical assessment and history are necessary, supplemented by the objective technique of tympanometry to verify presence of fluid. Fluid presence characterises a particular day of test rather than the long-term susceptibility that is clinically most relevant. For this reason, tympanometry alone as a screen on asymptomatic children leads to too many referrals or repeat tests, leading to eventual discredit. Similarly, the performance technique of audiometry (which here can be classed as objective) summarises the current presence and severity of OME, but reflects only indirectly the past and future impact on the child's development. This limitation may explain why facilities for audiometry in primary care have remained scarce2.
The diagnosis of OME is more complex than for AOM as the condition is often only semi-symptomatic and the signs can be subtle. OME fluctuates, and in its early stage under review in general practice it has a high rate of spontaneous remission. Thus a history from the parent may be in complete accord with the diagnosis that would have been given in the preceding month, if not on the test day. The main signs/symptoms triggering consultation in primary care are hearing loss, or behaviours reflecting the consequent breakdown of communication. In short, the diagnosis is not complex in the usual sense, but it does require alert questioning and awareness of any AOM history.
Case-finding and referral
The prevalence is high from one to about seven years of age, and is seasonal, rising abruptly in mid-late autumn. Screening at school entry brings in a peak of cases around 4.5 to 5.5 years old.
GPs have to decide tactically based on local knowledge of waiting times and number of physical symptoms between making a referral based on clinical questioning alone versus requesting a hearing test from community audiology services, where available, or referring directly to ENT.
Unfortunately, there are no convenient biological markers to show where in the usual progression a child currently is, or how long it will take him to get through it. In the future, bio-markers of genetic make-up should be able to indicate the likelihood of any incident becoming a persistent case. It is quite likely they will become low-cost enough for use in primary care, leading to a justified greater degree of selectivity in referral and treatment. Meanwhile, careful questioning on history and presentation, and the stimulation of parental observation to guide
re-consultation are informative.
A history of ear infections along with speech delay or behaviour problems and any parental suggestion of hearing problems should always trigger suspicion of longer-term OME. In the older child with OME a parent can detect inattention, misunderstood communication and disengagement, more commonly referred to as being 'out of it' or 'ignorant' in some local dialects.
In tapping into such family observations, the GP needs to compensate for the degree to which a parent may be personally or culturally inclined to over- or under-react to impaired communication. This is more difficult with mobile populations and in the under-threes. For example, withdrawn behaviour is one adaptive response to poor hearing and the 'shut-in' feeling of OME; some parents may view this as untroublesome, therefore desirable!
In the under-threes, some speech delay from OME is not uncommon, but because of wide normal variation the effects on language development from OME are mostly modest and transitory. Referral to speech therapy may be locally discouraged to avoid swamping an overloaded service. But if problems are severe and the child also belongs to a learning-disabled category or suffers other susceptibility it may be wiser to assure total effectiveness of the auditory channel via ENT referral with explicit mention of apparent language impact.
Evidence to guide practice
Relatively little research and development on OME has been oriented to management in primary care. Nevertheless secondary care research can throw light on primary care referral issues. In recruitment for our multicentre randomised tria · 3 we showed an interesting contrast between referrals from GPs and from community paediatricians in audiology (CPA).
Not surprisingly, CPA with its further stage(s) of watchful waiting using audiometry plus tympanometry instrumentation had a better hit rate at ENT for hearing problems. We also found GPs have the better hit rate for the physical health problems which accompany OME and are part of its causation and overall impact. For example, 80 per cent of our trial sample were below the population 50th percentile of weight-for-age. The symmetry between professionals is pleasing, especially as the hearing loss is helped by ventilation tubes (VTs 'grommets') whereas the physical health and hearing are both helped by an adjuvant adenoidectomy.
Any proactive case-finding for children's middle-ear conditions in primary care is problematic because of the high prevalence. Universal screening for OME with a hearing test is not a sensible proposition, because of the condition's fluctuating nature and its high remission rate. Secondary prevention to reduce developmental disability at first seems also not to be promising. This is because there is no objective predictor of persistence: we are forced to use delays and the longer-term impact of past disease as a prognostic sign of it.
This unsatisfactory impasse creates a challenge to avoid complacency within the policy of 'watchful waiting'. With watchfulness insufficiently emphasised in relation to waiting, the policy could allow OME to go on for many years, on the perverse grounds that eventually a child will grow out of it. Children who need help and can benefit from it need to be distinguished in the course of this process from the many who do not or cannot.
So vigilance with systematic questioning (see box below) is required to establish severity, persistence or recurrence, and thus justification for referral. This poses a further challenge for an under-resourced and hence largely reactive service. But meeting that challenge is the only way to find the severe and persistent cases before the impact upon them has become unjustifiably long and potentially irreversible.
Developing the role of primary care in OME
Immediately, GPs as gatekeepers can inform their referrals with the knowledge that the clock is in some ways being turned back from an over-emphasis on language disability to a more balanced emphasis on diverse presentations including physical problems.
This underlines the need to see AOM and OME as overlapping diagnoses. For example, we find the measurable behaviour problems in OME are roughly doubled in the subset of children whose parents additionally report intercurrent AOM. There is not a trial specifically on such children.
To help meet GPs' information needs for cost-effective referral we have developed a video for GPs (see 'useful information' below) and a checklist of clinical questions and risk factors (left) for documenting or predicting the severity/persistence of a history of middle ear disease.
For communication with other professionals and parents, GPs also need an awareness of the differences between the way they view children's ear problems and the perspectives of the other parties.
In a study examining parent/professional differences in understanding and concern about the impact of OME, our colleague Josie Higson4 showed GPs attributed less negative impact arising from hearing and concentration difficulties than parents did, but they invoked speech/language problems more. Language is a dominant theme in the older literature, but not one accompanied by a strong evidence base.
GPs also tended to ascribe lower impact to hearing problems than other medical professional groups examined in the study, but gave a higher impact to physical signs. This finding probably arises from the higher proportion of children in the GP caseload presenting with acute symptoms.
It also cautions that the differences in dominant aspect seen by different professions (for such reasons of stage in the clinical pathway at which children are typically seen) are large enough to generate failures of inter-professional communication.
Lest we trigger another cycle of fashion, by seeming to consider hearing unimportant, we finish by noting that parents' major single concern in OME is hearing. Whether watchful waiting or referral is adopted, GPs can provide parents with information about simple tactics for managing communication problems.
Checklist for systematic questioning on history
questioning on history
Parts of the following list have been formally evaluated and are suitable for use in
primary care. A quantitative scoring with precise cut-off is not yet available but the cumulative history reflected by an increasing number of positive answers is strongly related to the need for referral to ENT.
1Has there been a previous operation to insert grommets?
The fact of a previous operation means persistence has already been established and if problems are beginning to recur this is a strong indicator for re-referral.But an ear infection within three months of surgery may indicate a localised infection around the grommet site, rather than a middle-ear problem. Where this is suggested by otoscopic examination, locally agreed protocol for management should be followed.
2Did the first ear or hearing problem start before the
age of two?
3In the past 12 months have any of the following occurred?
a) Three or more ear infections, excluding this visit
b) frequent mouth-breathing
c) frequent snoring
d) frequent coughs, colds or sore throats
e) frequent earaches
f) frequently sounding as though the nose is blocked
4Have any close family members ever needed treatment by a hospital ear specialist for similar problems?
5Is the child often in an environment where someone
6In recent months, has the child?
a) had difficulty hearing when in a group
b) often asked for things to be repeated
c) mis-heard words when not looking at the speaker
d) had hearing described as a problem (by anyone)
7Has anyone had concerns about the child's communication or behaviour that might be due to poor hearing?
1 Bennett Ket al. Improving referrals for glue ear from primary care: are multiple interventions better than one alone? J Health Serv Res Policy
2001; 6 No 3: 139-44
2 Bennett KE et al. Do GPs have the techniques for 'watchful waiting' in glue ear? Brit J Gen Prac 1998;48:1079-80
3 MRC Multi-centre Otitis Media Study Group. Selecting persistent glue ear for referral in general practice: a risk factor approach. British Journal of General Practice 2002;52:249-553
4 Higson JM. Parent and professional beliefs about otitis media with effusion. PhD Thesis University of Nottingham 1999. Publications also in press
Williamson Ian. Otitis Media with Effusion. Clinical Evidence, V
ol II (2):206-12. BMA Publishing Group, London, December 1999
The parent information leaflet 'Glue ear a guide for parents' was favourably evaluated some years ago in a survey by the King's Fund and is available from the research charity Defeating Deafness, freephone 0808 808 2222, fax 020 7278 0404, e-mail firstname.lastname@example.org or visit www.defeatingdeafness.org.
A checklist of questions to quantify the severity of a history and related written material are also available from Defeating Deafness.
Mark Haggard is leader of the MRC external scientific staff team in children's middle-ear disease, Cambridge and of the MRC multi-centre otitis media study group Mary Gannon is health services researcher in the MRC multi-centre Otitis Media Study Group
Next week The treatment of glue ear