6.1 Tonsillectomy Rates

Tonsillectomy was the most common NHS operation in Scottish children in 1990, with a two-fold variation in rates across Health Boards.⁵⁸

Data from the Information and Statistics Division of the Common Services Agency of the NHS in Scotland shows that, between 1990 and 1996, the rate for tonsillectomies in children aged 0-15 declined from 602 per 100,000 (6,152 operations) to 511 per 100,000 (5,256 operations). 44% of patients were male. 54% had their adenoids removed and 13% had surgery to drain the middle ear at the same operation. In adults aged 16 years and over, the tonsillectomy rate increased from 72 per 100,000 in 1990 (2,919 operations) to 78 per 100,000 in 1996 (3,200 operations). 32% were male. In 1996, 0.8% of children and 3% of adults were treated as day cases. 54% of children and 61% of adults had a two-night stay.

It is likely that current practice has reduced the length of inpatient stay following tonsillectomy throughout Scotland, but there is as yet no data to confirm this.

Emergency re-admissions within four weeks of discharge after tonsillectomy and/or adenoidectomy in children under 16 years (from April 1992 to March 1995) are one of the Scottish Office clinical outcome indicators. 1.3% of Scottish patients (340) were re-admitted. This varied from 0.5% to 3.4% across the different Trusts.⁵⁹

The Scottish Tonsillectomy Audit, carried out by the Audit Subcommittee of the Scottish Otolaryngological Society and funded by the Clinical Resource & Audit Group (CRAG) looked at tonsillectomy activity throughout Scotland over a 12 month period from February 1992. A number of differences in management of recurring sore throat by ENT surgeons were highlighted across the country. These included variation in the rate of operation by area and differing management of children and adults. Outcome was measured by the response to a questionnaire at six months and one year after surgery, and indicated a high satisfaction rate among patients of 97%, with a 75% response rate at six months and 45% at one year.²

6.2 Evidence for Surgery in Recurrent Sore Throat

The literature on surgery for sore throat is scanty, out of date and lacking in scientific validity. Most published studies refer to a paediatric population. The current widely accepted criteria for surgery are of the order of seven episodes of tonsillitis in the preceding year, five episodes in each of the preceding two years or three episodes in each of the preceding three years, and have been arrived at arbitrarily.⁶⁰ They take no account of whether the condition is worsening or improving and make no distinction between children and adults, in whom the disease may behave differently. The small amount of information about adult sore throat and the effect of tonsillectomy is not scientifically robust by current standards but suggests that surgery is beneficial.⁶¹

It seems reasonable to assume that recurrent acute attacks of tonsillitis can be prevented by tonsillectomy, but tonsillectomy will not prevent recurrent sore throats due to other reasons. Hence, before considering tonsillectomy, the diagnosis of recurrent tonsillitis should be confirmed by history and clinical examination; and, if possible, differentiated from generalised pharyngitis.

The natural history of tonsillitis is for the episodes to get less frequent with time, but epidemiological data is lacking in all age groups to allow a predication of this to be made in individual patients.

Tonsillectomy requires a short admission to hospital and a general anaesthetic, is painful, and is occasionally complicated by bleeding. Return to usual activities takes on average two weeks, with a corresponding loss of time from education or work.

Four randomised controlled trials of tonsillectomy against non-surgical management in children have been reported.^{60, 62-64} All were designed before 1972 and none would satisfy current criteria for a well designed, controlled and analysed study. In the most quoted reference in particular, randomisation was not balanced in frequency of episodes or socio-economic group.⁶⁰ In this study, the number of episodes of sore throat post-tonsillectomy was significantly fewer than in the control group, although when the number of days of illness with sore throat was taken into account, including those associated with surgery, benefit from tonsillectomy was less evident. No randomised controlled studies have been reported in adults. Evidence level Ib

The following are recommended as reasonable indications for consideration of tonsillectomy in both children and adults, based on the current level of knowledge, clinical observation in the field and the results of clinical audit. Patients should meet all of the following criteria: sore throats are due to tonsillitis five or more episodes of sore throat per year symptoms for at least a year the episodes of sore throat are disabling and prevent normal functioning. Grade C recommendation extrapolated from evidence level Ib, IIb and III

Cognisance should also be taken of whether the frequency of episodes is increasing or decreasing.

Note that, in considering whether a patient meets these criteria, the general practitioner may have difficulty in documenting the frequency of episodes because patients do not always consult when they have an episode. There may also be uncertainty about whether the sore throats are due to acute tonsillitis.

6.4 Otolaryngological assessment

Patients referred will rarely be seen by a specialist during an acute episode of sore throat, so the diagnosis of recurrent acute tonsillitis rests with the referring doctor. Questioning the patient about the appearance of the throat, the degree of systemic upset, and the presence of tender neck lymph nodes can help confirm the diagnosis.

The specialist should also confirm the frequency of occurrence of the episodes and assess the associated disability. If the criteria set out above are confirmed, the management options should be discussed and the benefits of tonsillectomy weighed against the natural history of resolution and the temporary incapacity associated with tonsillectomy. This information may be reinforced by means of an appropriately designed patient information leaflet (see example at Annex 3). The rate of re-admission for bleeding should also be stated as part of informed consent.

In some cases this will be the first discussion the patient or parents have had which takes into account all factors for and against operation. In addition the frequency of episodes is often an impression rather than fully documented. Under these circumstances a period of watchful waiting of at least six months, during which the patient or parent can more objectively record the number, duration and severity of the episodes, may be suggested (see example proforma at Annex 5). This would allow a more balanced judgement to be made as to the likely benefit or otherwise of tonsillectomy. This could either be reported to the GP after six months, who would then re-refer if appropriate, or be reported by the patient at a pre-arranged review hospital appointment.

A six month period of watchful waiting is recommended prior to tonsillectomy to establish firmly the pattern of symptoms and allow the patient to consider fully the implications of operation.

Once a decision is made for tonsillectomy, this should be performed as soon as possible to maximise the period of benefit before natural resolution of symptoms may occur