Health costs and consequences: have UK national guidelines had any effect on tonsillectomy rates and hospital admissions for tonsillitis?

Abstract

Tonsillectomy is a common therapeutic option in the management of recurrent tonsillitis. In 1999, the Scottish Intercollegiate Guidelines Network (SIGN) introduced SIGN 34 outlining appropriate indications for tonsillectomy. Following concerns of increasing hospital admissions for tonsillitis, in 2009 ENT UK suggested that too few tonsillectomies were being undertaken. This study analyses the effect the SIGN guidelines have had on trends in population rates of tonsillectomy and hospital admissions for tonsillitis and peritonsillar abscess in England, Scotland and Wales. A retrospective study was undertaken using the health databases of England, Scotland and Wales between 1999 and 2010. Tonsillectomy, acute tonsillitis and peritonsillar abscess were identified using national classification codes. Changes in rate of tonsillectomy and hospital admissions for tonsillitis and peritonsillar abscess were assessed using a linear regression model. 699,898 tonsillectomies were undertaken in the three national cohorts over the study period. Linear regression analysis suggested that implementation of SIGN 34 significantly reduced the population rate of tonsillectomy in England (p = 0.005) and Wales (p = 0.003) but not in Scotland (p = 0.24), and indicated there had been an increase in hospital admissions for acute tonsillitis in all cohorts (England p = 0.000008, Scotland p = 0.03, Wales p = 0.000005) and peritonsillar abscess in England (p < 0.05) and Wales (p = 0.03). SIGN 34 has reduced tonsillectomy rates in England and Wales but not in Scotland. This finding is associated with increasing hospital admissions for acute tonsillitis in all national cohorts, which may suggest that the current stipulated guidelines miss patients who would benefit from surgical intervention.