EDITOR,--We envisage that databases for evidence based medicine would answer the question “What is best practice?” and that hospital based audit would check that best practice is being achieved. Our audit programme has examined pain relief for tonsillectomy. In the 58 children audited, nine tonsillar blocks were performed; additional oral pain relief was required earlier in these nine children than in the children who had alternative intraoperative analgesia (n=9) (3.3 h v 7.6 h).

A conflict therefore arises over whether we should persist with tonsillar block and improve our performance …