OBJECTIVE: To analyse the incidence and timing of postoperative haemorrhage that requires re-operation following tonsillectomy or adenotonsillectomy, and to formulate a recommendation regarding the optimal duration of postoperative clinical observation. DESIGN: Retrospective. METHOD: Data were collected on the incidence of haemorrhage that required exploratory surgery in all patients who underwent adenotomy (n=3508) or tonsillectomy or adenotonsillectomy (n=4909) in the period 1996-2002 at the Sint Lucas Andreas Hospital in Amsterdam, the Netherlands. The Sluder technique was used on an outpatient basis in 2439 of the 4909 patients who underwent tonsillectomy or adenotonsillectomy; these patients were aged <10 years. The remaining 2470 patients were aged > or =10 years and underwent conventional dissection followed by 24 hours of clinical observation before being discharged. RESULTS: Postoperative haemorrhage was recorded in 0 of the 3508 patients who underwent adenotomy, 12 (0.5%) of the 2439 patients aged <10 years who underwent tonsillectomy using the Sluder technique and 43 (1.7%) of the 2470 patient aged > or =10 years who underwent conventional dissection. Of the 43 cases of haemorrhage after conventional dissection, 31 (75%) occurred within 24 hours, including 2 cases that occurred between 12 and 24 hours (2/2470; 0.08%; 95% CI: 0.001-0.29). CONCLUSION: Reducing the postoperative observation period from 24 hours to 12 hours would have unfavourable consequences in 1 (95% CI: 0-3) per 1ooo patients operated. In regard to the risk of postoperative haemorrhage, outpatient tonsillectomy and adenotonsillectomy appear to be justified for patients aged 10 years or over, provided that surgery is performed in the morning and the patient is observed in a ward that is open until the evening.
OBJECTIVE: To analyse the incidence and timing of postoperative haemorrhage that requires re-operation following tonsillectomy or adenotonsillectomy, and to formulate a recommendation regarding the optimal duration of postoperative clinical observation. DESIGN: Retrospective. METHOD: Data were collected on the incidence of haemorrhage that required exploratory surgery in all patients who underwent adenotomy (n=3508) or tonsillectomy or adenotonsillectomy (n=4909) in the period 1996-2002 at the Sint Lucas Andreas Hospital in Amsterdam, the Netherlands. The Sluder technique was used on an outpatient basis in 2439 of the 4909 patients who underwent tonsillectomy or adenotonsillectomy; these patients were aged <10 years. The remaining 2470 patients were aged > or =10 years and underwent conventional dissection followed by 24 hours of clinical observation before being discharged. RESULTS:Postoperative haemorrhage was recorded in 0 of the 3508 patients who underwent adenotomy, 12 (0.5%) of the 2439 patients aged <10 years who underwent tonsillectomy using the Sluder technique and 43 (1.7%) of the 2470 patient aged > or =10 years who underwent conventional dissection. Of the 43 cases of haemorrhage after conventional dissection, 31 (75%) occurred within 24 hours, including 2 cases that occurred between 12 and 24 hours (2/2470; 0.08%; 95% CI: 0.001-0.29). CONCLUSION: Reducing the postoperative observation period from 24 hours to 12 hours would have unfavourable consequences in 1 (95% CI: 0-3) per 1ooo patients operated. In regard to the risk of postoperative haemorrhage, outpatient tonsillectomy and adenotonsillectomy appear to be justified for patients aged 10 years or over, provided that surgery is performed in the morning and the patient is observed in a ward that is open until the evening.