History of recurrent sore throat as an indication for tonsillectomy. Predictive limitations of histories that are undocumented.

As part of a prospective study of indications for tonsillectomy and adenoidectomy, we followed closely 65 children with histories of recurrent throat infection that seemed impressive (at least seven episodes in one year, five in each of two consecutive years or three in each of three consecutive years), but lacked documentation. During the first year of observation, only 11 children (17 per cent) had episodes of throat infection with clinical features and patterns of frequency conforming to those described in their presenting histories. Of the remaining 54 children, 43 (80 per cent) experienced no, one or two observed episodes each, and most of the episodes were mild. We conclude that undocumented histories of recurrent throat infection do not validly forecast subsequent experience and hence do not constitute an adequate basis for subjecting children to tonsillectomy.