I read with interest the commentary by Doust in the 2004 Sep/Oct issue.¹ I concur that the laboratory evaluation of vaginitis, in general, is superior to individual historical and physical findings. However, 3 additional points deserve emphasis. Firstly, evaluating combinations of clinical findings for calculating likelihood ratios (LRs) would be helpful, as combinations have been evaluated in at least 1 other clinical entity.² Secondly, evaluating LRs …