Prospective study of the accuracy of the surgeon's diagnosis in 2000 excised skin tumors.

Expeditious yet efficacious removal of skin tumors is a common responsibility for the plastic surgeon. The need to minimize potential risks for mortality or morbidity from undue or excessive surgical resections and to control costs by avoiding unnecessary procedures behooves us to make a precise clinical diagnosis preceding any decision even for such "minor" surgery. Just how accurate these decisions can be expected to be for a typical surgical practice was scrutinized by means of this prospective 4-year study involving the resection of 2058 skin lesions. Each lesion was initially assigned a clinical diagnosis after a brief gross examination and then compared with the pathology report, which was always considered to be the correct answer. Within these parameters, only 65 percent of all tumors were identified correctly preoperatively. Two-thirds of all lesions were benign. Three-quarters of benign lesions were as assumed, and 92 percent of all presumed benign lesions were benign even if incorrectly identified initially, whereas fortunately only 3 percent proved to be malignant. On the other hand, only three-fifths of malignant lesions were identified correctly clinically, yet only 11 percent were benign, implying that most such lesions properly deserved excision anyway. Therefore, approximately 90 percent of all lesions whether benign or malignant were removed appropriately without compromising the patient, but to expect a clinical acumen of 100 percent in this setting may not be realistic. The accuracy of the surgeon in identifying lesions as probably benign was certainly high enough that cost-containment mechanisms designed to deny authorization for their removal probably would be justifiable and difficult to appeal. Any suspicious or equivocal lesions still will require mandatory intervention despite such constraints, because often only histologic examination will allow a definitive diagnosis.