The development of a new operation for the treatment of a serious disease must progress through a number of preliminary stages, involving animal and cadaver investigation, before it may be offered to the diseased human being. When the operation is finally ready for trial in the diseased patient, particularly if it carries an expected high risk, it must be offered at first only to patients so desperately ill that they are willing to accept any risk for the sake of possible benefit. Above all, the diagnosis in these patients must be beyond doubt. In the case of coronary endarterectomy as devised by Longmire, Cannon, and Kattus,1 the chief medical problem was to establish the diagnosis of coronary occlusive disease beyond doubt. Of secondary importance was the problem of obtaining some form of quantitative or at least quasiquantitative evaluation of the patients which might provide a baseline from which postoperative