Predicting the Outcome From Hypoxic-Ischemic Coma

To the Editor.—  In the Sept 6, 1985, issue of The Journal, Dr Black¹ expressed major reservations about applying the scheme for predicting the outcome of hypoxic-ischemic coma suggested by Levy et al.² One important objection was that the scheme is based on retrospective data. Other serious methodological drawbacks are as follows: (1) The data analyzed in retrospect concern patients seen on occasion at three medical centers with different clientele and treatment strategies,^3,4 thereby making the possibility of interobserver variability very high. (2) The neurological findings are graded in a "worst to best" order and the greatest predictive weight is assigned to the response considered "best" at random examination time. However, any of the signs explored may be a facet of the orderly recovery. (3) The outcome categories are defined in the same way as are the end points of head trauma,⁵ including eventual "dependency" on