The wide variation in severity of cardiac infarction is well known. At one extreme the patient is admitted in severe irreversible shock, cold, clammy, and dazed with a rapid feeble pulse, low or perhaps immeasurable blood pressure, and widespread changes of severe degree in the cardiogram: his chances of surviving the acute stage of the illness are indeed slender. At the opposite extreme is the patient with cardiac pain, perhaps felt only on effort, of good colour, without shock or breathlessness, with normal pulse and blood pressure, and with limited cardiographic changes: such a patient is most unlucky if he does not survive the acute stage. Between these all possible gradations are met. As a consequence very large numbers are required (preferably 200 or more) if two groups on different therapeutic regimes are to be compared. With smaller groups, even though sampling has been strictly random, one is often left with the impression that one group has contained a larger number of the more serious cases.It seemed to us that a numerical system might be devised that would express the severity in an individual case, on lines similar to the "diagnostic score" advocated for thyrotoxicosis (Crooks, Murray, and Wayne, 1959). Such a system has previously been proposed by Schnur (1953a andb and) but we feel that in some respects this was too detailed while in others it left too much latitude for individual opinion; for example, the complication of diabetes might be allotted anywhere between 10 and 25 on his scoring system. We have attempted to devise a system in which the number of factors to be taken into account is kept to the minimum compatible with providing a reasonably close correlation between the total score and the mortality expectation. We have tried to limit the latitude allowed to the observer by defining strict criteria for the award of a given score for each factor. We have aimed at producing a method that can be easily memorized and rapidly applied, and one where the possibility of observer error is minimized. We fully realize that it is impossible to eliminate observer error completely: for example, in a borderline case one observer might well regard a patient as mildly shocked while another would classify him as having no shock. The more strictly we define the conditions qualifying for "black marks," the less room there will be for such differences of opinion.A study of data collected since 1930 has convinced us that the important factors covering the immediate prognosis (i.e. the prognosis for the first four weeks) after cardiac infarction are age, sex, previous history, degree, and severity of shock, prese4ce and severity of heart failure, cardiac rhythm, and the nature and extent of cardiographic signs. We shall discuss these factors individually.
METHOD AND MATERIALInitially we drew up a purely arbitrary score for each factor, based on our general clinical impression of its importance for prognosis. We
