It has been established that 60 to 80 per cent of deaths from acute myocardial infarction occur within 12 hours of the onset of symptoms (Yater et al., 1948; Bainton and Peterson, 1963). The majority of these deaths occur before the patients reach hos-pital (McNeilly and Pemberton, 1968). The efforts directed towards lowering the mortality from acute myocardial infarction which are most likely to suc-ceed are those that aim at getting intensive care conditions to the patient as soon as possible after the onset of symptoms. To achieve the rapid institution of intensive care, a mobile intensive care unit has been described (Pantridge and Geddes, 1966, 1967). The general adoption of the mobile intensive care unit will mean that a greater number of patients will survive to reach hospital. This may result in problems connected with the availability of hospital beds. These problems might be solved in part if a significant number of patients with acute myocardial infarction were discharged after a shorter period in hospital than is usually advocated (Brit. med. J., 1968). A unique opportunity of assessing the effects of early discharge presented itself, since it had been the practice of one general medical unit in this hos-pital to discharge patients much earlier than is usually considered advisable. PATIENTS AND METHODS With the exception of one patient whose record could not be traced all records were examined of those patients aged 70 years or less who were admitted to this unit with an acute myocardial infarction during the period June 1962 to December 1964. In all, 185 patients sur-vived their period in hospital. The diagnosis of myo-cardial infarction was established by the following criteria: (a) pathological Q waves on the electrocardio-gram accompanied by a rise of the ST segment and sub-sequent T wave inversion, or (b) changes in the ST
A.A.J. Adgey (1969) studied this question.