Incidence of serious complications of corticosteroid therapy in respiratory disease. A retrospective survey of patients in the Brompton hospital.

A retrospectively controlled study is described of mortality and serious morbidity in 550 patients treated with corticosteroids and 499 controls. The overall incidence of side-effects was similar in both groups, but gastro-intestinal bleeding, diabetes, and mental disturbance early in treatment occurred more frequently in the corticosteroid group. Other individual side-effects did not occur significantly more often than in the controls, but there was an overall increased frequency of complications with increased dose, and also an increased mortality with increase in the maintenance dose. Gastro-intestinal bleeding was not related to a past or present chronic peptic ulcer, and was not commoner in treated patients. Reactivation of tuberculosis and exacerbation of mental disturbance did not occur. Weight gain occurred in 29% of patients, but appeared to be related closely to a satisfactory response to treatment. Hypertension occurred in 4% of the treated group. Overall mortality in the two groups was similar. Mortality due specifically to corticosteroids was difficult to assess: two patients died of haematemesis and three died with severe asthma during withdrawal of treatment, but there was no increase in mortality following physical stress in the treated series. Sudden death in asthmatics was not more frequent in corticosteroid-treated patients. All seven control patients who died suddenly with asthma were middleaged. Increase in duration of treatment was not associated with increased risk. Indeed the long-term, low-dose group had a lower incidence of complications than the rest and a mortality similar to that of the controls. A trial of corticosteroids for less than one month carried no risk in 41 patients.