Objectives: Since there have been no recent reviews of Addison's disease, we have undertaken a retrospective case-notes review of all identifiable cases in Nottingham to define the prevalence, incidence and causes of Addison's disease. We have also reviewed the criteria for interpretation of the short Synacthen test in diagnosis.
Design: A retrospective study of all patients coded for the diagnosis and admitted to Nottingham's hospitals between 1 April 1987 and 31 March 1993, identified by the hospital Information Services.
Patients: A total of 86 cases were identifiable of whom 66 were still alive and living in the town.
Results and conclusions: The calculated prevalence was 110 per million population. The cause was attributed to autoimmune destruction of the adrenal cortex in 81 (93%). There were two cases of metastatic malignancy and three unrelated cases of late onset adrenoleukodystrophy, but none were attributable to tuberculosis. Twenty-one new cases were diagnosed between 1987 and 1993. The calculated incidence was 5.6 per million per annum. The biochemical basis of the diagnosis was reviewed in these 21 patients and as a result firm criteria are suggested for the interpretation of the short Synacthen test; criteria for normality being baseline cortisol >250 and 30 minute peak >600 nmol/l, taking into account clinical circumstances.