| Type of AE | Median: (25th to 75th percentiles) (AEs per 100 patient years) |
| Cardiovascular (dyslipidemia, water and electrolyte imbalance, oedema, renal and heart dysfunction, hypertension) | 15 (3 to 28) |
| Infectious (viral, bacterial, skin infections) | 15 (3 to 15) |
| Gastro-intestinal (peptic ulcer disease, pancreatitis) | 10 (4 to 20) |
| Psychological and behavioural (minor mood disturbances, steroid psychosis) | 9 (2 to 236) |
| Endocrine and metabolic (glucose intolerance and diabetes, fat redistribution, interference with hormone secretion) | 7 (3 to 34) |
| Dermatological (cutaneous atrophy, acne, hirsutism, alopecia) | 5 (2 to 80) |
| Musculoskeletal (osteoporosis, osteonecrosis, myopathy) | 4 (3 to 9) |
| Ophtalmological (glaucoma, cataract) | 4 (0 to 5) |
This table summarises reported AEs in studies (n = 18) of the general search of patients using GCs (n = 963) for a rheumatic disease. Only those studies of patients who were using GCs up to 30 mg prednisolone or equivalent and reporting dichotomous AE outcomes were included in the data of the table, which was used as introductory information for the taskforce. Raw data, not corrected for disease activity, co-morbidity and the frequency of AEs in the contrast group, if present, were used. So, not all AEs can be specifically attributed to the use of GCs; common events may be overestimated and less common ones underestimated. For instance, cardiovascular events are poorly correlated with GC-use. Types of AEs were divided into different groups (as has been published before25) and per group AEs per 100 patient years were derived by dividing the number of AEs by the duration of follow-up in years, times 100. The mean daily GC-dose was 8 mg, and the average duration of the studies was 19.6 months.