Purpose: To quantify risk for the occurrence of hyperglycemia requirin
g initiation of hypoglycemic therapy in patients treated with oral glu
cocorticoids. Patients and Methods: A case-control study of enrollees
in the New Jersey Medicaid program 35 years of age or older. The 11855
case patients had newly initiated treatment with a hypoglycemic agent
(oral or insulin) between 1981 and 1990. The 11855 controls represent
ed a random sample of other Medicaid enrollees. Results: In patients u
sing oral glucocorticoids, the estimated relative risk for development
of hyperglycemia requiring treatment was 2.23 (95% confidence interva
l, 1.92 to 2.59) as compared with nonusers. Risk increased with increa
sing average daily steroid dose, in hydrocortisone-equivalent milligra
ms; the odds ratio was 1.77 for 1 to 39 mg/d, 3.02 for 40 to 79 mg/d,
5.82 for 80 to 119 mg/d, and 10.34 for 120 mg/d or more. The estimated
effects persisted after adjustment for a variety of potentially confo
unding demographic, health service utilization, and medication use var
iables. Conclusion: The findings of this population-based study quanti
fy the risk of developing hyperglycemia requiring hypoglycemic therapy
after oral glucocorticoid use. The magnitude of risk increases substa
ntially with increasing glucocorticoid dose. These findings demonstrat
e the utility of large-scale health claims databases in defining the r
isk of important adverse drug effects.