Diabetes during diarrhea-associated hemolytic uremic syndrome: a systematic review and meta-analysis.

OBJECTIVE

To quantify the incidence of diabetes during the acute phase of diarrhea-associated hemolytic uremic syndrome (D + HUS) and to identify features associated with its development.

RESEARCH DESIGN AND METHODS

A systematic review and meta-analysis of articles assessing diabetes during D + HUS was conducted. Relevant citations were identified from Medline, Embase, and Institute for Scientific Information Citation Index databases. Bibliographies of relevant articles were hand searched. All articles were independently reviewed for inclusion and data abstraction by two authors.

RESULTS

Twenty-one studies from six countries were included. Only 2 studies reported a standard definition of diabetes; 14 defined diabetes as hyperglycemia requiring insulin. The incidence of diabetes during the acute phase of D + HUS could be quantified in a subset of 1,139 children from 13 studies (1966-1998, age 0.2-16 years) and ranged from 0 to 15%, with a pooled incidence of 3.2% (95% CI 1.3-5.1, random-effects model, significant heterogeneity among studies, P = 0.007). Children who developed diabetes were more likely to have severe disease (e.g., presence of coma or seizures, need for dialysis) and had higher mortality than those without diabetes. Twenty-three percent of those who developed diabetes acutely died, and 38% of survivors required long-term insulin (median follow-up 12 months). Recurrence of diabetes was possible up to 60 months after initial recovery.

CONCLUSIONS

Children with D + HUS should be observed for diabetes during their acute illness. Consideration should be given to long-term screening of D + HUS survivors for diabetes.