Laboratory and clinical variables to predict outcome in hemolytic-uremic syndrome
P. L. Havens, P. P. O'Rourke, J. Hahn, J. Higgins and A. M. Walker
Department of Pediatrics, Children's Hospital, Boston.
To develop a guide to the prognosis of children with hemolytic-uremic
syndrome, we reviewed the medical records of 78 patients with this
diagnosis seen at The Children's Hospital, Boston, from 1976 through 1986.
Two outcome groups were defined as follows: a "good outcome" group, which
contained 6 patients with no serious sequelae at hospital discharge, and a
"bad outcome" group, which contained 12 patients who died, had chronic
renal failure, or had central nervous system sequelae at hospital
discharge. Differences between the two groups in routine laboratory tests
available within 48 hours of admission were identified by bivariate
analysis. Using serum calcium (less than or equal to 2 mmol/L) plus urine
output (less than 0.4 mL/kg/h) over a 24-hour period as a test to predict
outcome, we identified patients who died, had chronic renal failure, or had
serious central nervous system sequelae, with 75% sensitivity, 98%
specificity, and 90% positive predictive value.