Methods
We analyzed data from 37 patients admitted from 1981 to 1991 who fulfilled four criteria (age >16 years, microangiopathic hemolytic anemia, creatinine level >150 µmol/L [>1.7 mg/dL], and no artificial heart valve). Three outcome variables were studied (survival vs death, recurrence vs no recurrence, and chronic renal failure vs no chronic renal failure).

Results
Eleven (30%) of the patients died, 10 (27%) needed dialysis, five (14%) developed chronic renal failure, and nine (24%) had recurrent episodes. Patients who presented with colitis did not die or have recurrences, but they developed chronic renal failure as often as other patients. Patients with hemolytic-uremic syndrome secondary to other diseases had the worst survival and the most recurrences. Those without any triggering factor (primary cases) were in between. In multivariate analysis, hemolytic-uremic syndrome secondary to colitis, a higher white blood cell count at admission, and a high maximum mean arterial pressure were associated with good survival prognosis.

Conclusions
The persistence of the trigger of adult hemolytic-uremic syndrome sets the stage for outcome. If the trigger is transient (such as Escherichia coli colitis), the disease will not recur and is rarely lethal. If no trigger is apparent (primary hemolytic-uremic syndrome) or the trigger persists (systemic lupus erythematosus and cancer), the syndrome has a high mortality and often recurs. We suggest a new classification: (1) extrinsic hemolytic-uremic syndrome: (a) toxic, (b) infectious; (2) intrinsic hemolytic-uremic syndrome: (a) primary, (b) secondary. The use of this classification, combined with simple data obtained at presentation and a further division of the cause as transient or persistent and irreversible, may improve the selection of therapy.

(Arch Intern Med. 1995;155:2077-2084)