Diagnostic sensitivity of extended renal and hematologic criteria to define hemolytic uremic syndrome.

Abstract

The usual definition of Shiga toxin-producing Escherichia coli hemolytic uremic syndrome (STEC-HUS) is based on the presence of anemia, thrombocytopenia, and elevated serum creatinine levels, with or without proteinuria and/or hematuria. The strict definition only considers elevated serum creatinine levels as a renal criterion. The extended definition maintains flexible renal criteria, although it replaces anemia with hemolysis and considers a sharp drop in platelet count as an indicator of platelet consumption. The objective of this study was to estimate and compare the diagnostic sensitivity of these definitions in patients with STEC-HUS as hospital discharge diagnosis. Retrospective review of medical records of HUS patients. Sensitivity and positive predictive value, with their corresponding 95% confidence intervals (CIs), were estimated for the 3 definitions based on a discharge diagnosis of STEC-HUS (reference diagnosis). The McNemar test was used. Out of 208 patients, 107 (51.4%), 133 (63.9%), and 199 (95.6%) were identified with the strict, usual, and extended definition, respectively. Sensitivity was lower for the strict definition (51.4%; 95% CI: 44.8-58.3), intermediate for the usual definition (63.9%; 95% CI: 56.9-70.4), and higher for the extended one (95.6%; 95% CI: 91.6-97.8); (p< 0.001). The different STEC-HUS definitions showed significant differences in diagnostic sensitivity. The extended definition reached a sensitivity above 95%, so its generalized use may help to reduce diagnostic delays.