Eugene S. Pearlman, MD (E-mail: czaroflabs@...); Melissa Haygood, MT(ASCP); Stanley Sprei, MD.
Department of Pathology, Lourdes Hospital, Paducah, Ky.
Context: The problem of postverification errors with wrongly reported results is a vexing one for the clinical laboratory. Efforts to quantify the problem depend on notification of the clinical laboratory by nonlaboratory personnel. We attempted an independent assessment by detecting discrepant results within serial troponin assays from inpatients undergoing cardiac catheterization.
Design: We evaluated data from the electronic medical records of 385 consecutive inpatients from the first quarter of 2007 who underwent caridac catheterization. Of these, 343 (89%) representing 1352 inpatient days had 2 or more troponin concentrations. Sampling was done from primary bar-coded tubes using an ultrasensitive troponin assay according to manufacturer's directions (Siemens, Tarrytown, NY). The laboratory had a delta check protocol in place. Expected variation was estimated from quality control data.
Results: Five patients had discrepant results in their precatheterization troponin results with an outcome that differed from a previous result by more than 5.5 SDs for a rate of 3.7 postverification error events per 1000 inpatient days. Regarding the patient days represented by this population as a sample of total patient days during the first quarter of 2007 and using a ratio estimation technique, the expected number of such events for this 3-month period (44 ± 7) greatly exceeded the actual number (8) coming to the attention of the laboratory during this time period.
Conclusions: The actual number of postverification errors may be significantly greater than the number communicated to the laboratory.
Invia email