Graphical Analysis of Laboratory Data in the Differential Diagnosis of Cholestasis: A Computer-Assisted Prospective Study
| Autor: | J. Baier, W. Nathusius, W. Gerhardt, M. Glocke, G. Börsch |
|---|---|
| Rok vydání: | 1988 |
| Předmět: |
medicine.medical_specialty
Pathology Bilirubin Clinical Biochemistry education Immunoglobulins Extrahepatic Cholestasis Gastroenterology Diagnosis Differential chemistry.chemical_compound Cholestasis Lactate dehydrogenase Internal medicine medicine Humans Diagnosis Computer-Assisted Prospective Studies medicine.diagnostic_test business.industry Clinical Laboratory Techniques Biochemistry (medical) General Medicine medicine.disease Enzymes chemistry Erythrocyte sedimentation rate Alkaline phosphatase Differential diagnosis business Viral hepatitis Blood Chemical Analysis |
| Popis: | Data on 15 laboratory analytes obtained in 145 prospectively investigated cholestatic patients with viral hepatitis, chronic intrahepatic cholestasis and extrahepatic biliary obstruction were submitted to a computer-based graphical evaluation using probabilistic test analysis. This revealed a marginal utility for alkaline phosphatase, gamma-glutamyltransferase and the direct/total bilirubin ratio at specific cut-off points for the exclusion of extrahepatic cholestasis (PVneg 90%-100%). Aspartate aminotransferase and alanine aminotransferase values with cut-off points at 200 U/l and 300 U/l, respectively, were powerful discriminators between acute viral hepatitis and the other disease categories, while lactate dehydrogenase, erythrocyte sedimentation rate and the ratios gamma-glutamyltransferase/alanine aminotransferase as well as total bilirubin/gamma-glutamyltransferase were useful at specific cut-off points indicating the absence of this diagnosis (PVneg 92%-100%). An aspartate aminotransferase/alanine aminotransferase ratio above 1.5 and serum gamma-globulin concentrations above 20 g/l strongly suggested cholestasis due to chronic parenchymal liver disease (PVpos 92% and 90%, respectively). This graphical approach to laboratory data analysis enhances the understanding of the interrelations between cut-off points and sensitivity, specificity and predictive values and also of the influence of disease prevalence on disease prediction. It also adds to present knowledge by demonstrating the clinical relevance of several readily available, albeit rarely utilized diagnostic analytes. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|