Discriminant function analysis applied to laboratory tests in patients with hepatobiliary disease

Statistical approaches to assessing single and multiple outcome measures in dry eye therapy and diagnosis

Dry eye is a multifactorial disease which would require a broad spectrum of test measures in the monitoring of its treatment and diagnosis. However, studies have typically reported improvements in individual measures with treatment. Alternative approaches involve multiple, combined outcomes being assessed by different statistical analyses. In order to assess the effect of various statistical approaches to the use of single and combined test measures in dry eye, this review reanalyzed measures from two previous studies (osmolarity, evaporation, tear turnover rate, and lipid film quality). These analyses assessed the measures as single variables within groups, pre- and post-intervention with a lubricant supplement, by creating combinations of these variables and by validating these combinations with the combined sample of data from all groups of dry eye subjects. The effectiveness of single measures and combinations in diagnosis of dry eye was also considered.

Classification of turkeys as low or high volume semen producers using discriminant analysis

Medium White turkeys in Generation 10 of divergent selection for semen ejaculate volume (SEV) were classified into lines that produce either low or high volumes of semen using a multivariate statistical technique, discriminant analysis. The discriminant function was based on records of hen fertility, embryonic mortality in the early and late incubation periods, and the incidence of pipped eggs. A hit ratio, the proportion of birds correctly categorized as low or high, of .60 was obtained. The discriminant function was subsequently used to classify birds in Generation 12. A misclassification rate of .28 and .48 was obtained for birds in the low and high lines, respectively. It was concluded that the relatively low rates of misclassification indicate a potential method for identifying male and female turkeys as belonging to low or high SEV lines, based on fertility and incubation records.

Methods for knowledge extraction from a clinical database on liver diseases

We performed exploratory data analysis (EDA) to examine the hidden structure in liver disease data. The purpose was to demonstrate the potential of statistical techniques for extracting knowledge from an active HIS (hospital information system) database with decision support. The goal is to give strong support to the creation of new rules or "tuning" of old rules in the knowledge base. This would facilitate utilization of large patient databases, now commonly available, to help build/update decision support systems for improved patient care. Several statistical techniques were investigated. Stepwise discriminant analysis was found to be a good method in discriminating among different disease classes. Results showed that classification strength of a few (3) variables was similar to all the available (19) variables. Other important issues in the work are treatment of missing values as well as atypical values in medical databases. In estimating missing values we utilized both statistical methods and artificial intelligence approaches. Both these approaches were promising in the estimation of missing values. The study showed that several statistical approaches are possible for knowledge extraction from clinical data collected retrospectively.

Advances in the application of biochemical tests to diseases of the liver and biliary tract: their role in diagnosis, prognosis, and the elucidation of pathogenetic mechanisms

Despite the biochemical complexity of the liver, few laboratory tests provide discriminatory diagnostic information in patients with hepatobiliary disease. Recent efforts have concentrated upon tests which assess the function of the liver, the severity of the disease state, and underlying pathological processes. Bile Acids: The emergence of facile technology and widespread application has brought the realization that these assays are not as sensitive in detecting liver disease as previously believed, although the cholate/chenate ratio may be useful in distinguishing cholestasis from chronic liver disease. The presence of unusual bile acids in serum or urine may be helpful in some cases. Drug Metabolism: A number of tests provide good evidence about liver function, hepatic blood flow and portal shunting, but the aminopyrine breath tests is the most useful, giving prognostic information in acetaminophen overdose and alcoholic liver disease. The antipyrine half-life identifies surgical cases at risk from poor hepatic function. Proteins and Immunochemical Tests: Interest has developed in plasma proteins such as prealbumin and retinol-binding protein to monitor hepatic protein synthetic function. Secretory IgA is more elevated in biliary tract disease, unlike the native protein which is increased principally in cirrhosis. Type III procollagen can be measured in serum, and correlates with the activity of collagen synthesis and the degree of fibrosis in biopsy samples. Reye's Syndrome: Biochemical tests play an essential role in diagnosis of this recently discovered disease. These will be presented and discussed.

Biochemical and haematological indicators of excessive alcohol consumption

Discriminant function analysis was used to determine the optimum combination of haematological and biochemical tests which gave the best discrimination between hospital patients with high and low alcohol intakes. We studied 265 patients with alcohol-related disease, 133 gastroenterology outpatients drinking less than 20 g of alcohol per day, and 104 patients with a variety of non-alcoholic liver disease. Values of mean cell volume (MCV), serum bilirubin, aspartate transaminase, serum alkaline phosphatase (AP) and gamma glutamyl transferase (gamma GT), serum albumin, serum globulin, and uric acid were determined in each patient. The best discrimination between the three groups of patients was provided by a combination of mean corpuscular volume, log10 gamma GT, and log10 serum alkaline phosphatase. In women, 92% of the high alcohol group, 100% of the low alcohol group, and 87% of the non-alcoholic liver disease were correctly allocated by the discriminant analysis. The corresponding figures for the men were 80%, 100%, and 71%. Thus, over 80% of patients with excessive alcohol intake were correctly allocated by the use of three simple laboratory tests.