Predictive models for chronic renal disease using decision trees, naïve bayes and case-based methods

Data mining can be used in healthcare industry to “mine” clinical data to discover hidden information for intelligent and affective decision making. Discovery of hidden patterns and relationships often goes intact, yet advanced data mining techniques can be helpful as remedy to this scenario. This thesis mainly deals with Intelligent Prediction of Chronic Renal Disease (IPCRD). Data covers blood, urine test, and external symptoms applied to predict chronic renal disease. Data from the database is initially transformed to Weka (3.6) and Chi-Square method is used for features section. After normalizing data, three classifiers were applied and efficiency of output is evaluated. Mainly, three classifiers are analyzed: Decision Tree, Naïve Bayes, K-Nearest Neighbour algorithm. Results show that each technique has its unique strength in realizing the objectives of the defined mining goals. Efficiency of Decision Tree and KNN was almost same but Naïve Bayes proved a comparative edge over others. Further sensitivity and specificity tests are used as statistical measures to examine the performance of a binary classification. Sensitivity (also called recall rate in some fields) measures the proportion of actual positives which are correctly identified while Specificity measures the proportion of negatives which are correctly identified. CRISP-DM methodology is applied to build the mining models. It consists of six major phases: business understanding, data understanding, data preparation, modeling, evaluation, and deployment.

Internal validity of the Swedish Maternal Health Care Register

Background: The Swedish Maternal Health Care Register (MHCR) is a national quality register that has been collecting pregnancy, delivery, and postpartum data since 1999. A substantial revision of the MHCR resulted in a Web-based version of the register in 2010. Although MHCR provides data for health care services and research, the validity of the MHCR data has not been evaluated. This study investigated degree of coverage and internal validity of specific variables in the MHCR and identified possible systematic errors. Methods: This cross-sectional observational study compared pregnancy and delivery data in medical records with corresponding data in the MHCR. The medical record was considered the gold standard. The medical records from nine Swedish hospitals were selected for data extraction. This study compared data from 878 women registered in both medical records and in the MHCR. To evaluate the quality of the initial data extraction, a second data extraction of 150 medical records was performed. Statistical analyses were performed for degree of coverage, agreement and correlation of data, and sensitivity and specificity. Results: Degree of coverage of specified variables in the MHCR varied from 90.0% to 100%. Identical information in both medical records and the MHCR ranged from 71.4% to 99.7%. For more than half of the investigated variables, 95% or more of the information was identical. Sensitivity and specificity were analysed for binary variables. Probable systematic errors were identified for two variables. Conclusions: When comparing data from medical records and data registered in the MHCR, most variables in the MHCR demonstrated good to very good degree of coverage, agreement, and internal validity. Hence, data from the MHCR may be regarded as reliable for research as well as for evaluating, planning, and decision-making with respect to Swedish maternal health care services.