Application domain: | Diagnosis-Therapy Index |
Source: | Infratest Burke |
Dataset size: | 180.000 facts |
Data format: | special code |
Systems Used: | Midos in Kepler |
Pointers: | S. Wrobel |
The data represents the results of questionnaires filled out at a sample of several hundred German hospitals. The questionnaires record for each patient general data (such as age, weight), historic pre-admission information, all the diagnoses that were made and when they applied, all treatments and when they applied, all other activities. In addition, information is available about the hospital and treatment success for each diagnosis.
The goal of the analysis is to discover unusual treatment or success patterns in the questionnaire data that could be used for optimization purposes.
The new algorithm Midos was applied to an initial sample of 1000 patients (30.000 facts) and uncovered irregularities in several interesting subgroups of the patient population that are now being investigated by Infratest.