Improving the Quality of Patient Care by Using a Clinical Expert System. (CLEOS)

Aim: To investigate the quality of history taking by physician and computer-based system. Patients: 100 inpatients presenting at the RBK for the first time and treated in the departments of nephrology and cardiology. Methods: The information obtained by the computer based system is compared with the information acquired by conventional history taking. Study endpoint is the comparison of historical data organized according to the elements in a standard medical history on a patient-by-patient basis. Study procedure History taking is performed by physicians according to the guidelines of the RBK. Within 2 days thereafter the patient is interviewed with help of the CLEOS system with the support of a study nurse.

Aim: To compare the completeness of history taking by physicians and by a computerized system interacting directly with the patients. Historical data recorded in the patient record, including the discharge letter, were compared with the history acquired by computer-based interview of each patient. Patients: 100 inpatients presenting at the RBK for the first time and treated in the departments of nephrology and cardiology. Methods: The information obtained by the computer based system is compared with the information acquired by conventional history taking. Study endpoint is the comparison of historical data organized according to the elements in a standard medical history on a patient-by-patient basis. Study procedure After informed consent of the patients the procedure of history taking is performed by the physician according to the guidelines of the RBK. Within 2 days thereafter the patient is interviewed with help of the CLEOS system [a computer-based, history-taking program] with the support of a study nurse. The physician will be informed about the results immediately after completion of the questionnaire. Data of medical significance were extracted from the official hospital chart and the computerized history by clinically experienced physicians and tabulated on a patient-by-patient basis. data elements were compared across each category of the history for each patient interviewed. The relevance of the differences of the information obtained by the two systems was rated by an independent reviewer.

All patients who were admitted to the departments of nephrology or cardiology in a tertiary hospital in Germany. The intervention was use of an expert system to acquire medical histories by direct interview of patients. Description of the software program - The program tested in this study consisted of a data acquisition [history-taking] component and a data analysis component. The data acquisition component was constructed on the basis of established principles of pathophysiology. Medical knowledge was formalized as software algorithms that were machine-readable by representing the knowledge as branched chain decision trees.

Methods The intervention was use of an expert system to acquire medical histories by direct interview of patients. Description of the software program - The program tested in this study consisted of a data acquisition [history-taking] component and a data analysis component. The data acquisition component was constructed on the basis of established principles of pathophysiology. Medical knowledge was formalized as software algorithms that were machine-readable by representing the knowledge as branched chain decision trees.

Detection of Medical Problems (by Number of Problems Reported by Computer Assisted History, That Were Not Reported by Physician Taken History)

Data were extracted from hospital charts by to experienced physicians.; data from the computer histories were extracted by a senior physician, who tabulated comparisons between the 2 sets of records. We used the number of problems reported by Computer Assisted History that were not reported by Physician. Nurses at Robert Bosch Krankenhaus do not take medical histories in regard to allergies or adverse drug reactions. Pharmacists make no entries into charts and have no separate records of drug allergies or history of adverse drug reactions. Data on these issues either are obtained only by physician interview of the patient.

Inclusion Criteria: All inpatients in a hospital environment Exclusion Criteria: Inability to give an informed consent