Ptah
A Decision Support System for Nosocomial Infection Therapy

Purpose

Nosocomial (or hospital-acquired) infection is a disease that develops during an admission to hospital and is a consequence of treatment, procedures of treatment or work of hospital staff. About 5 to 10% of patients admitted to hospital develop a nosocomial infection, and 3% of them die of consequences. Particularly dangerous initiators of nosocomial infections are aggressive bacteria that are developed and transmitted within the hospital, where they may reach a high level of resistance to antibiotics. Such bacteria have to be identified as soon as possible, and their further development must be discontinued by appropriate procedures.

Ptah is a decision support system, designed as an on-line tool for physicians. It performs four types of analyses of microbiological findings regarding bacterial resistance to antibiotics and effectiveness of antibiotics. The analyses provide information that supports the physician in making decisions about the therapy of nosocomial infections.

Ptah is aimed at improving the quality of decision making and therapeutic procedures, and possibly decreasing the costs of antibiotic treatment. For this purpose, it

provides an overview of bacterial resistance in the hospital,
allows an early detection of highly resistant bacteria in different organizational units and/or related to different therapeutic activities,
compares microbiological findings of a patient with similar cases in the past, decreasing the possibility of an inappropriate decision,
provides an overview of the effectiveness of antibiotics, and
allows an early identification of ineffective antibiotics.

Functionality

Ptah is based on a chronologically organized database of infections and therapies. The system extracts microbiologic findings of one or more patients from the database and constructs time series of bacterial resistance vectors. The time series is constructed according to the requirements defined by the physician. The requirements include the selection of bacteria and antibiotics, and some constraints, such as date of sample, patient's date of birth, and type of therapeutical activity.

Ptah provides four methods for the analysis of time series:

These methods are relatively simple, but flexible. They facilitate the investigation of the decision-making problem from various viewpoints, and provide information that is both relevant and comprehensible for the physician. Special attention is given to:

Visualization:: The results of all the analyses are presented graphically.
Comprehensibility:: The methods were carefully selected so that they involve only data and concepts that are familiar to and commonly used by physicians.
Simplicity and uniformity:: To facilitate easy learning and use of the system, the methods are controlled by a small and uniform set of parameters.

Development

Ptah was developed at the Department of Knowledge Technologies, Jožef Stefan Institute, within the R&D project Computerized Information System for Supporting Therapeutic Activities in Health Care (1994-96). The project was co-financed by

Infonet, d.o.o., Kranj and
the Ministry of Science and Technology of the Republic of Slovenia.

Medical expertise and databases were provided by the General Hospital Jesenice.

The principal developer of the system was Marko Bohanec.

Implementation

Ptah is implemented in Borland Delphi 1.0 and runs under Microsoft Windows 3.1, 95, and NT. The user interface is in the Slovene language.

Publications

Bohanec, M., Rems, M., Urh, B.: Snovanje informacijskega sistema za podporo terapevtskih aktivnosti pri hospitalnih infekcijah, Računalniška analiza medicinskih podatkov CADAM-95 (ed. Lavrač, N.), IJS Scientific Publishing IJS-SP-95-1, ISBN 86-80023-23-X, pp. 64-75, Bled, 1995.

Bohanec, M., Rems, M., Slavec, S., Urh, B.: Decision support of nosocomial infection therapy, Medical Informatics Europe MIE 96 (eds. Brender, J., Christensen, J.P., Scherrer, J.-R., McNair, P.), Amsterdam: IOS Press, ISBN 90 5199 278 5, pp. 599-603, 1996.

Bohanec, M., Rems, M., Slavec, S., Urh, B.: Ptah: A system for supporting nosocomial infection therapy, 12th European Conference on Artificial Intelligence ECAI 96, Workshop on Intelligent Data Analysis in Medicine and Pharmacology IDAMAP 96, pp. 7-10, Budapest, 1996.

Bohanec, M., Kramar, Z., Rems, M., Slavec, S., Urh, B.: Podpora terapevtskih aktivnosti pri hospitalnih infekcijah, Zbornik prispevkov s Kongresa Društva za medicinsko informatiko Slovenije MI-96, ISBN 961-90394-0-8, pp. 121-128, 1996.

Rems, M., Bohanec, M., Urh, B., Kramar, Z.: Implementation problems of decision support system for nosocomial infection, Informatica Medica Slovenica 3(1,2,3), pp. 97-100, ISSN 1318-2129, 1996.