AI systems in clinical practice

Laboratory systems
GermWatcher
Infection Control surveillance
developed by clinical domains keywords
Washington University, St. Louis Infection control in hospitals Rule-bases systems, alerts
location commissioned status
Barnes Hospital, Jewish Hospital, St. Louis, Missouri February 1993 In routine clinical use
description
The GermWatcher/GermAlert family of expert systems is designed to support Infection Control specialists in detecting, tracking and investigating infections in hospitalized patients.

GermWatcher is an expert system that monitors microbiology culture data from a hospital's laboratory system, identifies those cultures which represent nosocomial infections and reports them to the US National Centers for Disease Control and Prevention (CDC). The system includes a rulebase modelled on modelled on local hospital infection control guidelines and the CDC National Nosocomial (hospital-acquired) Infection Surveillance System (NISS) culture-based definitions for nosocomial infections. The NNIS provides a set of national standards for nosocomial infections so that national infection rates can be monitored and minimised. Nnosocomial infections represent a significant cause of prolonged inpatient days and additional hospital charges.

GermWatcher has been deployed at Barnes and Jewish Hospital, a large tertiary-care teaching hospital in St. Louis, since February 1993. It was later deployed at neighboring Jewish Hospital in July 1995.

Languages/Shells Used: generalized expert system shell (the GermWatcher Engine), CLIPS, Sybase ISQL scripts, Bourne shell scripts.

references

Kahn MG, Steib SA, Dunagan WC, Fraser VJ. Monitoring expert system performance using continuous user feedback. J Am Med Inform Assoc. 1996 May-Jun;3(3):216-23.

[PubMed]   [PubMed Central]

" OBJECTIVE: To evaluate the applicability of metrics collected during routine use to monitor the performance of a deployed expert system. METHODS: Two extensive formal evaluations of the GermWatcher (Washington University School of Medicine) expert system were performed approximately six months apart. Deficiencies noted during the first evaluation were corrected via a series of interim changes to the expert system rules, even though the expert system was in routine use. As part of their daily work routine, infection control nurses reviewed expert system output and changed the output results with which they disagreed. The rate of nurse disagreement with expert system output was used as an indirect or surrogate metric of expert system performance between formal evaluations. The results of the second evaluation were used to validate the disagreement rate as an indirect performance measure. Based on continued monitoring of user feedback, expert system changes incorporated after the second formal evaluation have resulted in additional improvements in performance. RESULTS: The rate of nurse disagreement with GermWatcher output decreased consistently after each change to the program. The second formal evaluation confirmed a marked improvement in the program's performance, justifying the use of the nurses' disagreement rate as an indirect performance metric. CONCLUSIONS: Metrics collected during the routine use of the GermWatcher expert system can be used to monitor the performance of the expert system. The impact of improvements to the program can be followed using continuous user feedback without requiring extensive formal evaluations after each modification. When possible, the design of an expert system should incorporate measures of system performance that can be collected and monitored during the routine use of the system. "

Kahn MG, Steib SA, Spitznagel EL, Claiborne DW, Fraser VJ. Improvement in user performance following development and routine use of an expert system. Medinfo. 1995;8 Pt 2:1064-7.

[PubMed]   []

" Hospital-acquired (nosocomial) infections represent a significant cause of prolonged inpatient days and additional hospital charges. In many hospitals, infection control nurses manually review positive microbiology culture results to monitor the incidence and prevalence of potential nosocomial infections. We have developed an expert system called GermWatcherTM, which uses the United States Centers for Disease Control and Prevention National Nosocomial Infection Surveillance criteria to classify microbiology results as potential nosocomial infections. In February 1993, we deployed GermWatcher at a large tertiary-care teaching hospital. In July 1993, we implemented a revised version of GermWatcher. With each version, we performed an evaluation of the program by comparing its electronic classification of positive culture results to the paper-based manual classification performed by three infection control nurses and one Infectious Disease specialist (gold standard). In the present study, we focus not on changes in the performance of the expert system, but on changes in performance among the infection control nurses. We found significant improvement in agreement and accuracy in the manual classification of cultures by the infection control nurses in the second evaluation compared to the first evaluation. We attribute this improved manual performance to the development of the expert system's rule base throughout the two evaluation phases and to the use of the expert system in the nurses' daily activities. "

Kahn MG, Steib SA, Fraser VJ, Dunagan WC. An expert system for culture-based infection control surveillance. Proc Annu Symp Comput Appl Med Care. 1993;:171-5.

[PubMed]   []

" Hospital-acquired infections represent a significant cause of prolonged inpatient days and additional hospital charges. We describe an expert system, called GERMWATCHER, which applies the Centers for Disease Control's National Nosocomial Infection Surveillance culture-based criteria for detecting nosocomial infections. GERMWATCHER has been deployed at Barnes Hospital, a large tertiary-care teaching hospital, since February 1993. We describe the Barnes Hospital infection control environment, the expert system design, and a predeployment performance evaluation. We then compare our system to other efforts in computer-based infection control. "

contact links

Washington University School of Medicine
Department of Internal Medicine
Division of Medical Informatics
660 South Euclid Campus
Box 8005 St. Louis
Missouri 63110 USA.

 bullet  Medical Informatics at Washington University in St. Louis  bullet  Barnes and Jewish Hospitals, St. Louis  bullet  BJC Healthcare
acknowledgements

Archive of AI systems in clinical practice previously administered by Enrico Coiera. Used with permission. Maintained and extended since 2001 by OpenClinical.

Entry on archive: November 15 1995
Last main update: November 15 1995
Search this site
 

 

Privacy policy User agreement Copyright Feedback

Last modified:
© Copyright OpenClinical 2002-2011