PhD

dissertationAt the Salt Lake Latter-Day Saints (LDS) Hospital, we developed an interface between the microbiology laboratory computer system and a central hospital computer system (HELP). This system contains an integrated data base including information from most medical areas. A translation program was created to transform the microbiology data from the laboratory file structure to a hierarchical data structure on the HELP system. The translated microbiology data could then be reviewed from any Nursing; division or ICU in the hospital within seconds after it was entered in the laboratory computer. We next developed a knowledge base, with the help of infectious disease experts, and linked it to a computer Infectious Disease monitoring system. The knowledge base is automatically activated when specific microbiology data are entered into the patient’s computer file (Data-driven). The Computer Infectious Disease Monitor (CIDM) was designed to detect patients with the following condition: 1) a hospital-acquired infection 2) and infection at a sterile body site, 3) an infection by antibiotic resistant organisms, 4) not receiving appropriate antibiotics, 5) could be receiving a less expensive antibiotic, 6a) a “reportable†disease and 7) receiving prophylactic antibiotics longer than necessary. A study comparing the nosocomial infections identified by the CIDM with the nosocomial infections identified by the Infection Control Practitioners (ICPs) at LDS Hospital during a 2 month period was conducted. A physician retrospective chart review (PRCR) was performed on 217 patients identified as having a nosocomial infection by either the CIDM or the ICPs. The CIDM identified 88% of the patients with nosocomial infections and the ICPs 74% (p<0.0002). The CIDM identified 78% of the total nosocomial infections and the ICPs 68% (p<0.032). By combining the infections found by the CIDM with those found by the ICPs, 88% of the infections were identified. The CIDM identified 84% of the infections with positive microbiology cultures and the ICPs 74%. By combining the infections identified by the CIDM with the ICPs, 95% of the infections with positive cultures were identified. Thus, the ICPs augmented by the CIDM identified. Thus, the ICPs augmented by the CIDM identified significantly more nosocomial infections (p