Table 5
Examples of Chicago Antimicrobial Resistance Project Electronic Surveillance Applications: Focus and Value
| Surveillance activities | |
| Rates of antimicrobial resistance | • Focus: Automated quantification of antibiotic-resistant isolates stratified by hospital, patient type, and time of acquisition |
| • Units of measure: Number of resistant isolates by month (using two different duplicate isolate rules) and by time of onset | |
| • Reports: On-demand trend report of resistant isolates, expressed per 1,000 patient days | |
| • Value: Before clinical data warehouse (CDW) 400 PH/yr required to trend five most common resistant organisms hospitalwide; with CDW, after the initial development of programing code (40 PH), clinical staff have on-demand access to reports | |
| Potential infections | • Focus: Automated identification of epidemiologically important positive microbiology data |
| • Units of measure: Daily list of individual patients including date of culture collection, organism, susceptibility, source of specimen, and patient location | |
| • Reports: On-demand daily line listing retrievable from intranet Web site | |
| • Value: Before CDW, 915 PH/yr required to create daily line listing; with CDW, after the initial development of programing code (80 PH), clinical staff have on-demand access to reports | |
| Blood stream infections | • Focus: Applied computer algorithms to categorize blood culture isolate as infection versus contamination and whether due to an infected intravascular device18 |
| • Units of measure: Number of patients with catheter-related or other bloodstream infection, expressed per 1,000 patient days | |
| • Reports: Incidence trend of these infections | |
| • Value: Before CDW, 452 PH/yr required to perform chart review of patients with positive blood cultures; with CDW, computer algorithm performs this decision-making task | |
| Central venous catheters (CVC) | • Focus: Identification of patient with central venous catheters (CVCs) using data mining software to predict patients by regression or decision tree analytic methods19 |
| • Units of measure: Number of patients with CVCs, patient characteristics | |
| • Report: Manuscript | |
| • Value: Demonstration of feasibility of electronic surveillance to predict presence of a CVC | |
| Antimicrobial consumption | • Focus: Automate quantification of antimicrobial use and stratified by hospital, unit, and route of administration |
| • Unit of measure20: Defined daily dose, therapy days, duration of therapy, and starts of new antibiotic courses | |
| • Report: On-demand access to trend of antibiotic use reports retrievable from intranet Web site | |
| • Value: Before CDW, 779 PH/yr would be expended to quantify use; with CDW, after the initial development of programing code (120 PH), clinical staff have on-demand access to reports offering unit-based analysis | |
| Quality improvement/patient safety activities | |
| Antimicrobial redundant prescribing practices | • Focus: Automated identification and access to information for clinical pharmacy staff to a daily listing of patient receiving two or more antibiotics classified as redundant (i.e., overlapping antimicrobial spectra)21 |
| • Unit of measure: Daily list of individual patients prescribed antibiotics; duplicate spectra antimicrobials highlighted | |
| • Report: On-demand daily line listing retrievable from intranet Web site | |
| • Value: Before CDW, surveillance for this type of medication error was not performed routinely; however, a pilot study of manual surveillance of one day's antimicrobial use took 10 PH (yearly projected time, 1.75 FTE) for case identification for one hospital; with CDW, 160 PH were used for design and development of application, which was then available on demand | |
| Cost accounting activities | |
| Cohort identification | • Focus: Identified cohort of patients with >5 discharge diagnoses (surrogate of severity of illness)22 23 |
| • Unit of measure: Population of discharge patients per year and count of individual patients' diagnoses | |
| • Report: List of patients meeting inclusion criteria | |
| • Value: Before CDW, we were unable to identify a cohort of patients with >5 discharge diagnoses (study inclusion criteria); with CDW, we are able to perform criterion-based queries in less than 15 minutes | |
| Resource utilization | • Focus: Identified line item use of resources for individual patients |
| • Unit of measure: Procedures, diagnostic tests, medications, patient bed occupancy | |
| • Report: Electronic list of patient encounters and resources utilized | |
| • Value: Estimated time saving of electronic vs manual abstraction of 1,500 medical records was 1,750 person hours | |
