Related Experiment Videos
Project IMPACT: results from a pilot validity study of a new observational database
Suzanne F Cook1, Wendy A Visscher, Connie L Hobbs
1Glaxo SmithKline, Research Triangle Park, NC 27709, USA.
Critical Care Medicine
|December 17, 2002
Summary
Project IMPACT database data accuracy was evaluated. Good agreement was found for discrete data, but poorer agreement for continuous and coded items, suggesting its value for intensive care unit (ICU) performance measurement.
Area of Science:
- Healthcare Informatics
- Clinical Data Management
- Intensive Care Unit (ICU) Performance Measurement
Background:
- Project IMPACT is a database system used by hospitals to measure and improve intensive care patient care.
- Data entered includes patient demographics, diagnoses, treatments, and outcomes.
- Assessing data accuracy is crucial for the database's utility in performance measurement.
Purpose of the Study:
- To evaluate the accuracy of data within the Project IMPACT database.
- To determine the usefulness of Project IMPACT for measuring ICU performance and patient outcomes.
Main Methods:
- A pilot study involved independent data abstraction from random samples of ICU patient records at two hospitals.
- Data from patient records were compared against existing data in the Project IMPACT database.
- Abstraction was conducted by an unaffiliated, trained abstractor onsite at each hospital.
Main Results:
- Good agreement (≥80%) was observed for discrete data items, such as patient type.
- Poorer agreement (<80%) was noted for continuous data (e.g., urine output) and coded items requiring judgment (e.g., admission reason).
- Agreement rates were high regardless of data capture method (electronic vs. manual), with higher agreement for electronically downloaded data.
Conclusions:
- The Project IMPACT database demonstrates good internal validity for most abstracted variables.
- The database is a valuable resource for ICUs to collect and evaluate treatment and patient outcome information.
- Electronic data capture enhances the accuracy of data within the Project IMPACT system.