Patient Safety Indicators: using administrative data to identify potential patient safety concerns.

Area of Science:

• Healthcare Quality Improvement
• Patient Safety Research
• Health Informatics

Background:

• In-hospital patient safety is a critical concern for healthcare quality.
• Existing methods for identifying patient safety events may be insufficient.
• There is a need for systematic tools to detect potential adverse events in hospitals.

Purpose of the Study:

• To develop Patient Safety Indicators (PSIs) for identifying potential in-hospital patient safety problems.
• To create a tool for targeted quality improvement efforts within healthcare institutions.

Main Methods:

• Utilized 2,400,000 discharge records from the 1997 New York State Inpatient Database.
• Developed PSI algorithms through systematic literature reviews and ICD-9-CM code book searches.
• Examined PSI event prevalence and associations with patient and hospital characteristics, length of stay, mortality, and charges.

Main Results:

• Developed 12 PSIs for distinct clinical situations and a summary measure.
• PSI event rates varied, with foreign bodies at 1.1 and birth traumas at 84.7 per 10,000 discharges.
• PSI events correlated with 2-3x longer stays, 2-20x higher mortality, and 2-8x higher charges.
• Associated factors included increasing age, high inpatient surgery volume, and high ICU bed percentage.

Conclusions:

• PSIs offer an efficient, user-friendly tool for identifying potential in-hospital patient safety issues.
• PSIs support targeted, institution-level quality improvement initiatives.
• PSIs can illuminate medical errors beyond mortality, aiding in the development of better error-reporting systems.