Patient Safety Indicator rates track potential in-hospital complications following surgeries and childbirth across New York State counties. The Agency for Healthcare Research and Quality developed the methodology using SPARCS inpatient data starting in 2009. New York State health authorities calculate observed, expected, and risk-adjusted rates with 95% confidence intervals.
Use Cases
- Analyze trends in Observed Rate and PSI Value for specific PSI Codes like surgical complications over time by County.
- Compare risk-adjusted performance between counties by calculating differences between Observed Rate and expected benchmarks.
- Identify counties with high Numerator counts for specific adverse events to target public health interventions.
- Model the relationship between Denominator (population at risk) and Numerator (event count) to estimate baseline risk by geographic area.
Strengths
- Data spans from 2009 onward, providing over a decade of longitudinal analysis
- Measures include Observed Rate, expected rate, and risk-adjusted rates with 95% confidence intervals for statistical reliability
- County-level aggregation allows for geographic comparison of patient safety outcomes
Limitations
- Exact row count and sample size are unknown, limiting assessment of statistical power
- Data is limited to New York State, reducing national generalizability
- Relies on administrative coding (ICD-9-CM) which can contain inaccuracies
Provenance
- Source
- New York State Department of Health via health.data.ny.gov
- Collection Method
- Calculated from Statewide Planning and Research Cooperative System (SPARCS) inpatient data and AHRQ-provided US Census denominators.
- Time Range
- Beginning 2009
- Freshness
- Updated as of April 2025.
- Geography
- New York State, aggregated by County