Measuring Hospital Performance Using Mortality Rates: An Alternative to the RAMR.

Area of Science:

• Healthcare quality assessment
• Biostatistics
• Hospital performance metrics

Background:

• Risk-adjusted mortality rate (RAMR) is widely used but complex due to hypothesis testing and confidence intervals.
• Unfamiliarity with hypothesis testing can lead to misinterpretation of hospital performance data.
• Current methods are insensitive to case volume, limiting their applicability.

Purpose of the Study:

• To propose a more defensible and understandable alternative to RAMR for hospital performance evaluation.
• To introduce a screening method that is easier to interpret and implement.
• To develop a performance metric sensitive to case volume.

Main Methods:

• Utilized upper-tail probability (UTP) for screening poor-performing hospitals and lower-tail probability for well-performing hospitals.
• Developed measures that do not require confidence intervals or hypothesis tests for computation or interpretation.
• Ensured the proposed measures are sensitive to the number of cases treated.

Main Results:

• Demonstrated the methodology using New York State Department of Health data for 10 Inpatient Quality Indicators (2009-2013).
• Found strong agreement between the proposed upper-tail probability (UTP) and the RAMR.
• Confirmed UTP as a viable alternative to RAMR for hospital performance screening.

Conclusions:

• The proposed UTP method is simpler to implement and easier to interpret than RAMR.
• The absence of confidence intervals enhances the method's interpretability for a wider audience.
• This screening approach is applicable to all hospitals, diseases, and conditions, irrespective of patient volume.