Measuring Hospital-Wide Mortality-Pitfalls and Potential

Area of Science:

• Healthcare Quality Measurement
• Health Services Research
• Medical Informatics

Background:

• Risk-adjusted hospital-wide mortality is a proposed indicator of system-level healthcare quality.
• The hospital standardized mortality ratio (HSMR) is one such measure, publicly reported internationally but not in the US.
• Existing risk-adjusted mortality measures have limitations for direct interhospital comparison.

Purpose of the Study:

• To review the potential uses of risk-adjusted hospital mortality measures.
• To evaluate the suitability of current methods for hospital quality assessment and payment.
• To provide recommendations for the effective application of mortality data in healthcare.

Main Methods:

• Literature review of risk-adjusted mortality measures and their applications.
• Analysis of the strengths and limitations of hospital-wide mortality indicators.
• Examination of disaggregation strategies for condition- and service-line-specific mortality.

Main Results:

• Available risk-adjusted mortality measures are not suitable for interhospital comparisons or rankings.
• These measures should not be used for pay-for-performance or value-based purchasing initiatives.
• Hospital-wide mortality is an imprecise quality indicator, but specific breakdowns can guide improvement.

Conclusions:

• Risk-adjusted mortality data can serve as an initial alert for significantly higher-than-expected mortality rates, prompting further investigation.
• Monitoring both observed and expected mortality rates over time is crucial to validate quality improvements and account for case-mix changes.
• Disaggregating mortality data by condition or service line is more effective for targeted quality improvement efforts than broad hospital-wide comparisons.