[Hospital standardised mortality ratios based on data from administrative registries. A pilot project]

Area of Science:

• Health Services Research
• Public Health
• Healthcare Quality Improvement

Background:

• Hospital standardised mortality ratios (HSMR) are integral to quality improvement initiatives.
• The applicability of HSMR computation using Danish administrative registries was previously unestablished.
• This study aimed to assess the feasibility of calculating HSMR from Danish registry data.

Purpose of the Study:

• To determine if HSMR can be computed using Danish administrative registries.
• To evaluate the variation in HSMR across Danish hospitals.
• To assess the utility of Danish registries for national healthcare quality monitoring.

Main Methods:

• Linked Danish hospital discharge and Civil Registration System registries.
• Identified 77 primary discharge diagnoses responsible for 80% of in-hospital deaths within 30 days.
• Calculated HSMR by comparing observed to expected deaths, stratified by diagnosis, age, gender, and admission type.

Main Results:

• Pneumonia (non-specified) was the leading diagnosis for 30-day in-hospital mortality.
• Crude mortality rates ranged from 5.7% to 6.3%.
• HSMR values showed minimal variation, ranging from 95 to 103 across hospitals.

Conclusions:

• Danish administrative registries are suitable for computing HSMR.
• HSMR demonstrated limited variability among hospitals with similar case-mixes.
• Findings support the use of HSMR for quality assessment in the Danish healthcare system.