Testing the construct validity of hospital care quality indicators: a case study on hip replacement

Area of Science:

• Healthcare Quality Measurement
• Orthopedic Surgery Outcomes
• Health Services Research

Background:

• Hospital quality indicators (HQIs) are increasingly used for care assessment.
• Dutch HQIs for hip replacements are primarily based on expert consensus and face validity.
• Limited understanding exists regarding the construct validity of these national HQIs.

Purpose of the Study:

• To evaluate the construct validity of national hospital quality indicators for hip replacements in the Netherlands.
• To assess the association between structure, process, and outcome indicators for hip replacement care.
• To determine if HQIs measuring the same clinical construct demonstrate expected correlations.

Main Methods:

• Utilized data from 100 Dutch hospitals over a two-year period.
• Assessed construct validity by analyzing relationships between structure, process, and outcome indicators.
• Employed chi-square statistics, bootstrapped Spearman correlations, and independent sample t-tests.

Main Results:

• Out of 28 hypothesized correlations, only three were significant in the expected direction.
• Hospitals with low wound infection rates had high scores for postoperative appointment scheduling and reduced homologous blood transfusions.
• Higher scores in complication meeting scheduling correlated with increased thrombosis prophylaxis provision.

Conclusions:

• Construct validity of national hospital quality indicators for hip replacement appears limited despite apparent face validity.
• Individual indicators may be valid, but overall conclusions from the indicator set require careful consideration.
• Potential factors contributing to limited construct validity include poor data quality, lack of case-mix adjustment, and statistical uncertainty.