Variation in Variation Measurement: A Mapping Review of Methods to Study Clinical Variation

Area of Science:

• Health Services Research
• Clinical Epidemiology
• Healthcare Quality Improvement

Background:

• Healthcare delivery exhibits variation across providers, practices, and systems.
• Understanding clinical variation is crucial for identifying overuse of low-value care and underuse of high-value care.
• Existing methods for describing and quantifying clinical variation are inconsistently applied.

Purpose of the Study:

• To systematically identify and characterize methods used to describe and quantify clinical variation in health care.
• To map the landscape of methodologies applied to the study of clinical variation.
• To inform improvements in measuring healthcare variation.

Main Methods:

• A mapping review systematically searched health care and health services literature for studies on clinical variation.
• Keyword analysis was used to identify graphical and statistical methods, specialties, settings, and performance areas.
• Included studies measured care processes or health outcomes at the person-level or higher using routinely collected data.

Main Results:

• Over 6,000 studies were analyzed, with most employing basic methods like regression or crude comparisons.
• Fewer than 1,000 studies utilized advanced multilevel models to quantify variation.
• Multilevel models were infrequently applied to the study of health care quality variation.

Conclusions:

• Current methods for studying clinical variation primarily identify its existence rather than quantifying or explaining its underlying causes.
• This review provides insights into the scope and application of existing methods for measuring variation.
• Improved measurement of clinical variation is essential for enhancing healthcare value and equity.