Inpatient Practice Variation and the Cost of Care: Is There A Business Case for Standardization?

Research Objective:

Researchers and policy experts have long pointed to regional variation in utilization and cost of care with little observable difference in outcomes to conclude that variation in health care practice patterns may lead to significant inefficiencies and waste. The presumption is that dissemination of best practices and practice standardization efforts could reduce ‘unnecessary’ variation and increase efficiency.

The objective of this study is to explore the relationship between inpatient practice variation and inpatient costs. Specifically, this research examines the variation in inpatient hospital practices within a hospital after controlling for patient and payer characteristics to provide insight to the extent non-standardization of practices contributes to increased costs. Research questions this study examines are:

• Do hospitals with a wider variation in practices and processes after controlling for patient characteristics have higher costs?
• Are there specific organizational structures associated with wider variation in practices?
• Do physicians who exhibit a wider variation in their own processes after controlling for patient characteristics have higher costs?

Study Design:

This study uses routine total hip and knee replacement surgeries to evaluate inpatient practice and process variation. The primary dependent variable is a standardized measure of inpatient cost of care as a proxy for physician practices. This is defined as total charges times the hospital specific cost to charge ratio then normalized to eliminate local price effects using the Medicare Geographical Adjustment Factor. Three-level hierarchical modeling is used to develop physician specific cost of care factors that control for patient, payer and hospital characteristics. We evaluate the association between physician specific cost factor spread within a hospital and the hospital’s mean cost of care. Primary data sources include patient encounter level data on charges, patient and payer characteristics from the HCUP State Inpatient Database and hospital characteristics from the AHA Hospital Survey.

Population Studied:

All patients in New York, Florida and Washington undergoing uncomplicated primary knee and hip replacement procedures (ICD 9 procedure codes 81.51 and 81.54) included in the HCUP State Inpatient Discharge database between 2011 and 2013.

Principal Findings:

Mean normalized hospital costs for knee replacements in our dataset ranged from $8,400 to $29,500 and from $6,700 to $31,300 for hip replacements. Costs are not evenly distributed, but increase rapidly at the top end of the distribution. The mean cost of a knee and hip replacement at hospitals in the top 20% of the total cost distribution are respectively 76% and 74% greater than the bottom eighty percent of hospitals. Preliminary results suggest that a wider spread in physician cost of care factors within a hospital increases the odds of being a ‘high cost’ hospital.

Conclusions:

This demonstrates the impact of variation in individual physician practice choices on overall cost of care, even after controlling for patient, payer and operating conditions.