It’s become more difficult for medical practices to remain profitable in today’s ever-changing health-care environment. It is also more difficult to care for patients and make improvements in care, while still containing costs and enhancing satisfaction. Additionally, medical practices are in the position of responding to the pressures purchasers have placed on keeping health plan premiums manageable.

Medical practice benchmarking has become a useful tool for practice management and documenting accountability. Historically, benchmarking meant comparing Medical Group Management Association (MGMA) cost survey data, or budgeted figures, to clinic results as a way to monitor the financial status of a medical practice. However, until recently, there were no comparatives involving the clinical aspects of a medical practice.

The Physician Profiling Project

The Center for Research in Ambulatory Health Care Administration’s (CRAHCA) Physician Profiling Project, supported by the Robert Wood Johnson Foundation, has developed quality improvement data for group practices. The purpose of the project was to develop, test, and demonstrate methods to compare clinical work profiles for physicians and practice specialties. The comparative features of the project’s reports may provide the first opportunity for many practices to review their own clinical operations, and compare their performance at specialty and physician levels with other practices. The project’s goals were clear.

• Construct a national database linking transaction data with practice and physician demographic information
• Profile physician work at the specialty and provider levels
• Educate practice teams in the use of profiling data
• Demonstrate the research applications of the project’s database

Dr. Neil Piland, CRAHCA research director, says profiling is designed to perform several functions.

• Describes patterns of care
• Identifies and describes variation in utilization of services and resources
• Investigates sources of variation in patterns of care and resource use
• Investigates the impact of profiling on reduction of variation and changing provider and system behavior

Determining measurements and collecting data

The primary indicators used to measure work variation were based on the clinical measures of procedures (American Medical Association’s Current Procedural Terminology (CPT-4)), ICD-9-CM Diagnosis Codes, Resource Based Relative Value Scale (RBRVS), Relative Value Unit (RVU) analyses, selected patient-mix measurements, and episodes of care analysis. Certain variables important to understanding the clinical work variation included practice organization, physician demographics, administrative data, and differences in patient mix. An important step in understanding the link between resource use and outcomes is connecting, by physician, clinical work profiles to specified indicators and patient mix.

The project collected data from over 70 practices, including sole practitioners and complex integrated systems, representing over 3,900 physicians. The administrative data was collected on an encounter-level basis. The data elements included the procedure code and modifier, beginning and ending date of service, diagnosis codes, patient identifier, patient date of birth and gender, and primary payer code. The practice demographic data included information on the type of practice (i.e., single versus multispecialty), practice specialty(ies), practice motivation, percentage of prepaid revenue, service area population, and managed care market penetration. Information collected on each participating physician included type of specialty, clinical full-time equivalency (FTE), date of birth, gender, state where residency was completed, and board certification/eligibility information.   

Reports were generated to integrate physician work (as measured by RVUs) and clinical data for use in the practice analyses. Comparative profiling reports were developed for each practice’s specialty and individual physicians to compare two items: specialty data to other practices with the same specialty, similar size, and setting; and individual physician’s data to other peer physicians in the same specialty.

The reports include the following indicators:

• Procedure frequency
• RVU ratio analysis which includes case intensity (RVU/patient)
• Procedure complexity (RVU/procedure)
• Productivity (RVU/clinical FTE)
• Diagnosis groupings
• Patient-mix measurements
• Episode of care analysis

The right benchmarking can make for better outcomes

The feedback reports generated from this data provide comparisons on specified indicators to identify clinical work variation among physicians, practices, and specialties. The project expects that practices will use these comparative reports to better understand the usefulness of RVUs in describing clinical work, to examine the use of procedure and diagnosis codes by specialty, and to aid in explaining clinical work variation. There does not appear to be a similar, published study that compares procedure, diagnosis, and patient-level data, and provides feedback across providers, specialties, and payers.

Measuring the provision of health care and its outcomes is quickly becoming the standard by which large health-care purchasers and patients evaluate prospective health-care providers. As health-care delivery systems evolve into the next century, purchasers and informed consumers will place greater emphasis on measurable processes and outcomes when selecting their health-care networks and providers. Benchmarking your data, and then developing clinical pathways, can result in modification of practice patterns and a reduction in costs. The end result should assist in increasing efficiency, competitiveness, and patient satisfaction.

A source for this article was the Center for Research in Ambulatory Health Care Administration’s Physician Profiling Project. Please contact Dr. Neil Piland at MGMA, 303.799.1111, ext. 834, for more information.