Managing provider networks through expert systems: use of patterns of treatment to address overutilization

Sometimes lost in the furor over rapidly rising health care costs are the components of these costs. Although administrative costs are widely cited as a major problem, most studies indicate that excessive utilization of services may be the most important contributor to accelerating costs. A variety of techniques have emerged in an attempt to control use of services over the past decade. Most involve hospital admissions and inpatient costs, including the DRG system, preauthorization criteria, and concurrent case review.

These methodologies have contributed to dramatic reductions in hospitalization rates in the United States, but relatively few efforts have addressed an equally important source of excess utilization: use of services by ambulatory patients. Indeed, the very success of efforts designed to limit hospital use has led to an increase in use of services among ambulatory patients, where few or no controls exist. Like the classic balloon metaphor, the costs bulge on the ambulatory side when squeezed on the inpatient side.

A number of types of excessive utilization occur in use of ambulatory services:

* Appropriateness. The service is not appropriate for the illness. For example, complex visual field testing on a routine health exam.

* Frequency. The service is appropriate for the illness, but is performed more frequently than necessary. For example, frequent ultrasounds during an uncomplicated pregnancy.

* Intensity. The service is appropriate and is performed with reasonable frequency, but a more complex code is used in billing. For example, a complex visit code, such as 99215, is used routinely for a diagnosis of hypertension, rather than the expected mix of complex and simple visit codes.

Problems of appropriateness and frequency can occur across the range of services, including physician visits, laboratory and x-ray testing, and procedures. Thus, efforts to control utilization should be comprehensive, considering the whole range of services offered by physicians and other health care providers. In addition, it is appropriate to focus on physicians, because they stand at the center of the ordering process for these services.

Among the first individuals to recognize excessive use of ambulatory services and to develop a system to address this problem was the late Donald C. Harrington, MD. Dr. Harrington was one of the fathers of managed care in the United States, among the founders of the first IPA (the San Joaquin Foundation) and of the predecessor of the American Managed Care and Review Association (AMCRA). In the mid-1950s, soon after beginning the San Joaquin Foundation, Dr. Harrington and colleagues developed the initial Patterns of Treatment criteria.

These criteria, developed by the now-common physician consensus panel methodology, defined levels of excessive use of ambulatory services in terms of appropriateness and frequency, as noted above. The criteria were designed to be used with claims data, facilitating an initial rapid review phase. Thus, a physician's use of services could be compared with the Patterns of Treatment criteria, and probable areas of excess utilization highlighted.

The Patterns of Treatment criteria have undergone tremendous expansion and revision since those early efforts. The underlying goals and framework remain the same: the criteria are designed to reflect a high level of practice, and the vast majority of physician practices are consistent with the criteria. However, in about 10 percent of instances, practice patterns conflict with the criteria; after careful review, most of these cases (greater than 80 percent) are judged to be excess utilization, and the remainder are inaccurate coding or, in rare cases, highly complex patients.

Use of the current version of Patterns of Treatment, based on revisions begun in the mid-1980s, involved manual review of claims or other data. However, some users soon realized the potential of the system to review large claims data sets when encoded in software. Commercial grade software encompassing the Patterns Treatment criteria were subsequently developed by Health Payment Review, Inc., which obtained ownership of the criteria in 1992.

Extensive use of Patterns of Treatment on literally hundreds of millions of claims since 1986 has involved both "prospective" and "retrospective" review:

* In prospective review, payers use Patterns of

Treatment to directly adjudicate claims after performance

of the service but prior to payment. Several

large payers report savings of more than 5 percent of

their outlays for physician fees using this methodology.

Metropolitan Life, for example, reports savings of $5 for

each dollar of program expense.

* In retrospective review, payers use Patterns of

Treatment after claims payment to profile physicians

and identify areas of likely overutilization. This information

can be returned to physicians, leading to

changes in practice by aberrant providers and support

for efficient, high-quality providers.

To remain current with changes in medical practice, Patterns of Treatment has incorporated additional information:

* Use of data from practice guidelines and outcomes

research as well as from consensus panels.

* Inclusion of a sophisticated case-mix adjustment algorithm

to allow comparisons despite price differences.

* Reporting based on time intervals as well as "episodes

of care" to allow clinical comparisons of resource use.

* Expansion of criteria for nonphysician services: chiropractic,

physical therapy, and podiatry.

How have physicians responded to the use of Patterns of Treatment? Medical directors at HMOs, PPOs, and other managed care organizations report that the information provided is critical to maintaining a managed care network. Most physicians' practice is consistent with Patterns of Treatment, because the bulk of excess utilization is attributable to 10 percent or fewer physicians. Many payers use these positive results to encourage or reward efficient, high-quality physicians. Among physicians whose claims are inconsistent with the Patterns of Treatment criteria, a substantial majority understand and accept the process. As such, Patterns of Treatment functions as an educational tool as well as a cost management system.

It is highly likely that the use of software to screen physician claims for utilization will increase, whatever the outcome of health care reform initiatives. Indeed, the information provided by such systems is critical to the success of any type of managed care; true managed care depends on practice pattern information. Thus, the use of these systems will continue to extend beyond the traditional range of payers to all parties interested in managing health care, including health coalitions, physician groups, and physician/hospital organizations. Expert systems offer a means to obtain, organize, and apply utilization information, so that quality can be maintained and costs can be controlled.

Lawrence G. Miller, MD, MPH, is Vice President, Clinical Affairs, and Medical Director, Health Payment Review, Boston, Mass. He is a member of the College's Societies on Insurance and Managed Health Care Organizations and its Forums on Cost Management and Health Policy.

COPYRIGHT 1994 American College of Physician Executives
COPYRIGHT 2004 Gale Group