Teaching clinic lowers radiology utilization: a look at how to achieve more cost-effective imaging.

Radiology is an increasingly complex, highly technical, computerized specialty. The days of plain radiographs have given way to a myriad of imaging technologies.

Yet, the rapid pace of change in diagnostic imaging often exceeds the non-radiologists ability to stay current, with new imaging software emerging frequently. The combination of these factors makes it difficult for physicians to keep current on best imaging choices for given clinical indications.

As a result, inappropriate radiology utilization plagues our health care system. It is estimated that the rate of incorrect ordering ranges from 10 percent to 65 percent.

According to Scott Mirowitz, MD, professor of radiology at University of Pittsburgh, "there is a consensus that the number is probably around 30 percent and the cost around $3 billion to $10 billion annually."

In 1993, The American College of Radiology (ACR) recognized the need to establish appropriateness guidelines for imaging. What emerged were the ACR Appropriateness Criteria (www.acr.org). The ACR criteria follows principles developed by the Agency for Health Care Policy and Research (AHCPR), relying heavily on validity, reliability, reproducibility, applicability, flexibility, clarity, multidisciplinary development, scheduled review and documentation.

Using this model, the ACR created the best tool available to guide the physician's hand when writing imaging orders. The ACR Appropriateness Criteria are intended for radiologists, referring physicians and even for patients.

The Providence Ambulatory Care and Education Center (PACE) in Oregon has demonstrated that a resident-driven educational program, in a resident-faculty teaching clinic, can decrease excessive radiology utilization.

By relying upon the evidence-based ACR Appropriateness Criteria when ordering diagnostic imaging, marked reductions in the utilization of magnetic resonance imaging (MRI) and computerized axial tomography (CAT) were achieved.

The medical setting

PACE is a medical clinic central to the internal medicine residency program at Providence Portland Medical Center. As an urban teaching clinic, PACE serves a large population of Medicaid (Oregon Health Plan) and charity patients for whom the sponsoring Providence Health System is at full financial risk.

In 2001, radiology utilization data provided by the InterHospital Physicians Association (IPA) showed that PACE had significantly higher severity adjusted utilization rates of MRI and CAT compared to that of other regional clinics.

The IPA represents more than 1,500 physicians in the Portland metropolitan area, including those at the PACE center. A rigorous interactive Web site is used to present utilization data to physicians in an actionable format.

Providence Health System administration and PACE faculty felt it was important to reverse the radiology utilization trend and improve clinic performance. PACE residents and faculty were informed of the situation and the need to take action. The group chose to address the issue by following an existing process for change, utilizing the population based health (PBH) rotation model. (1)

The PBH model introduces continuous quality improvement (CQI) principles to interns who then incorporate learnings to improve physician practices, with the first year residents taking the leadership role. Our PBH rotation was previously successful at improving pharmacy utilization and increasing cost-effective prescribing. (2)

This CQI process utilizes the plan, do, study, act (PDSA) model popularized by the Institute for Health Care Improvement. (3) The fundamental principles of CQI are particularly applicable in resident education clinics because they encourage development of a practical approach to the challenges of clinical practice variance. (4)

[ILLUSTRATION OMITTED]

The model also serves several of the new educational competencies promoted by the Accreditation Council for Graduate Medical Education (ACGME)--namely patient care, systems-based practice and practice-based learning.

Taking action

The intervention began with education of the faculty and residents about the clinic's performance on diagnostic imaging and the need to reduce our apparent over-utilization. Presentations were scheduled with the faculty and residents to increase awareness of the ACR Appropriateness Criteria Web site (www.acr.org).

Faculty and residents were shown data from the IPA Web site depicting their performance in faculty development and pre-clinic educational sessions.

To further encourage use of the site, a laminated pocket card was produced and distributed to house staff and faculty summarizing preferred imaging modalities for the most common clinical entities residents were likely to encounter. Utilization monitoring was provided quarterly by the IPA.

[GRAPHIC OMITTED]

Clear improvement in utilization

Physician utilization of diagnostic imaging was measured by comparing severity adjusted RVU's/1,000 for MRI and CAT in the Medicaid line of business including Oregon Health Plan (OHP).

The modalities of MRI and CAT were chosen for intervention because of their complexity and expense. The figure shows the results comparing the pre-(2001) to post-(2002) intervention period.

There was a significant decrease in radiology utilization for the procedures described when compared to the IPA average for other regional clinics. PACE achieved an 18 percent reduction in utilization while the IPA average increased by 11 percent. PACE utilization continues to be greater than the IPA average though that margin has decreased by 35 percent since the inception of the project.

This project was such a resounding success that it drew the attention of senior leadership from Providence Health System and the Providence Health Plan. As a result, there are now system-wide education initiatives to promote the use of the ACR Appropriateness Criteria between physicians and residents in our other residency programs and clinics.

A link to the ACR web site has been placed on the start page of all computers throughout the system, since we have computer terminals located at all clinical sites, including ambulatory facilities. The goal is to encourage all providers to consult the ACR Appropriateness Criteria for decision making when ordering diagnostic imaging.

This project demonstrates that physicians in a residency practice setting can achieve significant reductions in utilization of diagnostic imaging by using a CQI format of education and the ACR Appropriateness Criteria.

Given the rate of inappropriate radiology utilization nationwide, with associated costs estimated to be as high as $10 billion annually, there is a potential for other health systems to benefit substantially by adopting these approaches.

IN THIS ARTICLE ...

Discover how an Oregon medical clinic made great strides decreasing radiology utilization after implementing a continuous quality improvement project that was embraced by physicians.

References

1. Patmas, M., Rosenberg, M. and Gragnola, T. "A Rotation in Population-Based Health for Internal Medicine Residents." Academic Medicine. 2001, 76(5): 557.

2. Patmas, M., O'Kane, N. and Reitzenstein, J. "Teaching Clinic Lowers Pharmacy Costs." The Physician Executive. 2003. Jan-Feb. 16-19.

3. Langley, G., Nolan K., and others. The Improvement Guide: A Practical Approach to Enbancing Organizational Performance. Jossey-Bass, 1996.

4. Friedmann, P. and Selbovitz, LG. "Continuous Quality Improvement and Physician Training." Quality Management in Health Care. 1992, 1(1).

Michael A. Patmas, MS, MD, MMM, FACP, CPE, FACPE, is the medical director of the Providence Ambulatory Care and Education Center in Portland, Ore. He can reached by phone at 503-215-6600 or by e-mail at [e-mail].

[ILLUSTRATION OMITTED]

Rowena Rosenblum is a project manager in the department of quality management at Providence Health Plans. She can be reached by phone at 503-574-6423 or by e-mail at [e-mail].

[ILLUSTRATION OMITTED]

By Michael A. Patmas, MS, MD, MMM, FACP, CPE, FACPE and Rowena Rosenblum