In 1994, a group of board-certified radiologists who were engaged in clinical practice in the state of New York saw the need for a system that could manage (and monitor) the provision of outpatient imaging services to managed care clients. In response, they developed New York Medical Imaging, PLLC, Wappingers Falls, to emphasize high-quality, cost-effective care. The organization coordinates the delivery of imaging services to the health-plan enrollees of self-insured employers, self-insured labor unions, and managed care organizations in New York, New Jersey and Connecticut, and was the first provider of single-source, full-service imaging in its service area. New York Medical Imaging has become New York State’s largest provider and manager of diagnostic imaging services; 3 million enrollees are served by the network. New York Medical Imaging also acts as an agent for utilization review for the departments of health of New York and Connecticut.

The network that serves as the core of New York Medical Imaging consists of 750 board-certified radiologists at 150 imaging sites who are members of 50 radiology practice groups. These include both academic and private practice groups and are geographically distributed across the state of New York. In addition, utilization management and quality assurance services are provided for managed care contracts that cover 2,000 radiologists at 500 imaging facilities.

SPECIALIST-ONLY ORDERING

In the course of furthering its goals, New York Medical Imaging undertook research intended to evaluate the effects of ordering restrictions on the management of CT and MRI utilization. By measuring the effectiveness of restricting examination-ordering privileges to designated specialists, however, the investigators understood that another cost-increasing factor might be left unexplored. For this reason, the study was also designed to determine whether ordering restrictions caused primary care physicians to refer more patients to the specialists who were permitted to order CT and MRI studies. Because such a response to the restrictions on primary care physicians could well increase the cost of providing care beyond that noted when no ordering restrictions were in use, it was important to measure any additional referrals during the study period.

For the purposes of this study, New York Medical Imaging restricted the ordering of CT and MRI imaging of the joints to orthopedists and rheumatologists. The ordering of plain-film musculoskeletal radiography by primary care physicians was unaffected by the study protocol. The study’s patient population consisted of 100,000 capitated enrollees in the managed care plans served by New York Medical Imaging. These individuals were not covered by the Medicare program, so the elderly were not heavily represented in the study population.

The referral standards already in use at New York Medical Imaging continued to apply throughout the study for purposes of comparison. The organization’s standard precertification program was also unchanged. Orthopedists, however, were engaged in the decision-making process for imaging to a greater degree than they had been before the study began. They determined not only the most appropriate test for each patient, but advised on the applicability of those tests across the network.

The study covered a 2-year comparison period. CT and MRI utilization for joint examination was measured for a year before ordering was restricted to orthopedists and rheumatologists; utilization measurement continued for an additional year after the restrictions barring referral by primary care physicians for joint imaging had been imposed. The number of orthopedic and rheumatologic consultations ordered by primary care physicians was monitored to determine the effects of limiting access to joint imaging (apart from radiography).

STUDY RESULTS

Figure 1. Total number of procedures performed for joint CT and MRI before and after specialist-only ordering.

During the first half of the study, before ordering restrictions were in place, 27 upper-extremity CT studies and 54 lower-extremity CT studies were ordered for joint examination, for a total of 81 joint CT examinations. Per 1,000 members of the studied population, there were 0.78 joint CT studies. During the second half of the study, when the ordering of joint CT was restricted to orthopedists and rheumatologists, there were 0.6 joint CT studies ordered per 1,000 study subjects (29 upper-extremity studies and 29 lower-extremity studies, for a total of 58 joint CT studies). These results indicated a 23% reduction in total orders for CT imaging of the joints (figure 1).

In the absence of restrictions on MRI ordering for joint examination, physicians ordered 311 upper-extremity MRI studies and 863 lower-extremity MRI studies, for a total of 1,174 joint MRI examinations (11.32 MRI joint studies per 1,000 members of the study population). After restrictions on ordering joint MRI were implemented, 221 upper-body MRI studies and 586 lower-body studies were ordered by the orthopedists and rheumatologists permitted to do so, for a total of 807 MRI joint studies (8.35 examinations per 1,000 study subjects). This constituted a reduction in total MRI imaging of the joints of 26% (figure 1).

The investigators, in an effort to increase the precision with which they could analyze the actual cost savings created by these reductions in CT and MRI ordering, also undertook an evaluation of the relative value units (RVUs) expended for joint CT and MRI before and after specialist-only ordering had been instituted (figure 2).

Figure 2. Total relative value units expended for joint CT and MRI before and after specialist-only ordering.

During the first half of the study, before ordering restrictions were in place, 188 upper-extremity CT RVUs and 367 lower-extremity CT RVUs were expended for joint examination, for a total of 555 RVUs. Per 1,000 members of the studied population, there were 5.3 RVUs expended for joint CT studies. During the second half of the study, when the ordering of joint CT was restricted to orthopedists and rheumatologists, there were 4.1 RVUs expended for the joint CT studies ordered per 1,000 study subjects (195 RVUs for upper-extremity studies and 199 RVUs for lower-extremity studies, for a total RVU cost of 394 for joint CT studies). These results indicated a 24% reduction in total RVUs expended for CT imaging of the joints.

Figure 3. Summary of total relative value units expended pre- and post-program.

In the absence of restrictions on MRI ordering for joint examination, imaging accounted for 4,159 RVUs for upper-extremity MRI studies and 11,530 RVUs for lower-extremity MRI studies, for a total RVU expenditure of 15,689 for joint MRI examinations (or 151.3 RVUs for MRI joint studies per 1,000 members of the study population). After restrictions on ordering joint MRI were implemented, 2,953 RVUs for upper-extremity MRI studies and 7,823 RVUs for lower-extremity MRI studies were expended, for a total RVU cost of 10,776 for MRI joint studies (or 111.5 RVUs per 1,000 study subjects). This constituted a reduction in total RVUs expended for MRI imaging of the joints of 26% (figure 3).

The overall utilization rate for MRI and CT imaging of the joints decreased 26% after the ordering of these examinations was restricted to orthopedists and rheumatologists. No increase in the number of consultations with these specialists ordered by primary care physicians was seen; the frequency of orthopedic and rheumatologic consultation remained at 0.019 consultations per 1,000 study subjects throughout both phases of the study (before and after the imposition of specialist-only ordering restrictions).

CONCLUSION

Primary care physicians, it seems, did not attempt to obtain the CT and MRI examinations of the joints that they were no longer permitted to order by referring their patients to the specialists who were permitted to do so. In fact, it appears that many patients were receiving imaging studies ordered by the primary care physicians and then were referred to the specialists regardless of the outcome of those examinations. Referral to the specialist first achieved a reduction in unnecessary imaging and a more appropriate decision regarding the imaging test employed.

Because the results of this study favored the provision of high-quality care at a reduced cost, the ordering restrictions imposed during the second half of the study remain in place. The specialist-only ordering program will be subjected to future modifications intended to promote further increases in efficiency and productivity. For example, New York Medical Imaging will streamline the precertification process for specialists who are ordering certain imaging studies that have been designated as specific to their specialties. The organization will also institute a program in which individual specialists will be profiled in order to review their imaging referral patterns. n

Andrew W. Litt, MD, is director of neuroradiologic magnetic resonance imaging, director of the division of neuroradiology, vice chair of the department of radiology, and associate professor of clinical radiology, New York University Medical Center, New York City

Donald R. Ryan is CEO, New York Medical Imaging. This article has been excerpted from The Effect of