This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

D A T A W A T C H : 
M A N A G I N G  C A R E
W E B  E X C L U S I V E   

18 June 2003  
  

Managing Care: Utilization Review In Action 
At Two Capitated Medical Groups

Prospective denials of coverage on grounds of medical necessity
are only a small part of the overall picture.


by Kanika Kapur, Carole Roan Gresenz, and David M. Studdert

ABSTRACT:

Despite widespread concern about denials of coverage by managed care organizations, little empirical information exists on the profile and outcomes of utilization review decisions. This study examines the outcomes of nearly a half-million coverage requests in two large medical groups that contract with health plans to deliver care and conduct utilization review. We found much higher denial rates than those previously reported. Denials were particularly common for emergency care and durable medical equipment. Retrospective requests were nearly four times more likely than prospective requests were to be denied, and when prospective requests were denied, it was more likely because the service fell outside the scope of covered benefits than because it was not medically necessary.

The management of health care use is a defining feature of managed care. However, managed care organizations (MCOs) use diverse strategies to control the volume and type of services delivered to enrollees. At one end of the spectrum are indirect controls, such as provider panels and financial incentives targeted at physicians; at the other end is direct oversight, most notably utilization review (UR) procedures that evaluate the coverage status of services case by case.¹

Spurred by widespread dissatisfaction with overt cost-control strategies, MCOs in some markets have begun moving away from authorization-based approaches.² However, many MCOs continue to rely on some form of UR.³ Thus, it should not be surprising that recent efforts to protect consumers in managed care have a strong focus on explicit coverage denials.⁴ State and federal policymakers have enacted or are considering reforms that arm enrollees with opportunities to have coverage denials reconsidered.⁵

Private and public attempts to “dull” the sharp end of utilization management in managed care have unfolded in the context of surprisingly little empirical information about the operational reality of UR.⁶ Studies have found relatively low coverage denial rates (3 percent or less), but these studies are limited by the self-reported nature of the data and the specificity of the service areas investigated.⁷

We examine the outcomes of coverage requests using administrative data from two large medical groups based in California. These groups deliver services almost exclusively within a “delegated” model—an organizational structure prevalent in California in which medical groups receive capitated payments from health plans and are responsible for care delivery and UR.⁸ Approximately forty million managed care enrollees nationwide receive care through this type of delivery model.⁹ This descriptive portrait reveals contentious service areas in coverage decisions and the most frequent bases for denials.

Study Methods

Data. We analyzed administrative data associated with nearly a half-million coverage requests in two large medical groups in California (MG1 and MG2).¹⁰ They are multispecialty groups, each consisting of several hundred physicians who provide care to more than 100,000 enrollees in multiple health care facilities. The data encompass all coverage decisions rendered for privately insured enrollees during three-year periods at MG1 (1 January 1997 through 31 December 1999) and MG2 (1 January 1998 through 31 December 2000). In total, the database comprised 146,997 coverage requests at MG1 and 329,382 requests at MG2.

Study definitions. We defined a coverage request as a claim made by or on behalf of an enrollee for approval of coverage for a service or group of services. The enrollee might have been seeking coverage for desired services (“prospective” request) or already have obtained services (“retrospective” request).¹¹

Counting conventions. Although the definition of coverage requests may appear straightforward, several issues merit careful consideration. First, coverage requests for multiple related services, or multiple visits to a particular clinician, were counted as one request. Second, we excluded cases where providers were questioning whether or not a particular service was covered under their negotiated capitation rate with the medical group. These “provider requests” were excluded so that the analysis could focus on patient coverage issues, not disagreements between physicians and the medical groups over contractual issues. Third, we excluded coverage requests and denials for drugs, vision, dental, and behavioral health services. Because many enrollees received these services through “carve-out” arrangements, a “denial” of coverage from the medical group typically involved a redirection to the appropriate decision-making entity rather than a true denial of services.

Study Results

Distribution and outcomes of coverage requests. Exhibit 1 shows the distribution of coverage requests by detailed service category, the proportion of coverage requests denied for each service type, and the distribution of denials by service. There were some discrepancies in the medical groups’ service categorizations, but we attempted to create comparable categories across the two groups to the greatest extent possible.

Distribution of coverage requests. Diagnostics and testing (22 percent at MG1; 19 percent at MG2) and emergency care (11 percent at MG1; 13 percent at MG2) were common subjects of coverage requests. At MG2 physician services (28 percent) and surgery (16 percent) also accounted for a sizable share of requests; at MG1 these services accounted for 11 percent and 8 percent of requests, respectively.

Rates of denial. Overall, denial rates at the two medical groups were quite similar (10 percent at MG1; 8 percent at MG2) (Exhibit 1). Durable medical equipment (DME) (23 percent at MG1; 15 percent at MG2) and emergency care (17 percent at MG1; 16 percent at MG2) had relatively high rates of denial at both medical groups.¹² Denial rates were also relatively high for several service subcategories, most notably laboratory/pathology, speech therapy, chiropractic services, and dermatology consultations (MG2 only). In contrast, requests for inpatient care, surgery, and obstetric care had low rates of denial at both groups.

Distribution of denials across services. Diagnostics and testing services and emergency care together accounted for more than one-third of denials at both medical groups (Exhibit 1). However, there were several marked differences in the distribution of denials across groups, most notably for DME (19 percent of denials at MG1; 5 percent at MG2) and physician services (11 percent at MG1; 24 percent at MG2). These differences partly reflect the fact that DME and physician services accounted for different shares of coverage requests at the two groups.

Type of coverage request: prospective versus retrospective. Information on the prospective versus retrospective status of coverage requests was available at MG1 only. Four-fifths of coverage requests at MG1 were prospective (Exhibit 2), with diagnostic/testing (24 percent) and ancillary health services (17 percent) being the most prevalent clinical categories (data not shown). These services accounted for a much smaller proportion of retrospective requests, nearly half of which related to emergency care.

Overall, retrospective requests were much more likely than prospective requests were to be denied (23 percent versus 6 percent). Among prospective requests, DME had the highest denial rate (24 percent), and obstetrical care (<1 percent), surgery (2 percent), and inpatient care (2 percent) had the lowest. Among retrospective requests, diagnostic/testing services, ancillary health services, and other care all had denial rates greater than 40 percent.

Reasons for denial. Forty-two percent of prospective requests at MG1 were denied on the grounds that the service did not fall within the scope of benefits covered by the enrollee’s health insurance policy, 29 percent were denied because they were judged not to be medically necessary, and 22 percent involved denials of coverage for the specific provider requested by the enrollee (Exhibit 3).

Most DME denials (86 percent) were because the requested service was not contractually covered. About two-thirds of ancillary health service and minor surgery denials were because the service was not deemed medically necessary. As expected, all emergency care denials were made on the basis of medical necessity.

With respect to retrospective requests (not shown in Exhibit 3), MG1’s denial reasons were much more uniform. The reason cited for almost every denial of emergency care services was that the enrollee’s medical condition was not deemed an emergency according to the “prudent layperson standard.”¹³ Among retrospective requests for other services, virtually all denials were because enrollees should have obtained preauthorization and did not.

Study limitations. Because these data come from two medical groups in one state, they are not necessarily generalizable nationwide. Our results reflect the administrative structure, policies, and contracts within the two medical groups analyzed. In addition, although we labored to ensure that data from the two groups were comparable, we still relied on existing taxonomies, which had been developed independently at each group. Finally, our analysis does not address the kind of “implicit” denials that stem from changes in physicians’ practice behavior.

Summary And Policy Implications

Rates of denial across the two medical groups we studied were remarkably similar, and much higher than those previously reported. Separation of requests and denials along three basic axes—services, type (prospective versus retrospective), and denial reason—revealed much heterogeneity in several noteworthy areas.

At the service level, denial rates for emergency services and DME were particularly high. Retrospective requests were nearly four times more likely than prospective ones were to be denied. Lack of medical necessity accounted for a relatively small share of denials of prospective requests; these requests were more likely to be denied because the medical group judged that services were not covered benefits or because the enrollee’s choice of providers was impermissible.

Unbundling of the UR process is a valuable exercise for several reasons. As policymakers continue to press accountability on MCOs in the name of consumer protection, there is growing interest in denials of coverage as a potential marker of inappropriate management of care. Indeed, the external review programs that now exist in more than forty states are premised on objective scrutiny of denials. More generally, disaggregation of denials into services and reasons helps to pinpoint the irritants in enrollee-plan relations. Also, direct analysis of UR data at their administrative source showed some definitional inconsistencies between the medical groups and permitted us to adopt uniform counting conventions to address them. Differing approaches to UR tracking may have an enormous impact on basic descriptive statistics about the UR process; hence, careful standardization is essential to any interorganizational comparisons of UR performance.

There has been considerable concern among policy researchers about the use of medical necessity as a tool to control use of services.¹⁴ We found that prospective denials of coverage on the grounds of medical necessity were a surprisingly small part of the overall UR picture—16 percent of denials at MG1. A substantial number of coverage denials were for emergency care. The usual reason for these denials, which were predominantly retrospective, was that medical care was not deemed to be emergency in nature. This suggests the need for more extensive dissemination of coverage rules for emergency care and possibly a reevaluation of the standards for emergency care coverage. Better dissemination of insurance rules also could be useful in guiding enrollees to seek required preauthorization for services—another important factor in coverage denials.

Denials made on contractual grounds—the largest share of denials—may call for both clinical and contractual expertise. Hence, they should ideally be made by personnel who are versant in both areas. There was some evidence of this sort of dual expertise being brought to bear on coverage decisions at the two groups we studied. However, for reasons of size or financial stress, this may be beyond the reach of many smaller medical groups that have assumed responsibility for UR.

As MCOs in many parts of the country continue their evolution away from a command-and-control approach to utilization management and toward cost sharing with purchasers and enrollees, the future role of UR remains unsettled. Although some form of UR seems certain to remain a fixture in many MCOs, the shift to consumer-centered strategies could alter the profile of both coverage denials and requests. For example, higher cost sharing may force overall declines in the volume of requests. In this environment, contractual coverage and medical-necessity issues that persist are likely to be for services that enrollees feel especially strongly about. Such consumer concerns, together with ongoing consumer protection agendas that include reforms such as guaranteed external review and right-to-sue provisions, mean that the policy importance of UR denials in managed care is unlikely to wane in the foreseeable future.

Funding for this study was provided by the U.S. Department of Labor (DoL). The authors are grateful to Nancy Campbell for valuable programming assistance. Any views expressed herein are solely those of the authors, and no endorsement by the DoL or RAND is intended or should be inferred.

NOTES

1. P.R. Kongstvedt, The Managed Care Handbook, 4th ed. (New York: Aspen, 2000).
2. D.A. Draper et al., “The Changing Face of Managed Care,” Health Affairs (Jan/Feb 2002): 11–23.
3. T.M. Wickizer and D. Lessler, “Utilization Management: Issues, Effects, and Future Prospects,” Annual Review of Public Health 23 (2002): 233–254; and L. Greenberg, G. Carneal, and M. Hattwick, Trends and Practices in Medical Management: 2001 Industry Profile (Washington: URAC, American Accreditation HealthCare Commission, 2002).
4. E.D. Kinney, Protecting American Health Care Consumers (Durham, North Carolina, and London: Duke University Press, 2002).
5. A.A. Noble and T.A. Brennan, “The Stages of Managed Care Regulation: Developing Better Rules,” Journal of Health Politics, Policy and Law 24, no. 6 (1999): 1275–1305; and T.E. Miller, “Center Stage on the Patient Protection Agenda: Grievance and Appeal Rights,” Journal of Law and Medical Ethics 26, no. 2 (1998): 89–99.
6. Wickizer and Lessler, “Utilization Management.”
7. D.K. Remler et al., “What Do Managed Care Plans Do to Affect Care? Results from a Survey of Physicians,” Inquiry 34, no. 3 (1997): 196–204; H.H. Schauffler et al., “Differences in the Kinds of Problems Consumers Report in Staff/Group HMOs, IPA/Network HMOs, and PPOs in California,” Medical Care 39, no. 1 (2001): 15–25; A. Koike, J. Unutzer, and R. Klap, “Utilization Management in a Large Managed Behavioral Health Organization,” Psychiatric Services 51, no. 5 (2000): 621–626; TM Wickizer, D. Lessler, and J. Boyd-Wickizer, “Effects of Health Care Cost-Containment Programs on Patterns of Care and Readmissions among Children and Adolescents,” American Journal of Public Health 89, no. 9 (1999): 1353–1357; US General Accounting Office, Medicare Part B: Inconsistent Denial Rates for Medical Necessity across Six Carriers, Pub. no. GAO/T-PEMD-94-17 (Washington: GAO, March 1994); and Texas Department of Insurance, An HMO Purchasing Guide for Texas Consumers (Austin: Department of Insurance, June 1997).
8. Responsibility for adjudicating enrollees’ appeals of coverage denials remains with the health plan.
9. M.B. Rosenthal, B.E. Landon, and H.A. Huskamp, “Managed Care and Market Power: Physician Organizations in Four Markets,” Health Affairs (Sept/Oct 2001): 187–193; and Henry J. Kaiser Family Foundation, Trends and Indicators in the Changing Health Care Marketplace—Chartbook, May 2002,
www.kff.org/content/2002/3161/marketplace2002_finalc.pdf (6 May 2003).
10. These medical groups are much larger than the average medical group in California, which has about ten physicians. The average size of multispecialty groups in the United States (23.4 physicians) is nearly four times that of single-specialty groups (6.4 physicians). American Medical Association, Medical Group Practices in the US (Chicago: AMA, 1999).
11. Almost all coverage requests in both medical groups are made by physicians on behalf of their patients. In rare cases—for instance, when the enrollee is outside the United States—the enrollee may directly submit a request for reimbursement.
12. While these differences most likely represent true variations in patterns of UR between the medical groups, differences in service coding between medical groups may be partly responsible.
13. Under this standard, health plans would be required to cover emergency services if the patient seeks care for symptoms that a prudent layperson, possessing an average knowledge of health and medicine, could reasonably expect to result in serious impairment to the patient’s health.
14. S. Rosenbaum et al., “Who Should Determine When Health Care Is Medically Necessary?” New England Journal of Medicine (21 January 1999): 229–232; and L.A. Bergthold, “Medical Necessity—Do We Need It?” Health Affairs (Winter 1995): 180–190.

Kanika Kapur is an associate economist at RAND in Santa Monica, California. Carole Roan Gresenz is an economist at RAND in Arlington, Virginia. David Studdert is an assistant professor of law and public health at the Harvard School of Public Health in Boston.

©2003 Project HOPE–The People-to-People Health Foundation, Inc.