To determine the underlying causes of, and potential ways to reduce, vulnerabilities associated with Medicare payments for chiropractic services.
In 1972, Congress passed Public Law 92-603, which amended section 1861(r) of the Social Security Act (the Act) to define chiropractors as physicians who are eligible for Medicare reimbursement, but only for manual manipulation of the spine to correct a subluxation, or malfunction of the spine. Federal regulations (42 CFR § 410.21(b)) further limit Medicare payment to treatment of subluxations that result in a neuromusculoskeletal condition for which manual manipulation is appropriate treatment. In addition to these specific provisions, sections 1862(a)(1)(A) and 1833(e) of the Act require that all services billed to Medicare, including chiropractic manipulations, be medically necessary and supported by documentation.
The Medicare Carriers Manual (the Manual) outlines additional coverage criteria for chiropractic services billed to Medicare.1 Pursuant to section 2251.2 of the Manual, the existence of a subluxation must be documented through an X-ray or physical examination and chiropractic services must be provided as part of a written plan of care that should include specific goals and measures to evaluate effectiveness. Section 2251.3 of the Manual states that chiropractic treatment “. . . must provide a reasonable expectation of recovery or improvement of function.” The same Manual section states that “. . . ongoing maintenance therapy is not considered to be medically necessary under the Medicare program,” and is therefore noncovered.
Chiropractic has experienced considerable growth in Medicare, from 11.2 million services and $255 million allowed in 1994 to 21 million services and $683 million allowed in 2004. In previous studies, published in 1986, 1998, and 1999, the Inspector General found that a significant vulnerability existed in connection with chiropractic services, particularly concerning maintenance care.2
To gain a deeper understanding of the underlying causes of these vulnerabilities and ways to reduce them, we selected a simple random sample of 400 Medicare services (total allowed amount = $12,638.38) submitted by chiropractors and allowed in 2001. We contracted with practicing chiropractors who reviewed each service according to a standard protocol, which was based on Medicare coverage guidelines and requirements. The review instrument solicited information about the beneficiary’s chiropractic treatment as a whole and about the individual sampled service in particular. This enabled the reviewers to determine if the services billed to Medicare were covered, coded correctly, and properly documented. In particular, it enabled the reviewers to determine the extent to which payments were made for maintenance services, which are not covered under Medicare. Based on the results of this review, we also determined the likelihood of services being noncovered depending on the number of services billed per episode of care. Knowing this, Medicare carriers can make informed choices regarding the level of effort to expend in reviewing questionable billings based on billing patterns.
Because we only reviewed services provided by chiropractors in 2001, our sample results cannot be extrapolated to other periods. Accordingly, we make no inferences to chiropractic error rates in subsequent years.
Maintenance services were the most common type of noncovered chiropractic services that Medicare paid for in 2001. Medicare carriers routinely deny all chiropractic claims that do not carry a code for spinal manipulation, which is, by law, the only treatment for which chiropractors may be reimbursed. Our medical reviewers found that although billed with an allowable code, 57 percent of these services did not meet Medicare coverage criteria (i.e., were noncovered). In addition, 16 percent were miscoded or billed at the wrong level of spinal manipulation, and 6 percent were undocumented. Twelve percent had multiple errors, yielding an overall error rate of 67 percent, resulting in $285 million in improper payments.
Medical reviewers determined that the majority of inappropriately paid services were maintenance treatments ($186 million in allowed payments), which Medicare defines as medically unnecessary, and are therefore not covered. Another 14 percent ($65 million) were found to be medically unnecessary for other reasons. Medicare also allowed $24 million for services billed with a spinal manipulation code that were actually extraspinal manipulations or non-manipulative treatment, such as massage. Apart from coverage issues, upcoding was also a significant problem, resulting in a $15 million overpayment.
Supporting documentation for chiropractic services rarely met all Medicare Carriers Manual requirements. The Manual requires that specific supporting documentation be present in the chiropractic record. Nearly 94 percent of chiropractic services, though, lacked at least one of the supporting documentation elements listed in section 2251.2 of the Manual (including those that were completely undocumented). The lack of one or more of these elements did not automatically lead us to conclude a service was noncovered, although these determinations were often related. For instance, 34 percent of chiropractic services were not supported by an evaluation that met the Manual’s specific requirements for documenting a subluxation. Most, but not all, of these services were also determined to be noncovered.
Lack of medical necessity is directly related to service volume. As chiropractic care extends beyond 12 treatments in a year, it becomes increasingly likely that individual services are medically unnecessary. The likelihood of a service being medically unnecessary increases even more significantly after 24 treatments. Accordingly, identifying and carefully scrutinizing services beyond a certain frequency threshold could result in significant savings. Although frequency-based controls are common among carriers and in the private sector, the Centers for Medicare & Medicaid Services (CMS) does not have a national policy addressing their use.
Carrier controls to prevent overutilization are inconsistent.
Although all carriers have some mechanisms to prevent and recoup improper payments for chiropractic services, a significant vulnerability surrounding this benefit persists.
Based on the volume of medically unnecessary, undocumented, and noncovered services allowed, chiropractic services represent a significant vulnerability for the Medicare program. Therefore, we recommend that CMS take the following actions:
Ensure that chiropractic services comply with Medicare coverage criteria. CMS should require that its carriers or Program Safeguard Contractors conduct service-specific reviews of chiropractic services to identify improper payments. CMS should also implement national frequency-based controls to target high-volume services for review, since our medical review identified a strong correlation between high service volume and lack of medical necessity. When conducting reviews of individual providers, it is imperative that reviewers collect the entire records associated with services selected as part of a service-specific review. Several records we reviewed would have appeared legitimate for any one particular day of service; however, that day’s documentation was repeated verbatim for the entirety of the patient’s treatment.
Require that its carriers educate chiropractors on Medicare Carriers Manual requirements for supporting documentation. Many chiropractors seem unaware of the specific documentation requirements outlined in the Manual. CMS should address this lack of knowledge by directing its carriers to issue provider bulletins reminding chiropractors of their responsibilities.
In addition to these recommendations, we have forwarded information on the noncovered, miscoded, and undocumented services identified in our sample to CMS for appropriate action.
Agency Comments and OIG Response
In its comments on our draft report, CMS agreed with our findings and recommendations. The agency has clarified its chiropractic coverage criteria and indicated that most carriers are taking steps to reduce chiropractic error rates, including targeted educational efforts and service-specific medical reviews. In addition, as of October 1, 2004, CMS has required that chiropractors use the –AT modifier to indicate that a service is not maintenance; only claims to which this modifier is attached are payable.
We appreciate CMS’s response to our report, and support the steps the agency is currently undertaking to help prevent paying for noncovered, miscoded, and undocumented services.
CMS noted in its comments that while this Office of Inspector General (OIG) study projected that 67 percent of the chiropractic services allowed by Medicare did not meet program requirements, CMS’s Comprehensive Error Rate Testing (CERT) program identified a claims paid error rate of approximately 16 percent for claims submitted by chiropractors in 2002. CMS further noted that differences in the methodological approaches accounted for the significantly different rates. CMS recognized that OIG’s review of a beneficiary’s entire course of treatment enabled us to determine that approximately 40 percent of all chiropractic services are attributable to maintenance care, and thus are not covered under Medicare. In contrast, the CERT paid claims error rate is based on a review of a single claim, which limits its ability to detect uncovered maintenance costs.
We agree with CMS and would like to emphasize that the purpose of this inspection was to determine the underlying causes of, and potential ways to reduce, vulnerabilities associated with payments for chiropractic services. It was not designed to reproduce, or to review, the CERT paid claims error rate. In addition to the different methodological approaches that are noted above, the CERT used 2002 data, whereas our data was drawn from 2001. Hence, our results cannot be compared directly to the CERT program results.
Furthermore, chiropractic payment errors, while a significant vulnerability, contribute only minimally to the overall CERT national paid claims error rate. Medicare allowed approximately $191 billion for Medicare fee-for-service claims in 2001. Chiropractic services accounted for $500 million, or 0.26 percent of this amount. Therefore, the chiropractic-specific error rate has minimal influence on the overall CERT error rate for fee-for-service claims.
Given that Medicare payments for chiropractic services have continued to increase since 2001, the need for a more effective way to eliminate inappropriate maintenance payments is crucial. We recognize that it may not be practical for the CERT program to expend its limited resources to collect the extensive documentation that we used in our review. Therefore, in the future, CMS may wish to conduct additional studies outside the scope of the CERT program to determine cost-efficient ways to address chiropractic payment errors.
- At the time of our study, the references to the Medicare Carriers Manual were accurate. The Centers for Medicare & Medicaid Services has since moved to web-based manuals. The Carriers Manual sections cited in this report are now found in the Medicare Benefit Policy Manual, Pub. 100-2, Chapter 15, sections 30.5 and 240.
- Utilization Parameters for Chiropractic Treatments (OEI-04-97-00496. Chiropractic Care: Controls Used by Medicare, Medicaid, and Other Payers (OEI-04-97-00490), and Inspection of Chiropractic Services Under Medicare (OAI-05-86-00002).
This article was posted on July 5, 2005.