Driven by the escalating costs of both health care delivery and research, the concept of health care value is of increasing interest to both payers and providers. In this respect, the randomized controlled trial (RCT) often falls short. RCTs do not always translate well into broad clinical practice, because they are based on narrowly defined populations of patients and often lack long-term follow-up. They also fail to address the cost-effectiveness of an intervention.
Comparative effectiveness studies (CES), on the other hand, utilize real-world populations more closely resembling those encountered in clinical practice and, because they compare multiple interventions, they are more amenable to long-term outcome assessments in real-world clinical settings. They are a critical tool in establishing broad, evidence-based assessments of the health care value of new interventions.
This was the rationale for the creation of the Patient-Centered Outcomes Research Institute (PCORI) as part of the Affordable Care Act. PCORI’s charge is to support comparative clinical effectiveness research that enables patients, providers, and other stakeholders to make evidence-based health decisions; PCORI-funded studies focus on “real-world clinical and diverse health-system settings.” However, while PCORI funding covers costs including database development, research stipends, and patient outreach, they do not fund the experimental clinical services—drugs, devices, or procedures—which are, in fact, the focal elements of CES. Such costs are also not covered by traditional Medicare or private insurer reimbursement, given their experimental status. So what is an investigator to do?
In this post, we report on the first-ever investigator initiated implementation of a “Coverage with Evidence Development” (CED) model where multiple private insurers cover experimental clinical services in a PCORI-funded study of the comparative effectiveness of breast cancer screening protocols (the Wisdom Study). We describe the history and basic requirements of CED models, the challenges and benefits of implementing this approach within Wisdom, and we explore the applicability of CED as a funding model for comparative effectiveness research in general.
The Wisdom Study
The Wisdom (Women Informed to Screen Depending On Measures of risk) Study is a five-year, 100,000 participant trial comparing risk-based mammographic screening recommendations to the current standard of care, annual screening. Risk-based mammographic screening utilizes genetic profiling and personal family histories to reduce the frequency of screening in low-risk individuals, while maintaining frequent monitoring of high-risk women. The goal of risk-based mammographic screening is to improve breast cancer detection for higher-risk women while reducing the unintended consequences (false positives, biopsy recalls) of screening practices for those at low risk. This approach is consistent with that prescribed by the U.S. Preventative Services Task Force. The Wisdom Study began initial enrollment in Fall 2016 for women who receive care at the University of California, San Diego and the University of California, San Francisco and will enroll patients throughout California as well as within the Sanford Health System in the Midwest during Q2 2017. The first interim analysis is planned for when Wisdom has enrolled 25,000 participants, which is projected to occur by the end of 2017.
Annual mammography (the current most common practice) carries significant costs to a health plan because it is applied across the entire population of women aged 40 and older. On the other hand, risk-based screening would result in fewer mammograms and fewer associated false positive biopsies, although this approach does carry the additional cost of having every woman undergo genetic testing. Because in recent years the cost of such genetic testing has been reduced to about the cost of a single mammogram, it is anticipated that risk-based screening will result in net cost-savings.
The goal of the Wisdom Study is to determine if risk-based screening does in fact yield savings while improving cancer detection overall. But there is a funding gap: The genetic profiling for participants in the risk-based screening arm of the study is expensive, at $225 per person for about 50,000 participants. Traditional mechanisms like PCORI grants do not cover clinical tests such as host genetic testing, under the assumption that such tests are covered by insurers. But insurers generally do not cover clinical tests without sufficient evidence of efficacy. Although the genetic profiling tests have clear evidence of efficacy for high-risk populations, they are not yet established as a standard of care at the population level. Hence the catch-22: in order to fully fund the trial, we need the very evidence that the trial seeks to establish. It is a situation that is not unique to Wisdom.
Medicare has recognized this problem in the past, establishing a mechanism called “Coverage with Evidence Development” (CED), in which Medicare will pay for an experimental clinical service within the context of a clinical study that is designed to generate the evidence required to make a definitive coverage decision. Medicare only applies to patients over 65, however, and in order for Wisdom to be a definitive trial on breast cancer screening recommendations, the study population needed to be broadly representative of the larger population. While certainly helpful, Medicare’s CED policy offers little “gap” relief to those studying broader populations. A generalizable, representative study population cannot be established if limited to enrolling only those receiving Medicare.
With these challenges in mind, and given that few studies have attempted to use CED with private insurers in the past, we asked a question: Can CED be applied to private insurers in a comparative effectiveness study?
Medicare’s Approach To Coverage With Evidence Development
Before we answer that question, it is worth taking a closer look at Medicare’s approach to Coverage with Evidence Development. It was first introduced in 2005 as a means of addressing growing concerns over the quality of the evidence that was often used as the basis for positive coverage decisions for new services. It was designed by Medicare to provide a conditional reimbursement mechanism to facilitate high-quality evidence development prior to issuance of more broad coverage determinations. Under CED, items and services can be evaluated for coverage based on patient utilization and impact data, with Medicare’s evidentiary guidelines based on the likelihood of improving health outcomes relative to the risk of the service or technology.
The Centers for Medicare and Medicaid Services (CMS) has used CED for conditional coverage for 22 cases over the past decade, including continuous positive airway pressure for obstructive sleep apnea, cochlear implantation, and transcatheter aortic valve replacement. Many of these trials are ongoing, but five have resulted in positive coverage determination.
Although CED clearly fills a gap in coverage decision-making, it does have its critics. Many point to the fact that few studies using CED have generated sufficiently specific data for informed coverage decision-making, as described in the Centers for Medicare and Medicaid Services’ original guidance. Others highlight the operational challenges of implementing CED in practice, such as insufficient funding and inadequate data collection systems to capture and store cost information.
Operational challenges such as these are, to a large degree, the reason why CED has remained exclusive to Medicare and has not been adopted by private insurers. Because the participation of private insurers was critical to the success of Wisdom, we embarked upon a two-year effort before the start of the study to engage private payers and address the barriers to CED implementation beyond Medicare. We are continuing to work with private payers as the study progresses in order to ensure that we are adequately addressing barriers that arise.
Engaging Private Payers
The pragmatic trial approach typical of PCORI-funded trials was well-suited to this task. Inherent in the approach is multi-stakeholder consultation, in which everyone participates and everyone benefits: patients, researchers, payers, commercial and self-insured employers, government bodies, technology partners, donors, policymakers, and scientific bodies.
In particular, defining a priori the metrics that will be collected during the study, what exactly constitutes success and how information is shared are all critical both for stakeholder engagement and to mitigate potential problems in interpreting the resulting data. This is particularly true with respect to results that will influence coverage decisions and that are required by regulatory bodies.
For Wisdom, payers’ motivations were already aligned with those of other stakeholders. While there was clearly an interest in the potential for long-term cost savings, payers also placed a high value on potential improvements to care and long-term outcomes, as well as on providing a definitive answer to a controversial women’s health issue. Wisdom’s emphasis on multi-stakeholder consensus and transparency with respect to clearly defined goals and measures of success was critical in reassuring private insurers that their needs could be addressed, not simply as payers, but as partners in the study.
Initial outreach performed by Wisdom investigators in Spring 2014 sought simply to build among payers a baseline understanding of the goals, potential benefits and consensus-focused model of the study. From these activities, Blue Shield of California (Blue Shield) emerged as one of the most enthusiastic payers. It agreed to act as an industry champion and representative and to coordinate communications and negotiations with the various payers, and a contract with Blue Shield was signed in early 2016.
Issues to Address With Payers
The vast majority of the issues that were addressed in negotiations with Blue Shield are common among private payers, and are likely to be similar for other investigators interested in establishing CED as a model for their study:
Although payer enthusiasm regarding potential impacts of the study was high, Wisdom underwent a rigorous and meticulous due diligence process that scrutinized every aspect of the study including supporting evidence, protocols, outcome measures and timelines, among others. In fact, Blue Shield contracted with a third-party expert in study design and technology assessment to vet the Wisdom Study. A key priority for payers was ensuring that the clinical protocol was aligned with current clinical workflows, to increase the ease of future dissemination and implementation.
Based on our experience, it’s clear that future investigators should also be prepared to incorporate and capture additional cost and value information that is made available to payers during and after the trial. For example, Blue Shield and others requested that the capture and routine analysis of additional data about cost and member specific enrollment be integrated into internal reporting metrics.
Payers supporting CED wanted to avoid changing the Evidence of Coverage (EOC) document, which is the legal contract between a health plan and individual members that governs the terms and conditions of coverage. Changing the EOC requires health plans to re-file products with California insurance regulatory agencies. Re-filing EOCs would delay recruitment and pose a substantial administrative burden for the health plan. It was also critical to avoid any change in the definition of “medical necessity” contained in the EOC document.
The solution that Blue Shield developed was to maintain the existing EOC document and medical necessity language, while modifying their Technology Assessment Policy and Procedure document, which defines the health plan’s policies and procedures for creating medical policies regarding new technologies and services. Blue Shield created a new kind of medical policy, called a Coverage with Evidence Development Medical Policy, for sociotechnical innovations that met the stringent criteria in the rewritten Technology Assessment Policy.
The study must be administered and claims auto-adjudicated such that they pay accurately, according to the various benefit designs of participating members. Insurers may choose to designate specific plans as eligible for study participation. For example, an insurer may choose to cover their fully-insured preferred provider organization (PPO) and health maintenance organization (HMO) population, but may be unable to cover members with self-insured plans.
Wisdom had the advantage of being able to pool its initial clinical sites under a single system, thus decreasing administrative overhead. In order to minimize the friction in the system that is created when billing from the five University of California medical centers, the Athena Breast Health Network was established as a single entity, encompassing the five medical centers, under the University of California Office of the President’s Center for Health Quality and Innovation (CHQI). CHQI’s goal is to promote the transformation of health care delivery by creating a culture of improved health care delivery, improved population health, and lowered costs via enhanced efficiencies. CHQI has created a “virtual office” for Athena through which Athena can deliver study services, thus streamlining the process of contracting and billing.
The University of California medical system created a novel provider identification for the Athena Breast Health Network, so that insurers could pay for these services using a negotiated fee schedule separate from the fee schedule embedded in a payer’s standard contracts with the University of California provider systems. Insurers can identify specific CPT codes to use for services that would otherwise not be covered (if, as with Wisdom’s population-based genetic testing, the services are not customarily covered on a population level), such that beneficiaries will have those services covered. Removing a preauthorization requirement for these codes avoids the cost and administrative overhead associated with negotiating coverage for each test. Insurers can also identify codes that would adjudicate with no member liability, so that insured members would have no fees associated with participation. Blue Shield established a CED policy, with the proviso that clinical trials covered by CED were subject to a third-party review and must be implementable with minimal additional operating costs.
Pre-approved (CED) services and their assigned billing codes are associated with the tax ID assigned to CHQI and submitted to the payers – thereby eliminating preauthorization issues and minimizing both systems’ friction and overhead. Athena was added to the Blue Shield system as a network provider through CHQI, and Athena billing claims could thus be submitted under a single unique provider ID. Blue Shield permits payment for the experimental codes only for the tax ID associated with CHQI. Trial-related services have been programmed to process at custom rates, and the unique tax ID facilitates detailed reporting on trial participation.
The pooling of clinical sites under a single tax ID may not be possible in every clinical trial, but it underscores the value of reducing administrative overhead. The identification of specific experimental codes associated with the relevant institutional tax IDs is a replicable model that facilitates the ability to refine the study and its risk model over time through an adaptive learning engine.
Coverage with Evidence Development (CED) is a promising solution to a continued problem in health care innovation: funding investigations of experimental services. As a result of the Wisdom experience, Blue Shield has established new standardized policies and evaluation criteria and procedures for the use of CED to support studies that evaluate experimental services that have potential to improve health care value. This is a promising development that, through the experience of Wisdom and additional studies, will gain greater use with private payers to address a critical funding gap for comparative effectiveness studies.
The authors thank Dr. Lisa Latts for her leadership on the coverage with evidence approach and her work with the entire team to bring this effort to fruition.