Over the past several years, third party payers have struggled to address coverage and payment issues arising as a result of increasing use of digital health technologies. In general, Medicare payment systems, which were designed well before the advent of digital health technologies, are not structured in a manner that is particularly well suited to accommodate these new technologies. In addition, the speed of innovation far outpaces the cumbersome processes used to update Medicare and other third party payer allowances, resulting in coverage determinations and payment allowances whose obsolescence has a tendency to precede their implementation.
The coverage and payment barriers facing digital health technologies vary, depending on how neatly the technology fits into a coverage category established by statute (in the case of Medicare); the Medicaid state plan (in the case of Medicaid); and the terms of individual and group health insurance policies (in the case of private payers). However, all payers appear to be struggling with three common issues, discussed below. Suggested paths forward to address each of these barriers to payment will be addressed in future articles.
What’s in a Name?
The first challenge that many payers face is one of definition. A common nomenclature for digital health technologies has yet to emerge, and, in the absence of a universally accepted definition of terms, payers struggle to fit digital health technologies into the coverage categories included in the governing statute and regulations (in the case of Medicare); the Medicaid State Plan (in the case of Medicaid), or insurance policy terms and conditions (private payers). The AMA has contributed a useful (if not entirely complete) definition of terms that classifies artificial intelligence technologies into the following classifications, as set forth in CPT® Appendix S: Artificial Intelligence Taxonomy for Medical Services and Procedures:
Assistive: The work performed by the machine for the physician or other QHP is assistive when the machine detects clinically relevant data without analysis or generated conclusions. Requires physician or other QHP interpretation and report.
Augmentative: The work performed by the machine for the physician or other QHP is augmentative when the machine analyzes and/or quantifies data in a clinically meaningful way. Requires physician or other QHP interpretation and report.
Autonomous: The work performed by the machine for the physician or other QHP is autonomous when the machine automatically interprets data and independently generates clinically relevant conclusions without concurrent physician or other QHP involvement. Autonomous medical services and procedures include interrogating and analyzing data. The work of the algorithm may or may not include acquisition, preparation, and/or transmission of data. The clinically relevant conclusion may be a characterization of data (eg, likelihood of pathophysiology) to be used to establish a diagnosis or to implement a therapeutic intervention.
The primary problem with this lexicon is that it conceptualizes digital health technologies based on, and in terms of their relationship with, a service traditionally provided by a physician or other health care practitioner. Defining a digital health technology in this manner generally leads to valuation of the technology in terms of whether, and to what extent, it effectively substitutes for a traditional physician’s service or service provided by another type of health care practitioner or provider. Yet, the primary value proposition for many new digital technologies is that they perform functions and contribute to health care outcomes in ways that are distinct from traditional healthcare services typically provided by physicians and other health care providers—often ways that cannot be easily characterized as a substitute for traditional services.
The barrier created by conceptualizing new technologies in terms of whether and to what extent they substitute for more traditional and familiar health care services is essentially built into today’s coverage and payment mechanisms. For example, under the processes used by the Medicare program to determine whether to add a new service to the list of covered telehealth services, preference is given not only to whether a service is similar to a physician’s (or other practitioner’s) service, but whether the telehealth service is similar to a particular type of physician (or NPP) service—specifically, an “evaluation and management” service, such as an office visit. State laws requiring third party payers to provide coverage for telehealth or other remote services to in-state enrollees often utilize similar language, restricting coverage to services that would otherwise be provided by a physician or NPP.
If coverage of digital health technologies is to expand beyond these types of restrictions, it is fair to ask what standards should be imposed to ensure that a digital health technology actually makes a positive contribution to patient care: While the services of physicians and NPPs are generally assumed to be reasonable and necessary in the diagnosis and treatment of illness and injury, how are payers to evaluate new technologies that do not—and do not purport to- substitute for these types of services but nonetheless make a claim to coverage and payment?
The Ticky Tacky Box Problem
Little boxes on the hillside,
Little boxes made of ticky tacky,
Little boxes on the hillside,
Little boxes all the same.
There’s a green one and a pink one
And a blue one and a yellow one,
And they’re all made out of ticky tacky
And they all look just the same.
A second (and related) problem facing digital health technology is its failure to fit into any of the ticky tacky boxes used by payers to define the categories of covered services. This problem is clearly illustrated in the case of two types of digital health technology whose popularity appears to be growing: health care “apps” used by patients with or without the involvement of health care professionals and monitoring devices used for initial diagnosis or treatment follow up.
The growing use of various apps used by patients to manage their own conditions—with or without the involvement of traditional health care providers—presents a challenge to established coverage and payment processes. While a number of large insurers have begun to grapple with the issues involved, public payers have not, and current statutory or regulatory provisions limit their flexibility to experiment outside of formally established demonstration projects.
Digital health technologies focused on monitoring patients—either for the purpose of initial diagnosis or for the purpose of following up on a treatment plan—also may be difficult to “fit” into an established coverage category. Some such technologies are not easily classifiable as either “diagnostic” or “therapeutic” and run the risk of being labelled “preventative” by third party payers—a label that has significant coverage implications, since health insurance contracts and public programs severely limit coverage of preventive health services. In addition, monitoring technologies that involve little or no physician or NPP involvement may be technologically superior to those that require a medical professional’s oversight but, ironically, the more successful the technology’s efforts to limit the need for a health care professional, the greater the reimbursement challenge, since the time and effort of the physician or NPP is often the “hook” for claiming payment under both private and public health plans.
Mo’ Money: The Problem of Valuation
The frustration of forging a path through the third party payer labyrinth is exacerbated by a slew of shifting payment policies and informal rules. Proponents of digital health technologies may find themselves facing an unsatisfying good news/bad news scenario: The good news is that the technology is found to be reasonable and necessary and eligible for coverage. The bad news is that the amount of payment is a whopping $0. This may be the case, for example, when payment for a new technology is “bundled” or “packaged” into the payment for an underlying service.
In addition, this situation may result because of the payer’s methodology for determining payment amounts for the particular setting involved. For example, most private payers as well as many Medicaid programs utilize the Relative Value Unit (RVUs) published in the Medicare Physician Fee Schedule to determine the relative payment allowances paid for services provided by physician practices. The methodology currently used to determine the amounts payable for physicians’ practice expenses generally treats both digital health technologies as overhead, rather than allocating these costs directly to the underlying clinical service involved. This results in chronic and systemic underpayment of services that require significant expenditures for digital health technologies. Likewise, to the extent that a hospital fails to establish a separate charge for digital health technology, the subscription or other costs involved are unlikely to be reimbursed by payers that reimburse hospitals on the basis of a percentage of charges, or appropriately paid by Medicare, which uses relative charges to establish inpatient hospital prospective payment system rates.
These three barriers to payment of new digital health technology—the characterization of digital health technology as substitutes for conventional health care services; the difficulty of fitting digital health technology into recognized coverage categories; and the problems posed for digital health technology of commonly used methodologies for determining payment amounts—will be addressed in future articles. Stay tuned.