Providing equal access to quality care is the central tenet of health equity, a concept which has gained considerable traction in the last two years. The pandemic has highlighted numerous differences in the availability, utilization, and quality of healthcare services for underserved populations.
Improving health equity is quickly becoming a priority for healthcare organizations. Back in 2007, the Institute for Healthcare Improvement released its Triple Aim of healthcare with the goals of optimizing performance, reducing costs, and improving patient care. The framework quickly grew to encompass a fourth aim of improving the clinician experience; many experts are now calling for a quintuple aim that adds the goal of health equity.
Earlier this year, CMS also launched the Accountable Care Organization Realizing Equity, Access, and Community Health (ACO REACH) model to bring the benefits of accountable care to Medicare beneficiaries in underserved communities. The ACO REACH model will require participating ACOs to identify underserved communities and reduce measurable health disparities within their beneficiary populations.
While much of the national conversation has been focused on eliminating systemic biases and addressing social determinants of health, the ability to access care can also be impeded by health plan’s utilization management (UM) processes to ensure medically necessary care. In particular, the burden of securing prior authorization (PA), a required approval from a health insurer before a member receives medical treatment, can act as a barrier to timely care.
Logic might suggest that this barrier impacts all members equally, regardless of race or socioeconomic status. However, PA can in fact have a disproportionate effect on predominantly poor or minority communities. Providers in underserved communities tend to operate on tight financial margins and have fewer registered nurses and ancillary staff members dedicated to time-consuming manual PA processes. When providers have limited human resources, the PA documentation process takes valuable time away from patients, and is more likely to result in human error and treatment delays.
While it is unknown how often the burden of requesting PA affects the physician’s decision about which therapies to prescribe, it seems clear that resource-strapped providers see PA as a barrier to care for their patients. One study found that 40% of physicians at cardiology practices with large underserved and minority patient populations believe that obtaining approval for newer pharmacological therapies is difficult. For one-third of these resource-strapped practices, staff spend seven or more hours a week completing PA requests and filing appeals.
Reducing the administrative burden on provider staff through an automated UM platform is a compelling way to accelerate the PA process and help reduce care access disparities. However, the right platform must meet several criteria to have the desired impact on health equity.
PA automation must combine digitization with clinical intelligence
For years, the push to fix PA has centered on the digitization of existing processes. Many health plans have implemented electronic PA systems, and there has been some movement to stratify providers and selectively apply authorization requirements based on past performance.
However, simply digitizing PA does not get to the root of the matter: providers fundamentally distrust how health plans have designed their UM programs, as they rarely have visibility into the health plan’s policies. Some states, like California, have prohibited commercial and managed care plans from making medical necessity decisions based on clinical criteria which are inconsistent with the criteria set forth by relevant, specialty- specific nonprofit professional associations.
To foster greater trust in the value and utility of PA, health plans should leverage a system that uses clearly defined, evidence-based clinical criteria—preferably criteria developed by national medical associations—to reflect current best practices. The rules must also be fully transparent, as physicians need to be assured that PA requests are fairly evaluated for the purposes of ensuring safe, necessary care. California is leading the way in this arena, as it requires state-regulated plans to “provide, at no cost, the clinical review criteria and any training material or resources to providers and health care service plan enrollees.”
Most importantly, a health plan’s UM must offer providers meaningful support to help achieve the fastest and best outcomes for its members, which is possible with the integration of artificial intelligence (AI) paired with evidence-based clinical intelligence. AI enabled PA platforms can leverage clinical data, including social determinants of health data, to offer physicians meaningful and timely advice to optimize a patient’s care.
For example, a platform can suggest high-value care choices that are specific to a patient’s diagnosis and demographics.
Such platforms can also support bundled authorizations and value-based care models, which may also help to reduce care variation and speed patient access to high-quality care. Some evidence indicates that bundled care models, such as Medicare’s Comprehensive Care for Joint Replacement (CJR) model, can positively impact existing racial disparities in care. In a five-year study of 1.5 million hip and knee replacements, implementation of a CJR program led to larger reductions in 90-day and 180-day readmission rates for Black patients than for white patients, although this may in part be due to lower utilization of postoperative care among this group.
Alleviate the PA documentation burden on strapped providers
Recent legislative attention on the complex PA process has focused on Medicare Advantage (MA) plans, as an April report from the Office of the Inspector General (OIG) declared that MA plans inappropriately denied 13% of PA and payment requests for covered healthcare services. However, the systemic challenges of PA—including the inconsistent application of health plan rules and criteria, as well as the overarching lack of transparency—also impact members of most commercial and government-funded plans, including Medicaid.
In 2019, a study released by the Kaiser Family Foundation found that in-network claim denial rates by Affordable Care Act Marketplace plans varied widely, with 34 of 122 major issuers denying less than 10% of claims; 45 denying 10-20% of claims; 32 denying 20 30% of claims; and 11 denying more than 30%. At least 9% of these denials reportedly stemmed from insufficient or non-existing authorization requests.
Given the complexity of the PA process, this is no surprise. As the OIG reported, one of the most common reasons for improper denials is missing or inadequate information within the PA request, despite sufficient documentation being present in the member’s medical records. In many cases, providers remain unaware of which services require PA in the first place, as health plans do not necessarily share this information.
Typically, providers exchange information with five or more health plans, each with its own authorization portal, fax process, 278 EDI process, and coverage policies. After physicians submit individual authorization requests for each care service or medication, the health plan can take up to two weeks to review the case and ask the provider for any missing documentation. Once the provider has faxed the requested information to the health plan, the documentation must then be manually attached to the correct authorization request before the case moves to clinical review.
An AI-enabled authorization platform can ensure that providers submit PA requests for all services that require advance approval, as per the health plan’s policies. In addition, machine learning models can parse requests as they are entered, triggering automated prompts when expected information is omitted. These prompts can help streamline the documentation process for providers while reducing downstream delays and denials.
Expedite more personalized member journeys
Understanding the member’s healthcare journey is crucial for improving outcomes, and also, for preventing denials. By taking a more holistic approach to care management, health plans can better anticipate, manage, and approve a member’s needs across an episode of care. An intelligent authorization platform that can extract member-specific data from the electronic health record and pair unstructured clinical notes with the correct authorization request can provide health plans with a more complete medical record.
Using physician input, member data, historical data sets, and social determinants of health, an intelligent platform can also determine longitudinal member journeys, or care paths, for a wide range of conditions. Care paths that consider patient access and out- of-pocket costs, based on patient-specific factors such as geographic location, could provide patients with greater transparency regarding their options for care, enabling higher care plan adherence.
A platform that automatically suggests additional services which might be appropriate for a bundled authorization is also beneficial, as it allows patients to embark upon complex medical care without delay. Instead of submitting several disconnected requests for one member, physicians can get multiple services approved in advance for an entire episode of care, reducing the time and cost of PA for both providers and health plans.
There are numerous factors that contribute to health care disparities in our country. The PA process shouldn’t be one of them. Health plans should strive to provide practices in underserved communities with the tools to intelligently automate their PA efforts, addressing the unintended consequences of manual processes on poor and minority members.
Gina Kim is the chief product officer at Cohere Health, a utilization management technology company that aligns patients, physicians, and health plans on evidence-based treatment plans at the point of diagnosis. She earned a Bachelor of Science degree in Mechanical Engineering, a Master of Science degree in Technology and Policy, and a Master of Science degree in Mechanical Engineering from the Massachusetts Institute of Technology.
