Measuring success in the value-based care era requires establishing a baseline and implementing new technology tools.
The U.S. healthcare system is currently shifting to an incentive structure where health plans and providers are rewarded for delivering higher quality care with better efficiency. This transition from fee-for-service to value-based care is not only changing how patients are cared for, but also how health plans and providers are measured and compensated for performance.
With an ever-growing number of patients covered under value-based care programs, it’s crucial for payers and providers to collaborate on risk-adjustment performance so that they can position themselves for financial success in risk-sharing arrangements and deliver value without disrupting existing work flow.
Importance of risk adjustment
Risk adjustment in payment models refers to the practice of accounting for the differences in the underlying risk (i.e., expected costs) of patient populations. It would be unfair to compare the costs incurred of a healthy member to that of a sick member without properly adjusting for the expected cost of each person based on her/his health status. However, risk adjustment is not just a payment model mechanism. Successful capture of risk enables obtaining a complete and accurate picture of your members’ health status, which is critical to ensuring proper reimbursements, effectively managing costs of your high-risk members, and delivering high-quality care.
The health plan view
Managed care organizations, through participation in programs such as Medicare Advantage, Managed Medicaid, and ACA, are well-aware of the importance of accurate risk adjustment. However, currently risk adjustment is a very manual and inefficient process at most health plans-medical record acquisition and record review for coding are very expensive processes and leverage very little technology. Additionally, most of the analytics used to stratify members based on risk only utilize claims data leading to ineffective gap recognition and physician abrasion during the risk capture process.
New technologies that integrate with EHRs to acquire data and provide natural language processing (NLP) and machine learning (ML) tools dramatically change the risk adjustment expense and return. By automating medical record acquisition, an expensive part of the risk adjustment process can be eliminated. Additionally, using NLP and ML, the risk adjustment team can focus on members that have outstanding risk factors and save precious capacity from being spent on reviewing members with accurate health statuses. Finally, using NLP and ML based analytics that rely on clinical evidence to surface suspected conditions to fuel prospective risk adjustment campaigns have a much higher likelihood of success with low physician abrasion.
Even though health plans have been dealing with risk adjustment for years, the technology is now available to lower costs and improve financial returns on investment by a factor of five to 10.
The provider view
Providers have recently begun entering risk-sharing arrangements through contractual vehicles such as Medicare ACOs as well as Medicare Advantage partnerships with health plans. As a result, effective risk capture is now a high priority for health systems and physician groups. Some of the key hurdles providers cite that prevent them from entering risk-sharing arrangements is the lack of access to administrative and claims data, and the required risk infrastructure to allow effective management of care and costs for patients. As organizations take on more risk burden across multiple patient populations, it is important to set up a comprehensive risk adjustment program that can cover all relevant populations.
For providers new to risk adjustment, a comprehensive technology platform that utilizes NLP and ML while integrating into their established EHR workflow is an essential step toward accurate risk capture. Such a platform would support starting with a post-encounter coding review often provides the highest immediate ROI because it enables risk capture from diagnoses that have already been documented, but not coded, without requiring physician work flow modification. By using technology and bringing the retrospective review more upstream for a concurrent work flow, providers can ensure that claims accurately reflect the care provided and diagnoses are adequately substantiated before they go out the door. Prior to putting any information in front of physicians, there is an opportunity to leverage NLP and ML analytics on historical patient data to identify prospective risk capture gaps pre-encounter. Providers can identify gaps such as patients with suspected conditions who lack the proper documentation to be submitted for reimbursement. Advanced analytics technology can also ensure that only the most accurate intelligence is being put in front of a physician at the point-of-care to increase trust of the providers and reduce any potential for physician abrasion.
During encounter with a patient, a point-of-care solution can deliver the gaps to the physician and allow the closure of gaps without departure from the physician’s EHR-centered work flow. A successful point-of-care solution must also facilitate the documentation of outstanding risk conditions through accurate and complete encounter documentation in the EHR. This allows providers to efficiently capture the risk of their populations, decrease the need for follow-up visits, and streamline coding review processes.
This encounter-centric framework and phased implementation approach allows organizations to develop a plan for continuous improvement and transformation. Once an organization has developed a more mature risk adjustment infrastructure, it can then layer in additional technology and analytics solutions to identify gaps prospectively and optimize their risk capture at the point-of-care.
Value of understanding the status quo
The best start to improvement requires that organizations understand where they currently stand and, accordingly, set a baseline from which progress can be measured. Technology platforms that utilize NLP and ML are now available to help payers and providers achieve accurate risk capture at a fraction of the cost of legacy solutions. Organizations will need to adopt a comprehensive but customized solution that is fit for their readiness, infrastructure, process, and people that will serve as the blueprint for achieving success.
Anand Shroff is cofounder and chief development officer of San Mateo, California-based Health Fidelity, a provider of technology solutions for healthcare organizations.