Ensuring Quality and Risk Flow Together at the Point of Care

The intersection of quality and risk at the point of care is essential for successful value-based care (VCB). The industry has what it needs to create this intersection. The key is using AI-enabled technology to ensure providers have the right information at the right time to deliver the best possible care.

First, quality assessment has evolved both in terms of structure and measurement. The advent of Healthcare Effectiveness Data and Information Set (HEDIS) measures has put the necessary structure in place so VBC organizations can choose specific measures as the indicators of population-level quality.

Second, shifting risk adjustment from a separate, retrospective function to a more integrated, proactive process. Under VBC arrangements, it's understood that patients have different levels of risk that require other interventions. Rather than assembling that information retrospectively from claims, giving providers the correct patient information at the right time to initiate the best actions and perform effective risk stratification enables a proactive approach.

What makes this possible is bringing together complete, relevant patient data for providers at the point of care. For example, for patients with diabetes, providers need a holistic view that includes appropriate lab results, such as A1C and GFR, recommended specialist visits, such as ophthalmologist and podiatrist visits, and more, to make timely, informed preventive care decisions. Doing that both efficiently and effectively requires both data and AI-enabled technology to reveal timely, useful insights from that data.

Data is not the missing link

The health care industry is drowning in data flowing from all directions. Different approaches have been undertaken to bring it together but have yet to solve this problem. Health information exchanges (HIEs), where billions of dollars have been invested over multiple decades, have become massive information repositories – but they're not curated well enough to wade through efficiently. Electronic health record (EHR) consolidators, prominent industry players who share information, can only make data available within their limited consortium. On top of that, there are other systems, such as lab systems, prescription networks, etc., adding to the silos of essential patient data.

Until recent advancements in AI technologies, organizations needed to manually dig through detailed patient data housed in multiple, disparate systems, which are typically organization- or hospital-centered. Even as AI technologies, such as generative AI, continue to evolve, proven technologies, such as natural language processing and machine learning, can pull data together, deduplicate it, curate it, and summarize it as a concise patient-centered care document that has all the relevant information providers need. That patient summary is where the integration of risk adjustment, risk stratification, and quality come together to drive better patient outcomes and stronger financial performance.

Increasing efficiency while improving provider and patient satisfaction

Fundamental to improving patient outcomes is making the care process as efficient as possible for providers. Providers can spend 20-25 minutes documenting for every 5-to-10 minutes spent with the patient. In addition, there is a well-known problem where providers treat patients for multiple chronic conditions, but typically only have time to document a couple. Consequently, the provider continues to experience burnout, the patient’s risk isn’t captured completely or accurately, and the health plan and provider receive less-than-appropriate reimbursement.

Giving providers comprehensive patient information at the point of care, including a list of all potential diagnoses with as much supporting evidence as possible, equips them to document thoroughly and efficiently. Now AI-enabled tools make it possible to provide the list of possible diagnoses with links to the specific supporting data in the patient’s record. For example, for a suggested diagnosis of chronic kidney disease stage 3a for a patient with diabetes, the summary links to the lab report that shows the diagnosis and supporting lab values. It streamlines the entire process from diagnosis through documentation.

When providers spend more time with patients and less time on documentation, their job satisfaction as well as patient satisfaction increases. This makes the whole cycle flow better – patients get better care, the documentation is better, and the reimbursement improves, which means money can be put back into optimizing care.

Moving from intervention to prevention

Improving the process and equipping providers with the necessary information supports the shift to proactive, preventive care. For example, the approach to care is very different for a patient with early-stage diabetes versus one whose condition is more advanced. Giving the provider a summary that shows the patient's problems, doctor visits, medications, lab results, etc, helps reveal trends over time. When providers have that insight at their fingertips before the patient arrives for their visit, they can quickly see the recommended care the patient has – and has not – received. The provider then focuses on closing the care gaps to help patients achieve their best outcomes.

Providers can also help patients fully engage in their own care using AI tools. A simple document can be produced for patients with all their relevant information, including what they've done well and where to take action to improve their health. When patients take ownership of their care, they are more likely to have better outcomes while preventing deterioration, ED visits, and hospital admissions. If an ER visit occurs, AI-enabled tools can alert the provider to ensure the patient follows through with follow-up care.

Prevention doesn't only apply to individual patients. Historically, population-level decisions are made based on claims data, without the ability to reliably integrate clinical data. Now, AI technology can create dashboards for different populations that integrate claims and clinical data. Dashboards can facilitate analysis, such as:

• Identifying high-risk and rising risk for more accurate risk stratification.
• Targeting patients with care gaps for follow-up care.
• Identifying patients who would benefit from home care, hospice care, or other services.

Many population health management decisions can be enhanced with a tool that can bring all the clinical data into one place.

Embedding regulatory compliance as part of the process

In recent years, hundreds of millions of dollars in fines have been levied on organizations over allegations of reporting incorrect patient data to increase payments from Medicare and Medicaid health plans. Organizations struggle to respond to Risk Adjustment Data Validation (RADV) audits because the claims were based on information captured on paper or via other inconsistent means. Using AI-enabled patient summaries can completely prevent these issues.

With a comprehensive patient summary at the point of care, claims submissions become inherently compliant by embedding regulatory requirements for provider documentation within the process. For example, the supporting clinical documentation is electronically linked for any diagnosis the provider selects for a patient. The combination of high-quality, integrated data that has been curated by AI tools will enhance an organization's ability to be compliant up front as well as audit-ready when required.

Conclusion

AI technologies are poised to transform value-based care by integrating critical quality and risk management functions, enabling a proactive approach to care. Providing access to all the relevant data – integrated, well-curated data – at the point of care drives risk stratification based on the quality metrics that are most important for each patient. The results are significant, sustainable improvements to patient outcomes, population health, financial performance, and patient and provider satisfaction.