Move Beyond "Chart Chasing": Why Prospective Risk Adjustment Should Be the Anchor of Your Program

In risk adjustment, hindsight isn't always 20/20. In fact, it often misses a big part of the clinical picture.

For years, the dominant approach for ensuring that members' diagnoses are accurately and correctly coded involved retrospective risk adjustment—employing legions of medical coders to scour the medical chart after a patient encounter. Like all risk adjustment solutions, this approach seeks to document and reflect the true disease burden of members and populations, so that appropriate resources can be directed to them. Additionally, better coding ensures that providers are reimbursed based on the severity of illnesses that they are treating.

Yet retrospective solutions have long suffered from a blind spot: Coders are at the mercy of what's in the chart. If the physician did not consider a likely diagnosis, the coder can’t go back to change that fact.

There is a better way to carry out risk adjustment than constantly staring in the rearview mirror. Prospective risk adjustment solutions look ahead to the patient encounter, using medical and pharmacy claims as well as other data sources (e.g., lab results, EMR problem lists, etc.) to identify conditions that are likely to exist but have not been coded. Then, at the patient visit, providers are prompted to confirm or deny those conditions, and document them accordingly.

Prospective solutions to risk adjustment have a greater degree of difficulty than retrospective approaches. They require advanced technology, as well as effective methods to spur provider behavior change and the culture of physician practices. But when they are done right, these tactics come together not only to improve the effectiveness of a risk adjustment program, but to transform it from a back-office, financial task to a member-focused activity with benefits beyond accurate reimbursement.

There are several reasons why prospective risk adjustment should anchor your program:

Increases the likelihood of capturing newly inferred conditions

In Evolent Health Service's experience, newly predicted conditions make up 25% of all confirmed coding gaps. These conditions have no prior coding history, but they have very strong correlations to the rest of the member's medical profile. Prospective risk adjustment can capture more of these diagnoses on the front end, using predictive technology to complement providers' clinical reasoning. Learning from the historical data of millions of members, algorithms notice patterns of data points that suggest a disease may be present. For example, a diagnosis of hypertension combined with a prescription for insulin, recent test results for kidney damage and blood glucose could strongly suggest that a member has diabetes. If that diagnosis is not documented, providers should be prompted to consider if the disease is present during a visit. Ideally, that prompt occurs through the electronic medical record to minimize provider workload.

Promotes stronger provider coding over time, reducing future gaps

A prospective risk adjustment program not only pushes suspect conditions to physicians at the point of care, but also provides regular analytics and feedback to help them track their performance. By making documentation issues more top-of-mind, prospective approaches encourage providers to reflexively consider whether there is a new diagnosis to document or an older one that needs reconfirmation. In risk adjustment solutions that are overly reliant on a retrospective approach, coding occurs in the shadows, providing scant opportunity for physicians to learn and improve how they chart. Plans relying on retrospective solutions are stuck paying coding professionals to constantly "chase the chart."

Ensures high-risk members are seen by providers

The plan members with the greatest number of potential coding gaps are often the ones who have multiple chronic conditions and need immediate medical care. Plans employing prospective risk assessment can analyze their entire patient panel and schedule patients proactively, reprioritizing over the course of the year. These visits prompt better care coordination by giving primary care specialists the opportunity to refer patients to specialists. They also provide an opportunity to close gaps in quality measures. Particularly complex members may be eligible for at-home visits, in which their diseases are thoroughly documented and care coordination begins.  

Brings increased alignment with care management and utilization management

The act of capturing chronic conditions through proper documentation has advantages that can enhance both population health and value-based care efforts. More timely, accurate and comprehensive coding improves the quality of data used to identify members as high-need candidates for care management programs. In addition, by providing a more accurate and complete member profile, utilization management teams can make better-informed determinations of medical necessity. All teams—providers, utilization management and care management—benefit from having a more complete, accurate and reliable source of truth.    

Makes risk adjustment a clinically relevant, member-centered activity rather than a purely financial exercise

The more that risk adjustment gets marginalized as a financial activity, the greater the risk of physician abrasion. Well-executed prospective risk adjustment solutions tap into physicians' calling as healers. To most providers, capturing and coding a new diabetes diagnosis is important because it helps identify the member as a candidate for additional resources, not because of incremental revenue it may bring about.

As long as risk adjustment exists, there will be a need for retrospective chart review. However, when more accurate coding occurs at the point of care, retrospective "chase lists" will narrow, decreasing the workforce required to close coding gaps.

So, if focusing on prospective risk adjustment is a superior approach, why doesn't everyone use it? Successful execution of a prospective-first program requires a varied set of tools and capabilities, as well as an ability to partner with physicians, that not all vendors can replicate. Certainly, it requires top-of-the-line predictive algorithms that flag suspect conditions with a high degree of confidence, so providers don't devote precious clinic time to swatting down irrelevant diagnoses. But it also demands an understanding of how to motivate provider behavior change and the resources to bring it about—for example, population health managers who visit physician practices to educate them, deliver performance reports, and help them work risk adjustment activities into their workflows.

Ultimately, prospective programs succeed when they respect the value of physicians' time. Traditional physician training for risk adjustment can be characterized as "How to Be a Medical Coder 101." In reality, small changes in coding and documentation practices can make a major difference to the accuracy of risk adjustment, while enhancing clinical value. A more effective approach demands only the right amount of physician insight necessary to support risk adjustment activity, and it smoothly integrates that effort into existing workflows.

It's not easy, but for plans willing to take on the challenge, the ultimate results will be more than worth it.  


About the Author

Brandon Barber

As Evolent Health’s Vice President of Risk Adjustment, Quality, and Data Science, Brandon Barber leads the strategy, operations, technology and analytics for the company’s efforts to drive coding accuracy and quality performance across our partners. He also leads projects to identify and improve operational efficiencies using data science for Evolent Health Services. Over his career, Brandon has worked to solve complex problems through innovative approaches using large data—in risk adjustment, quality, utilization management, care management, auto-adjudication and partner strategy.

More Content by Brandon Barber
Previous Flipbook
7 Reasons Data Science Should Be Part of Your Risk Adjustment Coding Strategy
7 Reasons Data Science Should Be Part of Your Risk Adjustment Coding Strategy

Big-data approaches can improve coding accuracy while relieving documentation burdens.

No More Articles