Why does it matter if the clinician notes the patient’s issue as diabetes or diabetes with complications if the amount billed will be the same either way? When healthcare providers move to value-based contracts, it is important that they are paid enough in the budget for the population to cover the sum of the estimated costs of care for each patient. Risk adjustment, or scaling an average budget up or down based on the documented health of the patients, is how budgets are changed to represent the overall health of that population. Risk adjustment models consider chronic conditions and demographic information that might be valuable for determining patients’ approximate health expenditures. When a clinician assesses a patient’s chronic condition, this information must be submitted on a claim so that provider organizations can be fairly reimbursed by insurers.

One method that supports risk adjustment, known as “chart scrubbing,” is the review of a patient’s electronic medical record (EMR or the patient’s “chart”) to proactively address discrepancies between what appears on the patient’s clinical problem list (e.g., diabetes, or diabetes with complications) and what has been billed in patient claims data. By ensuring that the problem list is accurate, chart scrubbing can also provide more complete information to clinicians so that they can provide the right preventive care to patients before they get sick or end up in the hospital.

When healthcare providers move to value-based contracts, it is important that they are paid enough in the budget for the population to cover the sum of the estimated costs of care for each patient

While some organizations employ certified professional coders for this chart scrubbing process, Atrius Health, the largest independent medical group in the Northeast, recognized that registered nurses have greater clinical training, experience and judgment. Therefore, three years ago we put together a team of registered nurses to perform this crucial responsibility. In doing so, our goal was to leverage their expertise to improve the quality of information presented on the problem list, including adding conditions that are missing and better specifying certain clinical issues. We wanted these nurses to recognize these clinical nuances, determine appropriate diagnoses, and reflect that in the EMR.

After getting the buy-in of our primary care providers to support this plan, we set out to build a team. At the time, it was challenging to find nurses who had a diagnosis coding background. So instead we looked for people with the right aptitude and interest in this work. While recruiting, we developed a test that mimicked the work nurses were going to do in this role, which allowed us to assess their clinical judgment and confidence. We presented the candidates with different cases of patient information. After they had time to review and think about it, we asked them to determine if it was appropriate to add certain conditions to the problem list and to explain their thought process. Overall, this was an effective screening tool to predict success in the role.

Once we assembled our team, we had a robust onboarding process with our coding manager, who taught them about diagnosis coding as it relates to risk adjustment. As nurses honed their IT skills and their understanding of ICD-10 diagnosis structure, over time we set up productivity and quality standards to ensure efficiency and trustworthy work.

To support nurses in achieving these goals, we developed standard work and clinical guidelines under the leadership of our medical director. We instituted a quality assurance process to ensure that our team was following the guidelines and we regularly review cases together so that our collective knowledge is up to par. All of these nurses work remotely; therefore, it is important to have weekly check-ins so that the team discusses progress, asks questions and notes observations, and has the context to appreciate the value of their work.

The feedback from our clinicians about the work of the nurses has been overwhelmingly positive. They appreciate that our nurse chart scrubbing team can save them steps in updating problem lists, and they’ve been pleased with the quality of the work. In addition to saving time for clinicians, this team has helped improve the quality of care we provide and our treatment of certain conditions. While reviewing and stratifying aggregate data from the nurse review, we identified patterns where clinicians experienced difficulty in diagnosing clinical issues. For example, we found that our primary care clinicians are under diagnosing peripheral vascular disease. As a result, we consulted with our podiatry department, identified barriers and are now partnering with them to develop improved workflows for testing and referrals.

While it is certainly an investment to pay for an RN instead of a coder to do this chart scrubbing work, there is so much more a nurse can do to save on time and health care costs and improve quality in the long run. By leveraging clinical expertise and embedding it within our technological and analytics capabilities, we can ensure that our patients are receiving the highest-value care.