Eliminate Risk in Risk Adjustment with Innovaccer’s Suspect Coder Assistant

To err is human. Under pressure or when faced with disorganized or hard-to-find information, we humans are even more likely to make errors. In other scenarios, missing a piece of information might have minor consequences, but in healthcare, it can significantly impact patient safety, operations, revenue, and even regulatory compliance. For payers and providers who support Medicare Advantage (MA) and Commercial Exchange plans, risk adjustment coding is highly susceptible to mistakes, not to mention a crucial operational task.

Risk adjustment was developed to predict future costs for a group of beneficiaries under a specified health plan. It simply equates the health status of a person to a numerical value called a risk score. The risk to a health plan insuring members with expected high healthcare use is adjusted by also insuring members with predicted lower healthcare costs. This ensures that cost-sharing of expenses is spread across all members to provide access to quality healthcare regardless of health status and history. Accurate risk coding is crucial to effective value-based care delivery. A correct reflection of the patient’s clinical condition and a proper assessment of their risk factors are paramount for ensuring both high-quality care and optimal financial outcomes. However, inadequate documentation of clinical records can lead to incorrect coding and risk mapping. This can lead to lower quality of care, a higher probability of claim denials, and reduced physician performance. Research on Medicare Advantage (MA) shows that accurate risk coding can increase risk scores by 7-10% and has led to revenue increments of $12 billion in 2020 alone for MA plans and providers.

What is the HCC Coding Model?

Risk adjustment is based on the Hierarchical Condition Categories (HCC) Coding model. In simple terms, HCC is a list of diagnoses with an assigned value for risk adjustment. The Department of Health and Human Services (HHS) updates the HCC list yearly, however, each risk adjustment payment model uses its own variation of the yearly updated HCC list. All conditions coded in ICD-10-CM are organized into diagnosis groups of body systems or disease processes and these groups are subdivided further based on similar cost patterns. The final HCC list (also referred to as HCC crosswalk as it shows which HCCs cross to which diagnosis codes) includes only the diagnoses that are likely to impact long-term healthcare costs related to clinical and/or prescription drug management, particularly to the demographics of the specific risk adjustment payment model.

What are the Challenges in Achieving Accurate HCC Risk Adjustment Coding?

A patient may visit multiple providers throughout the year with each submitting a claim for services. Not all providers will document details of a condition identically and this is where the hierarchical system comes into the picture. The HCC value of the most severe condition in the hierarchy would be used for risk score calculation for that member. In addition to diagnoses, demographic factors such as age, sex, socioeconomic status, disability status, Medicaid eligibility, and institutional status is also used to accurately calculate risk scores. It seems a very well-defined methodical approach to capturing codes. But even for an expert medical coder risk adjustment coding is not easy, here’s why:

Voluminous Healthcare Data

Let’s say a provider has identified 1,000 patients for Medicare Advantage. On average, each patient has 10 charts and each chart has 20 pages. Ensuring accurate coding would require administrative work of reviewing a total of 200,000 pages. Even if the work is spread across a team, it still is a considerable amount of work. Due to the pressure and time constraints of the clinical environment, coding can often be too aggressive or passive, leading to missed codes and improper use of modifiers.

A Multi-Step Process Involving Multiple Stakeholders

For a streamlined and accurate risk adjustment process, physicians must ensure all the relevant information related to diagnoses is captured accurately. Based on the information captured, designated personnel must ensure that claims are coded correctly. This is followed by the addition of HCC code mapping by medical coders as per the HCC crosswalk. Any error at any stage can impact the rest of the chain. If a physician fails to add accurate diagnosis information, medical coders cannot map and submit final HCC codes.

The industry needs solutions that can help clinicians and staff enable accurate coding at the point of care so that all supporting documents and evidence are collected and recorded before the patient is discharged. This is the path to reducing denials, increasing revenue, and improving care. The right technology can help providers address these challenges by analyzing dropped or suspect codes to identify prospective coding opportunities and claim reimbursements for the delivery of appropriate, quality care. Moreover, automating a large amount of manual effort to capture and update codes regularly gives clinicians more time for patient encounters, further improving care quality, improving job satisfaction, and helping to reduce burnout. Improving coding accuracy with automation also helps providers get a comprehensive, accurate view of the patient population, helping them enhance risk stratification and improve outcomes from their risk-based contracts.

Error-Proofing Risk Adjustment Coding Right from the Outset: How Can Innovaccer Help?

Innovaccer’s Suspect Coder Assistant, equipped with advanced AI analytics, performs continual retrospective analysis of historical data to capture missed and suspect codes. It identifies coding opportunities based on claims data and suggested clinical markers, helping avoid incorrect data submission and subsequent audits, repayments to CMS, penalties, and other legal consequences for incorrect coding. It seamlessly integrates into providers’ population health management, analytics, and physician engagement solutions to enable point-of-care coding support.

The effective use of Suspect Coder Assistant can lead to an estimated 10% increase in coding accuracy through increased dropped and suspect code capture, an approximately 20% decrease in physician coding time at the point of care, and a 30% reduction in total coding time spent by physicians and coders through automation.

Powered by Innovaccer’s #1 Best-in-KLAS Data & Analytics platform, Suspect Coder Assistant leverages unified patient records, Machine Learning (ML), and Natural Language Processing (NLP) to analyze unstructured data and identify ancillary suspect conditions or diagnoses for comprehensive risk capture. It also gives users the ability to customize the solution to suit their organization’s specific and unique goals and priorities.

To learn more about how Suspect Coder Assistant can help you optimize coding accuracy for improved value-based performance, schedule a demo with our experts.