5 Tips for Selecting Your Second-Level Chart Review (2LR) Vendor Infographic Jul 9, 2020
Second-Level Review with NLP-Enabled Technology Will Boost Coding Accuracy Machine learning technology and advanced natural language processing can detect and extract HCC codes that were missed or miscoded during first-level reviews. Missed codes identified on a second-level review can result in millions of dollars in additional reimbursement for health plans while also improving RAF score accuracy.
The year 2020 is not like any other. For health plans invested in risk adjustment chart reviews, fewer patient visits, challenges to access patient charts, and emerging CMS regulatory changes could elevate the business risk to capture accurate member risk scores. And this could result in a significant revenue loss.
Therefore, it’s imperative for health plans to use expert tools and advanced processes now available to fully capture HCC conditions and accurately document risk scores for their members.
Second Level Review with NLP-Enabled Technology will Boost Coding Accuracy
Machine learning technology and advanced natural language processing can detect and extract HCC codes that were missed or miscoded during first-level reviews. Missed codes identified on a second-level review can result in millions of dollars in additional reimbursement for health plans while improving RAF score accuracy.
NLP and machine learning supplement the second-level review process whether through an outsourced coding staff or using the health plan’s internal coding resources.
Finding the right vendor to conduct a second-level pass could have a huge impact on risk scores and revenue. Below are the key qualifications to look for when researching a second-level pass vendor.
1. Return on Investment
Return on Investment (ROI) for a second-level review project is a function of finding HCC codes that were overlooked in the initial chart review and accurately reflect these additions in member risk scores. Health Plans using vendors with advanced NLP-enabled technology typically see an ROI between 8-12X on their second-level review (2LR) project and show much higher returns than manual review methods. The new incremental HCC codes would likely point to problem areas in the first-level coding, which in turn would provide the analytics to improve your future first-level coding practices. Accuracy matters – make sure your 2LR vendor has a history of finding missing codes and providing an ROI that is well worth the investment.
2. Throughput – Scalability
Time is of the essence when conducting a 2LR and a good vendor should have a Machine Learning and NLP system that is horizontally scalable to handle large volumes of data (including both member charts and claims) quickly. The vendor should also have a coding services staffing model that allows for a quick turnaround of NLP processed chart reviews, and be able to supply both added and deleted HCC codes. As an example, the 2LR vendor should be able to process hundreds of thousands of member charts and produce chart review output files within weeks of receipt.
3. Value Added Insights – Reporting
Data is crucial when identifying inaccurate coding patterns, validating assumptions around member risk scores, and conducting annual ROI analysis. Ask your vendor to provide real-time reporting visibility into project progress, coding operations, and QA percentages/accuracy across the coding team.
Drill-down reporting capabilities in a second-level review expose coding gaps, uncover miscoded HCCs, and improve future first-level coding initiatives. Look for reporting that includes provider coding patterns and insufficiently documented codes that can support your provider education initiatives. Miscoded recommendations in reporting can also help health plans process deletes appropriately to bolster coding compliance.
When selecting a 2LR vendor, make sure reporting is detailed enough to make improvements to your overall coding operation, provider training, and coder training.
4. Adherence to Customer-Specific Coding Guidelines
Be sure that you evaluate the vendor’s ability to configure 2LR functionality to identify codes according to your coding guidelines beyond CMS standard guidelines (i.e., M.E.A.T.). Otherwise, you could be subjected to a high number of false positives and negatives.
Engaging with a second-level review vendor should be fast, and straightforward. Once the data is sent to the vendor, work should begin immediately to find HCC codes missed in the first pass. Ask your 2LR vendor if they also have the right support staff to manage the project as well as technical personnel if the need arises. From contract signing to having patient charts loaded into the review system shouldn’t be more than a few week’s time.
An additional consideration when selecting a 2LR vendor is an audit-defensible balanced approach while conducting a second-level coding review. A good vendor looks for both adds and deletes and this helps to minimize the risk of an audit.
Benefits of Using the Right Vendor for Second-Level Reviews
There are several benefits health plans experience when using the right NLP and machine learning technology vendor for second-level reviews.
Talix’s NLP and machine learning coding platform can extract tens of thousands of additional codes from patient data and dramatically increase coding accuracy. To read case studies and learn more about Talix’s second-level review, visit Talix.com.
Improved coding accuracy associated with the right NLP vendor has both short and long-term benefits. In the short-term, missed HCC codes are uncovered and risk scores are adjusted to increase revenue. The data found in a 2LR can impact future first-level reviews by mitigating inaccurate coding patterns, educating providers for better accuracy, and adjusting member risk scores accordingly.
Second-level reviews help to improve compliance and reduced audit risk by having both add and deleting capabilities in one, unified workflow
Value-added insights provide more visibility into project progress and accuracy, but also areas of improvement like CDI and provider documentation patterns.