Bringing Order to Chaos: Augmenting Clinical Documentation with AI
The healthcare industry is undergoing a massive overhaul with US regulations mandating affordable care. In effect, providers are shifting from a service-based payment model to a value-based one, allowing payers to receive better care at lower costs. But this has put pressure on providers to find a balance between quality and cost efficiency. In this situation, providers are turning toward various technologies to streamline and optimize the reimbursement process.
Healthcare finance is closely intertwined with key data metrics. These are heavily reliant on physician notes, patient data, and other electronic health records. Since these affect metrics such as case mix index (CMI), Discharge Not Final Coded/Billed (DNFC/B), and others, it is critical to have meticulous clinical data records. For hospitals and health systems this continues to remain an uphill battle. In fact, not very long ago, we had a client, who before they came to us, had a recorded CMI of 1.11 and a 7.2-day DNFC rate.
Thanks to cloud and artificial intelligence (AI) enabled technologies, the healthcare industry has the opportunity to achieve effective care management with error-free visibility into the reimbursements. Clinical documentation improvement (CDI) services are integral to delivering this promise of accurate reimbursements.
Changing Healthcare Models and Clinical Documentation
With the introduction of ICD-10, CDI programs are becoming more important to healthcare providers (HCPs). ICD-10’s new clinical cataloging system has added more classification options, which the legacy CDI systems are unable to capture. As a result, it offsets productivity, leading to considerable loss in revenue. Our research found that returned and improper claims accounted for $329 million in productivity losses in 2015.
The new CDI program uses a user and process centric approach that integrates the efforts of the front end CDI staff with the back end coders. As a result, CDI services are using technology to streamline the documentation process that is ICD-10 compliant and enhance productivity across all stakeholders.
The Case of Halifax Regional Medical Center
Our engagement with Halifax Regional Medical Center is a great example of how technology and CDI services worked together to deliver a successful solution.
Halifax’s existing CDI program required executives to manually input data into the system. This led to introduction of lots of errors. In addition, the encoder did not support ICD-10 which led to issues with updating the code sets. As a result, the teams working with the CDI program faced regular challenges with the system’s connectivity, responsiveness, and uptime.
When Halifax partnered with us, we wanted to give them a solution that would resolve their current issues and be sustainable and scalable for their future needs. We determined that the best way forward was to build a platform in house that catered to all their needs. Our technology stack consisted of AI-driven, NLP-based technologies that allowed transparency and flexibility in operations and automation for all routine tasks. The solution was scalable and fully integrated with our client’s existing infrastructure.
Our solution generated several benefits for Halifax: an improved case review to query ratio, customized worklists, and Meditech query integration, among others. These improvements helped Halifax’s bottom line as well, and yielded substantial financial gains over time. A one-year bottom line revenue impact was $1.39 MM. Overall, Halifax was able to:
- Improve CMI by 6.04% due to better CC/MCC capture rates
- Enhance CDI coverage by almost 25%
- Reduce time to code by about 30%
Catching up to CDI
Clearly, NLP-enhanced CDI has the potential to transform documentation systems in the healthcare industry. And the ICD-10 is all set to take things to the next level. Consider these statistics: where ICD-9 had close to 4,000 procedure codes, ICD-10 has over 71,000. Also, diagnosis codes have increased from 14,000 to almost 70,000.
Such technological advancements have paved the way for better provider reimbursements. Advanced NLP has the ability to make highly nuanced distinctions. This enables HCPs to record data in greater detail, thereby mitigating reimbursement-related risks.
But, as is often the case with ambitious, disruptive technology, institutions haven’t quite cottoned on. A survey shows that 94% of surveyed hospitals failed to implement a future-forward CDI strategy. The need is to amalgamate new technologies with NLP, such as cloud and machine learning to enhance CDI features. It is only then that priority worklists and seamless physician query workflow into and out of the EMR directly will be successful.
According to a 2016 survey, CDI programs implemented in more than 90% of hospitals with a bed capacity of 150 or more have witnessed significant increases in revenue and reimbursements. HCPs partnering with agencies and investing in implementing an advanced CDI strategy can evidently experience improvements in their revenue collection.
To find out more about how AI-based Clinical Documentation Improvement Software can help streamline clinical documentation workflow and improve value-based care reimbursement, request a demo.
To get the new post update in your inbox, subscribe.
This post was originally published on ezDI.com/blog