Upping the CDI Game with Enterprise AI, NLP, and Analytics

5 min readJul 17, 2019

Data is the future of the healthcare industry. A recent study projects healthcare data to grow at 36% compounded annual growth rate through 2025. This vast repository of healthcare data can be efficiently leveraged for clinical documentation improvement (CDI).

Electronic health record (EHR) systems have been serving the healthcare industry for years now, but with data growing exponentially, there is a need for better tools and demand for CDI technology. With the implementation of the CDI program, coupled with strategically selected AI and NLP tools, healthcare organizations will truly transform.

Typically, AI and NLP tools require dashboards and analytics, so that professionals in the industry can access actionable data in real-time. With tools of CDI in place, doctors will be able to positively impact patient care through improved key performance indicators (KPIs), models of root cause analysis, and a plethora of information at their disposal. It will offer the crucial competitive edge that healthcare organizations need, to be relevant in today’s tech-savvy ecosystem. After all, clinical documentation is critical for healthcare service providers not only to stay ahead of their competitors but also to improve the entire gamut of patient care.

Documentation demands in the EHR sector have increased substantially. However, manual processing entails poor quality records, redundancy in information, and errors. Not just that, a conventional CDI approach which takes 16.9 minutes to complete an activity can be realized in 5.2 minutes by an NLP model.

Evidently, CDI solutions can help overcome operational roadblocks present in existing clinical data management systems.

Traditional CDI is Not Enough

The fundamentals of medical coding dictate that only what is documented can be coded. CDI fills in the gaps in the data which have been recorded to ensure effective coding and hospital reimbursements. Such specific data promotes relationships with providers both within and outside the healthcare system and acts as the key factor in improving patient care.

Another factor is the growing impact of clinical validation on CDI initiatives. The pitfalls range from financial risk associated with potential denials to operational challenges stemming from the lack of clinical data. These factors, combined with the need for efficient documentation in diagnosis, are urging leaders in the healthcare industry to look beyond traditional CDI solutions.

The Role of AI, NLP, and Analytics

Natural Language Processing and Artificial Intelligence in healthcare are assuming increasingly important roles in the applications that assist CDI professionals. Text- and voice-based NLP can efficiently convert an Electronic Health Record (EHR) data repository into structured and searchable content. This could aid coders to sift through huge data banks with tons of information in very little time and extract only the relevant information. Automating a time- and resource-intensive process to efficiently document data, is the primary advantage of NLP.

In addition to NLP, CDI specialists are also benefiting greatly from elements of artificial intelligence. AI is helping prioritize cases for review while identifying cases with the greatest probability of documentation gaps. Applying AI to meet the requirements of clinical validation can not only streamline the process of detecting inadequate case documentation but also save time and resources that can be devoted elsewhere. AI identifies the gaps faster, reduces chances of error, and enables CDI professionals to smoothly carry out the clinical validation process.

Clinical Data Analytics can help further enhance the documentation and coding process, eliminate errors, and capitalize on opportunities. Real-time analytics can help a healthcare organization to not just extract actionable insights from their data repository, but also locate opportunities to augment coding and data auditing in their CDI operations. These capabilities put together make analytics the final element in the combination of NLP and AI with CDI — the ideal amalgamation for obtaining maximum efficiency in healthcare operations.

ezDI and the Healthcare Industry

In 2014, Halifax Regional Medical Center (HRMC) implemented a CDI program. The processes and query tracking spreadsheets were manual, and queries were handwritten. Moreover, the HRMC staff and the CDI coders experienced challenges with the encoder. The encoder was often unresponsive, causing connectivity problems between encoder and EHR, and frequent downtime in the manual process of updating codes. This led to a wide range of errors during upgrades.

To overcome these set of challenges, HRMC decided to partner with ezDI for implementing an NLP driven CDI software solution. As part of this collaboration, ezDI implemented a full-site NLP-based workflow automation software through computer-assisted Coding and CDI. We then leveraged a completely integrated solution, our proprietary coding compliance, and coding QA software (ezAssess). Consequently, comprehensive CDI-enabled solutions paved the way for precise determination of treatments for improved economic and health outcomes.

Post-deployment, HRMC experienced a 6% increase in the case-mix index (CMI), due to a 12% increase in CC/MCC capture rate. Additionally, there was an enhanced CDI coverage of 24%, which in turn led to a 29% decrease in time to code ED cases and a 19% decrease in time to code inpatient charts. There was also a 13% hike in physician query response rate and CDI Query Agreement rates were a consistent 100%. ezDI helped HRMC optimize their revenue cycle process, where the bottom line impact after a year was more than a million dollars.

Healthcare organizations across the globe can now transform their big data assets into actionable insights and implement them to enhance patient care and minimize the cost to care. Enterprise AI, NLP, and Clinical Analytics in association with CDI can offer a wealth of information and are key to efficient communication of clinical data across care teams. ezDI can help your organization to achieve the same and reinvent your healthcare process for the future.