Workforce shortages, regulatory complexities and economic challenges have put hospitals and health systems under immense pressure to improve operational efficiency, accelerate payments and reduce costs. In 2022, over half of hospitals operated at a loss, and the situation in 2023 wasn’t any better due to a 12.4% economy-wide inflation surge that outpaced Medicare reimbursements for hospital inpatient care [1]. Failure to confront these issues could lead to hospital closures, further limiting access to care, a fate that has already befallen many hospitals across the nation.
Intelligent healthcare technology, particularly medical coding automation, holds immense promise in solving ongoing challenges faced by healthcare providers, and that makes it crucial for healthcare leaders to understand these technologies and their potential. So, let’s examine the inner workings of this promising technology.
Modern medical coding automation technology, using artificial intelligence (AI), has proven to significantly improve operational efficiency and financial performance while reducing reliance on human labor. AI-based medical coding utilizes various machine learning (ML) models, including deep learning (DL) and natural language processing (NLP), to analyze medical records. Done correctly [2], coding automation has the ability to derive and apply from over 200,000 medical codes – ICD-10, CPT, HCPCS, HCC, MIPS and Modifiers – ensuring accurate, efficient and consistent application while remaining compliant with payer contracts and healthcare regulations.
ML involves learning from large datasets, extracting knowledge from these datasets and training algorithms to make predictions and decisions without explicit programming. Traditional programming works well when the algorithms that produce outcomes from the input are known. However, ML is more effective when there are many variables involved. The ML model continuously refines itself through feedback, and its accuracy is influenced by the size of the dataset and the learning algorithm used. Often, ML, by itself, is not sufficient to derive the final codes and may require further algorithmic reduction, combination, calculation, sequencing and validation with provider and payer policies. Therefore, a complete and comprehensive AI-based automated coding platform requires a judicious combination of ML and traditional software comprising algorithms, rules and tables.
In healthcare, ML models predict outcomes and automate tasks by extracting key information from the clinical record using natural language processing (NLP). Large language models (LLMs) improve this process by understanding and generating text based on vast data sets, leading to enhanced reliability and accuracy.
Autonomous coding systems derive codes without requiring user intervention. Adding automation means that the autonomous codes are reliable enough to proceed directly to billing without human intervention or review, referred to as “touchless.” Using an AI-driven coding automation platform means hospitals can increase the speed of their entire revenue cycle, improve coding accuracy with little to no human intervention, and lower administrative costs.
Adopting coding automation technology, like PULSE Coding Automation Technology™ by CorroHealth, increases coding productivity by 7x, processing millions of charts daily. The system’s AI models, all discussed here, complete coding tasks in seconds and the entire process in minutes, reducing workload, accelerating revenue cycles and improving cash flow. This improvement allows hospitals to manage resources for scalable coding operations, moving coders into higher-level roles like auditors or system trainers.
PULSE can go fully touchless, depending on the customer’s desire, trust and confidence in the automated outputs. The level of automation varies depending on specific requirements and confidence thresholds set by healthcare providers. Regardless, after 90 days of use, PULSE has proven to achieve an impressive accuracy rate of up to 97% and offers modules for hospital facility-level coding, such as emergency, observation and risk-based coding. It also offers professional coding such as specialties, diagnostic imaging and MIPS, making it adaptable to diverse coding scenarios.
The impact of PULSE is illustrated in a recent case study involving a large healthcare system, which experienced significant improvements in its revenue cycle after implementing PULSE to address coding operations challenges. This included a 23% increase in net patient revenue, translating to an additional $504 million. Additionally, coding accuracy improved from 85.5% to 98% in E/M and observation, with 90% of all charts fully automated. Turnaround times for completion of charts were reduced by over 70%, and there was a 5% decrease in missing stop time rate for injections and infusions.
Partnering with CorroHealth offers long-term strategic benefits beyond just technology. The company offers a wealth of clinical and revenue cycle expertise, advanced healthcare analytics and tailored strategic solutions to improve hospitals’ financial health. They are dedicated to innovation and continuous improvement, as shown by their collaboration with a leading tech university to research, develop and improve its AI-driven technology platforms. Healthcare providers who partner with CorroHealth gain a dedicated, clinically led ally that continuously drives innovation and delivers results hospitals and health systems can count on.
- https://www.aha.org/system/files/media/file/2024/05/Americas-Hospitals-and-Health-Systems-Continue-to-Face-Escalating-Operational-Costs-and-Economic-Pressures.pdf
- Note: Effective autonomous coding does not result directly from AI but rather from the complete and correct implementation of AI, leveraging datasets, AI models, supporting software and implementation that respects customers’ special needs.
Adithya Mathai Yana Tyler Brian Ruswinkle ScottMadden