Globally, the United States spends almost twice as much as countries such as Canada, Scotland and France, where almost a third of each “medical dollar” goes toward medical billing and other administrative costs. Without a doubt, this figure is increasing annually.
The reality of increasing administrative costs for the US healthcare system is compounded by the fact that most physicians don’t process their own billing. Instead, outsourcing this aspect of their business to other companies, thus allowing insurance companies to pay out claims based on their own actuarial tables and algorithms.
The consequences in a system where the insurance company holds all of the cards results in an all too familiar scenario: both physicians and hospitals will bill accordingly to an expected amount known to ensure adequate payout and livelihood. However, this expected amount set by insurance companies may be potentially disastrous for cash pay patients or individuals without insurance.
While medical coders have benefitted from the job growth at a rate of more than 200%, most medical bills still remain inaccurate. This may be partly due to the update in ICD-10 (International Statistical Classification of Diseases and Related Health Problems) requiring more specific coding.
Sources of human errors, resulting in this miscoding, include:
– Duplicate charges
– Upcoding: a code inaccurately is changed to a different code representing a more severe diagnosis or treatment)
– Unbundling: an over encompassing code is changed and listed as several
With new coders, codes, and an ever constant backlog, the potential for errors in compliance and decreased revenue is easily apparent.
Treating a symptom but not the root cause comes with the advent of artificial intelligence and machine learning. Full automation of medical billing is years away but the current computer assisted coding (CAC) serves to aid coders in the ability to perform their jobs.
Natural language processing allows machines to recognize words and decipher their relationship and meaning with other words within medical context. Ultimately aiding coders in coding, auditing, and efficiency.
At the end of the day, both coders and providers should be looking forward to an artificial intelligence (AI) based solution.
For initial billing, AI works to find an initial and accurate set of codes by cross referencing a physician’s consultation notes with an ontological dictionary, allowing coders to better spend their time on complex tasks which computers are unable to perform.
At the end of the day, artificial intelligence does not replace coders. It provides a platform for coders to catch potential errors, well in advance of recovery audits.
Errors may still occur with Artificial Intelligence, however, the overall improvements will ultimately be an advantage for both coders and providers regardless.
DocCharge is developing an AI based platform to improve coding specificity within the mobile application and portal. The goal is to deliver coding specificity and medical necessity information to physicians and coders prior to submitting the claim. This ensures reduction of denials and improved revenue for the practice.