Medical billing and coding is an essential component of the health care industry, providing a standardized format for determining the costs of different treatments and procedures. However, while these procedures are necessary for providing fair prices to medical procedures, billing and coding is a practice rife with potential for error, which can affect patients, providers, and insurers alike.
Considering the scope of medical billing and coding, it’s no surprise that mistakes are common. With over 70,000 ICD and CPT codes to work with, many of which refer to combinations of services and procedures, keeping all the different designations straight is no simple task. The upcoming changes to standardized coding systems will likely also add to the potential complications.
Despite these challenges, billing accuracy is essential to the healthcare industry. After all, one misfiled code can lead to claim denials and increasing costs of both treatment and insurance. The importance of this accuracy has led to many innovations. Utilizing artificial intelligence is one improvement that has been building steady traction.
Artificial Intelligence as It Relates to Medical Billing and Coding
Artificial intelligence (AI) has worked its way into many different industries, as machine learning technologies become more advanced. The ability to analyze large amounts of data and then make predictions and decisions that become more accurate over time is an increasingly invaluable tool to the medical billing industry.
Medical billing and coding involve a great deal of data input, analysis, and generation. Medical coders must analyze treatments and procedures, determine what codes apply, and put those into the system to generate patient bills. It’s a rich ground for AI to assist and automate processes.
Text Processing
Many in-development AIs related to medical billing and coding are those capable of analyzing texts. By identifying keywords in medical reports, these programs can then identify which codes relate to provided treatments. By generating the relevant codes from reports, the margin of error related to selecting and inputting code decreases.
The level of text processing can vary among different programs, depending on whether it’s analyzing digital medical records or scanning in hard-copy files. Some programs can even analyze handwriting, taking away the need to transcribe handwritten files into digital format. This reduces the risk for transfer errors, thereby maintaining accuracy.
Some programs capable of analyzing handwriting can also accurately read cursive writing, which is much less straightforward than its print counterpart. The more comprehensive text processing AIs become, the more medical coders will be able to identify the proper codes for treatment.
Increasing Accuracy
One of the issues that leads to declined insurance claims is the inaccuracy of billing codes. Insurance companies are cautious about improper billing procedures, and one error can cause a claim to receive reduced, or no payout. Higher accuracy, thanks to AI assistance, would help reduce this problem.
As one might guess, improved accuracy is one of the major draws to bringing AI into the medical billing and coding industry. However, accuracy doesn’t just stop at analyzing texts and records to determine which procedures require billing.
Out of the extensive amount of ICD and PCT codes, not all of them refer to individual treatments or services. Some codes are intended for cases when a health care professional administers multiple procedures at once. These “bundled” codes are meant to provide greater ease of use over the practice of entering several different codes to refer to the same thing.
AI would be able to recognize when these bundled codes appear and update them to the appropriate single code. Besides helping with accuracy, this would also benefit processing times and help coders keep track of which treatments go together.
Improved Adaptation to Industry Changes
The list of medical codes doesn’t have large-scale changes often, but there are still updates that better reflect the needs of health care providers. In 2019 alone, there will be 392 new codes, 216 deleted ones, and eight revised titles. Complete revisions of the coding standards can also occur, which can involve substantial adjustments to new and existing codes.
While the medical industry can adjust to these changes as it has before, it does take time. And in that time – especially right after an update – the learning curve can lead to more errors as coders become used to the new codes and procedures.
AI systems could help better facilitate these adjustments in many ways. One example could be in suggesting new codes or preventing entering codes that no longer exist, as well as reminding users of any changes a code has gone through. Guided prompts would speed up the process of coders adapting to whatever changes may arise in the future.
Decreased Processing Times
One of the other complications the medical billing and coding industry faces, besides accuracy issues, is in processing time. Currently, it takes a long time for a procedure to be billed, and then additional time for insurance companies to accept that bill and distribute payment. These backed-up payments are a problem for patients, providers, and insurance companies alike.
Improved processing times begin with the coders. All the processes we’ve mentioned above, especially accurate text processing, can save coders time in compiling procedure and billing information. The time saved with this increase in coder efficiency can then speed up the rest of the process, benefitting everyone.
When less time and resources go into compiling bills, medical professionals can focus more of their efforts into maintaining accuracy in more complicated coding scenarios and providing higher-quality patient care. These outcomes can potentially improve the cost of health care services and health insurance.
Further Development of AI for Medical Billing and Coding
Many companies are working on developing AIs suited for medical billing and coding, including 3M, A2iA, and EMscribe, to name a few. It’s likely that even more companies will enter the field as more medical facilities bring AI into their medical billing and coding procedures.
Considering the growing interest in these prospects, further development and research will likely lead to new innovations for coding AIs. Artificial intelligence has the potential to address many of the challenges in the world of medical billing and coding, and only time will show us its true impact.
Sources:
www.techemergence.com/artificial-intelligence-medical-billing-coding/
www.healthitoutcomes.com/doc/unravelling-the-mysteries-of-medical-billing-with-artificial-intelligence-0001
www.pmd.com/blog/post/the-artificial-intelligence-takeover-could-medical-billing-be-next
www.osplabs.com/insights/how-to-boost-medical-billing-business-using-artificial-intelligence/
www.beckersasc.com/asc-coding-billing-and-collections/cms-publishes-procedural-coding-changes-392-new-codes-for-2019.html
www.aapc.com/blog/41446-three-medical-coding-changes/