Contributors: Dr Sarah Gebauer, Kelly Ayotte, Jasmine Balloch, Dr Michael Chuang, Dr Tanner Dean, Dr Srikanth Mahankali, Dr Mishaal Ali, Dr Shoreh Irani, Dr Justin Graham, Davide Locatelli
In this white paper, we discuss major issues related to billing in the United States and provide physician insight from a recent summit focused on this topic. The summit brought together clinicians and former clinicians from across the United States alongside a professional coder from Boston.
We discovered three common themes:
- Common issues in billing
- Gaps in physician knowledge and desire for ease of use
- Future directions and challenges for AI use
Common Issues in Billing
E/M Levelling
E/M (Evaluation and Management) levelling is a common source of incorrect billing and physician confusion. Physicians often misunderstand the requirements or don’t have enough time to thoroughly review the criteria to level visits. The requirements for E/M levelling changed in 2021 and 2023, which has led to significant changes in the way physicians level their visits. For example, the levelling for office/other outpatient visits and inpatient/observation hospital care are now levelled based on either medical decision-making or total time. Further, the criteria for levelling each of the medical decision-making elements and adding up total time have been updated.
Another aspect that confuses physicians is that the verbiage used in the guidelines can have different meanings from a coding perspective versus a clinical perspective. Physicians may have communicated everything that is clinically necessary, but it may not reflect all the services or the specificity that is needed. An example of this is the levelling of each E/M visit being reliant on the verbiage within the note. If the physician independently reviewed and interpreted imaging, that clarification along with their findings need to be explicitly included within the progress note to get Category 2 data management credit when not separately reported. This often ends up being the deciding factor between a level 3 or a level 4 visit.
Increasing Importance ICD-10-CM Codes with Value Based Care Leading
Physicians are also concerned about quality reporting due to changing quality structures and managing risk for professional services. Physicians historically have been compensated on a fee-for-service model and as a result were not financially motivated to focus on diagnosis coding. With quality reporting measures, more emphasis has been placed on the case mix and complexity of the cases, promoting the importance of coding completeness and accuracy including ICD-10-CM in the United States. It is important to ensure that our evolving healthcare IT systems helping to support quality improvement initiatives are compliant, respect patient privacy, and align with contracted payer requirements. However, this can be complicated as physicians may have different perspectives on what constitutes appropriate documentation and services.
Coding and sequencing ICD-10-CM codes to reflect diagnosis codes is another challenge physicians face in medical billing. There are rules as to how to select the code of highest specificity and how the codes should be properly sequenced, and there are instances when certain codes cannot be coded together or a more appropriate code should be reported. Coders do not expect physicians to understand all the nuances of coding guidelines, but they do require ongoing physician engagement to ensure the completeness and accuracy of the health record.
Physician Knowledge
Lack of Physician Billing Education Leads to Confusion
Most physicians are taught billing on the job, if at all. Multiple physicians reiterated that they had limited knowledge about billing and were never given any resources or education about billing with regards to financial or legal issues. In addition to starting with minimal knowledge, when the billing changes occurred in 2021 to decrease the copy and paste phenomenon, physicians reported they had cursory notification and no education about how to comply with the new regulations. One physician noted that providers would need to “unlearn bad habits” from the previous billing requirements, and another noted that it seemed like there were “secret rules” about billing that they didn’t understand.
Physicians desire ease of use and real-time interventions
Real-time prompts and integration with clinical notes can significantly enhance the ease of use for physicians. Real-time prompts can provide immediate feedback and guidance during the documentation process, ensuring accurate and complete billing information. Integration with clinical notes allows for seamless incorporation of billing information into the overall patient record, reducing administrative burden and improving efficiency. Highlighting these features can demonstrate how technology can simplify medical billing processes. One physician suggested that AI could read unstructured data and integrate it into ICD-10-CM codes as well as meeting supporting documentation requirements. One example is supporting documentation for the CMS 2-midnight rule, which is used to determine whether admission to outpatient or inpatient status is considered to be reasonable and supported by medical necessity. Another example would be validating supporting documentation for a condition that is “extensive and progressive” in order to make the appropriate code selection.
One potential barrier to improved AI coding and billing suggestions is the compliance requirement that coders and billers provide education and guidance for cases without leading the physician to a particular response or outcome, as this could promote fraudulent billing practices. For example, it would not be appropriate for a coder to tell a physician what to document in a patient’s note in order to qualify for a higher level for that encounter. The coder would need to follow the physician query process to ask questions of physicians case-by-case to identify whether there is an opportunity, and then advise the physician on whether a documentation amendment or addendum is appropriate to support the correct coding. It is critical that coders, billers, and other health information management staff provide physician education and training on the rules and regulations from a data quality perspective. Improving data quality does not always lead to higher reimbursement, but it does protect and justify reimbursement.
Future Directions
Professional coders to take on an auditing role
Advancements in automation and AI will modify the role of professional coders. It is likely that the profession will take an audit approach to coding, one in which AI facilitates the practice of coding individual cases whilst coders conduct both randomised and data-driven case reviews, proactively and retroactively. Coders envisage a system which flags complex cases and outlier instances of physicians billing at higher or lower levels than others within their speciality. AI could also play a role in facilitating prior authorisation acquisition; hospitals may work with vendors to obtain prior authorisation which incurs AI-solvable challenges such as vendor relationship management and staff turnover.
Capturing Data via Ambient Documentation
As we progress into an era in which ambient scribes are commonplace, physicians expressed hope that the information necessary to generate an accurate ICD-10-CM code, an accurate problem list, and an E/M code would be completed by AI without physician input. One physician noted that he would like to see AI help with the risk documentation for inpatient stays instead of having to justify the medical risk. For example, an AI system could flag homelessness as a factor for medical risk, which physicians may be likely to forget: in the event of a domicile address, an AI co-pilot could alert the physician to consider how this might impact coding.
Wearables
Wearables and remote patient monitoring (RPM) present exciting opportunities and challenges for medical billing. The increasing use of wearables allows for continuous monitoring of patient health data, which can inform treatment decisions and improve patient outcomes. However, integrating wearables into existing billing systems may require careful consideration of reimbursement models and data privacy concerns, and the CPT coding and billing for consumer wearables are still new and subject to change.
Remote Patient Monitoring
One physician noted that they had integrated remote patient monitoring into their clinic’s care but it had required three physician informaticists, a hospital executive, and multiple support staff to facilitate the billing aspect. Multiple physicians noted the importance of being able to bill to “justify” new clinical interventions on a cost basis. In addition to billing issues, remote patient monitoring also poses challenges related to data management, privacy, and ensuring accurate reimbursement for remote services. Addressing these challenges will be crucial for successful implementation of wearables and remote patient monitoring in medical billing.
