EXCLUSIVE: The Wide-Ranging Effects of 2021, 2023 E&M Code Changes 

In 2021, the Centers for Medicare & Medicaid Services (CMS) released a completely revamped version of the evaluation and management (E&M) codes, primarily for the office visit codes (99201-99205 and 99211-99215). This update represented the most comprehensive revision since the 1997 guidelines were released.

While the changes were limited to those office visit codes, they set the stage for the significant changes now seen in the 2023 guidelines. According to CMS, the purpose was “to update coding and payment for E&M visits so that they better reflect the current practice of medicine, are less administratively complex, and are paid more accurately,” adding that “this work is critical to help reduce practitioner burnout out in general, especially in light of the COVID-19 pandemic.”

For 2021, the most significant change focused on the history and exam, which were no longer considered to be “key components” when determining the appropriate office visit code. Rather, medical decision-making (MDM) became the driver for the code level, and was separated into three categories:

1. Category 1: review/ordering of tests, labs, studies, and assessment, requiring independent historian;
2. Category 2: independent interpretation; and
3. Category 3: discussion with an external provider.

We also saw greater flexibility in the time rules. Specifically, the provider no longer needed to document that counseling and coordination of care were responsible for 50 percent of the visit time. Rather, time spent before and after the patient encounter also counted towards the time used to support an E&M code, if the time was spent on the same calendar day. In addition, code 99210 was deleted. 

Summary of 2023 Changes

For 2023, we saw many of the rules implemented for 2021 applied to other E&M codes. This included codes representing services for hospital inpatient stays (99221-99223 and 99231-99233), emergency department visits (99281-99285), nursing facility visits (99307-99210), and all consultation services.

We also saw a more robust consolidation of several code sets: 

• Hospital observation codes 99217-99220 were deleted, and observation services were reported using hospital inpatient codes 99221-99223 and subsequent hospital care codes 99231-99233.
• Observation discharge was reported using hospital discharge management codes 99238/99239.
• CPT® codes for domiciliary, rest home, and custodial care services codes (99324-99328, 99334-99337, 99339, and 99340) were all deleted in 2023. Such visits instead are billed using the home or residence services codes (99341-99345) and 99318 for the lowest-level nursing facility visit code.

Even though Medicare no longer pays for outpatient consultation services, they do maintain those codes within their database. For 2023, level 1 codes (99241 and 99251) have been deleted. Additionally, the codes for “prolonged service with direct patient contact, except with office or other outpatient services” were deleted. CPT 99354 and 99355 are gone, and prolonged services as part of home or residence visits, or cognitive assessment and care plan, are billed using existing add-on code +99417. On the inpatient, observation, and nursing facility side, codes 99356 and 99357 were deleted, and you report prolonged services in these settings using the new code 99418.

Finally, code 99418 was added as a new code, and describes prolonged inpatient or observation time with or without direct patient contact, beyond the required time of the primary services.

Analysis and Findings

Near the end of 2021, we released a study of chart audits that were conducted on office visits to better understand how the 2021 changes impacted both coding levels (as well as to offer validation of those levels). In summary, we found that while there was a significant increase in some of the higher-level new and established office visit codes, they were, in fact, justified, based on the audits we reviewed. We also performed the same analysis on those code sets impacted by the 2023 changes, with particular emphasis on the inpatient hospital visits and emergency department services, as they make up most of the payments for these associated changes.

To better understand the relationship between audit results and utilization, we compared the total utilization of inpatient visits and ED services for the first nine months of 2022 to the first nine months of 2023. Data was taken from our CRA clients, so this may not be representative of the general population, as several of our clients are tertiary care facilities. In total, I looked at 5.2 million visits for 2023 and around 4.5 million visits for 2022. Of these, our clients reported a total of just over 50,000 actual audits, for which we have detailed results. 

Table 1:  Utilization comparison for 2022 vs. 2023

Hospital Visits

In total, our clients reported 1.1 million initial hospital visits (IHVs) for 2023 and 750,000 IHVs for 2022. We saw a substantial reduction in the reporting of 99221 (12.9 to 9.0 percent) and a substantial increase in the reporting of 99222 (35.2 to 40.7 percent), while the use of 99223 shifted downward by only 3 percent. We subsequently looked at 5,519 actual audits for IHV procedures. Approximately 77 percent were reported as properly coded, 7.4 percent were reported as under-coded, and almost 12 percent were reported as over-coded. Notably, code 99221 was reported as under-coded 46.1 percent of the time; this meant a definite opportunity for increased revenue, had those been coded appropriately.

Chart 1: Initial Hospital Visits

For subsequent hospital visits (SHVs), our clients reported utilizations of 3.1 million for 2022 and 3.4 million for 2023. As with IHV, we saw a significant increase in the use of the lowest-level code (99231), from 8.9 percent in 2022 to 11.0 percent in 2023. But unlike the IHV trend, we saw a significant decrease in the use of the 99233 code: from 36.1 percent in 2022 to 31.7 percent in 2023. 

I reviewed 8,900 subsequent hospital visit audits, and 78.5 percent were adjudicated as being coded properly, which was very close to what I found with the initial hospital visits. Of those that were determined to have been improperly coded, 3.5 percent of the total were under-coded, while 15.6 percent were determined to have been over-coded. Note that the incidence of under-coding for SHV was approximately half of what it was for IHV.

Chart 2: Subsequent Hospital Visits

Emergency Department Visits (99282-99284)

The final category I examined in detail was emergency department (ED) services (99282-99285). Our clients reported 693,000 claims submitted for ED services. There was a significant shift from the lower codes to the higher codes between 2022 and 2023 here. For codes 99282 and 99283, we saw utilization decreases of 56.9 and 20.3 percent, respectively. For the highest two code levels (99284 and 99285), we saw a right shift of 25.3 and 2.8 percent, respectively. 

From an audit perspective, our clients reported 1,531 completed audits. Of these, 92.8 percent were adjudicated as properly coded. Of those audits remaining, the majority were found to be under-coded (3.4 percent of the total), with only a small percentage found to be over-coded (1.31 percent).

Chart 3: Emergency Department Visits

Table 2: Audit results for 2023

Summary

In general, hospital utilization distribution changes appeared to be more random than patterned. For example, for IHV, the shift was higher for 99221 and lower for 99223, resulting in an increase in the use of 99222. For SHV, it was an increase in the lowest-level code (99231) and a decrease in the highest-level code (99233). And for ED services, the changes appeared more deliberate, with a definite shift from the lower two codes to the higher two codes. Regarding audit results, about three-quarters of audits for hospital visits were determined to have been coded properly, which would put the error rate at close to 25 percent for both. But within that error rate was a noticeable rate of under-coding, which would indicate a possible financial opportunity. For ED services, the pass rate was around 93 percent, which would suggest that either the changes were less impactful, or coding teams were more well-versed on those changes. In any case, these results should provide a benchmark for the compliance department moving forward.