Addressing Gaps in Coding for Virtual Care

By Susan Chapman, MA, MFA, PGYT for For the Record

After the COVID-19 pandemic began in 2020, patients and providers turned to virtual care to continue to seek and provide care safely. The United States Government Accountability Office (GAO) reported that in the early days of the pandemic, millions of provider visits were postponed or canceled. In many cases, these typically in-person visits were instead conducted by phone or video conference.¹

Medicaid patients, who number some 76 million, already had access to telehealth before the pandemic, but Medicare, which serves roughly 64 million older Americans and individuals with certain disabilities, offered limited access.1 However, with the pandemic’s advent, Medicare relaxed its restrictions, allowing all of its beneficiaries across the United States, new as well as established patients, to receive telehealth services. Clinicians were also permitted to provide new and established virtual check-in services and services by telephone.²

To help states adopt telehealth more efficiently, CMS created a toolkit and encouraged states to “consider telehealth options as a flexibility in combatting the COVID-19 pandemic and increasing access to care. States are encouraged to facilitate clinically appropriate care within the Medicaid program using telehealth technology to deliver services covered by the state.”²

As telehealth restrictions loosened, the use of virtual care services increased dramatically. According to the GAO, telehealth use under Medicare “increased tenfold from about 5 million services (in April–December 2019) to more than 53 million services during the same months in 2020. Spending on telehealth services also boomed—increasing from about $306 million to about $3.7 billion during those time periods.”¹

Despite patients’ and providers’ increasing use of virtual care, the medical coding infrastructure has yet to keep pace with this new trend in health care. Stakeholders—including policymakers, payers, and providers—have been unable to track the proliferation of e-visits, particularly when patients do not have relationships with their providers and when short messaging service (SMS), or secure text messaging, is used to offer patient care.

Zach Gaumer, principal at Health Management Associates, is the lead author of the study “Gaps in the Coding Infrastructure Limit Innovation of Virtual Care Services,” which addresses this coding challenge. From their research, Gaumer and his team determined that innovations in the infrastructure of CPT and HCPCS were necessary.³

Specifically, the team found that the infrastructure of CPT and HCPCS codes lacked a modifier code that could differentiate secure text messaging from other asynchronous modalities. Not having that level of assessment prevented payers and policymakers from establishing accurate payment rates for such services. They also were unable to assess the quality of care delivered through them. The report states, “[W]ith this granularity of information, researchers may assess if new innovative virtual care services, like secure SMS, are capable of increasing access, reducing costs, and improving quality.”³

Another area identified as having a coding gap was centered around e-visits and virtual check-ins for new patients. There, the team found that the infrastructure of CPT and HCPCS code sets lacked codes to define them. Because such codes did not exist, payers and providers were unable both to extend that type of care to new patients and also track the use of that care and its quality for existing patients. This was a notable concern for the researchers, who recommended that in spite of the gap they revealed, it was important for all patients, but especially those with limited access to care, to be able to quickly connect with clinical guidance through secure text messaging.

Overall, the team recommended that CMS and the coding authorities that maintain the CPT and HCPCS code sets address the current gaps by “creating a new set of time-based e-visit CPT and/or HCPCS codes specific to patients without an existing relationship with a provider, and creating a new unique modifier code to report that care was rendered via secure SMS communication.”³

The Study’s Origins

Along with his colleagues, Gaumer, a policy analyst with nearly 20 years of health policy experience, became curious about what he observed during the COVID-19 pandemic and began having conversations with policymakers about what had been happening in the field, trying to determine where there were gaps. “I think, at one point, we got into different modalities of telehealth and realized that coding for telehealth was fairly incomplete,” he explains. “Modalities were ‘bundled up,’ something that didn’t enable payers or researchers to identify different types of telehealth that were taking place.”

Over the years, there have been different iterations of virtual care. For example, five years ago, in the Medicare payment system, it was more difficult to tell where telehealth services were taking place and in precisely what form they were being delivered. “What was the modality at that time?” Gaumer asked. “Was it audio or video? Was there someone present in the room helping a person conduct the telehealth visit? All of those details were very difficult to determine.”

From the questions Gaumer and his fellow researchers were asking grew the need to identify gaps currently in coding that have prevented researchers and payers from knowing what was happening. The researchers were also curious about the potential existence of payment differences based on the type of virtual care being delivered.

To that end, the team looked at such delivery options as SMS, email, and audio-video communications. Additionally, they investigated whether the provider was keeping records of the modalities they used. The team also wanted to know if the person who was being treated was a new or existing patient. “All of those details have payment implications, and there might be payment differences because there are cost differences in how those services are conducted,” Gaumer says. “The impetus for our investigation was to make the process more transparent so that the payment could be more appropriate.”

Gaumer clarifies that the team’s goal of transparency was not necessarily with the aim that the payer would pay less. “Our thinking was that in order to pay for services appropriately, CMS and other payers had to get the coding right, and this needed to be corrected as soon as possible,” he says. “Accurate and detailed coding practices could then benefit everyone in the environment, whether it was payers, providers, or most importantly, patients.”

Impact on Patient Care

While changes to the coding system itself would affect all areas of health care relatively equally, understanding how virtual care was being delivered could have an impact on access to virtual services for those patients who typically would not have access. “If the recommendations we suggested were to be implemented, it would expand access to patient care,” Gaumer says. “In areas where there are not enough providers, such as in some rural and urban areas, virtual care would represent another touch point that patients could use to gain access to providers, particularly for new patients.”

He adds that rural environments and other areas that have limited or no access to broadband could also benefit. “One could argue that making changes to the coding infrastructure would allow patients in those areas to have greater access, through cell service, to some of the more simplistic forms of virtual health care, including SMS, email, and even online patient portals.”

Gaumer points out that when the COVID-19 public health emergency (PHE) expires, under Medicare, those patients who are without a current regular primary care doctor could lose access to virtual care. “When the PHE ends on May 11, 2023, Medicare technically will not be covering e-visits and virtual check-ins for patients without an existing relationship with a provider, and it would take a change in the law or regulation for that to happen. So, this type of coding would be the kind of foundational change that would be necessary to at least track if payers decided that they wanted to cover services for those patients,” Gaumer explains. “That change would be good for older Americans, pediatric patients, anyone that needs to have access to a new provider, or anyone that would need to have access through SMS or online, portal-based services.”

One of the trends Gaumer and his colleagues noticed was that payers had been establishing virtually focused insurance plans that encouraged patients to use virtual health care as much as they would like. “If the payers are going to go heavy on virtual care, there needs to be a solid infrastructure to begin to track the different forms of virtual care that are taking place,” Gaumer says.

Creating the Appropriate Codes

Gaumer notes that the American Medical Association (AMA) and CMS have been efficient in adding new codes when necessary. “They do a lot of hard work and do their work efficiently. But when a new code or a new service comes along, it takes the AMA about two years to get a new code fully in place,” he explains. “Mainly, they have to create a value for the code, and that takes data. In order for them to value the code, services would have to start, and they then would take the research from those cases to value the code relative to all other types of services.”

What Gaumer and his team propose would take less time. “A new code for new patients, essentially a carbon copy of the codes used for existing patients, would take about a year because they’re using an existing code and changing the description. This requires a year to navigate the regulatory process for approvals,” he says. “The second thing that we’re proposing, which is a new modifier code, takes less time because all that entails is a code that gets added to a claim, indicating a modification to some detail about the service, in this case, the mode of communication. It would take less time to do. In fact, we see CMS creating new modifier codes all the time, and they do it within the regulatory cycle, which is not as time-consuming.”

An issue with the regulatory cycles, though—which are in place to protect patients and providers—is that innovation happened very quickly during the COVID-19 PHE, Gaumer says. “Those innovations happened faster than the regulatory cycle could keep up, and we ended up with a number of temporary rules that were coming out of CMS. But, eventually, we’ll catch up, and we’ll get the types of codes we’ve recommended inserted.”

Guardrails in Place

Gaumer and his team suggested that payers build guardrails around new services to control for misuse or fraud. Use of such guardrails would accelerate the researchers’ solutions while also mitigating fraud risk. However, once patients have had the widespread experience of virtual health care, it may be harder to prove that the risk of fraud could be minimized. “With any hospital or physician service, there is a risk of fraud, and it usually takes the regulatory community or the policymakers a little while to get comfortable with new services, new technologies. Prior to the PHE, concern about fraud related to virtual care resulted in payers limiting their coverage of these services. However, given the increased level of comfort and convenience patients and providers have with virtual care in 2023, I suspect coverage of virtual care will not be scaled back to 2019 levels,” Gaumer says. “I also believe that those who audit claims have become savvy about identifying virtual care-related fraud, and this mitigates some of the fraud risk. What’s happening right now is we’re trying to figure out what is fraud and what is not. We’re getting smarter collectively about that, but I don’t think fraud risk will resolve completely.

“As it relates to virtual care,” Gaumer continues, “there is always going to be some kind of risk. I’ve heard people argue that any fraud that takes place with regard to virtual care, which is essentially low-paying services, virtual primary-care services, which are among the more inexpensive services in health care, those are necessary simple services. If there is some fraud, it won’t bankrupt the system. But my counter to that is that we have auditing systems that can catch fraud more easily than we had before, or better than we had before, and we’re just getting better at tracking and clamping down on it. So, it’s going to take a little bit of an adjustment, but I don’t think we’re going back to where we were in 2019. I think some of the innovations ushered in as a part of the PHE are here to stay, and the providers are very wary of submitting claims that might be perceived as fraudulent.”

The Future of Virtual Health Care

Gaumer believes that the use of virtual care in the years ahead will look very different than it did in the years prior to 2020. “Patients, providers, and payers are likely to identify more efficient models of care delivery that blend virtual care with in-person care,” he says. “This is likely to mean greater use of virtual care services such as care through SMS, online portals, email, and audio only, which will increase the touchpoints between patient and provider. In the next five years, I suspect we will see more rapid development of these modalities in areas such as mental health, substance use disorders, basic primary care, and care for patients with chronic conditions.”

References
1.Telehealth in the pandemic—how has it changed health care delivery in Medicaid and Medicare? United States Government Accountability Office. https://www.gao.gov/blog/telehealth-pandemic-how-has-it-changed-health-care-delivery-medicaid-and-medicare. Published September 29, 2022. Accessed January 31, 2023.

2. Centers for Medicare & Medicaid Services. State Medicaid and CHIP telehealth toolkit. https://www.medicaid.gov/medicaid/benefits/downloads/medicaid-chip-telehealth-toolkit.pdf. Accessed February 2, 2023.

3. Gaumer Z, Bassano A, Glossa J; Health Management Associates. Gaps in the coding infrastructure limit innovation of virtual care services. https://www.elevancehealth.com/content/dam/elevance-health/articles/ppi_assets/56/20221107_Gaps%20in%20the%20Coding%20Infrastructure_
HMA_final4.pdf. Published November 8, 2022.