Shotgun approach to remote monitoring driven by a small fraction of physicians during pandemic: study

Growth of remote patient monitoring during the pandemic was driven by a small fraction of primary care physicians using untargeted methods for patient selection, a new study revealed.

The study, published in the September issue of Health Affairs, tracked billing for RPM for patients with chronic illness diagnoses. The tool has been touted for its potential to effectively improve the treatment of patients with poorly managed chronic diseases. Billing in this area increased fourfold and was largely driven by a small handful of primary care physicians, the study revealed. Physicians who implemented the digital health tool showed little evidence of targeting patients with greater disease burden or worse disease control.

“Remote patient monitoring could be really great for folks for whom the status quo of chronic disease management isn’t working for them,” said Ariel Stern, study co-author and associate professor at Harvard Business School, in an interview. “So, one way to think about that: To the extent that’s true, you would want to see more targeting of people with poor diabetes control. We're not seeing that. What this study can’t tell us is whether remote patient monitoring might still be desirable for people that have good diabetes control if it’s allowing them to continue to manage their health, but see their doctor less frequently or if it's somehow more patient-centered in reducing travel time or allowing for the earlier detection of future complications, but we’re not able to get that data in this project.”

In the early stages of the COVID-19 pandemic, Congress approved emergency authorization to encourage the use of RPM, leading to its skyrocketing usage by some physicians. Prior to the pandemic, the tool was used sparsely and primarily for patients with poor disease management, complications and those unable to regularly seek in-patient care. With the expansion of digital health guidelines, some physicians began using RPM in a “shotgun approach,” the study showed.

The study mined de-identified claims data from January 2019 to March 2021 to track growth and targeting. The data were used to follow claims volume, persistence, concentration among providers and use rate of patients treated by high-volume providers. Data revealed that there were only modest differences of patient usage between patients with low and high complexity of illness.

Patients diagnosed with one chronic condition had a use rate of 16.5%, while those with five or more chronic condition diagnoses had use rates of 24.1%. Patients with uncomplicated hypertension used RPM at a rate of 21.7% as compared to 23.8% for patients with complicated hypertension. Those with “good diabetes control” had a use rate of 22.1% versus 21.9% for those with “poor diabetes control.”

“What we find is that across these providers’ patient base they’re getting remote patient monitoring at a pretty high rate, but we’re not seeing big differences between these groups: those with more disease burden versus less disease burden,” said Mitchell Tang, study co-author and doctoral student at Harvard Business School, in an interview. “There are some differences, but it’s only a few percentage points. These are not big differences. We’d be cautious to say the wrong people are getting remote patient monitoring, but this is not strongly indicative of a lot of targeting.”

The study also showed clustered usage of RPM within the time of early expansion and high concentration among providers. A puny 0.75% of providers had any general RPM claims, and 0.1% of that group accounted for 69% of all general RPM claims. Conversely, the top 0.1% of primary care providers for outpatient telemedicine claim volume accounted for 6.1% of telemedicine claims.

This could speak to the upfront costs required by such programs. Establishing an RPM program demands notable investment from providers, including the cost of monitoring devices, according to the study.  

“Whether you think of it as a sunk cost or just a fixed cost that’s already been paid, the marginal cost of adding additional patients in terms of time, effort and actual money is quite low once a practice is up and running with remote patient monitoring,” Stern said. “I suspect that what we’re seeing is a bit of the sort of ‘why not’ mentality. It’s not that these patients have nothing to gain here, it’s just that we’re not seeing the type of targeting that would indicate that doctors are being discriminating in whom they are targeting for remote patient monitoring in their patient panels.”

Increased spending in RPM could lead to overall less spending by payers like Medicaid if the result is better chronic illness management. However, the study’s co-authors agree that not enough is known about the results of RPM programs.

“We can easily say what the costs are, but the benefits are a bit more ambiguous, so the next study is really trying to say, 'what is the impact, how is it manifesting?'” Tang said.

The researchers encouraged further study into clinical outcomes, patient use, spending, access to care and convenience of care. The goal of such research would be to better delineate when RPM is most effective, how to incorporate it into other forms of care and how various healthcare professionals may be employed in its usage.  

As physicians have been using the technology broadly, so have researchers been measuring its effects. Targeted use cases for isolated patient groups, such as remote monitoring of congestive heart failure using data from implanted devices, show more revealing data, according to the study. Focused patient-centered care programs including concentrating on patients who need the tool the most are more likely to be high value.

“The shotgun approach might be teaching us a lot about what does and doesn’t work in this context so we can actually make some data-driven decisions about how we can better target remote patient monitoring moving forward,” Stern said.