Trends in Remote Patient Monitoring Use in Traditional Medicare

Abstract

This cross-sectional study uses traditional Medicare claims data to assess trends in general remote patient monitoring from January 2018 through September 2021.

Remote patient monitoring (RPM), the collection by patients of physiological measurements that are automatically sent to their health care practitioners, has been touted as a promising tool for improving chronic disease management. Interest in RPM has grown because of technological advancements, pandemic-related increases in virtual care, and expanded reimbursement. In 2019, Medicare expanded RPM coverage through new billing codes facilitating monthly payment for monitoring physiological data of any kind (termed general RPM).¹ However, given a lack of robust evidence on the clinical benefits of RPM and which patients benefit from RPM, some have raised concerns about potential overuse.² This cross-sectional study quantified trends in general RPM use in traditional Medicare.

Methods

Using 100% of traditional Medicare claims from January 1, 2018, through September 30, 2021, we identified general RPM use with Common Procedural Terminology (CPT) codes, including the codes introduced in 2019 (eTable 1 in the Supplement). General RPM use was measured as monthly claims volume per 100 000 enrollees. Enrollee counts included all beneficiaries enrolled in Medicare Parts A and B in a given month. General RPM was compared with continuous glucose monitoring (CGM), a more specific RPM use case with its own previously established CPT codes. This study was approved by the Harvard Medical School institutional review board. Informed consent was not required because this was secondary use of administrative data.

To understand how general RPM was used during the pandemic, general RPM services provided from March 1, 2020, through September 30, 2021, were grouped based on practitioner specialty and the first 3 characters of the service’s primary ICD-10 diagnosis code (primary diagnosis). Specialties and diagnoses were aggregated into broader categories to improve interpretability (eTable 2 in the Supplement). For each specialty–primary diagnosis combination, we calculated service volume as a percent of total volume. R, version 4.1.1 and SAS, version 9.4M7 were used for analysis.

Results

From February 2020 to September 2021, general RPM use per month increased from 91 to 594 claims per 100 000 enrollees (increase, 555%) and CGM use increased by 42% (Figure 1).

Figure 1. Monthly Claims for General Remote Patient Monitoring (RPM) and Continuous Glucose Monitoring (CGM).

General RPM claims include any claim with at least 1 general RPM service; CGM claims include any claim with at least 1 CGM service. Volumes were aggregated by month and normalized by enrollee counts. Enrollee counts included all beneficiaries enrolled in Medicare Parts A and B in a given month. Vertical dashed line indicates February 2020 as the last prepandemic calendar month.

During the pandemic, 63.1% of general RPM services were provided by primary care practitioners (Figure 2). The next most common specialties were cardiology (19.7%) and pulmonology (4.1%). The dominant primary diagnosis for general RPM services was hypertension (62.5%). The next most common were diabetes (8.3%), sleep disorders (3.9%), hyperlipidemia (3.5%), and heart disease (3.2%). No other primary diagnosis accounted for more than 3% of services.

Figure 2. General Remote Patient Monitoring (RPM) Use Across Specialty–Primary Diagnosis Combinations During the COVID-19 Pandemic.

General RPM services provided from March 2020 through September 2021 were categorized according to the practitioner specialty and primary diagnosis code. Each cell in the heatmap represents a unique specialty–primary diagnosis combination. Rows reflect the top 9 diagnoses by volume, with all remaining diagnoses included as “other.” Columns reflect the top 6 practitioner specialties by volume, with all remaining specialties included as “other.” Values in each cell reflect the general RPM service volume within that combination as a percent of total general RPM service volume. The cells in the “total” row reflect the total volume for each specialty group, summing across all other rows in the corresponding column. Similarly, the cells in the “total” column reflect the total volume for each diagnosis, summing across all other columns in the corresponding row.

Within specialties, primary diagnoses varied. For pulmonologists, sleep disorders and respiratory disorders accounted for 76.4% of general RPM services. Across all specialty–primary diagnosis combinations, hypertension monitoring by primary care practitioners accounted for 42.7% of services.

Discussion

General RPM use in traditional Medicare increased substantially during the COVID-19 pandemic, reaching more than 6 times the prepandemic levels by September 2021. Most general RPM services were for hypertension or diabetes. Although general RPM use was relatively small, if its trajectory continues, the cost implications could be substantial. In 2019, more than 20 million traditional Medicare enrollees had diagnoses for hypertension or diabetes,³ and a patient can accrue $2270 in general RPM costs annually.⁴

Costs must be balanced with RPM’s potential benefits, such as reducing hospital admissions. Randomized clinical trials of RPM showed mixed results overall, but some targeted use cases showed promise.⁵ Further research is necessary to identify clinical scenarios in which RPM is most beneficial and to understand which patients are using it and whether there are groups facing access issues.

A study limitation is that we categorized RPM use by diagnosis codes but could not observe which physiological measures were transferred. We believe our study provides valuable insights into how RPM use has increased and is used.

Supplement.

eTable 1. General RPM and Continuous Glucose Monitoring Common Procedural Terminology (CPT) Codes

eTable 2. Practitioner Specialty and Primary Diagnosis Categorization