For many medical conditions, there are parameters that can be assessed at some frequency to determine how well a care plan is working. Patients with hypertension can track their blood pressure continuously or at any interval they prefer. Patients with hypercholesterolemia can have their lipid levels checked every year, every third year or every five years. Even patients who are overweight can vary how often they weigh themselves. When a patient tracks a physiologic parameter to facilitate behavior change, they can use whatever frequency makes the most sense to maintain a specific behavior or trigger a new behavior. Some individuals may track themselves daily to maintain the habit while others may use some version of a statistical process control process to adjust their measurement interval based on the parameter’s recent trend in relation to the expected target range.
What if the measurement frequency has an implication on healthcare costs and quality? “Established patient” encounters cost around $100 in many markets. Before the COVID-19 pandemic, these “return office visits” were the most expensive line item when considering all costs for a particular chronic condition. Reviewing a single in-office blood pressure reading or reviewing lab results and adjusting medications have been the mainstay of many primary care offices for decades. The sporadic nature of these evaluations could lead to over- or under-utilization of healthcare resources if they do not happen frequently enough to identify decompensations or happen so frequently that they do not lead to a change in the patient’s treatment plan. The sudden interest in telemedicine could lead to additional virtual encounters to continue chronic disease management with nearly the same cost structure.
In November 2018, the Centers for Medicare and Medicaid Services (CMS) updated its rules around remote patient monitoring (RPM), including CPT codes for:
- Initial technology set-up and patient education (two separate codes),
- Remote monitoring of multiple physiologic parameters (e.g., weight, blood pressure, pulse oximetry, respiratory flow rate) with daily recordings or programmed alert transmission, each 30 days and
- 20 minutes or more of qualified professional’s time in a calendar month requiring interactive communication with the patient/caregiver during the month.
The following year, CMS relaxed the qualified professional supervising requirement from direct supervision to general supervision. CMS also added another CPT code for an additional 20 minutes of monitoring time. In cases where a provider needed to reach out to a patient based on the RPM readings, clinicians can bill for a 5-10 minutes virtual check-in separately.
These changes have spurred technology companies to offer their services to outpatient practices. At least one company suggests a practice monitoring 500 patients could generate over $1.5 million annually. Many states have established eligibilty criteria to limit who in their Medicaid populations might quality for RPM. If these dollars led to meaningful improvements in healthcare value, most payers would whole-heartedly agree with shifting more dollars into RPM.
Over the last five years, 150 RPM randomized clinical trial manuscripts were published on PubMed. Here’s a summary of the most recent publications:
|Trial (condition, publication year)||Population||Intervention||Outcome||Results|
|CHROMED (COPD, 2018)||Global Initiative for COPD grades II-IV with history of exacerbation in the previous year and at least one nonpulmonary comorbidity||Self-assessed lung mechanics daily||Time to first hospitalization, change in EuroQOL EQ-5D utility index score||No change in either outcome|
|BEAT-HF (Heart failure, 2016)||Hospitalized patients 50 years and older who received active treatment for decompensated heart failure||Health coaching telephone calls and daily telemonitoring (blood pressure, heart rate, symptoms and weight)||All-cause readmission at 180 days||Intervention rate 50.8%, control rate 49.2% (adjusted hazard ratio 1.03 (95% CI 0.88-1.20)|
|(Type 2 diabetes, 2015)||Patients 30-70 years, not using insulin, with A1C levels 7.5-10.9%||Daily health sessions through a tablet informed by glucometer readings with focus on self-care behaviors||A1C reduction||1.11% reduction in intervention group, 0.70% reduction in usual care group (0.41% difference [p=0.009])|
The displayed results are a sampling of the trials included. RPM may have been approved on the basis of promising cohort studies or a persuasive theoretical model. The randomized, controlled trial results do not suggest that the technology, at least as it is currently deployed, will make a meaningful improvement in healthcare outcomes. On a related note, a 2018 Danish randomized trial of 24-hour ambulatory blood pressure monitoring versus usual care found no difference in recorded blood pressures after 12 months.
A different approach to remote patient monitoring might link the patient’s measurement of clinical parameters to a reduction in office visits. In 2019, CMS approved two virtual communication codes for a check-in (about $15) and reviewing digitally stored images (about $12). The codes are only available for established patients and subject to co-insurance and deductibles. Most clinicians do not even bother to submit these claims as the work to prepare these claims is perceived to be more than the reimbursed amount. But in a value-based virtual monitoring context, assessing a patient’s blood pressure every month would be more cost-effective than either RPM or quarterly in-office visits. Here’s a set of cost assumptions for 10,000 patients annually with some disease that could be managed by virtual check-ins compared to face-to-face encounters (usual care) and remote patient monitoring:
|Patient pay||Usual care||Three visits a year at blended 99213/99214 rate||$477,000 (assuming 20% copay)|
|From payer to provider||Usual care||Three visits a year at a 1:1 blended 99213/99214 rate||$1.9 million|
|Patient pay||Remote patient monitoring||12 months of monitoring with 15% of encounters requiring an additional 20 minutes of evaluation time||$1 million (assuming 20% copay)|
|From payer to provider||Remote patient monitoring||12 months of monitoring with 15% of encounters requiring an additional 20 minutes of evaluation time||$4 million|
|Patient pay||Virtual check-ins||Quarterly check-ins at $15/check-in||$90,000 (assuming 20% copay)|
|From payer to provider||Virtual check-ins||Quarterly check-ins at $15/check-in||$360,000|
I did not account for the costs of the technology (e.g., home blood pressure cuffs, weight scales) or the healthcare labor resources to manage remote patient monitoring or virtual check-in patient-generated health data, but those costs would presumably be much less than the difference between usual care and virtual check-ins. RPM’s current reimbursement model does not appear to be cost-effective in any scenario unless the technology is more effective than other assessment methods at reducing readmissions. Some patients may have multiple chronic conditions that could be assessed with virtual check-ins, increasing the difference between the interventions even more.
For provider practices overwhelmed with patient demand, moving from usual care to virtual check-ins would open up their schedules to see higher-acuity patients. Although remote patient monitoring might garner additional dollars, practices would need to invest in more technology and staff to accrue those funds. From a patient perspective, the virtual check-in approach should be very appealing, unless patients feel that a virtual check-in is lower quality than a face-to-face encounter.
Many chronic diseases that can be monitored remotely. If we all parties in healthcare are profit-maximizing, then paying more for additional monitoring will continue. Alternatively, if patients, providers and payers are looking to maximize value in monitoring chronic disease, optimizing virtual check-ins may be a more prudent approach. The virtual check-in allow providers to customize the monitoring interval over a span of weeks instead of days, a more meaningful unit of time before suggesting a treatment plan change.