Remote Patient Monitoring (RPM)
Remote patient monitoring (RPM) is the use of connected devices, such as blood pressure cuffs, scales, pulse oximeters, and glucometers, to collect patient health data between visits and transmit it to the care team for review. In home health, RPM extends surveillance of high-risk patients beyond the two or three visits a week a clinician can physically make.
How RPM works in a home health episode
The agency identifies a patient at risk for decompensation, commonly heart failure, COPD, or uncontrolled diabetes, and places devices in the home at or near the start of care. Readings transmit automatically or after patient entry, and monitoring staff review a dashboard with alert thresholds tied to plan-of-care parameters, for example a 3-pound overnight weight gain for a heart failure patient. Out-of-range readings trigger a call, a PRN visit, or physician contact for an order change. The clinical logic is early detection: catching fluid overload on day 2 costs a phone call and a diuretic adjustment, while catching it on day 5 costs an emergency department visit.
How Medicare pays for RPM in home health, and how it does not
Under the Home Health Prospective Payment System, RPM is not separately reimbursed and remote readings do not count as visits for payment or LUPA purposes. CMS allows agencies to report RPM costs on the Medicare cost report as allowable administrative costs, and since 2023 agencies report the use of remote monitoring on home health claims with G-code G0322. That reporting is mandatory when the technology is used and is on the plan of care, but it does not change the period payment. The business case therefore rests on outcomes: fewer hospitalizations, better HHVBP performance, stronger referral relationships, and more efficient use of visits, not on direct revenue.
Where RPM earns its keep
RPM concentrates value in a few places:
- Rehospitalization reduction: acute care hospitalization is a heavily weighted claims-based measure in HHVBP and star ratings
- High-risk diagnoses: heart failure, COPD, and post-surgical patients with narrow physiologic margins
- Visit efficiency: monitoring data helps target in-person visits where they matter instead of spreading them evenly
- Referral positioning: hospitals and ACOs favor agencies that can demonstrate monitoring programs for their highest-risk discharges
Implementation pitfalls
Programs fail in predictable ways. Devices get deployed without clear alert response protocols, so readings pile up unreviewed, which is worse than not monitoring because it creates documented, unactioned risk. Thresholds set too tight generate alert fatigue; too loose and they miss decompensation. Patients with cognitive impairment or low technology comfort need simpler cellular-connected devices, not app pairing. And RPM data that lives outside the EHR fragments the record. Before launch, define who reviews alerts, within what timeframe, what actions they take, and how every reading and response is documented in the clinical record.
Frequently asked questions
Can RPM readings count as home health visits?
No. Medicare counts only in-person visits for payment and LUPA threshold purposes. Remote monitoring supplements the visit pattern on the plan of care but cannot replace ordered in-person visits or contribute to the visit count.
Does Medicare pay home health agencies separately for RPM?
Not under the home health benefit. RPM costs are reportable as allowable administrative costs on the cost report, and agencies report RPM use on claims with G0322, but there is no separate fee-for-service RPM payment to the agency under HH PPS. Physician practices bill RPM under separate CPT codes, which is a different program.
Which patients should an agency prioritize for RPM?
Patients with conditions where early physiologic change predicts hospitalization: heart failure, COPD, hypertension with recent instability, and complex post-acute patients with prior admissions. Risk-stratifying the census and targeting the top tier usually beats broad deployment on both cost and outcomes.