Digital Health Outcomes

The Trajectory evaluation model helps demonstrate to key stakeholders how a digital health solution such as a remote patient monitoring (RPM) program is making a difference. And if the program is not fully optimized, we can show the client – using Trajectory generated evidence and consulting expertise – how to make real time, evidence-based tweaks to improve the outcomes they are targeting.

Digital health represents the convergence of digital technologies with care interventions to improve the efficacy and efficiency of the healthcare delivery system. Its footprint is broad based and is constantly changing.

For example, RPM interventions have been gaining traction in recent years, including an uptick due to the reimbursement codes issued by CMS for several types of RPM services. The concept often includes some form of telecommunication or digital technology to collect medical and other forms of health data from individuals in one location and electronically transmit that information securely to health care providers in a different location for clinical assessment, disease monitoring and analysis for potential clinical intervention with the patients care team.

However, digital health applications like RPM devices can mean different things to different people. So, it is imperative that we study these new emerging care treatment trends to promote the best evidence-based pathways.

With digital health (like wearable monitoring devices), the evaluation solution should use comparative evaluation techniques. These can be of three (3) types, listed in order of evaluation power:

  • Compare to a reference population: This can be classic evaluation between a referent group without the RPM program intervention and the test group with the intervention using claims or billing data. An alternative is to compare different “dose” levels among all those who received the RPM services; and adjust for baseline factors.
  • Compare to a benchmark. When a peer-review article or a generally accepted benchmark is deployed, this becomes a reference instead of a claims-based real-time reference.
  • Compare to self-expressed utilization by the patients. This is based on the “intention” of the patient before the RPM intervention suggested course of action compared to the stated action of the patient after the intervention suggested a course of action. If claims are available, the actual course of action can be verified.
Case Example: Does a Wearable “Safety Alert” Device for Seniors Reduce Medical Costs?

A remote patient monitoring (RPM) company contacted a Medicare Advantage Plan about using a wearable safety alert device to promote patient safety and reduce cost of emergency room visits and inpatient stays. In response, the Plan’s strategic management team asked if the RPM company could work with an independent 3rd party to evaluate the efficacy of the device in a population with congestive heart failure (CHF).

Study Design

Trajectory designed a prospective study, with a randomized, risk-matched CHF population into two groups, one cohort that used the device and one that did not. The RPM company believed (hypothesized) the device would: a) bring peace of mind to the seniors and family members; b) increase the likelihood of re-enrollment into Medicare Advantage; and c) lower utilization costs, especially inpatient costs. The study was designed to last 9-12 months; the first 3 months were used to install the device on those selected to receive the device; the start date for each person was the date the device was installed, the start date for the reference was randomly selected from all the start dates of the intervention group. Each patient and their age-, sex-matched control were followed for 6 months (due to delayed claims run out the team needed the full 9 months to get data on everyone). After the study was completed a survey was generated for a random sample from each group. A report was generated month-by-month (after the 3-month lag) and published after 6 months of data was available and the survey was completed.


The survey found that the “peace of mind” hypothesis was a major factor among RPM users. The overall satisfaction with the health plan – as a marker of likelihood to re-enroll—was higher in the RPM users than the comparator group. The overall cost of care was lower in the wearable device population, but it was not statistically significant. The inpatient costs were lower in the wearable device population for the 1st three months (p<0.05) and still lower (but less so) for the entire 6 months (p=0.54).


Inpatient cost savings were superior in the RPM cohort but did appear to drop off after the first three months and then some more after the second three months (as did use of the RPM device). This provided an important feedback loop to both the RPM Company and the Medicare Advantage Plan. If the parties can make sure that the “safety alert” device continues to be used over time, it can continue to improve clinical outcomes and promote savings. But the ongoing success is tied to finding updated strategies to promote patient engagement after six months. This could include creating incentive programs for patients to continue actively wearing the device, among other options.

Scroll to Top