When remote monitoring ends, what happens to patient outcomes?

They do not maintain improvement, just as we thought. A just-published study fills a significant gap in telehealth knowledge. The study published in the Journal of Medical Internet Research (JMIR) analyzed the long-running Partners HealthCare-Massachusetts General Hospital Connected Cardiac Care Program (CCCP) for patients with heart failure. They matched 174 patients participating in the 120-day program between January 2008 and August 2012 with an equal number of control patients, all with heart failure, previously hospitalized and receiving care at MGH. The program patients were provided with a ViTelNet monitoring hub plus devices–a digital weight scale, a blood pressure cuff and meter, and a pulse oximeter–as well as receiving patient education plus unscheduled education intervention from nurse care coordinators when results trended out of pre-set ranges.

During the 120 day program, the program patients (predictably) had consistently lower hospitalization rates versus the control group at the 30, 60, 90 and 120 day marks, though the differential narrowed over time. But after program completion, in the study’s follow up eight months after the end of the four-month program (see one year below), the program group’s hospitalization rates increased to exceed the control group’s. Mortality rates remained lower though not statistically significant:

[grow_thumb image=”https://telecareaware.com/wp-content/uploads/2015/04/Partners-CCCP-study-hosp-rate.jpg” thumb_width=”400″ /]

The study authors, including Dr Joseph Kvedar of Partners, qualify that the patients were not matched for disease severity, (unknown) out-of-system hospitalizations were not included, or hospitalizations were due to other diseases.

The authors also speculate on reasons why rates increased post-program (Editor’s notes in parentheses):

• Progressive nature of disease (they may have worsened regardless after one year)
• Disproportionate deaths of sicker control patients (affecting the mortality rate)
• Dependency on remote monitoring and access to care coordinators (including their proactive intervention)
• Lack of sufficient competency in self-monitoring after program ended (transitioning was too early)
• Through education, they may have greater awareness of ‘warning signs’ of disease and will seek help (including hospitalization) earlier

Conclusions were: further study is needed to evaluate program duration (longer may be required), risk stratification to determine optimal monitoring length (one size may not fit all) and determine better transitioning to lower intensity long-term monitoring. Another alternative they did not mention was self-monitoring with automated virtual coaching. JMIR (full access). Also iHealthBeat, Mobihealthnews. Hat tip to reader Paul Costello of Viterion Digital Health.