Large NIH study fails to show any benefit to “telemedicine” program for heart failure patients – closer look at study methods reveals more

As our recent coverage of the mHealth Summit shows, medicine has developed some pretty lofty expectations for mobile technology in healthcare. A big part of that promise has been in telemedicine, where the opportunity to remotely provide care on a wide range of platforms, including the iPhone, could be a huge asset in chronic disease management. With that hope in mind, a group of researchers at the National Heart, Lung, and Blood Institute (NHLBI) launched a large, multicentered randomized controlled trial to evaluate the benefits of a telemedicine system called Tel-Assurance from Pharos Innovations – basically an automated phone-in system which they used to monitor patients at home for signs and symptoms of worsening disease. Their study showed that within the approximately 1600 patients randomized, there was no significant benefit in terms of death, hospitalization, or length of stay. However, a closer look at the study and its shortcomings suggests that these results may fall far short of being a nail in the coffin for modern telemedicine in heart failure.

The system that was utilized in the study was basically an automated phone-in system – patients were instructed to dial in and use the phone keypad to answer a series of preset questions including their weight, symptoms of heart failure, and so on. Responses were then downloaded to a secure, central system. The data were then presumably reviewed every work-day by a clinician who then used their own judgment to intervene as needed (change a medication dosage, have the patient come in, etc). In addition, the system would automatically flag certain concerning trends and responses that prompted required clinician calls to patients for assessment and intervention.

The study population was randomized to either this telemonitoring system or the more vague “usual care.” There was no significant difference in the baseline characteristics of the patients (age, race, comorbidities,severity of disease, etc). Nor was there a significant difference in any of the measured outcomes – death, hospital admission rates or time to admission, length o stay if hospitalized. Quality of life was not reported though it was asked about in the telemonitoring system – I doubt however that a minor to moderate benefit in the rather vague QOL endpoint would sway anyone for or against this technology, especially in when cost-savings are the hot topic.

While this study was large and randomized, there were several shortcomings which I believe limit what we can take away from its results. The two biggest –

1. They didn’t quantify usual care. We assume that the patients enrolled in the study and assigned to the control group are reflective of what all heart failure patients get. I find it hard to imagine that they got anything short of the standard of care, given that every clinician caring for a “control” patient knew there was someone watching over their shoulder.

2. This intervention just doesn’t seem that great to me. It falls short in one big way – it creates a new burden on the patient rather than reducing the burden for managing their disease. With this system, about 50% of the patients using the telemonitoring system were still using it by six months – and “using it” in this study meant three times per week. The real potential lies in interventions that are easily integrated into a patient’s day – take the fall detection app for Android or the mirror from MIT that monitors vital signs. Similarly, a less onerous intervention may be a weight scale that automatically uploads daily weights to the patients computer via Bluetooth or a Wifi network. If we want to realize the benefits of telemedicine, we’ve got to be a bit more imaginative than the touch-tone phone system my cable company uses.

So what this study tells me is that we need to be more inventive in how we care for our patients. Managing diabetes, heart failure, COPD, and other chronic diseases is hard work for patients even with the things we try to have them do now (try keeping to 2 g salt diet some time). Adding new responsibilities to a patient’s already full plate isn’t the way we should go as we turn telemedicine from an idea into a real intervention. Interventions that integrate easily into a patient’s life seem to me to have a far better chance of success.