Sunday, December 12, 2010
Telemonitoring: Sounds Great. But it doesn't work.
However, a rigorous study has just been published in the New England Journal of Medicine that throws big doubts on the usefullness of telemonitoring. The study was led by Dr. Sarwat Chaudhry of Yale University. Chaudhry randomized 1653 patients with heart failure as they were discharged from the hospital. This was an excellent test for telemonitoring, as heart failure defines a condition for which telemonitoring should work, and the time after hospital discharge is a very high risk period as many patients decompensate shortly after they go home and need to be readmitted.
Patients randomized to telemonitorinng received a device that allowed them to report weight and heart failure symptoms on a daily basis via a toll free number. The information was transmitted to their clinicians for review. Concerning information (variances) were flagged for attention.
The benefits of the intervention? Absolutely nothing. Those who got telemonitoring had the same rate of mortality and readmission as those who got usual care.
There are many possible reasons for the ineffectiveness of telemonitoring in heart failure. It may be that collecting lots of data on patients and feeding it back to providers is not as useful as we want to believe. Perhaps what's needed is a lesser flood of information and more sophisticated thinking about how to use this information.
But the bigger question: Is all this enthusiasm for telemonitoring justified? I don't think so. This study certainly should raise major concerns in the VA, which now needs to consider the possibility that its investment in telemonitoring is not the wisest use of resources.
Is it possible that other approaches to telemonitoring might be more effective? Certainly. But the burden of proof is now on the advocates of those approaches. Medical practice should be driven by evidence, not wishful thinking.
This study is an important example of why new technologies need to be subjected to rigorous evaluation before they undergo widespread implementation.
by: [Ken Covinsky]