I’d like to congratulate Dr. Anand and colleagues on performing an innovative and thought provoking trial I think that this trial illustrate a number of important issues they even go beyond the specific details presented here.
First - most of the studies that we will hear about today and at most of our meetings will be studies of therapeutic interventions. Trials of diagnostic strategies and approaches are just as important, especially when these involve new technologies. There is no other way to test these strategies in a real world setting and too often diagnostic strategies are tested in single centers leading to a false sense of utility.
Second – this trial represents a trend that I believe we are going to see more of in the next several years – the merging of medical technologies and consumer technologies. These ubiquitous i-devices that we all carry around with us, or the X-box we have in our living room, will allow us to communicate not just with our family, friends, and other random people on the internet, but with our doctors.
Third – all of us who have taken care of patients with heart failure are well aware of how difficult assessment of fluid status can be – even when the patient is in front of you. We need to recognize that in fact we can use the help of novel technologies.
Finally, we are all acute aware that communication between patients and caregivers. This communication is generally either dependent on a patient taking initiative, which we know doesn’t always happen, or very expensive programs in which a care giver, generally a nurse, communicates with patients directly. Surely, a non invasive ambulatory approach to determining fluid status and transmitting that information to caregivers certainly be a worthy goal.
Yet this trial illustrate a number of problems both with new technologies and with this particularly vexing diagnostic problem.
The amount of technical difficulty that this trial illustrated is at best disappointing, and at worst unacceptable for a diagnostic strategy. Dropped calls on my iphone are bad enough, but in this trial nearly 30% of patients could not be included because of missing data and over 40% of patients with acute decompensated heart failure events, had missing data preceding their event. What can we make of the false negative rate when so much data is missing. Well, we are likely underestimating the true value of the system, if it worked. So in effect, these are false false negatives.
Indentification of the cutoff was designed to minimize false positives – After all a remote device that “cries wolf” won’t be useful and will quickly desensitize the very caregivers who need to act on the information obtained. Yet we could ask... was this rate too low and will it reduce the sensitivity.
Dr Anand showed us that combining the fluid index measure with other measures improved the prognostic utility. They included breath index and “personalized fluid status” a measure that indexes fluid status based on individual parameters. That’s great, but why not other readily obtainable measures, such as heart rate, patients weight. Are we certain that they wouldn’t help improve the diagnostic accuracy. Or even better, how the patient feels? Surely a patient whose fluid index was discordant with their symptoms would be more likely to be a false positive than one in whom both of these parameters tracked together.
Finally, we all understand the importance of protecting the intellectual property of companies that dip their feet into this dangerous world of heart failure. But as physicians, we need to understand the methods used to diagnose our patients.
So exciting concept, definitely. Unproven technology, surely. We need as a community to support and applaud these diagnostic clinical trials.