Posted on 08/04/2011

A watch-sized device worn on the wrist successfully detected loss of pulse in an early-stage trial, a key step in developing a practical, noninvasive way to monitor patients at risk for sudden cardiac death, researchers said. In most patients participating in the phase I trial, the investigational device correctly signaled artificially induced pulselessness, though it had a false-positive rate of about 10%, according to John Rickard, MD, of the Cleveland Clinic, and colleagues.

Called the Wriskwatch, the device incorporates a piezoelectric disk strapped snugly against the wrist, such that it detects the arterial swelling at the radial pulse point. It can therefore identify episodes of pulselessness or ventricular fibrillation. It transmits a Bluetooth radio signal when loss of pulse occurs, contacting emergency medical system and "appointed bystanders in a community, old age home, or nursing home," Rickard and colleagues explained.

The device also included a motion sensor that blocked the alert unless the wrist was still. In practice, to keep down the number of false alarms, the developers are seeking to have the device activate only when the wearer is unconscious. In the current trial, the researchers tested the device in 24 patients hospitalized for a variety of reasons -- including 10 with atrial fibrillation and 14 with normal sinus rhythm -- and in 10 patients seen at the Cleveland Clinic for implantable cardioverter device (ICD) placement and testing.

Hospitalized participants were to wear the devices for 20 minutes while keeping the wrist as still as possible. In all but four of these patients, a blood pressure cuff was inflated for random 15-second intervals to cut off circulation to the arm where the device was placed. The latter procedure was omitted in four patients, all with normal sinus rhythm, who served as controls.

A somewhat different protocol was followed in the ICD patients, in whom ventricular fibrillation was to be induced as part of the pacemaker testing. In this case, the test was whether the Wriskwatch device could successfully detect pulselessness caused by the arrhythmia, so it was placed on patients' uncuffed arms and the ICD testing was conducted in the normal fashion.

The device failed to operate successfully in five of the 34 participants -- two because of inadequate signals, two because patients were unable to keep their arms sufficiently still, and one in which the device fell off during ICD testing. The remaining 29 patients included 21 in the hospitalized group (seven with afib, 10 with normal rhythm having pulselessness induced, and the four controls with normal rhythm) and eight in the ICD group.

Among the hospitalized patients, the device successfully identified the timing of pulselessness in all seven with afib and nine of the 10 with normal rhythm. It also correctly recorded no pulselessness in three of the four controls, but incorrectly flagged a loss of pulse in one. Timing of ventricular fibrillation was detected correctly in seven of the eight ICD patients.

The researchers calculated that the sensitivity was 99.9%, on the basis of 15-second intervals in which pulselessness occurred during the 561 minutes of testing. Specificity was 90.3%. Positive and negative predictive values were also 99.9% and 90.3%, respectively, they wrote.

Rickard and colleagues suggested that the device, or something like it, could reduce the number of deaths resulting from out-of-hospital cardiac arrest, by shortening the period between onset and medical response. "The primary determinant of survival following out-of-hospital cardiac arrest is time to defibrillation such that small delays can have a profound negative impact on outcomes," they wrote.

Because the Wriskwatch correctly monitored pulse status in 26 of the 29 evaluated patients in the study, "this technology shows promise for the early detection of out-of-hospital cardiac arrest and could provide a tool to hasten response times," they added. The researchers noted that the device design needs improvement before phase II testing would be appropriate. In particular, they pointed to snugness of fit as the issue underlying three of the five device failures in the current study. In addition to the incident where the device fell off, the two in which signal quality was the problem also probably resulted from "poor wrist contact." The snugness issue could stem from the fact that only one size of watchband was available, Rickard and colleagues indicated, pointing to a simple solution -- providing different sizes of watchband.

Another problem was the 10% rate of false positives, they suggested, which in practice could be a burden on emergency personnel. The researchers indicated that future versions will include an accelerometer that better detects motionless, along with an audible signal that allows the wearer to deactivate the unit in case of false alarm before it calls the paramedics. Rickard and colleagues noted that the study was limited in a number of ways: small patient numbers, the laboratory setting, and the lack of data on whether such a device could actually improve response times and resuscitation rates.

The study was supported by Emergency Medical Technologies, which is developing the Wriskwatch. Two co-authors were employees of Emergency Medical Technologies. They provided the devices and determined participants' pulse status on a blinded basis, but they had no role in the study's interpretation or manuscript writing except to help describe the technology. Other authors declared they had no relationship with the development company.

SOURCE: Rickard J, et al "The utility of a novel watch-based pulse detection system to detect pulselessness in human subjects" Heart Rhythm 2011; DOI: 10.1016/j.hrthm.2011.07.030, as reported in MedPage Today