ICD Software Predicts Earlier Device Failure

Researchers at the Minneapolis Heart Institute at Abbott Northwestern Hospital in Minneapolis, Minn. have reported that commercially available implantable cardioverter-defibrillator (ICD) monitoring software could identify problems with ICDs earlier than current monitoring practices.

Modern ICD devices are capable of sophisticated sensing and data logging techniques which aim to both optimize therapies and monitor device performance. This logged data can easily be read by a clinician in an outpatient setting using existing ICD monitoring software.  The study, which was published in Circulation: Cardiovascular Quality and Outcomes, aimed to identify whether this existing data could be used to predict device failures at an earlier stage.

From the announcement:

“Current monitoring approaches aimed at reducing harm from malfunctioning medical devices rely largely on voluntary reporting of adverse events by manufacturers, possibly leading to missed warning signs and delayed responses to the problems, such as late recalls,” said Robert G. Hauser, M.D., lead study author and senior consulting cardiologist at the Minneapolis Heart Institute at Abbott Northwestern Hospital in Minneapolis, Minn. “We looked at whether using an automated software program to monitor large databases of ICD patients might help us detect potential device-related problems earlier.”

Hauser and colleagues used a commercially available software surveillance program to compare data from about 1,000 patients with recalled leads to about 1,600 patients implanted with ICD leads still on the market. Patients in both databases had their ICDs implanted between 2001 and 2008.

Using the surveillance software, researchers simulated what occurred years earlier. The software detected problems with the recalled leads at least a year before the company had recalled them.

This is quite a significant finding and another example of a growing need for data aggregation and mining techniques to help us keep on top of the ever-growing heap of medical sensor data.

Press release: Computer software monitoring detects ICD malfunctions sooner…

Abstract in Circulation: Cardiovascular Quality and Outcomes: Early Detection of an Underperforming Implantable Cardiovascular Device Using an Automated Safety Surveillance Tool CIRCOUTCOMES.111.962621