Subjects: Quality and safety
Part of the Big O’s plans to improve clinical effectiveness involves setting up registries to monitor patient outcomes after interventions like chemotherapy or the insertion of a medical device.
Come to think of it, that’s how Robert Hauser first detected problems with Medtronic’s mission-critical heart defibrillator a few years back.
Using data from a small registry at the Minneapolis Heart Institute, Hauser deduced that cables connecting the Sprint Fidelis defibrillator to the heart tended to crack causing the charge-box to either deliver shocks at inopportune times or fail to discharge at the moment of truth.
In early 2007, Hauser raised the matter with Medtronic, which recalled the product 8 months later. The company indicated at the time that 5 deaths may have resulted from the problem.
Last week, Hauser and a colleague published another registry-based study suggesting the problem was worsening for patients who still have the hardware in their chests, and that physicians should consider replacing the cables proactively.
Medtronic disputes the findings, claiming its own analysis is based on a much larger sample and the ruckus is likely to continue for the foreseeable future.
A properly funded, well-designed national registry would settle the matter in a heartbeat, not to mention detecting the problem more quickly than Hauser was able to manage, according to cardiologist Alan Kadish of Northwestern.
Funny thing though, in 2004 Medicare required formation of a defibrillator registry as a condition for approving the devices for use in a new class of patients in whom only equivocal evidence suggested they were helpful.
But Medicare never put a dime towards funding it and support from the device companies tailed off after the first year.
Now, hospitals pay for a stripped-down version of the registry and get marginally useful benchmarking data in return.
The mini-registry can’t merge its data with Medicare claims, meaning among other things it can’t track longer term outcomes, which is pretty much the whole point.