Mostly Ignored CRT Device Diagnostics Could Predict Events, Says Team

September 13, 2012

September 13, 2012 (Seattle, Washington) — Clinicians who are in an interrogative mood could find, hiding in the implanted pacemakers and defibrillators of patients going home after hospitalization for heart failure, a lot of data useful for predicting who will have to return within a month [1].

Dr Roy S Small

That was the message from Dr Roy S Small (The Heart Group, Lancaster, PA), who with his colleagues showed that a composite score based on predischarge device diagnostic data can stratify such patients for risk of 30-day rehospitalization. He presented the findings here at the Heart Failure Society of America 2012 Scientific Meeting.

The analysis was based retrospectively on >1500 patients with defibrillating cardiac-resynchronization therapy (CRT-D) implants hospitalized for heart failure in hospital case files and from several Medtronic-sponsored trials such as PARTNERS-HF, FAST, and CONNECT. Criteria that were collectively predictive of 30-day readmission were burden of atrial fibrillation (AF), nighttime heart rate, amount of CRT pacing, and intrathoracic impedance.

Typically in practice, he said to heartwire , the devices are either not interrogated predischarge or the information is ignored. "A lot of these patients are seen by cardiologists who don't even look at the data," according to Small. Moreover, "most heart-failure patients aren't cared for by heart-failure specialists. They're cared for by internists or even hospitalists who don't know the first thing about these devices. That's a huge population of patients."

At his practice, he said, "we don't calculate the score for everyone at discharge, but we look at the information every time we see patients, and we spend a lot of time talking [with them] about it. It's already there; it's very useful. I don't know why physicians don't look at it more."

There were 34 HF rehospitalizations within 30 days in 27 of the patients; they didn't differ significantly from those not rehospitalized in terms of mean age, LVEF, prevalence of EF <35, distribution among NYHA classes, or prevalence of CAD, hypertension, diabetes, MI, or AF.

The presence of at least two of the four risk markers in the device-downloaded data was highly predictive of 30-day repeat HF hospitalization.

Predictive value of parameters obtained at predischarge device interrogation in patients hospitalized with heart failure


HR (95% CI) for 30-d HF readmission


Criteria contributing to risk score*


Intrathoracic impedance >8 ohms

2.38 (1.04–5.43)


AF burden >6 h

1.59 (0.49–5.21)


Nighttime heart rate >80 bpm

2.99 (1.30–6.87)


Prevalence of CRT pacing <90%

1.53 (0.78–2.99)


Score based on number of criteria met






2.19 (0.99–4.86)



4.13 (1.45–11.78)


* vs criterion not met

Of the four criteria contributing to the risk-prediction score, intrathoracic impedance is the only one not necessarily long a standard feature of CRT-device diagnostics. The Medtronic devices used by patients in the current analysis had the company's proprietary OptiVol Fluid Status Monitoring feature. But other companies have recently incorporated similar functions; for example, St Jude Medical's version is called CorVue Congestion Monitoring and Biotronik's simply its Thoracic Impedance (TI) feature.

Small and three coauthors disclose receiving honoraria and consulting fees from Medtronic, of which three other coauthors are employees. The clinical trials contributing data to the analysis were sponsored by Medtronic.