Women With ICDs: Higher Mortality, More Hospitalizations Than Men

November 16, 2012

LOS ANGELES — It's recognized that the complication risk related to implantable cardioverter-defibrillator (ICD) placement is higher in women than men, but they also don't fare as well clinically in the following months, according to data from the National Cardiovascular Data Registry (NCDR) [1]. They suggest that women's mortality and risk of heart-failure hospitalization six months after getting primary-prevention ICDs are significantly greater than in men.

The findings point to a whole tapestry of questions about why there would be such differences in ICD outcomes between women and men, including whether there is ICD referral bias based on sex, whether physicians see ICDs as being as effective in women as in men, the overall extent to which heart failure in women is adequately treated, and other issues, according to Dr Andrea M Russo (Cooper University Hospital, Camden, NJ), who headed up the analysis presented here last week at the American Heart Association 2012 Scientific Sessions.

By the time of primary-prevention ICD implantation, Russo told heartwire , "there are some features that are clearly different in women. They are sicker by the time they get ICDs, so then their outcome is dictated by things other than what the ICD can impact."

In the current analysis, women made up about one-fourth of the approximately 39 000 first-time recipients of ICDs from 2006 to 2009. Those with biventricular devices were excluded; 62% of women and 64% of men received dual-chamber devices, with the remainder getting single-chamber ICDs.

At baseline, women were more likely to have heart failure or diabetes, to have nonischemic cardiomyopathy, to be in NYHA class 3-4, and to have an LVEF <30% (p<0.001 for all differences). Women were also less likely to be getting aspirin or statins (p<0.001 for both). Men were significantly more likely to have atrial arrhythmias, ischemic heart failure, and a history of MI (p<0.001 for each).

The findings don't necessarily show--but are consistent with the common perception--that by the time of ICD implantation, Russo proposed, "men have more ventricular arrhythmias and less progressed heart failure, and therefore the ICD saves them more often, while women have more progressed heart failure and tend to die from heart failure."

But, "even after adjusting for all that, women still had a higher complication rate, higher rehospitalization rate for heart failure, and higher mortality," Russo said. So beyond those issues, "there's really a difference between men and women, and we just need to do more research to learn why there is that difference."

Rates of Device-Related or Procedural Complications and Clinical End Points by Sex

End point Women, n=9750 (%) Men, n=38 912 (%) p
Hematoma at 30 d 0.37 0.25 <0.05
Tamponade at 30 d 1.39 0.45 <0.01
Mechanical complicationsa at 90 d 2.40 1.71 <0.01
Any complication b 5.94 3.85 <0.01
Clinical end points      
Mortality at 30 d 1.45 1.05 0.001
Mortality at 6 mo 6.5 5.6 <0.001
HF readmission at 6 mo 14 10 <0.001
All-cause readmission at 6 mo 37.2 31.7 <0.001

a. Includes lead dislodgement

b. No significant differences in pneumothorax or infections

Dr Paul Varosy (VA Eastern Colorado Health Care System and University of Colorado, Denver), who presented the NCDR analysis at the sessions in lieu of Russo, who could not attend, said the results "make me wonder whether there are barriers or biases in the care of women such that they're receiving ICD therapy at a much later stage of their care than are men, [and] whether there should be greater efforts to make sure that [all] patients who meet the indications for ICD therapy are receiving it." Varosy, who has otherwise been closely involved in NCDR-based ICD studies, was not part of the current one.

"What I think would be a mistake in interpreting these data is that somehow this should mean that women should not be receiving ICD therapy because they're high risk," he said. "I think this just suggests that women are a higher-risk population, at least of the patients who are receiving [ICDs]."

Varosy also cautioned that the current analysis "represents a small fraction of the entire [NCDR] ICD registry cohort." He said "a little over 900 000" patients had entered the registry since 2006, and about 12 000 to 15 000 are joining it every month. "So this is a relatively limited subset, and generalization of information from here to the rest of the population, I think, has its limitations." That said, he added, the analysis is important in that it can help inform the conversation between clinicians and female ICD candidates regarding periprocedural risk and outcomes.

Odds Ratio (OR) for Device-Related and Clinical End Points in Women vs Men

End point OR (95% CI)* p
Device-related complications 1.40 (1.27–1.54) 0.001
6-mo mortality 1.11 (1.00–1.22) 0.041
6-mo HF readmission 1.34 (1.25–1.44) <0.001
6-mo all-cause readmission 1.22 (1.17–1.29) <0.001

*Adjusted for demographics, medical history, risk factors, diagnostics, and ICD type

The guidelines don't make distinctions between men and women in their recommended indications for ICD therapy, "so it's not clear to me why women would have worse heart failure by the time they are referred for and get an ICD," Russo said.

"One of the possibilities is that women are not being referred [as much], or they're being referred later because they aren't being picked up early enough." That might occur if clinicians don't adequately consider the possibility of heart disease in women until they are older, "so they're treating them for bronchitis or treating them for other causes that are not cardiac, until they get sick--that's one possibility."

Also, some earlier studies have been interpreted as suggesting women don't benefit as much from ICDs, Russo pointed out. "So do some [clinicians] think we shouldn't refer them to begin with? It might be that."

She added, "I would suggest that we try to focus on educating the public but also educating our physicians to think of heart disease earlier in women. Maybe catching it earlier will be able to change some of these differences [between women and men] that we see."

As moderator for the session featuring the NCDR analysis, Dr Matthew R Reynolds (Beth Israel Deaconess Medical Center, Boston, MA) proposed that some of the women in the NCDR cohort might have been candidates for cardiac resynchronization therapy (CRT) devices--that is, biventricular pacemakers with or without ICD capability.

The mean QRS interval in the women was about 110 ms, "but the standard deviation of that mean suggests that some of those patients actually might have been CRT-eligible [with QRS intervals >120 ms]." Receiving CRT would likely have improved heart-failure symptoms and sudden-death risk in many of them.

"I wonder if some of the patients who actually received single- or dual-chamber devices may have had unsuccessful attempts at CRT, something that's not totally straightforward to recognize in the registry data," he said.

The NCDR registries are supported by the American College of Cardiology Foundation. Russo discloses participating in trials with involvement by Medtronic and Cameron Health and receiving honoraria from or consulting for Medtronic, Boston Scientific, St Jude Medical, Cameron Health, and Biotronik. Varosy had no disclosures. Reynolds discloses consulting or serving on an advisory board for Medtronic.