Boston, MA - The risk of infection related to the insertion of arrhythmia-management devices climbed sharply during a recent eight-year period, rising far faster than the increase in device implantations, according to an analysis of US data [1]. Although the risk grew more dramatically for implantable cardioverter defibrillators (ICDs) than for pacemakers, numerically pacemakers accounted for many more cases of infection.

"Interestingly, we found that increasing age and female gender were actually protective against infection," Dr Andrew Voigt (University of Pittsburgh Medical Center, PA) said when presenting the study at the Heart Rhythm Society 2006 Scientific Sessions. The analysis, based on a partial database of device-related infections designed to be extrapolated to the national level, also found that such infections tripled the risk of death independent of demographics and clinical factors such as diabetes and renal failure.

A single-center case-control study also presented at the meeting identified poor LV function, use of a central venous catheter, and multiple operators at device implantation as significant risk factors for infection related to pacemakers or ICDs [2]. Three quarters of such infections over 13 years involved the pocket created for the generator and 17% manifested as myocarditis.

Estimated US national rates and risk factors . . .

Voigt and colleagues compared device implantation with infection numbers and trends between 1996 and 2003 in the National Hospital Discharge Survey [3]. The patients with such infections were compared with a control group of device-implanted patients who were discharged without documented evidence of infection.

Estimated numerical increases in device implantations and infection, by device type (Voigt et al)

Years and differences	All devices	Pacemakers	ICDs
Implantations
1996 (n)	159 585	137 870	21 715
2003 (n)	237 720	180 284	57 436
Increase (%)	149	131	264
Infections
1996 (n)	4120	3767	353
2003 (n)	12 619	10 487	2132
Increase (%)	306	278	604

During the study period, hospitalizations for infections related to pacemakers or ICDs seemed to rise about twice as fast as the total number of device implantations, Voigt observed. A similar proportional increase was seen for pacemaker infections and insertions. But infections related to ICDs alone multiplied by six over the eight-year period. Still, pacemaker infections might represent the greater clinical problem overall, because, even by 2003, there were more than three times as many insertions of pacemakers as ICDs, and the pacemaker-related infection rate was higher.

Estimated infection rates in first and last analysis years

Years	All devices	Pacemakers	ICDs
1996 (%)	2.58	2.73	1.63
2003 (%)	5.31	5.82	3.71

Exploring potential independent predictors of device-related infections and mortality, Voigt and colleagues found not only that increasing age and being female lowered the infection risk, larger hospital size was associated with more infections. "This is likely secondary to referral bias," he said, observing that patients referred for device explantations often are sent to larger, more experienced centers.

Multivariate predictors of infection and mortality, OR (95% CI)

Parameter	Device infection	p	In-hospital mortality	p
Age (per 20-year increase)	0.62 (0.56-0.68)	<0.001	1.63 (1.41-1.89)	<0.001
Female sex	0.63 (0.51-0.76)	<0.001	0.86 (0.72-1.03)	NS
Renal failure	1.17 (0.75-1.81)	0.49	1.76 (1.26-2.47)	<0.001
ICD vs pacemaker	0.79 (0.59-1.04)	0.09	0.78 (0.55-1.11)	NS
Hospital size (per 100-bed increase)	1.44 (1.33-1.57)	<0.001	1.04 (0.97-1.12)	NS
Device-related infection	—	—	2.41 (1.58-3.66)	<0.001

With his presentation's subtitle (An epidemic on the rise?) suggesting that the findings point to an impending "epidemic" of device-related infections, Voigt speculated on possible causes. "One could be that we are now implanting a more vulnerable population of patients and that perhaps patients who receive ICDs for primary prevention are in some way more compromised and more vulnerable to having an infection," he said. "Alternatively, the explosive rise in the use of [such devices], particularly ICDs, could increase the proportion of implantations by lower-volume operators and facilities," which have been linked to increased complication rates. Or, "as the medical community accepts device therapy, perhaps we're more aware." Detecting and reporting device-related infections could also have improved over the eight years of the analysis, he said.

. . . And a single-center experience

Dr Akbar H Khan (Emory University, Atlanta, GA) and associates compared the 58 patients with pacemaker- or ICD-related infections who were implanted and followed up at the Mayo Clinic (Rochester, MN) between 1991 and 2003 with twice as many case-controls matched by age, sex, and year of hospitalization. About half the infections were related to first device insertions and the rest to generator replacements or lead procedures.

Multivariate predictors of implantable-device-related infection, single-center experience

Parameter	OR (95% CI)	p
LVEF (every 10-point decrease)	1.80 (1.38-2.36)	<0.001
Central venous catheterization	7.01 (1.37-35.85)	0.02
Two operators vs one*	3.74 (1.23-1.39)	0.02
Number of device-related procedures	1.55 (1.09-2.22)	0.02
Preimplantation temporary leads	3.19 (1.05-9.68)	0.04
Three operators vs one*	6.24 (0.83-47.11)	0.08

A lack of antibiotic prophylaxis was also a predictor, but not an independent one. In addition, the group observed, "purported" risk factors in the literature—including obesity, diabetes, cancer, renal insufficiency, skin diseases, corticosteroid use, and disorders associated with a compromised immune system—did not predict infection in this analysis. The most common offending organisms were Staphylococcus aureus and coagulase-negative staphylococci.

That reduced LVEF was a risk factor might stem from vulnerability related to clinical status, according to the group. Consistent with that, chronic obstructive pulmonary disease emerged as a univariate risk factor, although not an independent one, Khan told heartwire.

"If they have heart failure, maybe with occasional exacerbations, and if they come into the hospital frequently, they get exposed to an infectious environment," Khan said. Sicker patients also tend to get "sticked" more often, to have blood drawn or for other reasons, he added, which could be a source of device-related infections.