Risk Factors Related to Infections of Implanted Pacemakers and Cardioverter-Defibrillators
Results of a Large Prospective Study
Background— The Prospective Evaluation of Pacemaker Lead Endocarditis study is a multicenter, prospective survey of the incidence and risk factors of infectious complications after implantation of pacemakers and cardioverter-defibrillators.
Methods and Results— Between January 1, 2000, and December 31, 2000, 6319 consecutive recipients of implantable systems were enrolled at 44 medical centers and followed up for 12 months. All infectious complications were recorded, and their occurrence was related to the baseline demographic, clinical, and procedural characteristics. Among 5866 pacing systems, 3789 included 2 and 117 had >2 leads; among 453 implantable cardioverter-defibrillators, 178 were dual-lead systems. A total of 4461 de novo implantations occurred and 1858 pulse generator or lead replacements. Reinterventions were performed before hospital discharge in 101 patients. Single- and multiple-variable logistic regression analyses were performed to identify risk factors; adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were calculated. At 12 months, device-related infections were reported in 42 patients (0.68%; 95% CI, 0.47 to 0.89). The occurrence of infection was positively correlated with fever within 24 hours before the implantation procedure (aOR, 5.83; 95% CI, 2.00 to 16.98), use of temporary pacing before the implantation procedure (aOR, 2.46; 95% CI, 1.09 to 5.13), and early reinterventions (aOR, 15.04; 95% CI, 6.7 to 33.73). Implantation of a new system (aOR, 0.46; 95% CI, 0.24 to 0.87) and antibiotic prophylaxis (aOR, 0.4; 95% CI, 0.18 to 0.86) were negatively correlated with risk of infection.
Conclusions— This study identified several factors of risk of device infection and confirmed the efficacy of antibiotic prophylaxis in recipients of new or replacement pacemakers or implantable cardioverter-defibrillators.
Received December 1, 2006; accepted June 25, 2007.
Currently >3 million implanted cardiac pacing systems and ≈180 000 implantable cardioverter-defibrillators exist worldwide.1 More than 50 000 devices are implanted each year in France. The rate of device implantations is increasing with aging of the general population and the formulation of new indications. Although conferring obvious benefits, the use of these implantable devices is associated with serious complications, including infection. The course of untreated infections is dismal, and their management has important economic implications.2 The estimated rate of infection after implantation of permanent endocardial leads is between 1% and 2%, although this rate varies in the literature from 0.13% to 12.6%.3–13 The files collected by the National Hospital Discharge Survey revealed that between 1996 and 2003, the rates of hospitalization for infections of implantable antiarrhythmic systems increased faster than the rates of system implantations.14 To limit the incidence of this major complication, a precise definition and understanding of associated risk factors are needed. The Prospective Evaluation of Pacemaker Lead Endocarditis (PEOPLE) study was a French, nationwide, multicenter prospective survey of the incidence and risks factors of device-related infections after transvenous implantations of pacing and implantable cardioverter-defibrillator systems, hereafter referred to as implantable antiarrhythmic systems.
Clinical Perspective p 1355
The overall study population consisted of 6319 patients (mean age, 73.4±13.9 years; 59.7% men) who underwent transvenous implantation of an antiarrhythmic system at one of the 44 participating medical institutions (see Appendix), including 15 university-based medical centers, 14 public hospitals, and 15 private medical institutions, between January 1, 2000, and December 31, 2000. The protocol of PEOPLE was reviewed and approved by the institutional ethics committee, and written informed consent was obtained from all enrolled patients.
Data were recorded prospectively at each enrolling medical center with respect to (1) patients’ demographic and clinical characteristics; (2) case load and facilities available; (3) cumulative experience and numbers of operators of support staff; (4) preoperative risk factors, including administration of anticoagulants or antiplatelet regimen, presence of cutaneous lesions, presence of signs or symptoms of infection, and use of a temporary pacing system; (5) procedural characteristics (de novo implantation, pulse generator or lead replacement, or local or general anesthesia); (6) type of device implanted, including number of leads; (7) vascular access and presence of an ipsilateral intravenous infusion; (8) use and type of antibiotic prophylaxis; (9) use of a wound-draining system; (10) duration of hospitalization; and (11) occurrence of early or late complications requiring reintervention or not. Patients were followed up by a physician of the implanting center at 3, 6, and 12 months after the index procedure for a physical examination, interrogation, and verification of the proper function of the implanted system, as well as recording of all device-related complications. The occurrence of a complication mandating a reintervention was the beginning of a new follow-up period, but patients experiencing a complication were recorded only once in the database, and the follow-up was closed at 12 months after the inclusion.
Information on the circumstances and modes of death was systematically retrieved from patient records when death occurred in hospital or by telephone communication with the patient’s primary physician, family, or both when a patient died in an ambulatory setting. Individual patient data, gathered by the implanting physicians at each center, were centrally collected by a coordinating center at the University of Lille (Lille, France). Audits were performed on 600 patient records selected randomly. All deaths and device complications were reviewed and adjudicated by the events committee, which could request additional information when needed.
The use and type of antibiotic prophylaxis and the management of postprocedural complications were left to the discretion of each participating center. Infectious complications were defined prospectively before initiation of the study and were described in the study protocol. Extensive investigations, including transesophageal echocardiography, were strongly recommended when a device-related infection was suspected.
The Duke criteria for diagnosis of endocarditis were applied to systemic infections related to the implantable antiarrhythmic systems.15,16 Nosocomial surgical site infections were defined according to the US Centers for Disease Control.17
Local infection was diagnosed when a fistula, abscess, or purulent collection had developed at the site of the implanted material or when impending or frank material erosion was associated with fever or with a major Duke criterion. Impending or frank erosion without apparent infectious manifestations was not included among infectious complications but was recorded in the database.
Reintervention, defined as a new procedure after the index implantation to manage a noninfectious complication, was defined as early when occurring before hospital discharge.
Immunosuppression was defined as a state in which the ability of the body’s immune system to respond was decreased. Immunosuppression may be caused by certain infections (AIDS), some drugs (corticosteroids, cytotoxic drugs, immunosuppressive drugs for transplantation), and radiation.
For descriptive univariate analyses, quantitative variables are presented as median values and interquartile range. Categorical variables were compared by the χ2 test. Variables recorded prospectively at each enrolling medical center and detailed in Methods were used for the univariate analysis. Variables associated with a value of P<0.15 in the single-variable analysis were entered into a multiple-variable logistic regression analysis. We used a backward stepwise procedure to select variables in the final model. Adjusted odds ratios (aOR) and 95% confidence intervals (CIs) were calculated with the use of estimated regression coefficients and their SEs in the logistic regression analysis. Because some situations may influence the prescription of antibiotic prophylaxis, the existence of interactions between antibiotic prophylaxis and use of a temporary pacing system, fever, and de novo implantation was examined. The statistical rules of independence of observations were applied, and each patient was included once in the data set. The adequacy of model to the data was tested with the Hosmer-Lemeshow goodness-of-fit test, and for the model fitting, we used the Akaikes information criterion.
Patients were recorded only once in the database, and the follow-up was closed at 12 months after the inclusion. Infections occurring after 12 months of follow-up were not taken into account in the analysis.
The null hypothesis was rejected in each test when values of P<0.05. The analyses were made with Epi-info version 6.4 (Centers for Disease Control and Prevention, Atlanta, Ga) and SAS (SAS Institute, Cary, NC) statistical software packages.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Demographic and disease characteristics of the study participants are summarized in Table 1. The types of index procedure performed and devices implanted are summarized in the Figure. At least 1 postoperative follow-up was available in 6134 patients (97%), and 12-month follow-up forms were completed for 5404 patients (85.5%). Antibiotic prophylaxis was used in 88% of procedures, consisting of a β-lactam antibiotic in 92.2%. Antibiotic prescriptions were in accordance with the international guidelines on antibiotic prophylaxis in 2294 patients. A single dose was given to 3380 patients; several doses were given to the remaining patients.
Among 633 deaths (10.1% of the overall population), 8 were related to an infectious process, including 1 joint prosthesis infection, 1 cellulitis, 1 renal abscess, 4 pneumonia, and 1 sepsis of undetermined origin. In 2 instances, a diagnosis of pacemaker-related infection could not be excluded.
Noninfectious complications (Table 2) occurred in 548 patients (8.9%), of whom 175 (31.9%) required a reintervention and 101 (18.4%) required early reinterventions.
Infections developed over 12 months in 42 patients, representing an incidence of 0.68 per 100 patients (95% CI, 0.47 to 0.89) or 2 per 105 patient-days (1.4 per 105 to 2.6 per 105). The incidence of infection was 0.56 per 100 patients (95% CI, 0.33 to 0.78) and 0.99 per 100 patients (95% CI, 0.54 to 1.45) after de novo implantation and non–de novo implantation, respectively. Important characteristics of the infected patients are presented in Table 3. Management of the infection required removal or replacement of the foreign material in all but 5 patients whose prognosis was poor because of concomitant noninfectious disorders. The median time to infection was 52 days (quartile 1 to 3, 24 to 162 days) between the index procedures and diagnosis, with 16 patients having acute or subacute presentations within 6 weeks of the procedure. The results of the single- and multiple-variable analyses are summarized in Table 4. Infections were positively correlated with fever occurring within 24 hours before system implantation, temporary pacing before implantation, and early reintervention for hematoma or lead replacement. Infections were negatively correlated with de novo implantation procedures and antibiotic prophylaxis.
Confirmed infections or impending or frank erosions were observed in 73 patients. The incidence of these complications was 1.19 per 100 patients (95% CI, 0.92 to 1.46) and were positively correlated with fever within 24 hours before system implantation (aOR, 4.8; 95% CI, 1.6 to 14.5; P<0.01) and early reintervention (aOR, 55.3; 95% CI, 33.1 to 92.6; P<0.0001) and negatively correlated with de novo implantation (aOR, 0.4; 95% CI, 0.2 to 0.7; P<0.0001) and antibiotic prophylaxis (aOR, 0.37; 95% CI, 0.20 to 0.70; P<0.01).
This large prospective, multicenter registry examined the 1-year incidence of infectious complications after the implantation of antiarrhythmic devices and the risk factors associated with this complication, which have important consequences from the standpoint of preventive measures. The value of antibiotic prophylaxis at the time of device implantation has not been established by a single definitive study, and a meta-analysis was needed to show a possible benefit conferred by systemic antibiotic administered during the procedures.18 In this study, an inverse correlation was observed between the development of infections and antibiotic prophylaxis. Most patients received β-lactam antimicrobial agents, and our observations confirmed the preventive efficacy of these drugs when delivered during the implantation procedure.
The patients in whom a temporary pacing system was present at the time of implantation of the permanent antiarrhythmic systems were more than twice as likely to develop device-related infections as patients who were not temporarily paced. This finding is concordant with previous studies that showed a relationship between temporary pacing wires and pacemaker-related infections.9,19 Although the use of a temporary pacing system could be a marker of urgent procedures, the duration of hospitalization before device placement was not related to the risk of infection. Other systemic factors have been reported to contribute to a higher incidence of pacemaker-related infections.20,21 Our study did not confirm the effects of diabetes mellitus, immunosuppression, the use of anticoagulants, and skin diseases; it did, however, confirm that another infection causing fever within 24 hours before implantation of the permanent system increased the risk of infection.
In the study by Aggarwal et al,9 the incidence of infections in patients with dual-chamber (6 of 573, 1%) compared with single-chamber devices (4 of 486, 0.82%) was similar, whereas in the study by Chauan et al,19 wound infections developed in 0.6% of patients with single-chamber versus 2.1% of patients with dual-chamber devices. In our study, although the infection rate between patients with dual- versus single-chamber devices was not significantly different, a trend was observed toward an increasing risk of infection with increasing number of leads implanted. The very small number of systems with 3 leads might explain that the 1.77% rate of infection in this group did not reach significance.
Secondary procedures such as pulse generator replacements are a risk factor for infection. The rate of infectious complications in patients who undergo multiple implantations of devices in their lifetime is inordinately high. For example, in a recent retrospective analysis, the infection rate in young patients who had undergone a median of 2 pacemaker implantations was 5.5%.13 Early reinterventions for hematoma or lead dislodgement were the leading risk factor of infection, associated with an aOR of 15.04. Therefore, the consequences of an early reintervention for a noninfectious complication can be considerably more serious than the original complication itself. Furthermore, it is likely that necessary secondary procedures such as replacements of a depleted pulse generator added to early reinterventions for noninfectious complications have a multiplicative effect on the risk of system infection. Although is obvious that hematomas or lead dislodgements must avoided, their occurrence should not be cause for systematic reinterventions because not all hematomas are unequivocally in need of evacuation and not all leads are indispensable for the patient’s comfortable survival. In such cases, therapeutic abstention might be preferable to exposing the patient to a high risk of infection.
With a cutoff value set at 100 implantations per year, our analysis did not show an effect of medical center activity on the incidence of infection. Because only 196 patients underwent the index procedure in a center performing <50 implantations per year, a cutoff of 50 implants per year could not be examined.
Because pacing and implantable cardioverter-defibrillator system infection lacks a precise definition, its diagnosis may be a challenge.16 Therefore, we adopted the definitions of surgical site infections proposed by the US Centers for Disease Control.17 Despite the inclusion of all cases corresponding to these definitions, the rate of infections might nevertheless have been underestimated. In addition, although the mechanical versus infectious origin of implanted material erosion remains controversial, it is generally believed that its occurrence in the absence of clinical or biological signs of infection is due to mechanical factors. However, in a recent study, an infectious process was found in most local device-related complications,22 including in the absence of associated manifestations of infection. The inclusion of erosions and impending erosions in our analysis changed only the incidence of infectious complications, without modifying the identification of risk factors. Finally, some infections develop several years after device implantation,1 and a longer follow-up is needed to include these late infections in the analysis.
The incidence of definite system-related infections was 0.68% at 12 months. Predictive factors were fever before implantation, temporary pacing before implantation, device replacement or revision, early reintervention, and absence of antibiotic prophylaxis.
The following investigators and institutions participated in PEOPLE:
Claude Barnay, Jérôme Taieb, Centre hospitalier du pays d’Aix, Aix-en-Provence; Daniel Galley, Centre Hospitalier d’Albi, Albi; Noel Labban, Clinique Ambroise Paré, La Bassée; Jean-Luc Rey, Centre Hospitalier Régional Universitaire d’Amiens, Amiens; Jacques Victor, Centre Hospitalier Régional Universitaire d’Angers, Angers; Etienne Bearez, Géry Hannebicque, Centre Hospitalier d’Arras, Arras; André Atallah, Centre Hospitalier Intercommunal Basse Terre, Basse terre; Bernard D’Hautefeuille, Hôpital de Béthune, Béthune; Marc Botte, Clinique Bois Bernard, Bois Bernard; Eric Fossati, Polyclinique du Bois, Lille; Jacques Clementy, Centre Hospitalier Régional Universitaire de Bordeaux, Bordeaux; Jean-Jacques Blanc, Centre Hospitalier Régional Universitaire de Brest, Brest; Jean-Michel Hethuin, Clinique de Cambrai, Cambrai; Jean-Michel Sommier, Clinique Saint Sauveur, Mulhouse; Claude Sussmann, Clinique du Diaconat, Mulhouse; Jean-Jacques Dujardin, Centre Hospitalier de Douai, Douai; Jean-Charles Aisenfarb, Hôpital de Dunkerque, Dunkerque, Daniel Gras, Philippe Ritter, Arnaud Lazarus, Clinique Georges Bizet, Paris; N’Guyen, Polyclinique de Grande Synthe, Grande Synthe; William Mekerke, Clinique de Gravelines, Gravelines; Manuel Font, Centre Hospitalier Henri Mondor, Aurillac; Robert Frank, Françoise Hidden, Hôpital Jean Rostand, Paris; Max Pecheux, Hôpital de Lens, Lens; Henri Jambert, Centre Hospitalier de Libourne, Libourne; Salem Kacet, Centre Hospitalier Régional Universitaire de Lille, Lille; Henri Bussilet, Hospices Civiles de Lyon, Lyon; Jean Cavallaro, Centre Hospitalier des Chanaux, Mâcon; Pierre Djiane, Assistance Publique, hôpitaux de Marseille, Marseille; Dominique Defaut, Centre Hospitalier Sambre Avesnois, Maubeuge; Bernard Dodinot, Clinique Saint-André, Vandoeuvre-lès-Nancy, Nancy; Etienne Aliot, Nicolas Sadoul, Hôpital de Brabois, Centre Hospitalier Universitaire de Nancy, Nancy; Gille Lande, Hôpital Guillaume et René Laennec, Nantes; JG Kanayakis, Lotfi Larouchi, Clinique Cardiologique Paulmy, Bayonne; Christophe Camier, Christian Caron, Polyclinique du Parc, Maubeuge; Marc Delay, Centre Hospitalier Régional Universitaire Purpan, Toulouse; Michèle Salvador-Mazenq, Centre Hospitalier Universitaire Rangueil, Toulouse; Dominique Pavin, Centre Hospitalier Universitaire de Rennes, Rennes; Stéphane Dennetière, Centre Hospitalier de Roubaix, Roubaix; Frédéric Anselme, Centre Hospitalier Universitaire de Rouen, Rouen; Charles Nguyen Tan Cuong, Clinique Saint-François, Mainvilliers; Dominique Huyart, Clinique de Saint-Omer, Saint Omer; Pierre Graux, Centre Hospitalier Saint-Philibert, Lomme; Jean Michel Hethuin, Clinique Sainte Marie, Cambrai; Michel Chauvin, Centre Hospitalier Universitaire de Strasbourg, Strasbourg; Arnaud Hubert, Polyclinique de Ternoise, St-Paul sur Ternoise; Eric Decoulx, Centre Hospitalier de Tourcoing, Tourcoing; Ben Agraou, Centre Hospitalier de Valenciennes, Valenciennes; Bruno Sivery, Centre Hospitalier de Wattrelos, Wattrelos.
Sources of Funding
This study was supported by the French Working Group on Cardiac Pacing of the French Society of Cardiology and by a grant from the French Ministry of Health (PHRC/1918)
Dr Clementy has received research support from Medtronic, the Sorin Group, and Biotronik and has served as a consultant for the Sorin Group and Guidant.
Harcombe AA, Newell SA, Ludman PF, Wistow TE, Sharples LD, Schofield PM, Stone DL, Shapiro LM, Cole T, Petch MC. Late complications following permanent pacemaker implantation or elective unit replacement. Heart. 1998; 80: 240–244.
Aggarwal RK, Connelly DT, Ray SG, Ball J, Charles RG. Early complications of permanent pacemaker implantation: no difference between dual and single chamber systems. Br Heart J. 1995; 73: 571–575.
Klug D, Lacroix D, Savoye C, Goullard L, Grandmougin D, Hennequin JL, Kacet S, Lekieffre J. Systemic infection related to endocarditis on pacemaker leads: clinical presentation and management. Circulation. 1997; 95: 2098–2107.
Da Costa A, Kirkorian G, Cucherat M, Delahaye F, Chevalier P, Cerisier A, Isaaz K, Touboul P. Antibiotic prophylaxis for permanent pacemaker implantation: a meta-analysis. Circulation. 1998; 97: 1796–1801.
Voet JG, Vandekerckhove YR, Muyldermans LL, Missault LH, Matthys LJ. Pacemaker lead infection: report of three cases and review of the literature. Heart. 1999; 81: 88–91.
Klug D, Wallet F, Lacroix D, Marquie C, Kouakam C, Kacet S, Courcol R. Local symptoms at the site of pacemaker implantation indicate latent systemic infection. Heart. 2004; 90: 882–886.
The rate of pacemaker and implantable cardioverter-defibrillator implantations is increasing with the formulation of new indications. Although conferring obvious benefits for patients, use of these implantable devices is associated with serious complications. To limit the incidence of major complications, a precise definition and understanding of associated risk factors are needed. This large prospective, multicenter registry examined the 1-year incidence of infectious complications after the implantation of antiarrhythmic devices and the risk factors associated with this complication, which have important consequences from the standpoint of preventive measures. Five risk factors have been identified with direct consequences in the management of patients requiring device implantation. Infections were positively correlated with fever occurring within 24 hours before system implantation and temporary pacing before implantation. When possible, implantation must be delayed in case of fever, and temporary pacing must be avoided. An inverse correlation was observed between the development of infections and antibiotic prophylaxis. The study confirmed the preventive efficacy of antibiotic when delivered during the implantation procedure. Infections were negatively correlated with de novo implantation procedures. Finally, early reinterventions for hematoma or lead dislodgement were the leading risk factors of infection. Therefore, their occurrence should not be cause for systematic reinterventions because not all hematomas are unequivocally in need of evacuation and not all leads are indispensable for the patient’s comfortable survival. In such cases, therapeutic abstention might be preferable to exposing the patient to a high risk of infection.