Abstract: Objectives. The purpose of this study was to investigate the relationship between operator volume and implantable defibrillator lead failure and patient mortality at a single large implanting center. Methods. This study analyzed the differences between high-volume and low-volume defibrillator implanters in the Pacemaker and Implantable Defibrillator Lead Survival Study (“PAIDLESS”) between February 1, 1996 and December 31, 2011 at NYU Winthrop Hospital. “High-volume” was defined as performing ≥500 implants over the study period, while “low-volume” was defined as performing <500 implants. Comparisons between the procedure volume groups were performed using Fisher’s Exact test, Wilcoxon rank-sum test, and Kaplan-Meier analysis as appropriate. Results. Eight operators participated in the study, four of whom were high-volume operators. Of 3801 patients, a total of 3149 (83%) were operated upon by high-volume operators. Low-volume operators implanted fewer recalled leads (12% vs 42%; P<.001) and more often obtained venous access through the cephalic vein cut-down approach (63% vs 38%; P<.001) than high-volume operators. Kaplan-Meier analysis revealed shorter time to lead failure in the low-volume group (P=.02). Time to mortality was not significantly different between the high-volume and low-volume groups (P=.18). When adjusted for lead recall status, patients of high-volume operators were 43% less likely to experience lead failure compared to patients of low-volume operators. Conclusions. High-volume defibrillator implanters selected a higher percentage of recalled leads, but their patients were less likely to encounter lead failure when adjusted for lead recall status compared to low-volume operators.
J INVASIVE CARDIOL 2017;29(12):E184-E189.
Key words: implantable defibrillator, complications, lead failure
Previously, in interventional cardiology, there has been an established relationship between operator procedure volume and outcomes. Studies involving percutaneous coronary interventions (angioplasty and stenting) have demonstrated that high-volume operators have the lowest complication rates and the best procedural outcomes.1-5 In addition, an association between higher defibrillator implant volume and a lower complication rate has been demonstrated in at least four large studies.6-9 However, these studies only examined implants over a relatively short period (<6 years) and were multicenter in design. Also, none of these studies focused in detail on defibrillator lead failure. The purpose of this substudy was to determine whether high-volume operators have better outcomes than low-volume operators with respect to implantable defibrillator lead failures and patient mortality in a large single-center study. A secondary objective was to investigate whether implant approaches differed by operator volume.
The Pacemaker and Implantable Defibrillator Lead Survival Study (“PAIDLESS”) is a large, single-center, retrospective analysis of pacemaker and implantable defibrillator leads implanted over approximately a 15-year span at NYU Winthrop Hospital.10 This study analyzed all pacemaker and defibrillator lead implants performed from February 1, 1996 to December 31, 2011 and was approved by the NYU Winthrop Hospital Institutional Review Board. In the present substudy, defibrillator implant procedures were categorized based upon operator volume. Patients were categorized into either the high-volume group (≥500 procedures) or low-volume group (<500 procedures), based on the number of implanting procedures completed by their operating physician during the study period. These groups were then compared in regard to patient characteristics, implant procedure information, lead survival, and mortality as determined by the Social Security Death Index.11 All information was recorded in a de-identified database, following regulations of the Health Insurance Portability and Accountability Act (HIPAA).12
Lead failure was defined according to the Medtronic System Longevity study, and included failure to capture and/or sense, abnormal pacing and/or defibrillator coil impedance, increased pacing thresholds, insulation defect, lead fracture, cardiac perforation, extracardiac stimulation, and/or dislodgment.10,13 The investigators responsible for analyzing the data were blinded to patient and operator identifiers.
Statistical analysis. In general, continuous variables are presented as mean ± standard deviation and the categorical data as frequency (percentage). The main endpoints were lead failure and mortality. Continuous variables were tested for normality using histograms, q-q plots, box and whiskers plots, and Kolmogorov-Smirnov tests. The effects of outliers were investigated by analyzing data with and without extreme observations. The interpretation of the data remained the same with or without the outliers. Hence, the final analyses included full data. Patient characteristics were compared between the high-volume operator group (≥500 procedures) and low-volume operator group (<500 procedures). Wilcoxon rank-sum tests and Fisher’s exact tests were used as appropriate for the comparisons.
Survival estimates and cumulative event rates were compared by the Kaplan-Meier method using the time to event approach for both lead failures and mortality. The Wilcoxon test was used to compare the Kaplan-Meier survival curves of the high-volume and low-volume operators. The Wilcoxon test puts more weight on the earlier time period compared to the later follow-up time. This fits our data well, as most of the events (ie, lead failure) occurred within the first 5 years. A multivariable Cox regression hazard model was developed for time to lead failure using procedure volume groups and recall lead status as the covariates.
All calculations were performed using SAS 9.4 for Windows (SAS/STAT version 13.1; SAS Institute) and results were considered statistically significant when P-value was <.05.
There were a total of eight operators who performed defibrillator implant procedures during the PAIDLESS trial period: seven electrophysiologists and one cardiothoracic surgeon. Four out of the seven electrophysiologists were categorized as high-volume operators, having completed ≥500 procedures. The remaining three electrophysiologists and the cardiothoracic surgeon were considered low-volume operators.
Table 1 shows the patient characteristics for the high-volume vs low-volume groups. High-volume implanters operated on 3149 patients while low-volume implanters operated on 652 patients. High-volume operators implanted a significantly greater proportion of leads in male patients (75.0% vs 68.9%; P<.001) and in older patients (70.8 ± 12.5 years vs 68.4 ± 13.0 years; P<.001). The follow-up time for patients operated on by high-volume operators was significantly longer than the follow-up time for those operated on by low-volume operators (4.4 ± 3.1 years vs 2.0 ± 2.0 years; P<.001). Significantly more patients of high-volume operators had ischemic cardiomyopathy, syncope, significant bradycardia, significant conduction disease, first-degree atrioventricular (AV) block, and non-sustained and sustained ventricular tachycardia (P<.001). Also, more patients in the high-volume group had coronary artery disease, prior pacemakers, supraventricular tachycardia (including AV node reentrant tachycardia), atrial flutter, and trifascicular block (P<.05). Significantly more patients operated on by low-volume operators had previous coronary bypass artery graft surgery, percutaneous coronary interventions, congestive heart failure, hypertension, diabetes mellitus, left ventricular hypertrophy, valve disease, atrial fibrillation, and long QT syndrome (P<.001). Additionally, more of these patients had mitral valve replacements, dilated cardiomyopathy, hypertrophic cardiomyopathy, prior implantable cardioverter defibrillators, left ventricular dysfunction, and ventricular fibrillation (P<.05).
A total of 4078 leads were implanted in 3802 patients by the eight operators during the PAIDLESS trial period. However, 1 patient was excluded from this subanalysis as information regarding the implanting physician was missing, resulting in an analysis of 4077 leads implanted in 3801 patients. Table 2 compares the type of leads implanted by high-volume vs low-volume operators by lead manufacturer. More than one-half of the leads implanted by the high-volume operators were Medtronic leads (52%), nearly one-half of which were recalled Sprint Fidelis leads (43%). In contrast, 85% of the procedures performed by low-volume implanters used St. Jude Medical leads, the vast majority of which were Durata leads (92%). High-volume operators implanted significantly more recalled leads than low-volume operators (42% vs 12%; P<.001). Figure 1 displays the differences in lead selection over time between high-volume and low-volume operators. The majority of recalled leads were implanted between 2003 and 2007 in both groups, although low-volume operators completed fewer procedures during this time. Failure rates for each lead family implanted by high-volume and low-volume operators are presented in Table 3. The high-volume group encountered a significantly higher incidence of failure during the study (4.0% vs 2.4%; P=.04).
There were notable differences in venous access approach between the two study groups. The majority of high-volume implanters obtained percutaneous needle access through the subclavian and/or axillary vein (62% of implants). In contrast, the majority of low-volume implanters used the cephalic vein cut-down approach (63%). The difference in implant approach between the two groups was statistically significant (P<.001).
Kaplan-Meier analysis of time to lead failure comparing high-volume and low-volume operators (Figure 2) revealed a statistically significant difference in which patients of low-volume operators experienced a shorter time to lead failure (P=.02). The mean (± standard error) time to lead failure in the low-volume group was 3.4 ± 0.02 years vs 12 ± 0.07 years in the high-volume group. Kaplan-Meier analysis of all-cause patient mortality revealed no significant difference in time to mortality between the high-volume and low-volume groups (P=.18).
A multivariable Cox regression hazard model for time to lead failure using procedure volume group and lead recall status as covariates revealed that when adjusted for lead recall status, patients operated upon by high-volume operators are 43% less likely to experience lead failure vs the patients operated upon by low-volume operators (hazard ratio [HR], 0.57; 95% confidence interval [CI], 0.34-0.97; P=.04). The HR (95% CI) for lead recall status in this model was 1.58 (1.13-2.2). It is noteworthy that the hazard ratio (95% CI) for the high-volume operator group from the univariate model is 0.63 (0.38-1.06), which lacks significance by a slim margin. All analyses were repeated excluding the cardiothoracic surgeon. However, the results of these analyses remained virtually unchanged.
This study demonstrated a lower risk of implantable defibrillator lead failure in patients of high-volume operators compared to those of low-volume operators when adjusting for lead recall status. This finding agrees with earlier interventional studies, which found that higher operator volume was associated with fewer complications and better outcomes.1-5 This same phenomenon has also been observed with respect to defibrillator implantation.6-9 Al-Khatib et al examined Medicare patients between 1999 and 2001 and identified that high-volume operators had a lower rate of defibrillator infections and complications.6 Additionally, Freeman et al examined the National Cardiovascular Data Registry (1463 hospitals, 4011 physicians, and 356,515 initial implants) and found not only a lower complication rate among high-volume operators but also a significant decrease in mortality in that subgroup.7 Another study that examined the ICD Registry of the National Cardiovascular Data Registry found higher procedural volumes to be associated with lower odds of cardiac perforation.8 An analysis of the Danish Pacemaker and ICD Register also revealed that low-volume operators had higher risks of cardiac perforation, infection, and minor hematoma compared to high-volume operators.9
The finding of a higher usage of recalled leads among our definition of high-volume operators may be related to the timing of this study with respect to the issuance of two major defibrillator lead recalls – Medtronic’s Sprint Fidelis and St. Jude Medical’s Riata and Riata ST leads. This study examined lead implants between February 1, 1996 and December 31, 2011. The Sprint Fidelis leads, commonly used by the high-volume operators in this study, were recalled on October 17, 2007.14 The Riata and Riata ST leads were recalled on November 28, 2011,15 but St. Jude Medical distributed a safety notice regarding both of these leads and stopped selling them in December 2010.16 Both high-volume and low-volume operators may have exhibited a preference for smaller diameter leads, which include the Sprint Fidelis, Riata ST, and Durata leads. Once a significant safety advisory or recall was issued, however, operators could no longer implant the affected leads. Since low-volume operators completed the majority of their implants toward the end of the study (Figure 1), the recalled Sprint Fidelis and Riata ST leads were unavailable for selection for most of these operators’ procedures. This may have influenced their use of the non-recalled Durata lead in 78% of their implants.
While Kaplan-Meier analysis revealed a shorter time to failure for low-volume operators (Figure 2), no instances of failure were observed beyond 5 years post implantation. Long-term data for low-volume operators were similarly sparse during analysis of time to mortality. This could be explained by differences in follow-up time between operator groups. Low-volume implanters had a shorter average follow-up time vs high-volume operators (2.0 years vs 4.4 years; P<.001), likely due to completing many of their procedures toward the end of the study (Figure 1).
This study revealed a significantly higher incidence of lead failure among high-volume operators vs low-volume operators (4.0% vs 2.4%) over the 15-year time period examined. However, this comparison may be misleading as it does not take time into consideration. Since the probability of lead failure increases with time, a shorter follow-up time, as observed among low-volume operators, may have contributed to the lower incidence of lead failure seen in this group. Additionally, PAIDLESS has previously demonstrated a strong, statistically significant association between recalled leads and lead failure as well as mortality.10 Thus, the higher incidence of lead failure observed in the high-volume group may also be attributable to more frequent use of recalled leads by high-volume operators (Table 2).
Patient characteristics differed significantly between the two groups with respect to age, gender, previous invasive procedures, and comorbidities. In addition, this study identified that the cephalic vein cut-down approach was more commonly used by the low-volume group to obtain venous access. This approach has a lower incidence of lead fracture from subclavian crush syndrome when compared to other approaches and could have served as a protective factor for low-volume operators.17
Study limitations. This study has several limitations. First, the definition of high-volume vs low-volume operators was based on an arbitrary volume of 500 implants, in order to create and compare two different groups with an equal number of operators. Alternatively, the data could have been analyzed using operator annual implant volume quartiles, similar to two large multicenter defibrillator trials.6,7 The study conducted by Al-Khatib et al placed physicians who performed 1-6 implants per year into the low-volume quartile and physicians who performed ≥18 implants per year into the high-volume quartile.6 Similarly, Freeman et al defined the low-implant quartile as ≤4 implants per year, and the high-volume implanter quartile as >37.25 implants per year.7 If we had applied the latter definition to our study, only two of our operators would have fallen into this low-volume category, significantly limiting our sample size. Thus, this PAIDLESS substudy used absolute procedural volume throughout the course of the study, rather than annual implant volume. Another limitation was the timing of this study, as it included two major lead recalls, and low-volume operators mostly completed their procedures toward the end of the study. These factors, along with disparities in patient characteristics, implant approach, and specific lead selection may have all influenced the findings of this study.
This study examined defibrillator lead failure and mortality among patients operated on by high-volume and low-volume defibrillator lead implanters. This study also investigated the differences in implant approach between the two groups. High-volume operators chose recalled leads and utilized the subclavian and/or axillary venous approach more frequently than low-volume operators. PAIDLESS previously identified recalled status as independently associated with lead failure and mortality.10 The current substudy demonstrated a lower risk of lead failure among patients of high-volume implanters when adjusted for lead recall status and a shorter time to lead failure for patients of low-volume operators. Further research is necessary to fully understand the relationship between lead failure, mortality, and operator experience.
1. McGrath PD, Wennberg DE, Malenka DJ, et al. Operator volume and outcomes in 12,998 percutaneous coronary interventions. J Am Coll Cardiol. 1998;31:570-576.
2. Kansagra SM, Curtis LH, Anstrom KJ, Schulman KA. Trends in operator and hospital procedure volume and outcomes for percutaneous transluminal coronary angioplasty, 1996 to 2001. Am J Cardiol. 2007;99:339-343.
3. McGrath PD, Wennberg DE, Dickens JD Jr, et al. Relation between operator and hospital volume and outcomes following percutaneous coronary interventions in the era of the coronary stent. J Am Med Assoc. 2000;284:3139-3144.
4. Hannan EL, Wu C, Walford G, et al. Volume-outcome relationships for percutaneous coronary interventions in the stent era. Circulation. 2005;112:1171-1179.
5. Kastrati A, Neumann FJ, Schomig A. Operator volume and outcome of patients undergoing coronary stent placement. J Am Coll Cardiol. 1998;32:970-976.
6. Al-Khatib SM, Lucas FL, Jollis JG, Malenka DJ, Wennberg DE. The relation between patients’ outcomes and the volume of cardioverter-defibrillator implantation procedures performed by physicians treating Medicare beneficiaries. J Am Coll Cardiol. 2005;46:1536-1540.
7. Freeman JV, Wang Y, Curtis JP, Heidenreich PA, Hlatky MA. Physician procedure volume and complications of cardioverter-defibrillator implantation. Circulation. 2012;125:57-64.
8. Hsu JC, Varosy PD, Bao H, Dewland TA, Curtis JP, Marcus GM. Cardiac perforation from implantable cardioverter-defibrillator lead placement. Insights from the National Cardiovascular Data Registry. Circ Cardiovasc Qual Outcomes. 2013;6:582-590.
9. Kirkfeldt RE, Johansen JB, Nohr EA, Jorgensen OD, Nielsen JC. Complications after cardiac implantable electronic device implantations: an analysis of a complete, nationwide cohort in Denmark. Eur Heart J. 2014;35:1186-1194.
10. Cohen TJ, Asheld WJ, Germano J, Islam S, Patel D. A comparative study of defibrillator leads at a large volume implanting hospital: results from the Pacemaker and Implantable Defibrillator Leads Survival Study (“PAIDLESS”). J Invasive Cardiol. 2015;27:292-300.
11. Social Security Death Index cross-referenced with manufacturer-supplied data to ensure up-to-date status of Out of Service leads that are due to the death of the patient (OOS-D) Social Security Death Index. http://www.genealogybank.com/gbnk/ssdi/
12. The United States Human and Health Services. Guidance regarding methods for de-identification of protected health information in accordance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule. Washington, DC: Government Printing Office, 2012.
13. Medtronic Criteria for Cardiac Rhythm Disease Management (CRDM) and System Longevity Study. http://wwwp.medtronic.com/productperformance/content/method_for_estimating_leads.html
14. FDA Safety Communication. http://www.fda.gov/Safety/MedWatch/SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm152658.htm
15. FDA Safety Communication. http://www.fda.gov/MedicalDevices/Safety/AlertsandNotices/ucm314930.htm
16. Chester KM. Important Product Information: St. Jude Medical Riata and Riata ST Silicone Endocardial Leads Models 1560, 1561, 1562, 1570, 1571, 1572, 1580, 1581, 1582, 1590, 1591, 1592, 7000, 7001, 7002, 7010, 7011, 7040, 7041, 7042. St. Jude Medical. December 15, 2010.
17. Roelke M, O’Nunain SS, Osswald S, Garan H, Harthorne JW, Ruskin JN. Subclavian crush syndrome complicating transvenous cardioverter defibrillator systems. Pacing Clin Electrophysiol. 1995;18:973-979.
From the Department of Medicine at NYU Winthrop Hospital, Mineola, New York.
Presented in part at Venice Arrhythmias 2015 (October 2015).
Funding: This study was submitted to each of the manufacturers listed in the manuscript (Medtronic, Boston Scientific, and St. Jude Medical); however, the study was only partially funded by Medtronic and Boston Scientific.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted January 26, 2017, provisional acceptance given March 29, 2017, final version accepted May 22, 2017.
Address for correspondence: Todd J. Cohen, MD, Director of Electrophysiology, NYU Winthrop Hospital, 212 Jericho Turnpike, Mineola, NY 11501. Email: firstname.lastname@example.org