Abstract
Background:
Increase in the number of patients treated with Implantable Cardioverter Defibrillator (ICD) requests more attention regarding its complications.
Objectives:
This study aimed to assess the generator- and lead-related complications at implantation and during follow-up in the patients who were treated with ICD for primary and secondary prevention reasons.
Methods:
We retrospectively reviewed 255 consecutive patients who underwent transvenous ICD implantation for the first time in a 7-year period and were followed-up for 3 years at Tehran Heart Center. The personal and clinical data of the patients as well as specific data on the ICD implantation were retrieved. The frequency of each of the complications was reported and the study variables were compared between the patients with and without complications using Student’s t-test and chi-square test where appropriate. P values less than 0.05 were considered as statistically significant.
Results:
Out of a total of 525 implanted leads and 255 implanted devices in 255 patients (mean age = 62.57 ± 13.50 years; male = 196 [76.9%]), complications leading to generator or lead replacement occurred in 32 patients (12.5%). The results revealed no significant difference between the patients with and without complications regarding gender and age (P = 0.206 and P = 0.824, respectively). Also, no significant difference was found between the two groups concerning the ejection fraction (P = 0.271). Lead fracture was the most frequent lead-related complication and was observed in 17 patients (6.6%). Besides, it was mainly observed in the RV leads. Generator-related complications leading to generator replacement were observed in 2 patients (0.7%).
Conclusions:
Despite considerable improvements in the ICD technology, the rate of the ICD complications leading to device replacement and surgical revision, especially those related to the leads, is still clinically important.
Keywords: Implantable Cardioverter Defibrillators, Complication, Implanted Electrodes
1. Background
The use of Implantable Cardioverter Defibrillator (ICD) in prevention of sudden cardiac death has been increased dramatically in the past few decades (1). The patients with structural heart disease and symptomatic sustained Ventricular Tachycardia (VT) or ventricular fibrillation are the prime candidates (2). Recent years have witnessed a dramatic rise in the number of patients receiving treatment with ICD and other cardiac devices and, thereby, complications necessitate timely identification and treatment (3). Currently, about 30% of the patients who receive an ICD experience at least one complication after the implantation (4). A study on the frequency and incremental cost of major complications among the patients with an ICD showed that 10.8% of the patients experienced one or more complications, and this was accompanied by a significant increase in the length of hospital stay and health service costs (5). Device related complications can be classified into generator- and lead-related complications. ICD lead failure is frequent and mostly stems from insulation defect or conductor disruption, and can affect the high-voltage or the pace-sense circuit of the lead (6, 7). The most common complications of lead failure in previous studies included oversensing electrical noise, undersensing ventricular tachyarrhythmias, inappropriate therapy, and lethal proarrhythmia (8, 9).
Up to now, several studies have been conducted on the ICD-related complications, such as inappropriate shocks, infection, and endocarditis. However, data on the complications directly related to the ICD leads or generator malfunction are limited.
2. Objectives
Hence, the present study aims to assess both the generator-related and lead-related complications at implantation and during follow-up in the patients who were treated with an ICD for primary and secondary prevention reasons at Tehran Heart Center, Tehran, Iran.
3. Patients and Methods
3.1. Patients
In this study, clinical data of 255 patients who underwent first-time transvenous ICD implantation between March 2003 and February 2009 at Tehran Heart Center and had at least 3 complete follow-up visits within 3 years following device implantation were retrospectively reviewed. General characteristics of the study population and its clinical data as well as specific data on the ICD implantation were recorded for each patient. All the patients signed a written informed consent in accordance with the protocols of Tehran Heart Center at the time of admission, granting approval for the anonymous use of their data in medical research. Also, the proposal of the present study was approved by the Research Board of Tehran Heart Center.
3.2. ICD Implantation and Follow-up
Indications for the ICD implantation were based on the contemporary guidelines and consisted of primary and secondary prevention (10). The implantation procedure was performed in the Electrophysiology Laboratory of Tehran Heart Center by five electrophysiologists who were experienced in device therapy. All the patients received multi-programmable ICDs with pacing capabilities and automatic intracardiac electrogram storage for the events triggering device therapy. The ICD system manufacturers included Medtronic, St. Jude, and Bostone Scientific. Lead implantation was performed via the transvenous approach through subclavian or axillary access using non-thoracotomy lead systems. In case of failure in transvenous insertion, the Left Ventricular (LV) lead was placed in the epicardial position in the operating room. The efficacy of pacing, sensing, and defibrillation was assessed after positioning the ICD lead. Moreover, defibrillation testing was performed using 10 - 15 joules of energy initially. If the arrhythmia failed to respond in the first step, the second step was to utilize 20 - 25 joule shocks and then with maximum energy as the last resort. The ICD evaluations during the follow-up visits involved routine clinically appropriate measurements, including interrogation of the device for tachyarrhythmia episodes, evaluation of sensing and pacing thresholds, lead impedance, and battery status. Data on the ICD-related complications were reviewed from the patients’ records.
3.3. Classification of Complications
In this study, the ICD-related complications were divided into two main categories: the complications related to the generator and those resulting from lead problems. The generator-related complication was generator malfunction. On the other hand, the lead-related complications included the following situations: dislodgement defined as X-ray-confirmed dislodgement of the lead combined with significant changes in sensing / pacing performance, failure to capture at practical device output with no visible change in the lead position or considerable impedance rise, loose set screw at the ICD connector, lead insulation defect, and lead fracture defined as changes in impedance with changes in sensing / pacing performance (intermittent or permanent) which could be optionally confirmed by X-ray study. Suboptimal and abnormal findings in the device analysis were defined as threshold > 2 V, Right Atrial (RA) lead impedance > 1500 Ω, Right Ventricular (RV) lead impedance > 2000 Ω, LV lead impedance > 1500 Ω, RV coil > 200 Ω, SVC coil > 200 Ω, R-wave < 5 mV, and p-wave < 0.5 mV.
3.4. Statistical Analysis
The statistical analyses were performed using the PASW 18.0 software package for Windows (SPSS, Inc., Chicago, IL, USA). The continuous variables were expressed as mean ± 1 Standard Deviation (SD) and were compared using Student’s t-test. Additionally, comparison of the dichotomous categorical variables and comparisons between the patients with and without complications were made using chi-square test. P values less than 0.05 were considered as statistically significant.
4. Results
In this study, the follow-up data of 255 consecutive patients (mean age = 62.57 ± 13.50 years; male = 196 [76.9%]) who underwent ICD implantation at Tehran Heart Center were reviewed. General characteristics of the patients have been summarized in Table 1. The indication of device implantation in almost half of the patients (54.1%) was secondary prevention. There were 8 patients (3.1%) in this study who died during the follow-up period. The characteristics of the implantable devices, including their type and manufacturer, are depicted in Table 2.
Table 1. General Characteristics of the Study Population .
| Variable | Study Population (n = 255) |
|---|---|
| Age, year (mean ± SD) | 62.57 ± 13.50 |
| Male gender, n (%) | 196 (76.9) |
| CAD, n (%) | 173 (67.8) |
| Diabetes mellitus, n (%) | 70 (27.5) |
| Hypertension, n (%) | 83 (32.5) |
| Hyperlipidemia, n (%) | 75 (29.4) |
| Smoking, n (%) | 58 (22.7) |
| CRF, n (%) | 90 (35.3) |
| EF, % (mean ± SD) | 29.55 ± 11.91 |
| EF < 35%, n (%) | 169 (66.2) |
Abbreviations: CAD, Coronary artery disease; CRF, Chronic renal failure; EF, Ejection fraction; SD, Standard deviation
Table 2. Characteristics of the ICDs in the Study Population .
| Variable | Study Population (n = 255) |
|---|---|
| ICD type, n (%) | |
| Single chamber device | 65 (25.4) |
| Dual chamber device | 110 (43.1) |
| Biventricular device | 80 (31.3) |
| ICD manufacturer, n (%) | |
| Medtronic | 140 (54.9) |
| St. Jude | 109 (42.7) |
| Bostone | 6 (2.4) |
| RA leads (n = 190) | |
| Capsure | 110 (57.8) |
| Tendril | 66 (34.7) |
| Isoflex | 11 (5.7) |
| Flextend | 3 (1.5) |
| RV leads (n = 255) | |
| Endotalk | 2 (0.7) |
| Sprint Quattero | 101 (39.6) |
| Riata | 79 (30.9) |
| Durata | 32 (12.5) |
| Sprint Fidelis | 41 (16.0) |
| LV leads (n=80) | |
| Attain | 47 (58.7) |
| QuickSite | 24 (30.0) |
| Epicardial | 5 (6.2) |
| QuickFlex | 4 (5.0) |
Abbreviations: ICD, Implantable cardioverter defibrillator; LV, Left ventricle; RA, Right atrium; RV, Right ventricle
Out of a total of 525 implanted leads and 255 implanted devices, complications leading to generator or lead replacement occurred in 32 patients (12.5%). The results showed no significant difference between the patients with and without complications regarding gender and age (P = 0.206 and P = 0.824, respectively). Also, no significant difference was found between the two groups concerning the ejection fraction (P = 0.271).
During the follow-up period, generator-related complications leading to generator replacement were observed in 2 patients (0.7%). One case was due to high-voltage circuit damage and the other one had generator software corruption following Magnetic Resonance Imaging (MRI). It should be noted that the type of the generator was not associated with the complications.
Lead-related complications included skin erosion due to lead, lead displacements; lead fractures, insulation defects, and loose set screw (Table 3). Among these complications, lead fracture was the most frequent one and was observed in 17 patients (6.6%). Besides, it was mainly observed in the RV leads. Lead fracture was more frequently observed with Sprint Fidelis® leads; albeit, it was not statistically significant (P = 0.287) (7 leads out of 40 Sprint Fidelis® leads were complicated). Riata® leads were the second most fractured leads with an overall number of 5 fractures. Only 30 patients (11.7%) underwent lead reimplantation and the rest were treated by changing the device mode or omitting the lead from the circuit. No cardiac perforation was recorded in the study population. Details of the findings in the device analysis are described in Table 4.
Table 3. Lead-Related Complications .
| Complication | N (%) |
|---|---|
| RA lead (n = 190) | |
| Skin erosion due to lead | 2 (1.0) |
| Lead fracture | 2 (1.0) |
| Displacement | 3 (1.5) |
| Total | 7 (3.6) |
| RV lead (n = 255) | |
| Skin erosion due to lead | 2 (0.7) |
| Lead fracture | 14 (5.4) |
| Displacement | 4 (1.5) |
| Insulation break | 1 (0.3) |
| Decreased R-wave amplitude | 1 (0.3) |
| Total | 22 (8.6) |
| LV lead (n = 80) | |
| Loose set screw | 1 (1.2) |
| Lead fracture | 1 (1.2) |
| Displacement | 1 (1.2) |
| Total | 3 (3.7) |
Abbreviations: LV, Left ventricle; RA, Right atrium; RV, Right ventricle
Table 4. Suboptimal and Abnormal Findings in the Device Analysis (Abnormal Findings Are Marked with Asterix) .
| Parameters | N (%) |
|---|---|
| RA-lead related (n = 190) | |
| RA threshold > 2 V | 5 (2.6) |
| RA impedance > 1500 Ω | 2 (1.0) |
| P-wave < 0.5 mV | 7 (3.6) |
| RV-lead related a (n = 255) | |
| RV threshold > 2 V | 20 (7.8) |
| RV impedance >2000 Ω | 10 (3.9) |
| R-wave < 3 mV | 29 (11.3) |
| RV coil > 200 Ω | 2 (0.7) |
| SVC coil > 200 Ω | 2 (0.7) |
| LV-lead related (n = 80) | |
| LV threshold > 2 V | 19 (23.7) |
| LV impedance > 2000 Ω | 4 (5.0) |
Abbreviations: LV, Left ventricle; RA, Right atrium; RV, Right ventricle; SVC, Superior vena cava
5. Discussion
This study showed that lead- or generator- related complications in almost 12.5% of an unselected population of the patients receiving ICD resulted in lead and device replacement. The risk of these complications was not influenced by any patient-specific factor, such as age and gender, or device-specific factors. Our findings are in line with those of other studies that have reported an almost similar risk of complications following ICD implantation (4).
Despite the growing number of the ICD implantations, modern ICDs have fewer complications compared to the first generations (11). Previous studies have reported a wide variety of complications after ICD implantation. They were not, however, consistent with respect to the rate of the events related to the lead or the generator.
The rate of generator-related complications in the present study was very low. It should be noted that we only described technical issues and other complications leading to device replacement, such as infection, were excluded from our analysis. The current evidence shows that the risk of generator-related complications is higher for ICD than that for the pacemaker (12). Prolonged charging periods and early battery depletion are the most frequent reported complications of the ICD generator, while prophylactic replacement after the manufacturer recall accounts for the majority of the cases with ICD replacement (4, 13, 14). The limited number of generator-related complications in the present study precluded the identification of any specific risk factor.
In the present study, most of the device replacement procedures were performed due to lead problems, which is in line with the previous studies (15, 16). Moreover, the principal reason for lead replacement was microfracture. However, it has been reported previously that lead failures mostly involved dislodgements (15-17). RV lead complications were observed more frequently and, therefore, comprised more cases of device replacement. This finding is consistent with those of other studies that discussed lead problems and may be related to the higher number of RV leads in this study (15-17). The prevalence of lead complications may imply the fact that leads are more susceptible to damage and injury and special consideration is required during lead insertion and fixation. Although age and gender have been reported as risk factors for defibrillation lead defects, their influence was not significant in the current study (18, 19). It should be noted that not every patient with abnormal or suboptimal analysis needed device reimplantation. Therefore, based on the electrophysiologist’s opinion, some patients were treated by reprogramming the device or omitting the malfunctioning lead.
Manufacturer and technical parameters of the leads have been shown to influence their function and durability (17, 20, 21). In this study, a higher rate of lead fracture was observed with the Sprint Fidelis® leads; albeit, it was not statistically significant (7 out of 17 cases, 41.1%). Sprint Fidelis® is a family of small-diameter leads released by Medtronic (Minneapolis, MN, USA) in 2004 with a 6.6 Fr lead body. In 2007, Medtronic Inc. announced withdrawal of the Sprint Fidelis® leads from the market due to malfunction and performance below the company’s standards (22). Since that time, studies have shown a higher rate of complications with Sprint Fidelis® leads (23-25). Despite these observations, the current data suggest that the decision to perform prophylactic lead replacement must include a weighing of the potential risks of lead revision against the risks of lead failure (25).
Our study provided a good overview of the generator and lead problems in the patients who received ICD within a 5-year period. Nevertheless, the study had a few limitations. First, we only included the patients at Tehran Heart Center and, consequently, our results may not be generalizable. Second, there is a minimal probability that a small number of the patients had also been admitted in other centers due to complications and their records were not available for this analysis. Finally, we only included hardware-related complications of the leads and generators; thus, complications, such as infection, were not within the scope of this work.
The rate of the ICD complications leading to device replacement and surgical revision is still clinically important in spite of the recent remarkable improvements in ICD technology. A large number of our patients had suboptimal function in the device interrogation and are, thus, probably prone to device failure in future. These findings are consistent with the previously reported rates of generator- and lead-complications. Further data on risk factors for generator- and lead-related complications and longer follow-up studies will be helpful in recognizing device complications with specific focus on the manufacturers, modifying the treatment guidelines, and improving the device technology.
Acknowledgments
The present study was extracted from the doctorate thesis written by Dr. Nader Soofizadeh for the degree of Cardiology subspecialty and was financed and supported by Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran.
Footnotes
Implication for health policy/ practice/ research/ medical education:This manuscript gives a detailed overview of complications related to ICDs in a 5 year period. Unlike previous studies, we focused on the leads and generator related complications rather than procedural complications. Therefore, we suppose that this study can provide good evidence not only for electrophysiologists but also all cardiologists.
Authors’ Contribution:Nader Soofizadeh: Data collection, Akbar Shafiee: Drafting, Ahmad Yaminisharif, Ali Kazemisaeid, Ali Vasheghani-Farahani: Case management, Supervision and final approval, Arash Jalali: Data analysis.
Financial Disclosure:There is no financial disclosure.
Funding/Support:This study was financed and supported by Tehran Heart Center, Tehran University of Medical Sciences, Tehran, Iran.
References
- 1.Connolly SJ, Gent M, Roberts RS, Dorian P, Roy D, Sheldon RS, et al. Canadian Implantable Defibrillator Study (CIDS) : A Randomized Trial of the Implantable Cardioverter Defibrillator Against Amiodarone. Circulation. 2000;101(11):1297–302. doi: 10.1161/01.cir.101.11.1297. [DOI] [PubMed] [Google Scholar]
- 2.Chao CC, Hu S, Shark KB, Sheng WS, Gekker G, Peterson PK. Activation of mu opioid receptors inhibits microglial cell chemotaxis. J Pharmacol Exp Ther. 1997;281(2):998–1004. [PubMed] [Google Scholar]
- 3.Poorbahador R, Shad B, Oraii S, Eftekharzadeh M, Mirmasoumi M, Sharif AG, et al. The National Survey of Cardiac Pacemakers and Cardioverter Defibrillators. The Journal of Tehran University Heart Center. 2006;1(2) [Google Scholar]
- 4.Alter P, Waldhans S, Plachta E, Moosdorf R, Grimm W. Complications of Implantable Cardioverter Defibrillator Therapy in 440 Consecutive Patients. Pacing and Clinical Electrophysiology. 2005;28(9):926–32. doi: 10.1111/j.1540-8159.2005.00195.x. [DOI] [PubMed] [Google Scholar]
- 5.Reynolds MR, Cohen DJ, Kugelmass AD, Brown PP, Becker ER, Culler SD, et al. The Frequency and Incremental Cost of Major Complications Among Medicare Beneficiaries Receiving Implantable Cardioverter-Defibrillators. Journal of the American College of Cardiology. 2006;47(12):2493–7. doi: 10.1016/j.jacc.2006.02.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Borek PP, Wilkoff BL. Pacemaker and ICD leads: Strategies for long-term management. Journal of Interventional Cardiac Electrophysiology. 2008;23(1):59–72. doi: 10.1007/s10840-008-9249-7. [DOI] [PubMed] [Google Scholar]
- 7.Dorwarth UWE, Frey B, Dugas M, Matis T, Fiek M, Schmoeckel M, et al. Transvenous Defibrillation Leads: High Incidence of Failure During Long-Term Follow-Up. Journal of Cardiovascular Electrophysiology. 2003;14(1):38–43. doi: 10.1046/j.1540-8167.2003.02305.x. [DOI] [PubMed] [Google Scholar]
- 8.Vollmann D. Patient AlertTM to detect ICD lead failure: efficacy, limitations, and implications for future algorithms. Europace. 2006;8(5):371–6. doi: 10.1093/europace/eul023. [DOI] [PubMed] [Google Scholar]
- 9.Vollmann D, Lüthje L, Vonhof S, Unterberg C. Inappropriate therapy and fatal proarrhythmia by an implantable cardioverter-defibrillator. Heart Rhythm. 2005;2(3):307–9. doi: 10.1016/j.hrthm.2004.11.019. [DOI] [PubMed] [Google Scholar]
- 10.Epstein AE, DiMarco JP, Ellenbogen KA, Estes NA, Freedman RA, Gettes LS, et al. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices) developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. J Am Coll Cardiol. 2008;51(21):e1–62. doi: 10.1016/j.jacc.2008.02.032. [DOI] [PubMed] [Google Scholar]
- 11.Grimm W, Menz V, Hoffmann J, Timmann UWE, Funck R, Moosdorf R, et al. Complications of Third-Generation Implantable Cardioverter Defibrillator Therapy. Pacing and Clinical Electrophysiology. 1999;22(1):206–11. doi: 10.1111/j.1540-8159.1999.tb00334.x. [DOI] [PubMed] [Google Scholar]
- 12.Poole JE, Gleva MJ, Mela T, Chung MK, Uslan DZ, Borge R, et al. Complication Rates Associated With Pacemaker or Implantable Cardioverter-Defibrillator Generator Replacements and Upgrade Procedures: Results From the REPLACE Registry. Circulation. 2010;122(16):1553–61. doi: 10.1161/circulationaha.110.976076. [DOI] [PubMed] [Google Scholar]
- 13.Bhatia A. The implantable cardioverter defibrillator: technology, indications, and impact on cardiovascular survival*1. Current Problems in Cardiology. 2004;29(6):303–56. doi: 10.1016/j.cpcardiol.2004.02.003. [DOI] [PubMed] [Google Scholar]
- 14.Gould PA. Complications Associated With Implantable Cardioverter-Defibrillator Replacement in Response to Device Advisories. Jama. 2006;295(16):1907. doi: 10.1001/jama.295.16.1907. [DOI] [PubMed] [Google Scholar]
- 15.Duray GZ, Schmitt J, Cicek-Hartvig S, Hohnloser SH, Israel CW. Complications leading to surgical revision in implantable cardioverter defibrillator patients: comparison of patients with single-chamber, dual-chamber, and biventricular devices. Europace. 2009;11(3):297–302. doi: 10.1093/europace/eun322. [DOI] [PubMed] [Google Scholar]
- 16.Maisel WH, Kramer DB. Implantable Cardioverter-Defibrillator Lead Performance. Circulation. 2008;117(21):2721–3. doi: 10.1161/circulationaha.108.776807. [DOI] [PubMed] [Google Scholar]
- 17.Gradaus R, Breithardt G, Bocker D. ICD Leads:. Design and Chronic Dysfunctions. Pacing and Clinical Electrophysiology. 2003;26(2p1):649–57. doi: 10.1046/j.1460-9592.2003.00112.x. [DOI] [PubMed] [Google Scholar]
- 18.Grimm W, Stula A, Sharkova J, Alter P, Maisch B. Outcomes of Elderly Recipients of Implantable Cardioverter Defibrillators. Pacing and Clinical Electrophysiology. 2007;30(s1) doi: 10.1111/j.1540-8159.2007.00623.x. [DOI] [PubMed] [Google Scholar]
- 19.Kleemann T, Becker T, Doenges K, Vater M, Senges J, Schneider S, et al. Annual Rate of Transvenous Defibrillation Lead Defects in Implantable Cardioverter-Defibrillators Over a Period of >10 Years. Circulation. 2007;115(19):2474–80. doi: 10.1161/circulationaha.106.663807. [DOI] [PubMed] [Google Scholar]
- 20.Hauser RG, Kallinen LM, Almquist AK, Gornick CC, Katsiyiannis WT. Early failure of a small-diameter high-voltage implantable cardioverter-defibrillator lead. Heart Rhythm. 2007;4(7):892–6. doi: 10.1016/j.hrthm.2007.03.041. [DOI] [PubMed] [Google Scholar]
- 21.Lee DS, Krahn AD, Healey JS, Birnie D, Crystal E, Dorian P, et al. Evaluation of Early Complications Related to De Novo Cardioverter Defibrillator Implantation. Journal of the American College of Cardiology. 2010;55(8):774–82. doi: 10.1016/j.jacc.2009.11.029. [DOI] [PubMed] [Google Scholar]
- 22.Goette A, Cantu F, van Erven L, Geelen P, Halimi F, Merino JL, et al. Performance and survival of transvenous defibrillation leads: need for a European data registry. Europace. 2008;11(1):31–4. doi: 10.1093/europace/eun301. [DOI] [PubMed] [Google Scholar]
- 23.Farwell D, Redpath C, Birnie D, Gollob M, Lemery R, Posan E, et al. Inappropriate implantable cardioverter defibrillator shocks in fractured Sprint Fidelis leads associated with 'appropriate' interrogation. Europace. 2008;10(6):726–8. doi: 10.1093/europace/eun113. [DOI] [PubMed] [Google Scholar]
- 24.Hauser RG, Hayes DL. Increasing hazard of Sprint Fidelis implantable cardioverter-defibrillator lead failure. Heart Rhythm. 2009;6(5):605–10. doi: 10.1016/j.hrthm.2009.02.024. [DOI] [PubMed] [Google Scholar]
- 25.Parkash R, Crystal E, Bashir J, Simpson C, Birnie D, Sterns L, et al. Complications Associated With Revision of Sprint Fidelis Leads: Report From the Canadian Heart Rhythm Society Device Advisory Committee. Circulation. 2010;121(22):2384–7. doi: 10.1161/circulationaha.109.924357. [DOI] [PubMed] [Google Scholar]