N M van Hemel

University Medical Center Utrecht, Utrecht, Utrecht, Netherlands

Are you N M van Hemel?

Claim your profile

Publications (195)643.08 Total impact

  • Source
    N M van Hemel
    09/2014;
  • Source
    N M van Hemel
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 02/2014; · 1.41 Impact Factor
  • Source
    N M van Hemel
    [Show abstract] [Hide abstract]
    ABSTRACT: Literature about pacemaker (PM) implantations shows that several clinical and technical factors determine the short- and long-term complications after the intervention. Annual hospital volume, however, does not negatively affect complications in contrast with the cumulative experience of the operator. In view of this observation, the current required number of 20 to 30 first PM implantations for cardiology training does not match standards for quality of care. In addition, concentration of implants and replacement of pacemakers to a limited number of operators per hospital to comply with the increasing demands of patients and other parties has to be seriously considered.
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 12/2013; · 1.41 Impact Factor
  • Source
    N M van Hemel, J M T de Bakker
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 11/2013; · 1.41 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Pacing technology and care have improved notably over the past decade, justifying an update on the long-term prognosis and pre-implantation determinants of prognosis of bradycardia pacemaker (PM) recipients. Prospective cohort study. 23 Dutch pacemaker centres PATIENTS: Pre-implantation characteristics of 1517 patients receiving a first bradycardia PM between 2003-2007 were studied in relation to survival. None; patients were followed up during routine clinical practice. Cause and time to death. At the end of a mean follow-up of 5.8 (SD 1.1) years, 512 patients (33%) died, mostly of non-cardiac cause (67%). There were two PM related deaths. Survival rates were 93%, 81%, 69%, and 61% after 1, 3, 5 and 7 years, respectively. PM recipients without concomitant cardiovascular disease at implantation showed survival rates comparable to age and sex matched controls. Predictors at time of implantation associated with cardiac mortality were: age, coronary artery disease (CAD), diabetes, heart failure, valve disease, and the indication for PM implantation. Predictors for all cause mortality were: male gender, age, body mass index, CAD, cardiac surgery, diabetes, heart failure, and maintained atrioventricular synchrony. A pre-implantation history of heart failure, CAD, and diabetes are the most important predictors for worse prognosis in PM recipients. Without baseline heart disease, survival rates equal that of the general population, suggesting that the prognosis of contemporary PM recipients is mainly determined by comorbid diseases and a bradycardia pacing indication as such does not influence survival. ClinicalTrials.gov Identifier: NCT00135174.
    Heart (British Cardiac Society) 08/2013; · 5.01 Impact Factor
  • Norbert M van Hemel
    Europace 03/2013; · 2.77 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: BACKGROUND: Health-related quality of life (HRQoL) values shortly after pacemaker (PM) implantation for bradycardia have been established, however little is known about long-term HRQoL. METHODS: Using the generic SF-36 and the PM specific Aquarel questionnaire, HRQoL was repeatedly measured during a 7.5year follow-up period in 881 bradycardia PM recipients included in the large scale nationwide Dutch FOLLOWPACE study. HRQoL over time, corrected for age, gender, diabetes, hypertension, heart failure, cardiovascular disease and AV-synchrony, was assessed with a linear mixed model. RESULTS: Increased scores both on overall SF-36 and on all SF-36 subscales were observed shortly after implantation. Although scores on SF-36 gradually declined over time, scores remained improved over the measured pre-implantation values. Also, scores for almost all subscales remained increased throughout the 7.5year observation period, except for physical functioning which showed a gradual decline several years after the initial rise. Additionally, higher scores on all Aquarel scales were observed after implantation. Scores on the arrhythmias and chest discomfort subscales improved and remained stable throughout follow-up (FU), whereas the dyspnea at exertion subscale showed a gradual decline during FU to reach pre-implantation values at 5years. CONCLUSIONS: Increased HRQoL is observed not only shortly after PM implantation, but also after long-term FU. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT00135174; http://www.clinicaltrials.gov/ct2/show/NCT00135174.
    International journal of cardiology 02/2013; · 6.18 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: AIMS: Remote follow-up (FU) enables to cope with the expanding number of pacemaker (PM) FU. Although remote FU offers comparable monitoring options to in-office FU, reprogramming of device settings is not available, thereby imposing a potentially important restriction to the applicability of remote FU. The aim of this study was to assess in a large cohort of bradycardia PM recipients, the incidence of PM reprogramming during long-term FU and its predictors, to judge the possibilities for remote FU.METHODS AND RESULTS: Between 2003 and 2010 all in-office FU of 1517 bradycardia PM recipients included in the FOLLOWPACE study were recorded. Only 24.5% of all 13 258 recorded FU visits >3 months after implantation were visits-with-reprogramming (VWRs), occurring in 1158 patients (79%). Fifty percent of patients were free of reprogramming at 9 months, and 29% at 24 months. Using multivariable binary logistic regression analysis, the following patient characteristics were predictive for frequent PM reprogramming, defined as >3 VWRs during 3 year FU: age, a history of atrial arrhythmias, PM complication <3 months after implantation, congestive heart failure, PM indication, and lead fixation method. This model had a receiver operating characteristic area of 0.66 (95% confidence interval 0.61-0.71).CONCLUSION: This study observed a low proportion of VWR (∼25%) during a mean FU of 5.3 years; however, those patients at high risk for PM reprogramming cannot easily be predicted. The vast majority of patients (>80%) do not need frequent reprogramming, suggesting a potential benefit of using remote FU to reduce the number of unnecessary in-office visits.
    Europace 02/2013; · 2.77 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: AIMS: Guidelines regarding pacemaker (PM) follow-up (FU) are not precisely defined. The study aim is to describe long-term routine in-hospital FU, evaluate compliance to guidelines, and assess the portion of visits-with-an-action (VWA).METHODS AND RESULTS: The multicentre prospective FOLLOWPACE study collected data in the period 2003-2010, regarding FU of 1517 patients with a first PM for bradycardia indications in 23 Dutch hospitals. A total of 15 472 visits were analysed with a median FU of 4.9 years, adding up to 6750 patient years. The median time to the first three visits was 35, 127, and 303 days, respectively. Thereafter the median interval between visits was 180 days. Most patients had 2 FU/year, 22% had 1 FU/year, and 18% had >3 FU/year. Seventy-three percent of patients with single-chamber PMs had at least 1 FU/year, whereas 36% of patients with dual-chamber PMs received at least 2 FU/year. During the first year, 52% of visits were VWA, as opposed to 17% after 6 years. Battery status was assessed in 98%, and stimulation and sensing thresholds in 90% and 77% of visits, respectively. Reprogramming markedly declined from ∼60% in the first FU visit to 10-20% after the fifth visit.CONCLUSION: Although the measurements during PM FU are according to guidelines, the frequency of FU is not. Moreover, in the vast majority of performed FU, PM programming is left unchanged. This suggests that a large portion of FU visits is redundant, such that their frequency after the first year can be diminished, or replaced by remote FU.
    Europace 10/2012; · 2.77 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The prognostic value of normal dual-isotope single photon emission computed tomography (SPECT), with technetium 99m-tetrofosmin for the stress images, is not well known. Furthermore, thallium-201 cross talk in the Tc-99m window may reduce the defect severity of the Tc-99m images. In a consecutive series of 610 patients, 246 patients with normal rest and stress SPECT images by means of visual semiquantitative analysis were included. The patients' pre-test likelihood of coronary artery disease was determined, based on age, sex, symptoms, and the results of stress electrocardiography. Quality of life was assessed by using a mailed self-administered general quality-of-life questionnaire. We compared the mean scores (8 dimensions) of our study population with the scores of 1063 control subjects, sampled randomly out of the inhabitant register of a Dutch city. The mean follow-up period was 25+/-3 months. Compared with the control group, patients in our study scored lower (less perceived health) for all dimensions of the SF-36 (P<.05), suggesting a selection of symptomatic or otherwise diseased patients. The primary cardiac event rate was 0.4% per year. The cardiac events occurred in patients with an intermediate-to-high pre-test likelihood of disease and negative or nondiagnostic exercise electrocardiographic results. In a subset of patients with a high pre-test likelihood of coronary artery disease (more than 85%), the primary cardiac event rate was 0.7% per year. Patients with normal results on dual-isotope myocardial perfusion scintigraphy, performed with Tc-99m-tetrofosmin for the stress images, have an excellent prognosis. Furthermore, our results suggest that Tl-201 cross talk in the Tc-99m window may be low and functionally and clinically unimportant.
    Journal of Nuclear Cardiology 04/2012; 7(4):333-41. · 2.85 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Remote monitoring of cardiac implanted electronic devices (CIED: pacemaker, cardiac resynchronisation therapy device and implantable cardioverter defibrillator) has been developed for technical control and follow-up using transtelephonic data transmission. In addition, automatic or patient-triggered alerts are sent to the cardiologist or allied professional who can respond if necessary with various interventions. The advantage of remote monitoring appears obvious in impending CIED failures and suspected symptoms but is less likely in routine follow-up of CIED. For this follow-up the indications, quality of care, cost-effectiveneness and patient satisfaction have to be determined before remote CIED monitoring can be applied in daily practice. Nevertheless remote CIED monitoring is expanding rapidly in the Netherlands without professional agreements about methodology, responsibilities of all the parties involved and that of the device patient, and reimbursement. The purpose of this consensus document on remote CIED monitoring and follow-up is to lay the base for a nationwide, uniform implementation in the Netherlands. This report describes the technical communication, current indications, benefits and limitations of remote CIED monitoring and follow-up, the role of the patient and device manufacturer, and costs and reimbursement. The view of cardiology experts and of other disciplines in conjunction with literature was incorporated in a preliminary series of recommendations. In addition, an overview of the questions related to remote CIED monitoring that need to be answered is given. This consensus document can be used for future guidelines for the Dutch profession.
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 02/2012; 20(2):53-65. · 1.41 Impact Factor
  • Source
    C C de Cock, J Elders, N M van Hemel
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 02/2012; 20(2):51-2. · 1.41 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Today quantitative information about the type of complications and their incidence during long-term pacemaker (PM) follow-up is scarce. To assess the incidence and determinants of short- and long-term complications after first pacemaker implantation for bradycardia. A prospective multicenter cohort study (the FOLLOWPACE study) was conducted among 1517 patients receiving a PM between January 2003 and November 2007. The independent association of patient and implantation-procedure characteristics with the incidence of PM complications was analyzed using multivariable Cox regression analysis. A total of 1517 patients in 23 Dutch PM centers were followed for a mean of 5.8 years (SD 1.1), resulting in 8797 patient-years. Within 2 months, 188 (12.4%) patients developed PM complications. Male gender, age at implantation, body mass index, a history of cerebrovascular accident, congestive heart failure, use of anticoagulant drugs, and passive atrial lead fixation were independent predictors for complications within 2 months, yielding a C-index of 0.62 (95% confidence interval 0.57-0.66). Annual hospital implanting volume did not additionally contribute to the prediction of short-term complications. Thereafter, 140 (9.2%) patients experienced complications, mostly lead-related complications (n = 84). Independent predictors for long-term complications were age, body mass index, hypertension, and a dual-chamber device, yielding a C-index of 0.62 (95% confidence interval 0.57-0.67). The occurrence of a short-term PM complication was not predictive of future PM complications. Complication incidence in modern pacing therapy is still substantial. Most complications occur early after PM implantation. Although various patient- and procedure-related characteristics are independent predictors for early and late complications, their ability to identify the patient at high risk is rather poor. This relatively high incidence of PM complications and their poor prediction underscores the usefulness of current guidelines for regular follow-up of patients with PM.
    Heart rhythm: the official journal of the Heart Rhythm Society 12/2011; 9(5):728-35. · 4.56 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The number of patients >80 years receiving pacemakers (PMs) is increasing. Little is known about survival and complications in this specific subgroup. We aim to determine predictors of long-term survival. Pacemaker-related complications and death occurring in patients receiving a first PM for conventional bradycardia indications were systematically documented (the FollowPace registry). This report describes 481 patients ≥80 years during a mean follow-up of 5.8 (SD 1.2) years. Within 2 months 54 PM complications occurred in 47 patients (9.8%). During follow-up, 35 adverse PM events were reported in 33 patients (6.9%). Complication rates in patients ≥80 years were comparable with those for patients <80 years. Survival rates were 86, 75, and 49% after 1, 2, and 5 years, respectively, and were comparable with survival for age- and sex-matched controls from the general Dutch population. Most patients died of non-cardiac causes. Age at the time of implantation, male gender, the presence of congestive heart failure, coronary pathology, and diabetes mellitus were independent predictors of all-cause mortality. This large study of long-term cardiac pacing for bradycardia in octogenarians and nonagenarians showed a cumulative 5-year survival of ~50%, which compares with that of age- and sex-matched controls. These data suggest a beneficial impact of bradycardia pacing, restoring life expectancy to previous levels. The occurrence of PM complications during long-term follow-up is not infrequent with 18.1% of patients experiencing a PM-related complication during a mean of 5.8 years follow-up. The complication rate was not higher than in younger PM patients.
    Europace 10/2011; 14(4):502-8. · 2.77 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Atrial fibrosis is related to atrial fibrillation but may differ in patients with mitral valve disease or lone atrial fibrillation. Therefore, we studied atrial fibrosis in patients with atrial fibrillation+mitral valve disease or with lone atrial fibrillation and compared it with controls. Left and right atrial appendages amputated during Maze III surgery for lone atrial fibrillation (n=85) or atrial fibrillation+mitral valve disease (n=26) were embedded in paraffin, sectioned, and stained with picrosirius red. Atria from 10 deceased patients without a cardiovascular history served as controls. A total of 1048 images (4-μm sections, 10-fold magnification, 4 images per appendage) were obtained and digitized. The percentage of fibrous tissue was calculated by quantitative morphometry. Irrespective of the presence or absence of atrial fibrillation or mitral valve disease, more fibrous tissue was present in right atrial appendages than in left atrial appendages (12.7%±5.7% vs 8.2%±3.9%; P<.0001). The mean amount of fibrous tissue in the atria was significantly larger in patients with atrial fibrillation+mitral valve disease than in patients with lone AF and controls (13.6%±5.8%, 9.7%±3.2%, and 8.8%±2.4%, respectively; P<.01). No significant differences existed between patients with lone atrial fibrillation and patients without a cardiovascular history (controls). Atria of patients with atrial fibrillation and mitral valve disease have more fibrosis than atria of patients with lone atrial fibrillation. However, patients with lone atrial fibrillation have an equal amount of atrial fibrosis compared with controls. These findings support the notion that fibrosis plays a more important role in the pathogenesis of atrial fibrillation secondary to mitral valve disease than in lone atrial fibrillation and potentially explains the relatively poor success of antiarrhythmic surgery in patients with mitral valve disease.
    The Journal of thoracic and cardiovascular surgery 10/2011; 144(2):327-33. · 3.41 Impact Factor
  • N M van Hemel
    Europace 08/2011; 13(12):1667-8. · 2.77 Impact Factor
  • Vera H M Deneer, Norbert M van Hemel
    [Show abstract] [Hide abstract]
    ABSTRACT: Aging is associated with electrical and structural changes of the myocardium. The response to catecholamines is also reduced and the baroreceptor reflex activity is blunted. These aspects conceivably affect the response to antiarrhythmic drugs in the elderly. Furthermore, physiological parameters change in older age, affecting the pharmacokinetics of drugs. In this article, the literature on the pharmacokinetics and pharmacodynamics of antiarrhythmic drugs in elderly subjects is reviewed with the purpose of improving their optimal and safe prescription. Pharmacokinetic studies of antiarrhythmic drugs in the elderly are sparse, and there are no data available for procainamide and propafenone. Mean dose reductions calculated for elderly patients relative to younger patients are 60% for digoxin, 19% for diltiazem, 32% for disopyramide, 31% for flecainide, 40% for metoprolol, 35% for quinidine, 29% for sotalol and 26% for verapamil. No dose reductions are required for dofetilide or dronedarone. The clearance of dofetilide is not affected by age after correction for renal function. The dosage of dofetilide is individualized according to an algorithm based on the corrected QT (QTc) interval and renal function. Although the area under the plasma concentration-time curve (AUC) for dronedarone is larger in elderly patients, the dose should not be reduced because the registered dose has specifically been studied in an elderly population. In elderly patients with renal insufficiency, hepatic impairment, heart failure or certain genetic variants, the pharmacokinetics of antiarrhythmic drugs might be affected to an even greater extent, meaning additional dosage adjustments are necessary. With increasing age, the number of prescribed drugs increases because of co-morbidity, making interactions between drugs more likely. Several drugs interact with antiarrhythmic drugs, leading to clinically relevant changes in drug concentrations or AUC values. Furthermore, several drugs with non-cardiovascular indications appear to have QTc prolonging effects. The combination of these drugs with antiarrhythmic drugs that affect the QTc interval increases the risk of developing torsades de pointes and should therefore be avoided. Altered effects of drugs in the elderly can also be the result of age-related changes in the cardiovascular system. For example, atenolol and sotalol show greater effects, i.e. reductions in heart rate and increased probability of adverse effects, at a given plasma concentration in older subjects compared with younger subjects. It remains unclear whether old age as such is a determinant for reduced or modified efficacy of antiarrhythmic drugs. In a randomized study it was found that patients aged ≥65 years with atrial fibrillation had better survival with rate control than with rhythm control. However, different treatment strategies were compared and the results cannot be extrapolated to indicate better survival with a specific antiarrhythmic drug. Antiarrhythmic drugs will remain the first-line approach in most patients for the prevention or suppression of atrial and ventricular arrhythmias. As a rule of thumb, a 50% reduction in the starting dose of antiarrhythmic drugs compared with younger patients appears a wise approach in elderly patients. However, this does not apply to dofetilide and dronedarone. The selection of antiarrhythmic drugs in the elderly is predominantly determined by factors such as the treatment target, assumed patient compliance, possible drug interactions, co-morbidity, and renal and liver function. Efficacy and safety monitoring should take into account symptoms, ECG findings, rhythm recordings, plasma drug concentrations and other laboratory parameters.
    Drugs & Aging 08/2011; 28(8):617-33. · 2.50 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: INTRODUCTION: Myocardial perfusion SPECT (MPS) is frequently used for cardiovascular risk stratification. The significance of MPS in patients with abnormal electrical ventricular activation is often questionable. This review assesses the value of MPS for risk stratification of patients with intrinsic left bundle branch block or that due to right ventricular apical pacing. METHODS: We reviewed the literature by a search of the MEDLINE database (January 1980 to September 2010). The terms prognosis or prognostic value were combined with SPECT and LBBB or pacing or pacemakers. MPS was categorised as low and high risk according to the original definitions. RESULTS: We identified 11 studies suitable for review. A low-risk MPS is associated with a low risk of cardiac events whereas high-risk MPS carries a 4.8-fold increased risk, 95% CI [3.2 - 7.2] (p < 0.0001). Despite secondary prevention and an improved medical and interventional care, these figures have hardly changed over time. CONCLUSION AND CLINICAL IMPLICATIONS: A low-risk MPS permits a policy of watchful waiting whereas a high-risk MPS requires further analysis and treatment. The persistent high cardiac death and acute myocardial infarction rate after a high-risk MPS suggest that the current management of these patients does not suffice and needs reconsideration.
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 06/2011; · 1.41 Impact Factor
  • Source
    N M van Hemel
    Netherlands heart journal: monthly journal of the Netherlands Society of Cardiology and the Netherlands Heart Foundation 01/2011; 19(1):3-5. · 1.41 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: cardiac resynchronization therapy (CRT) may improve prognosis in patients with chronic right ventricular (RV) pacing, and optimal lead position can decrease nonresponders. We evaluated the clinical and echocardiographic response to CRT in patients with previous chronic RV pacing, using pressure-volume loop analyses to determine the optimal left ventricular (LV) lead position during implantation. In this single-blinded, randomized, controlled crossover study, 40 patients with chronic RV apical pacing and symptoms of heart failure, decreased LV ejection fraction (LVEF) or dyssynchrony were included. During implantation, stroke work (SW), LVEF, cardiac output, and LV dP/dt(max) were assessed by a conductance catheter. Clinical and echocardiographic response was studied during a 3-month period of RV pacing (RV period, LV lead inactive) and a 3-month period of biventricular pacing (CRT period). At the optimal LV lead position, SW (37 ± 41%), LVEF (16 ± 13%), cardiac output (29 ± 16%), and LV dP/dt(max) increased (11 ± 11%) significantly during biventricular pacing compared to baseline. Additional benefit could be achieved by pressure-volume loop guided selection of the best left-sided pacing location. RV outflow tract pacing did not improve hemodynamics. During follow-up, symptoms improved during CRT, VO(2,max) increased 10% and significant improvements in LVEF, LV volumes, and mitral regurgitation were observed as compared to the RV period. CRT in patients with chronic RV pacing causes significant improvement of both LV function as measured by pressure-volume loops during implantation and clinical and echocardiographic improvement during follow-up. Pressure-volume loops during implantation may facilitate selection of the most optimal pacing site.
    Journal of Cardiovascular Electrophysiology 12/2010; 22(6):677-83. · 3.48 Impact Factor

Publication Stats

2k Citations
643.08 Total Impact Points

Institutions

  • 1994–2014
    • University Medical Center Utrecht
      • • Department of Cardiology
      • • Julius Center for Health Sciences and Primary Care
      Utrecht, Utrecht, Netherlands
  • 2012
    • VU University Medical Center
      • Department of Cardiology
      Amsterdam, North Holland, Netherlands
  • 2007–2012
    • Utrecht University
      • Department of Cardiology
      Utrecht, Utrecht, Netherlands
  • 1986–2012
    • St. Antonius Ziekenhuis
      • • Department of Cardiology
      • • Department of Nuclear Medicine
      Nieuwegein, Provincie Utrecht, Netherlands
  • 2010
    • Maasstad Ziekenhuis
      Rotterdam, South Holland, Netherlands
    • Bronovo Hospital
      's-Gravenhage, South Holland, Netherlands
  • 2007–2010
    • Isala Klinieken
      • Department of Cardiology
      Zwolle, Provincie Overijssel, Netherlands
  • 1987–2009
    • Academisch Medisch Centrum Universiteit van Amsterdam
      • • Heart Failure Research Center
      • • Department of Experimental Cardiology
      • • Department of Cardiology and Cardio-thoracic Surgery
      Amsterdam, North Holland, Netherlands
  • 2008
    • St. Lucas Andreas Hospital
      Amsterdamo, North Holland, Netherlands
  • 2006–2008
    • Catharina Ziekenhuis
      • Department of Cardiology
      Eindhoven, North Brabant, Netherlands
  • 2005
    • Deborah Heart & Lung Center
      Philadelphia, New York, United States
    • Antonius Ziekenhuis
      Sneek, Friesland, Netherlands
  • 2004
    • Maxima Medical Center
      Veldhoven, North Brabant, Netherlands
  • 2000–2003
    • Leiden University Medical Centre
      • Department of Cardiology
      Leiden, South Holland, Netherlands
  • 1996–1999
    • London Health Sciences Centre
      • Department of Surgery
      London, Ontario, Canada
  • 1990–1993
    • University of Amsterdam
      • • Faculty of Medicine AMC
      • • Department of Clinical and Experimental Cardiology
      Amsterdam, North Holland, Netherlands
  • 1992
    • University of Groningen
      Groningen, Groningen, Netherlands
  • 1988–1990
    • Universitair Medisch Centrum Groningen
      • Department of Cardiology
      Groningen, Groningen, Netherlands