Risk Factors for Occlusion of Infrainguinal Bypass Grafts

Eur J Vasc Endovasc Surg 20, 118–124 (2000) doi:10.1053/ejvs.2000.1125, available online at http://www.idealibrary.com on

Risk Factors for Occlusion of Infrainguinal Bypass Grafts M. J. D. Tangelder†1, A. Algra2, J. A. Lawson3 and B. C. Eikelboom1 on behalf of the Dutch BOA Study Group∗ 1

Department of Vascular Surgery, University Hospital Utrecht, 2Julius Center for Patient Oriented Research and Department of Neurology, PO Box 85500, 3508 GA Utrecht, and 3Department of Surgery, Ziekenhuis Amstelveen, Amsterdam, The Netherlands Objectives: to identify risk factors for infrainguinal bypass occlusion and quantify the predictive value of data available before and after surgery. Design: prospective study of 2650 patients who participated in a randomised trial of oral anticoagulants or aspirin after infrainguinal bypass surgery. Materials and methods: risk factors were determined by univariate Cox regression analysis, and entered in multivariate analyses which distinguished two models: analysis of factors available from history and clinical examination, completed by radiological and surgical data in the second model. To compare the information content of the two models, receiver–operator characteristic (ROC) curves were computed. Results: in all patients female gender, critical ischaemia, femorocrural bypass grafting and non-venous graft material were independent risk factors. In patients with femoropopliteal bypasses female gender, critical ischaemia, poor run-off and non-venous graft material, the latter even in patients with supragenicular bypasses, were independent risk factors. The only significant risk factor in patients with femorocrural bypass grafts was use of a non-venous graft. The information contained in the first model was poor, whereas the second model had a higher predictive value. Conclusions: the major risk factor, even in above-knee bypasses, is non-venous graft material. The venous bypass graft should be offered to patients whenever possible. Key Words: Infrainguinal bypass; Peripheral bypass surgery; Femoropopliteal bypass; Femorocrural bypass; Graft occlusion; Risk factors; Graft material; Multivariate analysis.

Introduction In patients not treated with antithrombotics postoperatively, 28–45% of all bypass grafts will occlude in the first year after operation, most of them in an early stage.1–4 Failure of infrainguinal bypass grafts remains, despite expert surgeons and meticulous surgical technique, a matter of concern for both patient and surgeon. Based on patient history, symptoms, pathological and anatomical findings and technical aspects of surgery, we endeavour to assess the probability of successful bypass surgery for each patient. Few prospective studies have addressed the identification of specific vascular risk factors for infrainguinal graft occlusion.5–7 Evidently, little is known about the predictive value of variables generally considered as risk factors for graft occlusion. ∗ Participating investigators are listed in the Appendix. † Please address all correspondence to: M. J. D. Tangelder, BOA Trial Office, Bolognalaan 30, 3584 CJ Utrecht, The Netherlands. 1078–5884/00/080118+07 $35.00/0  2000 Harcourt Publishers Ltd.

The aim of this study, therefore, was to identify risk factors for graft occlusion and to quantify the predictive value of data available before and after surgery in a large cohort of patients who were recruited into a multicentre randomised trial of oral anticoagulants and aspirin after infrainguinal bypass grafting.

Materials and Methods Patients and baseline characteristics All patients studied were included in the Dutch BOA Study (Bypass, Oral anticoagulants or Aspirin Study), a multicentre randomised trial to compare the effectiveness of oral anticoagulants with that of aspirin in preventing occlusions of bypass grafts and other thrombotic events. Background, design and results of the trial have been reported elsewhere.8 The study was

Risk Factors for Occlusion of Infrainguinal Bypass Grafts

approved by the Ethics Committees of all participating hospitals and by the Dutch Health Insurance Council. All patients who were enrolled in the study gave written informed consent. Between April 1995 and March 1998, a total of 2650 patients were randomised. Demographic factors (age and gender), medical history (angina pectoris, myocardial infarction, transient ischaemic attack (TIA) or stroke, and vascular intervention), vascular risk factors (current smoking, hypertension, diabetes mellitus and hyperlipidaemia), indication for surgery (intermittent claudication or critical ischaemia, i.e. rest pain or tissue loss), brachial and ankle blood pressures, arterial run-off (Ζone versus two or three arteries), site of distal anastomosis (popliteal artery above or below knee versus crural or pedal artery), graft material used (autologous vein or non-venous, i.e. prosthetic, homogenous or composite graft material), and postoperative antithrombotic medication were all recorded at baseline. Vascular intervention comprised thrombolysis, PTA, or surgery in any arterial section. Hypertension, diabetes mellitus and hyperlipidaemia were present if a patient was treated for it. In 97% of the patients vein grafts consisted of the greater saphenous vein (63% reversed and 35% in situ), and in 3% of all patients other venous grafts were used. Inherent to the design of the trial, half of the patients were allocated to postoperative oral anticoagulant treatment (INR 3.0–4.5), whereas the other half of the patients received aspirin (80 mg daily).

Follow-up All patients were evaluated 3 and 6 months postoperatively and every 6 months thereafter. Graft patency was assessed by clinical examination and by Doppler or duplex scanning and arteriography when indicated.

Data analysis The object of analysis was to find the combination of baseline characteristics that most accurately predicted the occurrence of graft occlusion. After analysis of all patients, subgroup analysis of patients with femoropopliteal and femorocrural bypasses was performed. In the univariate analysis we calculated the hazard ratios with corresponding 95% confidence intervals (CI) of the considered risk factors by means of the Cox proportional hazards model.9

119

The aim of the multivariate analysis was to find independent predictors of graft occlusion. Variables with a significant predictive value (95% CI does not include 1) in the univariate analysis were sequentially entered in the multivariate analysis by a stepwise forward conditional approach. The default probability value for inclusion of possible independent risk factors in the model was 0.05. Factors were removed from the model if the probability value exceeded 0.10. The multivariate analysis distinguished two models. In the first model (the ‘‘office only’’ model), variables available from patient history and clinical examination only were analysed. In the second ‘‘complete’’ model these factors were completed by data from radiology and surgery. To compare the information content of the two models, receiver–operator characteristic (ROC) curves were computed. The more an ROC curve is located in the upper left corner of the graph, the higher the information content of the model; that is, the greater sensitivity and specificity of the model for predicting graft occlusion. Information content of the models is quantified by calculation of the areas under the curve. An area under the curve of 0.5 indicates no prognostic value at all, whereas an area under the curve of 1 means a sensitivity and specificity of the model of 100%.

Results The mean length of follow-up was 21 months (range, 0–45 months) during which 630 grafts occluded (23.8%); 469 (22.1%) in the femoropopliteal group (n= 2119) and 161 (30.3%) in the femorocrural group (n= 531).

Univariate analysis Higher ankle blood pressures were associated with a lower risk for graft occlusion, whereas female gender, critical ischaemia, femorocrural bypass and use of non-venous graft material were related to a higher probability of graft occlusion (Table 1). In the subgroup analysis of patients with femoropopliteal bypasses, poor (one or no patent) run-off artery appeared to be a risk factor as well, whereas ankle blood pressure just failed to reach statistical significance. Infragenicular distal anastomosis was not associated with graft occlusion. Analysis of the patients with femorocrural bypasses revealed use of non-venous graft material as the only risk factor for occlusion. Eur J Vasc Endovasc Surg Vol 20, August 2000

120

M. J. D. Tangelder et al.

Table 1. Results of univariate analysis of risk factors for infrainguinal graft occlusion. All patients, n=2650 % with HR variable or mean±s.d.

Variable Demographic factors Female gender Age Medical history Angina pectoris Myocardial infarction TIA/stroke Vascular intervention∗ Vascular risk factors Current smoking Hypertension‡ Diabetes mellitus‡ Hyperlipidaemia‡ Brachial blood pressure# Ankle blood pressure# Ankle–brachial index# Critical ischaemia Run-off ≤1 artery Below-knee anastomosis Femorocrural bypass Non-venous graft Aspirin vs. anticoagulants

95% CI

Femoropopliteal grafts, n=2119

Femorocrural grafts, n=531

% with variable or mean±s.d.

% with HR variable or mean±s.d.

HR

95% CI

95% CI

36 69±10

1.21 0.996

1.03–1.42 0.989–1.005

35 68±9.8

1.31 0.994

1.09–1.57 0.985–1.003

38 72±10

0.95 0.994

0.69–1.31 0.980–1.009

17 18 12 45

0.94 0.99 0.94 1.03

0.76–1.17 0.80–1.22 0.72–1.21 0.88–1.21

16 17 11 45

0.90 0.96 0.93 1.03

0.70–1.16 0.75–1.23 0.68–1.27 0.86–1.24

19 18 15 45

1.02 1.04 0.86 1.05

0.69–1.52 0.70–1.54 0.54–1.38 0.77–1.43

54 39 26 16 159±31

1.03 0.85 1.12 0.81 0.998

0.88–1.20 0.72–1.00 0.94–1.33 0.65–1.01 0.995–1.001

57 39 24 17 159±31

1.10 0.82 1.10 0.81 0.998

0.91–1.32 0.68–1.00 0.89–1.36 0.62–1.04 0.995–1.001

43 40 37 13 158±32

1.01 0.94 0.98 0.90 0.999

0.74–1.37 0.68–1.29 0.71–1.36 0.56–1.44 0.993–1.005

85±38 0.55±0.34 49 — — 20 42 50

0.997 0.78 1.35 — — 1.54 2.10 1.05

0.995–0.999 0.58–1.05 1.15–1.58 — — 1.29–1.85 1.79–2.46 0.90–1.23

88±37 0.57±0.33 40 23 42 — 46 50

0.998 0.84 1.21 1.30 0.97 — 2.23 1.04

0.995–1.000 0.61–1.16 1.01–1.45 1.06–1.60 0.80–1.16 — 1.85–2.69 0.86–1.24

70±42 0.45±0.32 82 — — — 24 52

0.998 0.82 1.29 — — — 2.97 1.05

0.993–1.003 0.43–1.58 0.83–1.98 — — — 2.16–4.08 0.77–1.43

∗ Any intervention in any arterial section. ‡ Variable is present when the patient was treated for it. # Blood pressures were available in 2160 of all patients, in 1785 of the patients with femoropopliteal grafts and in 375 patients with femorocrural grafts. s.d.=standard deviation. HR=hazard ratio. CI=confidence interval. Table 2. Results of multivariate analysis of risk factors for infrainguinal graft occlusion (all bypasses, n=2650). Office only model

Complete model

Variable

HR

95% CI

HR

95% CI

Female gender Critical ischaemia Femorocrural bypass Non-venous graft

1.22 1.29

1.02–1.46 1.08–1.53

1.29 1.76 2.45

1.07–1.56 1.40–2.23 2.03–2.95

“Office only” model: history and clinical findings only. Complete model: history, clinical, radiological and surgical findings. HR=hazard ratio. CI=confidence interval.

Multivariate analysis The ‘‘office only’’ model, developed with data available from patient history and clinical examination, revealed female gender and critical ischaemia as independent prognostic factors for graft occlusion (Table 2). In the ‘‘complete’’ model, in which data from radiological and operative findings were added, femorocrural bypass and use of non-venous graft material proved to be additional independent risk factors. The results of subgroup analysis of the patients with Eur J Vasc Endovasc Surg Vol 20, August 2000

femoropopliteal bypasses are shown in Table 3. In the ‘‘office only’’ model only female gender appeared to be associated independently with graft occlusion. The ‘‘complete’’ model indicated female gender, critical ischaemia, poor run-off and use of non-venous graft material as independent predictors of graft occlusion. Additional analysis Because the use of non-venous graft material appeared to be the strongest independent risk factor of graft

Risk Factors for Occlusion of Infrainguinal Bypass Grafts

121

Table 3. Results of multivariate analysis of risk factors for femoropopliteal graft occlusion (n=2119). Office only model

Complete model

Variable

HR

95% CI

HR

95% CI

Female gender Critical ischaemia Run-off Ζ1 vessel Non-venous graft

1.31

1.09–1.57

1.21 1.22 1.26 2.29

1.01–1.45 1.01–1.48 1.02–1.56 1.89–2.77

“Office only” model: history and clinical findings only. Complete model: history, clinical, radiological and surgical findings. HR=hazard ratio. CI=confidence interval.

occlusion, notably in femoropopliteal bypass grafts, an additional analysis of patients with supragenicular femoropopliteal bypasses (n=1222) was performed. Univariate analysis demonstrated that increasing brachial blood pressure, treated hypertension and hyperlipidaemia were associated with a lower risk for graft occlusion, whereas critical ischaemia, poor run-off and use of non-venous graft material were related to a higher probability of graft occlusion. Multivariate analysis using the ‘‘office only’’ model indicated increasing brachial blood pressure as the only independent factor associated with a lower risk of graft occlusion (hazard ratio 0.994; 95% CI 0.990–0.999). The ‘‘complete’’ model revealed brachial blood pressure (hazard ratio 0.994; 95% CI 0.990–0.999), hyperlipidaemia (hazard ratio 0.66; 95% CI 0.45–0.97), poor run-off (hazard ratio 1.45; 95% CI 1.06–1.99) and non-venous graft material, which was used in 64% of the patients (hazard ratio 2.44; 95% CI 1.18–3.34), as independent prognostic factors for graft occlusion.

Information content The ROC curve (Fig. 1a) of the “office only” model for all patients is only slightly elevated above the dotted line of no prognostic value. The corresponding area under the curve is 0.55 (95% CI 0.53–0.57). The “complete” model contained significantly more information, the area under the curve (Fig. 1b) is 0.64 (95% CI 0.62–0.66). In patients with femoropopliteal bypass grafts (Figs 1c and 1d) the information content of the “office only” model is also poor (area under the curve 0.54; 95% CI 0.48–0.60), whereas the “complete” model has a significantly higher predictive value (area under the curve 0.64; 95% CI 0.62–0.66).

Discussion This prospective study in a large group of patients who were included in a multicentre clinical trial gave

us the opportunity to calculate precise estimates of the prognostic importance of vascular risk factors for infrainguinal bypass graft occlusion. Only a few independent risk factors were identified. The information content of the ‘‘office only’’ model, containing data available from preoperative interview and physical examination, was extremely poor. Adding radiological and surgical data resulted in a more informative, but far from perfect, ‘‘complete’’ model. This model identified critical ischaemia, femorocrural bypass and use of non-venous graft material in all patients, supplemented by female gender and poor run-off in patients with femoropopliteal bypasses, as independent risk factors for occlusion. Other variables which are generally considered to be vascular risk factors did not predict graft occlusion. The higher risk for graft occlusion in women was also seen in a recent review of outcome of infrainguinal bypass surgery.10 Although men are more prone to atherosclerotic disease than women, atherosclerosis appears to be more advanced by the time women need surgical treatment. In the present study, women were, on average, 4 years older than men, but had less vascular interventions in the past (36% versus 51%, p<0.001). Women more often had surgery because of critical ischaemia (56% versus 45%, p<0.001), and in 45% of the women a non-venous graft was inserted compared with 40% of the men (p=0.01). The latter difference may be due to a smaller proportion of graftable greater saphenous veins in women, caused by, for example, varicose or small diameter. However, female gender is a risk factor for femoropopliteal graft occlusion independent of these factors. Critical ischaemia, femorocrural or femoropedal bypass grafting and, in patients with femoropopliteal bypasses, poor arterial run-off are generally assumed to be risk factors for graft occlusion. These factors, however, if present, or not avoidable. The most pronounced independent risk factor for occlusion, use of non-venous graft material is, in the event that a suitable vein is present, susceptible to intervention. Based on several studies,11–13 most of us agree on the preferred Eur J Vasc Endovasc Surg Vol 20, August 2000

122

M. J. D. Tangelder et al.

Fig. 1. ROC curves for all patients: the “office only” model (a) and the “complete” model (b), and ROC curves for patients with femoropopliteal bypass grafts: the “office only” model (c) and the “complete” model (d). The dotted line represents no prognostic value at all.

use of the greater saphenous vein for below-knee bypasses. However, whether or not we should preferably use the autogenous vein for above-knee bypass grafting still remains a matter of debate. Advocates of use of prosthetic materials in this position found upon similar patency rates of supragenicular prosthetic grafts and vein grafts. Advantages of use of prosthetic grafts are preservation of the greater saphenous vein for later use and shorter operation time. In a prospective study in 310 patients who received aboveknee autologous saphenous vein bypass and 132 Eur J Vasc Endovasc Surg Vol 20, August 2000

patients with a polytetrafluoroethylene (PTFE) supragenicular graft, multivariate analysis did not demonstrate any effect of the graft materials on patency.5 The study did show, however, that only 7% of the patients needed a secondary below-knee bypass in the same limb during 4 years’ follow-up. So far, we do not know how many patients in our study, particularly those with supragenicular prosthetic grafts, have needed their saphenous vein for revascularisations carried out for progression of atherosclerotic disease. In the present study, 64% of the above-knee bypasses

Risk Factors for Occlusion of Infrainguinal Bypass Grafts

were constructed of non-venous material (29% Dacron, 26% PTFE, 8% biograft and 1% composite materials). Our results indicate non-venous graft material as the most clear risk factor for occlusion, even in aboveknee bypasses; the relative risk compared with venous grafts was 2.4 in these patients. The Dutch BOA study was a large trial in which, for practical reasons, only easily obtainable and the most relevant information was registered. For that reason, we do not have any information about the prognostic value of blood constituents (such as clotting and rheological factors) which may carry additional information with regard to future graft occlusion. For the same reason we also have to do without information on the number of operations that were reoperations. Therefore, we were unfortunately not able to estimate the prognostic value of this variable. Most of the patients who underwent bypass surgery in The Netherlands during the inclusion period and did not have exclusion criteria (the most important ones being absolute indication or contraindication for oral anticoagulants of aspirin) were enrolled in the Dutch BOA Study. The results presented here can therefore be generalised to most patients requiring infrainguinal bypass surgery. In conclusion, this large prospective study has revealed only a few independent predictors for infrainguinal bypass-graft occlusion. The prognostic information content of variables available after interviewing and examining the patient is poor. Because length of the bypass, arterial run-off (in femoropopliteal bypasses) and graft material are the most pronounced risk factors for graft occlusion, the predictive value of a model containing all variables is significantly better. The major risk factor, even in above-knee bypasses, is the use of non-venous graft material. As this is also the risk factor which can be avoided if a suitable autogenous vein is available, the venous bypass graft should be offered to patients whenever possible.

References 1 Szilagyi DE, Hageman JH, Smith RF et al. Autogenous vein grafting in femoro-popliteal atherosclerosis: the limits of its effectiveness. Surgery 1979; 86: 836–849. 2 Kacoyanis GP, Wittemore AD, Couch NP, Mannick JA. Femorotibial and femoroperonaeal bypass vein graft. A 15-year experience. J Vasc Surg 1981; 116: 1529–1534. 3 Kretschmer G, Wenzl E, Wagner O et al. Influence of anticoagulant treatment in preventing graft occlusion following saphenous vein bypass for femoropopliteal occlusive disease. Br J Surg 1986; 73: 689–692. 4 McCollum C, Alexander C, Kenchington G, Franks PJ,

5 6 7 8

9 10 11

12 13

123

Greenhalgh RM. Antiplatelet drugs in femoropopliteal vein bypasses: a multicenter trial. J Vasc Surg 1991; 13: 150–162. Berlakovich GA, Herbst F, Mittlbock M, Kretschmer G. The choice of material for above-knee femoropopliteal bypass. A 20year experience. Arch Surg 1994; 129: 297–302. Sayers RD, Thompson MM, London NJ et al. Selection of patients with critical limb ischaemia for femorodistal vein bypass. Eur J Vasc Surg 1993; 7: 291–297. Sayers RD, Thompson MM, Dunlop P, London NJ, Bell PR. The fate of infrainguinal PTFE grafts and an analysis of factors affecting outcome. Eur J Vasc Surg 1994; 8: 607–610. The Dutch Bypass Oral anticoagulants or Aspirin (BOA) Study Group. Efficacy of oral anticoagulants compared with aspirin after infrainguinal bypass surgery (The Dutch Bypass Oral anticoagulant or Aspirin Study): a randomised trial. Lancet 2000; 355: 346–351. Cox DR. Regression models and lifetables. J R Stat Soc Ser B 1972; 34: 187–220. Magnant JG, Cronenwett JL, Walsh DB et al. Surgical treatment of infrainguinal arterial occlusive disease in women. J Vasc Surg 1993; 17: 67–76. Veith FJ, Gupta SK, Ascer E, White-Flores S et al. TI – Six-year prospective multicenter randomized comparison of autologous saphenous vein and expanded polytetrafluoroethylene grafts in infrainguinal arterial reconstructions. J Vasc Surg 1986; 1: 104–114. Tordoir JH, Van der Plas J, Jacobs MJ, Kitslaar PJ. Factors determining the outcome of crural and pedal revascularisation for critical limb ischaemia. Eur J Vasc Surg 1993; 7: 82–86. Panayiotopoulos YP, Reidy JF, Taylor PR. The concept of knee salvage: why does a failed femorocrural/pedal arterial bypass not affect the amputation level? Eur J Vasc Endovasc Surg 1997; 13: 477–485.

Accepted 14 February 2000

Appendix Participating centres (with number of patients randomised and principal investigators): Medisch Centrum Alkmaar (21; P.J. van den Akker, H.A. van Dijk, R.W. Meijer); Twenteborg Ziekenhuis Almelo (20; J.G. van Baal, J.W. van den Heuvel); Ziekenhuis Amstelveen (60; P.J. van Aken, J.A. Lawson); Academisch Medisch Centrum Amsterdam (79; R. Balm, M.J.H.M. Jacobs, C. Kox, D.A. Legemate, S.M.M. van Sterkenburg); Academisch Ziekenhuis Vrije Universiteit Amsterdam (16; A.J.C. Mackaay, J.A. Rauwerda, E.G.J. Vermeulen, C. van der Waal); Slotervaartziekenhuis Amsterdam (9: B.J. Dwars, T.O.M. Nagy); St. Lucas Ziekenhuis Amsterdam (15; A. Voorwinde); Ziekenhuiscentrum Apeldoorn (9; P.H. Rutgers); Ziekenhuis Rijnstate Arnhem (44; T.I.F.M. Bloemen, W.H.A. Govaert, I.M.C. Janssen, J.H.G. Klinkenbijl, W.R. de Vries); Wilhelmina Ziekenhuis Assen (5; J.A.G. de Groot, H.A.M. Heikens); Medisch Centrum Molendael Baarn (22; M.S. Verweij); Stichting Ziekenhuis Lievensberg Bergen op Zoom (2; T.H.A. Bikkers); Ziekenhuis Gooi-Noord Blaricum (18; J. Greebe, G.T. The); Ignatiusziekenhuis Breda (23; H.G.W. de Groot, R.A.E. Wirtz); Ziekenhuis De Baronie Eur J Vasc Endovasc Surg Vol 20, August 2000

124

M. J. D. Tangelder et al.

Breda (131; R.M.P.H. Crolla, F.J.M. van Geloven, G.A.M. Kokke, P.M. Raams, W.J. van Remortel); IJsselland Ziekenhuis Capelle a/d IJssel (35; I. Dawson, C.L. Koppert); Reinier de Graaf Gasthuis Delft (89; W.B.J. Jansen, J. Koning); Gemini Ziekenhuis Den Helder (9; J.J.M. Jutte, H.D. Tjeenk Willink); Stichting Deventer Ziekenhuizen Deventer (19; P.J. van Elk, D. van Lent); Merwedeziekenhuis Dordrecht (3; J. de Gruyl, M.T.C. Hoedt, P.R. Schu¨tte); Catharina Ziekenhuis Eindhoven (84; J. Buth, Ph.W.M. Cuypers, J.M.M.P.H. Herman, H.J.T. Rutten); Diaconessenhuis Eindhoven (62; W.J. Prakken); Medisch Spectrum Twente Enschede (77; R.J. van Det, R.H. Geelkerken, H.J. Mulder, P. de Smit); St. Anna Ziekenhuis Geldrop (38; P.R.M. De Bevere, F.Th.P.M. van der Linden); Oosterscheldeziekenhuizen Goes (2; H. Bruins Slot, A.J. de Nie); Groene Hart Ziekenhuis Gouda (19; B.J.L. Kothuis, J.C. Melse); Bronovo Ziekenhuis ’s-Gravenhage (59; A.B.B. van Rijn, H.J. Smeets); Rode Kruis Ziekenhuis ’s-Gravenhage (2; J.H. Allema, P.J. Breslau); Westeinde Ziekenhuis ’s-Gravenhage (13; J.C.A. de Mol van Otterloo, A.C. de Vries); Ziekenhuis Leyenburg ’s-Gravenhage (66; C.M.Z. Bruijninckx, B. Knippenberg, J.C. Sier); Academisch Ziekenhuis Groningen (37; J.J.A.M. van den Dungen, J. Oskam, R. van Schilfgaarde, E.L.G. Verhoeven); Martini Ziekenhuis Groningen (30; H.R. van Dop, A.J. Julius, A. van der Tol); Kennemer Gasthuis locatie Deo Haarlem (12; E.J. Boerma, J.A.R. Coosemans); Kennemer Gasthuis locatie Elisabeth Haarlem (10; H.L.F. Brom, A. Jansen, A.R. Koomen); Ziekenhuis St. Jansdal Harderwijk (22; W.L. Akkersdijk, A.C. van der Ham, A.K. Marck, M. Scheuer); Atrium Medisch Centrum Heerlen (20; E.C.M. Bollen, R.J.Th.J. Welten); Streekziekenhuis Midden Twente Hengelo (24; P. van der Sar); Bosch Medicentrum locatie Groot Ziekengasthuis ’s-Hertogenbosch (22; J. Wever, F.G.J. Willekens, R.J. de Wit); Bosch Medicentrum locatie Willem-Alexander Ziekenhuis ’s-Hertogenbosch (60; R.M.M. van Loenhout, T.J.M.J. Schiphorst, J.C. Wissing); Carolus-Liduina Ziekenhuis ’s-Hertogenbosch (22; I.P.T. van Bebber, S. Bouwer, E.J. Carol, J.H. Duppen, F.T.T. Liem); Ziekenhuis Hilversum (46; F.R.S. van Asperen de Boer, N.A. Koedam); Ziekenhuis Bethesda Hoogeveen (1; H.P. Dahler); Westfries Gasthuis Hoorn (103; D.L. Brands, M.W.C. de Jonge, J.W.D. de Waard); Medisch Centrum Leeuwarden (38; D.C. Busman, R. Leemans, J.A. Zijlstra); Academisch Ziekenhuis Leiden (25; G.J.M. Akkersdijk,

Eur J Vasc Endovasc Surg Vol 20, August 2000

J.M. van Baalen, J.H. van Bockel, G.W.H. Schurink); Diaconessenhuis Leiden (21; B.J. Hornstra, H. Stigter); Rijnland Ziekenhuis Leiderdorp (12; S.A. da Costa, J. Dubbeld, P.A. Neijenhuis, J.F.W.B. Rijksen); IJsselmeer Ziekenhuizen Lelystad (8; J.E.L. Cremers); Academisch Ziekenhuis Maastricht (72; P.J.E.H.M. Kitslaar, J.H.M. Tordoir); St. Antonius Ziekenhuis Nieuwegein (6; F.L. Moll, H.D.W.M. van de Pavooordt, R.P. Tutein Nolthenius); Academish Ziekenhuis Nijmegen (41; A.P.M. Boll, F.G.M. Buskens, J.A. van der Vliet); CanisiusWilhelmina Ziekenhuis Nijmegen (71; W.B. Barendregt, J.A.M. Hoogbergen, L.A.A. van Knippenberg, A.G.A. Spelde); Waterland Ziekenhuis Purmerend (67; Th.A.A. van den Broek, G.A. Vos); St. Laurentius Ziekenhuis Roermond (11; E.F.A. de Haan, C.M. Nuyens); Academisch Ziekenhuis Dijkzigt Rotterdam (42; L.M.C. van Dortmond, N.A.J. Dubois, M.R.H.M. van Sambeek, H. van Urk); Ikazia Ziekenhuis Rotterdam (56; R.U. Boelhouwer, H.F. Veen); St. Clara Ziekenhuis Rotterdam (21; J.M. Hendriks, A.A.E.A. de Smet, T.I. Yo); St. Franciscus Gasthuis Rotterdam (31; W.B. van Gent, C.H.A. Wittens); Maaslandziekenhuis Sittard (62: A.G.M. Hoofwijk, W.H.V.M. L’Ortije, H.J.G. Stroeken); Merwedeziekenhuis Sliedrecht (10; P.J. Mouthaan); St. Elisabeth Ziekenhuis Tilburg (85; D.P. van Berge Henegouwen, J.C. Breek, J.F. Hamming); Tweesteden Ziekenhuis Tilburg (100; S.J. Brenninkmeijer, G.P. Gerritsen, S.E. Kranendonk); Academisch Ziekenhuis Utrecht (66; J.D. Blankensteijn, B.C. Eikelboom, B.H.P. Elsman, R.W.H. van Reedt Dortland, J.J.F. Steijling, Th.J.M.V. van Vroonhoven); Ziekenhuis Overvecht Utrecht (54; B.C.V.M. Disselhoff); St. Joseph Ziekenhuis Veghel (36; C.J. Broers, H.A.P.A. de Geus); St. Joseph Ziekenhuis Veldhoven (24; M.H.M. Bender, J.A. Charbon, Th.J. van Straaten); St. Maartens Gasthuis Venlo (8; P.F. Verhagen); Holy Ziekenhuis Vlaardingen (57; E.R. Snijder); Streekziekenhuis Koningin Beatrix Winterswijk (2; A.A. Vafi); Hofpoort Ziekenhuis Woerden (26; E. Bakker, W. van Eesteren); Kennemer Gasthuis locatie Zeeweg IJmuiden (3; J.J. Petit); Stichting Ziekenhuis De Heel Zaandam (2; R.A. Cohen, R.P. Strating); Streekziekenhuis Zevenaar (33; R.F. de Haan, P.E. Reenalda, C. Sieswerda, J. van Wijk); ’t Lange Land Ziekenhuis Zoetermeer (8; M.T. Sjer, L.J.J.P. Speetjens); Het Nieuwe Spittaal Zutphen (15; R.H. Schreeve); St. Sophia Ziekenhuis Zwolle (41; P.J.G. Jo¨rning, M.C. Kerdel); Ziekenhuis De Weezenlanden Zwolle (36; A.M. Blomme, E. A. Kole).