Advertisement

Editor's Choice – Validation of the Management of Aortic Graft Infection Collaboration (MAGIC) Criteria for the Diagnosis of Vascular Graft/Endograft Infection: Results from the Prospective Vascular Graft Cohort Study

  • Author Footnotes
    ‡ These authors contributed equally to the manuscript.
    Alexia Anagnostopoulos
    Correspondence
    Corresponding author. University of Zurich, Institute of Epidemiology, Biostatistics and Prevention, Hirschengraben 84, 8001, Zurich, Switzerland.
    Footnotes
    ‡ These authors contributed equally to the manuscript.
    Affiliations
    Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Author Footnotes
    ‡ These authors contributed equally to the manuscript.
    Fabienne Mayer
    Footnotes
    ‡ These authors contributed equally to the manuscript.
    Affiliations
    Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Bruno Ledergerber
    Affiliations
    Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Judith Bergadà-Pijuan
    Affiliations
    Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Lars Husmann
    Affiliations
    Clinic for Nuclear Medicine, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Carlos A. Mestres
    Affiliations
    Clinic for Cardiac Surgery, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Zoran Rancic
    Affiliations
    Clinic for Vascular Surgery, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • Barbara Hasse
    Correspondence
    Corresponding author. University Hospital Zurich, Division of Infectious Diseases and Hospital Epidemiology, Raemistrasse 100, 8091, Zurich, Switzerland.
    Affiliations
    Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland
    Search for articles by this author
  • the VASGRA Cohort Study
  • Author Footnotes
    ‡ These authors contributed equally to the manuscript.
Open AccessPublished:June 14, 2021DOI:https://doi.org/10.1016/j.ejvs.2021.05.010

      Objective

      The timely management of vascular graft/endograft infection (VGEI) is crucial to a favourable outcome, yet can be challenging as there is no validated gold standard diagnostic test. Recently, a new case definition has been proposed by the Management of Aortic Graft Infection Collaboration (MAGIC) to close the diagnostic gap. The aim of this study was to validate the MAGIC criteria as a suggested diagnostic standard for the diagnosis of suspected VGEI in the prospective Vascular Graft Cohort study (VASGRA).

      Methods

      VASGRA is an open, prospective, observational cohort study. Prospective participants in VASGRA between 2013 and 2019 were included (257 patients; 137 with VGEI). The accuracy of the MAGIC criteria for a diagnosis of VGEI was evaluated retrospectively by calculating the sensitivity and specificity vs. the consensually adjudicated VASGRA infection status.

      Results

      The VASGRA cohort categorised 137 (53.3%) patients as “diseased” and 120 patients as “not diseased”; using the MAGIC criteria, 183/257 (71.2%) patients were considered to be “diseased”. Thus, for the MAGIC criteria, a sensitivity of 99% (95% confidence interval [CI] 96–100) and a specificity of 61% (95% CI 52–70) were calculated. Considering suspected VGEI according to the MAGIC criteria as “not diseased” achieved congruent assessments of the VASGRA team and the MAGIC criteria, with a sensitivity of 93% and a specificity of 93%. The accuracy of the MAGIC criteria for the different graft locations were also compared. If the suspected VGEIs were assigned to the “not diseased” group, VGEIs of the thoracic aorta seemed to have a poorer sensitivity (86%; 95% CI 73–95) than the other graft locations.

      Conclusion

      The current MAGIC criteria offer good sensitivity and specificity in the context of true infections but a reduced specificity for a possible VGEI.

      Keywords

      The Management of Aortic Graft Infection Collaboration (MAGIC) criteria have been proposed as a novel diagnostic test for vascular graft/endograft infection (VGEI). The criteria were validated retrospectively in a prospective cohort of patients with definite and suspected vascular graft infections. For a definite VGEI diagnosis, the criteria had a good sensitivity but reduced specificity, owing to suspected VGEI. To improve the accuracy, further modifications of the criteria should be evaluated.

      Introduction

      Vascular infections involving prosthetic graft material cause substantial morbidity, lethality, and high healthcare costs.
      • Hasse B.
      • Husmann L.
      • Zinkernagel A.
      • Weber R.
      • Lachat M.
      • Mayer D.
      Vascular graft infections.
      ,
      • Chakfe N.
      • Diener H.
      • Lejay A.
      • Assadian O.
      • Berard X.
      • Caillon J.
      • et al.
      Editor’s Choice – European Society for Vascular Surgery (ESVS) 2020 Clinical Practice Guidelines on the Management of Vascular Graft and Endograft Infections.
      The timely and accurate assessment and diagnosis of vascular graft/endograft infection (VGEI) seems to be crucial for a favourable outcome. A VGEI may be obvious in patients with bacteraemia and abscess formation around a vascular graft. However, a definite VGEI diagnosis is challenging, and usually involves multiple findings rather than one gold standard diagnostic test. Physicians often rely on a diversity of clinical symptoms, descriptive radiological imaging, and ambiguous or even missing microbiology, laboratory, or histopathological results. In the absence of a formal case definition, either the rather non-specific Fitzgerald criteria are applied to abdominal or peripheral VGEIs,
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      or the modified Duke criteria to thoracic VGEIs with composite grafts.
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      Recently, criteria for the diagnosis of VGEI have been proposed by the multidisciplinary Management of Aortic Graft Infection Collaboration (MAGIC),
      • Lyons O.T.
      • Baguneid M.
      • Barwick T.D.
      • Bell R.E.
      • Foster N.
      • Homer-Vanniasinkam S.
      • et al.
      Diagnosis of aortic graft infection: a case definition by the Management of Aortic Graft Infection Collaboration (MAGIC).
      and the European Society for Vascular Surgery suggests the use of the MAGIC criteria as a diagnostic standard for all kinds of VGEI (thoracic, abdominal, and peripheral arteries).
      • Chakfe N.
      • Diener H.
      • Lejay A.
      • Assadian O.
      • Berard X.
      • Caillon J.
      • et al.
      Editor’s Choice – European Society for Vascular Surgery (ESVS) 2020 Clinical Practice Guidelines on the Management of Vascular Graft and Endograft Infections.
      The aim of this study was to validate the MAGIC criteria for VGEI diagnosis in the prospective Vascular Graft Infection Cohort study (VASGRA). The accuracy of different VGEI locations was also assessed using the MAGIC criteria.
      • Lyons O.T.
      • Baguneid M.
      • Barwick T.D.
      • Bell R.E.
      • Foster N.
      • Homer-Vanniasinkam S.
      • et al.
      Diagnosis of aortic graft infection: a case definition by the Management of Aortic Graft Infection Collaboration (MAGIC).

      Materials and methods

      Study design and data collection

      VASGRA is an open, prospective, observational cohort study, which enrols patients aged ≥ 18 years who are receiving vascular graft implantations at the University Hospital Zurich, Switzerland. In VASGRA, standardised data collection forms covering demographic, surgical, clinical, and laboratory information, are completed every three months by physicians and study nurses.
      • Hasse B.
      • Husmann L.
      • Zinkernagel A.
      • Weber R.
      • Lachat M.
      • Mayer D.
      Vascular graft infections.
      Patients are followed prospectively in the vascular surgery outpatient clinic of the hospital with contrast enhanced computed tomography (CECT), and laboratory tests. VGEI incidence and post-operative complications have been documented since 2013. Patients who return to hospital with the suspicion of a VGEI, receive blood and/or tissue cultures, serologies if the culture is negative, laboratory and imaging assessments, 18Fluorodeoxyglucose (FDG; positron emission tomography [PET]/CT), CECT, and a combined imaging approach (contrast enhanced PET [CEPET]/CT). The percentages of CEPET/CECTs among VASGRA suspected/confirmed VGEI and VASGRA rejected VGEI are 89% and 100%, respectively. The Ethics Committee Zurich, Switzerland, approved the study (KEK-ZH-Number 2012-0583). All participants gave written informed consent.

      Study patients, internal VASGRA adjudication of infection status, and adjudication by the use of the MAGIC criteria

      VASGRA patients presenting with suspected VGEI between April 2013 and September 2019 were included, reviewed, and consensually grouped, in an unblinded manner, into definite, suspected, and rejected VGEI by the multidisciplinary VASGRA team, which includes infectious diseases specialists, cardiac and vascular surgeons, radiologists, microbiologists, and pathologists. For further validation, a group of control patients was also included. For each patient with a VGEI, a control with the same graft location site, a clinic visit, laboratory tests, and an imaging procedure at the same ± 2 number of weeks between the initial graft surgery and the diagnosis of a VGEI in the respective patient was searched for. Controls were selected only once. Laboratory controls were found for all patients, but information on C reactive protein (CRP) and leucocytes was only available in 80% of the controls.
      In detail, the adjudication process was done by two infectious diseases specialists, a nuclear medicine physician/radiologist, and cardiac and vascular surgeons. The following parameters for the VGEI adjudication process were used, whereby a precondition of a diagnosis of VGEI was the presence of at least one clinical, imaging, and laboratory parameter in each category.

      Clinical

      Pus around the graft; open wound with exposed graft or fistula; graft insertion in an infected site (e.g., in case of endovascular treatment of an infected aneurysm).
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      ,
      • Legout L.
      • Sarraz-Bournet B.
      • D’Elia P.V.
      • Devos P.
      • Pasquet A.
      • Caillaux M.
      • et al.
      Characteristics and prognosis in patients with prosthetic vascular graft infection: a prospective observational cohort study.
      ,
      • Wilson W.R.
      • Bower T.C.
      • Creager M.A.
      • Amin-Hanjani S.
      • O’Gara P.T.
      • Lockhart P.B.
      • et al.
      Vascular graft infections, mycotic aneurysms, and endovascular infections: a scientific statement from the American Heart Association.

      Imaging

      CEPET/CECT showing a focal FDG activity combined with at least one CT criterion (e.g., perigraft fluid on CT scan ≥ 3 months after insertion, perigraft gas on CT scan ≥ 7 weeks after insertion; fat stranding; and contrast enhancement).
      • Sah B.R.
      • Husmann L.
      • Mayer D.
      • Scherrer A.
      • Rancic Z.
      • Puippe G.
      • et al.
      Diagnostic performance of 18F-FDG-PET/CT in vascular graft infections.
      • Husmann L.
      • Huellner M.W.
      • Ledergerber B.
      • Anagnostopoulos A.
      • Stolzmann P.
      • Sah B.R.
      • et al.
      Comparing diagnostic accuracy of (18)F-FDG-PET/CT, contrast enhanced CT and combined imaging in patients with suspected vascular graft infections.
      • Kim S.J.
      • Lee S.W.
      • Jeong S.Y.
      • Pak K.
      • Kim K.
      A systematic review and meta-analysis of (18)F-fluorodeoxyglucose positron emission tomography or positron emission tomography/computed tomography for detection of infected prosthetic vascular grafts.
      • Reinders Folmer E.I.
      • Von Meijenfeldt G.C.I.
      • Van der Laan M.J.
      • Glaudemans A.
      • Slart R.
      • Saleem B.R.
      • et al.
      Diagnostic imaging in vascular graft infection: a systematic review and meta-analysis.
      No cutoffs (i.e., for standard uptake value measurements) were defined. Readers scored their confidence in the diagnosis using a previously published four point score.
      • Husmann L.
      • Huellner M.W.
      • Ledergerber B.
      • Anagnostopoulos A.
      • Stolzmann P.
      • Sah B.R.
      • et al.
      Comparing diagnostic accuracy of (18)F-FDG-PET/CT, contrast enhanced CT and combined imaging in patients with suspected vascular graft infections.

      Laboratory

      Positive cultures for bacteria, fungi, or mycobacteria from intra-operative specimens/biopsies/explanted grafts and/or positive blood cultures. To rule out potential contaminant pathogens (e.g., coagulase negative staphylococci, Cutibacterium acnes, and Corynebacterium spp.), two positive intra-operative specimens or blood cultures, or one positive intra-operative specimen and one positive blood culture were required.
      • Legout L.
      • Sarraz-Bournet B.
      • D’Elia P.V.
      • Devos P.
      • Pasquet A.
      • Caillaux M.
      • et al.
      Characteristics and prognosis in patients with prosthetic vascular graft infection: a prospective observational cohort study.
      Positive broad range polymerase chain reaction (PCR) and/or species PCR for Brucella spp., Coxiella burnetii, Bartonella spp., and Tropheryma whipplei.
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      ,
      • Ajdler-Schaeffler E.
      • Scherrer A.U.
      • Keller P.M.
      • Anagnostopoulos A.
      • Hofmann M.
      • Rancic Z.
      • et al.
      Increased pathogen identification in vascular graft infections by the combined use of tissue cultures and 16S rRNA gene polymerase chain reaction.
      ,
      • Bisharat N.
      • Minuhin I.
      Prosthetic vascular graft infections between blood and concordance of graft culture pathogen.
      Serological evidence of infection with microorganisms consistent with VGEI diagnosis (C. burnetii, Bartonella spp., and Brucella spp.).
      • Hagenaars J.C.
      • Wever P.C.
      • van Petersen A.S.
      • Lestrade P.J.
      • de Jager-Leclercq M.G.
      • Hermans M.H.
      • et al.
      Estimated prevalence of chronic Q fever among Coxiella burnetii seropositive patients with an abdominal aortic/iliac aneurysm or aorto-iliac reconstruction after a large Dutch Q fever outbreak.
      ,
      • Million M.
      • Raoult D.
      Recent advances in the study of Q fever epidemiology, diagnosis and management.
      Microorganisms demonstrated by microorganism staining or histopathological examination of perigraft material and, if applicable, postmortem examinations (immunohistology, haematoxylin and eosin stain, Ziehl–Neelsen stain, Grocott stain, periodic acid–Schiff stain.
      • Mehta R.I.
      • Mukherjee A.K.
      • Patterson T.D.
      • Fishbein M.C.
      Pathology of explanted polytetrafluoroethylene vascular grafts.
      ,
      • Van Damme H.
      • Deprez M.
      • Creemers E.
      • Limet R.
      Intrinsic structural failure of polyester (Dacron) vascular grafts. A general review.
      Elevated inflammatory markers (CRP, white blood cell count).
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      ,
      • Wilson W.R.
      • Bower T.C.
      • Creager M.A.
      • Amin-Hanjani S.
      • O’Gara P.T.
      • Lockhart P.B.
      • et al.
      Vascular graft infections, mycotic aneurysms, and endovascular infections: a scientific statement from the American Heart Association.
      The following VGEI locations were distinguished: “thoracic aorta” (intrathoracic part of the aorta, aortic root, and intrathoracic branches; and supra-aortic trunks); “abdominal aorta” (intra-abdominal part of the aorta, including iliac branches); and “peripheral arteries” (groin and lower extremity). The VGEI was further classified according to the Samson classification.
      • Samson R.H.
      • Veith F.J.
      • Janko G.S.
      • Gupta S.K.
      • Scher L.A.
      A modified classification and approach to the management of infections involving peripheral arterial prosthetic grafts.
      In VASGRA, a VGEI was suspected if there were elevated inflammatory markers and an unexplained fluid collection without focal FDG uptake around the vascular graft (> 3 months after insertion) or positive blood cultures (two for contaminant pathogens; one for “non-contaminant pathogens”) in a patient with a vascular graft. A VGEI was discounted if, in due course, a diagnosis other than VGEI was confirmed.
      In addition, infection status was retrospectively assessed and adjudicated by applying the MAGIC criteria to suspected and confirmed VGEI.
      • Lyons O.T.
      • Baguneid M.
      • Barwick T.D.
      • Bell R.E.
      • Foster N.
      • Homer-Vanniasinkam S.
      • et al.
      Diagnosis of aortic graft infection: a case definition by the Management of Aortic Graft Infection Collaboration (MAGIC).

      Statistical analysis

      Patient and procedure related variables were assessed overall and by infection status using non-parametric tests (Fisher’s exact test or the Kruskal–Wallis test, whichever was appropriate). The accuracy of the MAGIC criteria for the diagnosis of a VGEI was evaluated by calculating sensitivity and specificity vs. the consensually agreed VASGRA infection status. Sensitivity and specificity estimates with exact binomial confidence intervals (CIs) were provided.
      All patients were further grouped based on their adjudicated infection status into “diseased” (definite and suspected VGEI) and “not diseased” (rejected VGEI and control patients) groups. The most challenging group in which no definite diagnosis/rejection of VGEI was possible was then looked at, leading to a status of “suspected VGEI”. Those cases were first assigned to the “diseased” group and then to the “not diseased” group, and the accuracy of the MAGIC criteria calculated and compared for both scenarios.
      Stata/SE Version 16.1 (StataCorp., College Station, TX, USA) was used for the statistical analyses.

      Results

      Study participants

      Overall, 257 predominantly male (83.7%) participants with a median age of 67 years (interquartile range 59 – 75 years) were included. The VASGRA cohort categorised 137 (53.3%) patients as “diseased” (135 definite VGEI; two suspected VGEI) and 120 patients as “not diseased”. The “not diseased” group included 35 patients in whom a VGEI was rejected by consensus and a control group of 85 control patients undergoing routine vascular surgery. Table 1 shows the patient and procedure related variables by infection status and by disease group. With regard to the surgical strategy in patients with VGEIs, 45% received debridement, 34% a total graft replacement, 9% a partial graft replacement, and 12% were managed conservatively.
      Table 1Characteristics of 137 patients with and 120 patients without vascular graft/endograft infection (VGEI) by internally adjudicated Vascular Graft Cohort Study (VASGRA) infection status
      VASGRA diseased (n = 137)VASGRA not diseased (n = 120)p value
      Definite VGEI (n = 135)Suspected VGEI (n = 2)Rejected VGEI (n = 35)Control patients (n = 85)
      Male gender114 (84.4)2 (100)31 (89)68 (80).67
      Age – y64 (57–73)71 (67–75)72 (59–78)70 (61–75).071
      Emergency at index surgery33 (24.4)0 (0)8 (23)4 (4.7)<.001
      Location of vessel
       Abdominal aortal52 (38.5)2 (100)31 (89)66 (78)<.001
       Thoracic aorta44 (33.6)0 (0)2 (6)10 (12)<.001
       Peripheral arteries40 (29.6)0 (0)2 (6)1 (1)<.001
      Samson classification
      Samson criteria:16 III = involvement of graft body but no anastomosis affected; IV = infection surrounding exposed anastomosis but no bleeding or bacteraemia; V = involvement of a graft to artery anastomosis, septicaemia, and/or anastomotic bleeding.
       III58 (43.0)1 (50)NANA
       IV28 (20.7)0 (0)NANA
       V49 (36.3)1 (50)NANA1.0
      Data are presented as n (%) or median (interquartile range). NA = not applicable.
      Samson criteria:
      • Mehta R.I.
      • Mukherjee A.K.
      • Patterson T.D.
      • Fishbein M.C.
      Pathology of explanted polytetrafluoroethylene vascular grafts.
      III = involvement of graft body but no anastomosis affected; IV = infection surrounding exposed anastomosis but no bleeding or bacteraemia; V = involvement of a graft to artery anastomosis, septicaemia, and/or anastomotic bleeding.

      Distribution of MAGIC criteria within the VASGRA definite vascular graft/endograft infection group

      Of the 135 VASGRA definite VGEIs, 102 (75.6%) had at least two major MAGIC criteria. Of these, 26.5% (n = 27) had all three MAGIC criteria categories, whereas the other patients either had combined clinical and laboratory major criteria (54.9%; n = 56), combined clinical and radiological major criteria (11.7%), or combined radiological and laboratory major criteria (6.9%). In 24 of 135 (17.8%) patients with VASGRA definite VGEI, one major and one minor MAGIC criteria were present. Of these, 14 had a clinical major criterion, while in six patients a radiological and in four patients a major laboratory criterion was present. The remaining eight (5.9%) patients with VASGRA definite VGEI had at least two minor criteria from different categories and were therefore considered to have a suspected VGEI according to the MAGIC criteria. One patient with a VASGRA definite VGEI was rejected as a VGEI according to the MAGIC criteria.
      The distribution of all MAGIC major and minor criteria by VASGRA infection status within the “diseased” and “not diseased” groups is shown in Table 2.
      Table 2Management of Aortic Graft Infection Collaboration (MAGIC) major and minor criteria by Vascular Graft Infection Cohort Study (VASGRA) infection status within the “diseased” group of 135 patients with a vascular graft/endograft infection (VGEI)
      VASGRA participants (n = 135)
      MAGIC major criteria
       Pus (definite by microscopy) around graft or aneurysm sac at surgery61 (45.2)
       Open wound with exposed graft or communicating sinus33 (24.4)
       Fistula development, e.g., aorto-enteric or aortobronchial30 (22.2)
       Graft insertion in an infected site, e.g., fistula, mycotic aneurysm, or infected pseudo-aneurysm31 (23.0)
       Perigraft fluid on CT scan ≥ 3 months after insertion43 (31.8)
       Perigraft gas on CT scan ≥ 7 weeks after insertion20 (14.8)
       Increase in perigraft gas volume demonstrated on serial imaging12 (8.9)
       Microorganism recovered from an explanted graft27 (20.0)
       Microorganism recovered from an intra-operative specimen92 (68.1)
       Microorganism recovered from a percutaneous aspirate of perigraft fluid12 (8.9)
      MAGIC minor criteria
       Localised clinical features of VGEI, e.g., erythema, warmth, swelling, purulent discharge, and pain69 (51.1)
       Fever ≥ 38°C with VGEI as most likely cause64 (47.4)
       Other, e.g., suspicious perigraft gas/fluid/soft tissue inflammation; aneurysm expansion; pseudo-aneurysm formation; focal bowel wall thickening; discitis/osteomyelitis; suspicious metabolic activity on FDG PET/CT; radiolabelled leucocyte uptake109 (80.7)
       Blood culture(s) positive and no apparent source except for VGEI51 (37.8)
       Abnormally elevated inflammatory markers with VGEI as the most likely cause, e.g., ESR, CRP, and white cell count126 (93.3)
      Data are presented as n (%). CT = computed tomography; FDG PET/CT = fluorodeoxyglucose positron emission tomography/computed tomography; ESR = erythrocyte sedimentation rate; CRP = C reactive protein.

      Comparison of the MAGIC and VASGRA infection status for definite and rejected vascular graft/endograft infections

      Table 3 shows the comparison of the assessment using the MAGIC criteria with the adjudicated VASGRA infection status. All but one rejected VGEIs according to the MAGIC criteria were either VASGRA rejected VGEIs or controls. Among infections, the MAGIC assessment was in line with the VASGRA infection status in 93.3% (n = 126), while in 3.7% (n = 5) and 2.2% (n = 3) the VASGRA team placed the patient either in the group of rejected VGEIs or in the control group, respectively. One patient with a MAGIC definite VGEI remained under suspicion by the VASGRA adjudication team.
      Table 3Comparison of infection status according to the Vascular Graft Infection Cohort Study (VASGRA) and Management of Aortic Graft Infection Collaboration (MAGIC) adjudication in 137 patients with and 120 patients without vascular graft/endograft infection (VGEI)
      MAGIC adjudicationVASGRA adjudicationTotal
      Confirmed VGEISuspected VGEIRejected VGEIControl patients
      Confirmed VGEI126 (93.3)1 (50)5 (14)3 (3)135 (52.5)
      Suspected VGEI8 (5.9)1 (50)25 (71)14 (16)48 (18.7)
      Excluded VGEI1 (0.7)0 (0)5 (14)0 (0)6 (2.3)
      Control patients0 (0)0 (0)0 (0)68 (80)68 (26.4)
      Total135 (100)2 (100)35 (100)85 (100)257 (100)
      Data are presented as n (%).

      Comparison of the MAGIC and VASGRA infection status for suspected vascular graft/endograft infection

      The group of suspected VGEIs based on the MAGIC criteria showed a heterogeneous picture. Only one of 48 patients in this group (2.1%) was equally assessed as having a suspected VGEI by the VASGRA team. The majority (52.1%; n = 25/48) were considered as not having a VGEI, while 16.6% patients were placed in the definite VGEI group; 29.2% of patients were part of the control group (Table 3).
      The current MAGIC criteria offer a reduced specificity for possible VGEI. While only 137 (53.3%) patients were counted as being part of the VASGRA “diseased” group, the MAGIC criteria considered 183 (71.2%) patients as being “diseased” (Supplementary Table 1, Panel A). Based on this association, a sensitivity of 99% (95% CI 96 – 100) and a specificity of 61% (95% CI 52 – 70) were calculated for the MAGIC criteria. The likelihood ratio of a positive test among definite/suspected VGEIs vs. rejected VGEI/controls was 2.53 (95% CI 2.03–3.17).
      Patients with a non-definite infection status were assigned to the “not diseased” group. In this case, the assessments were congruent (Supplementary Table 1, Panel B). This is also reflected in improved accuracy, as calculated by a sensitivity of 93% (95% CI 88–97) and a specificity of 93% (95% CI 87–97). The likelihood ratio of a positive test among patients with a definite VGEI vs. those with a suspected and rejected VGEI and controls was 12.7 (95% CI 6.7–23.8).

      Sensitivity analyses

      The accuracy of the MAGIC criteria were also compared for different graft locations. Assigning a suspected VGEI to the “diseased’” group showed a poor specificity for VGEI of the thoracic aorta (42%; 95% CI 15 – 72), while sensitivities were comparable among the different locations and overall (Table 4). If suspected VGEIs were assigned to the “not diseased” group, VGEIs of the thoracic aorta seemed to have a poorer sensitivity (86%; 95% CI 73 – 95) than other graft locations. For VGEIs of the peripheral arteries, the specificity was substantially lower (67%; 95% CI 9.4 – 99) (Table 4).
      Table 4Accuracy of Management of Aortic Graft Infection Collaboration (MAGIC) criteria by graft location depending on the composition of “diseased” and “not diseased” groups in 257 patients with or without vascular graft/endograft infection (VGEI)
      Sensitivity (95% CI) – %Specificity (95% CI) – %
      “Diseased” = definite and suspected VGEI; “not diseased” = rejected VGEI and controls
      Overall99 (96–100)61 (52–70)
      Intracavitary abdominal VGEI100 (93–100)62 (51–72)
      Thoracic aorta VGEI98 (88–100)42 (15–72)
      Peripheral arteries VGEINAN/A
      “Diseased” = definite VGEI; “not diseased” = suspected VGEI, rejected VGEI, and controls
      Overall93 (88–97)93 (87–97)
      Intracavitary abdominal VGEI94 (84–99)92 (85–96)
      Thoracic aorta VGEI86 (73–95)100 (74–100)
      Peripheral arteries VGEI100 (91–100)67 (9.0–99)
      CI = confidence interval; NA = not applicable (as there were no patients with a MAGIC score of 0).
      Further sensitivity analyses after the exclusion of VASGRA suspected cases (n = 2) or the exclusion of VASGRA control patients are provided in the supplementary material.

      Discussion

      The MAGIC criteria were assessed retrospectively and validated as the proposed new diagnostic standard for VGEI diagnosis in the prospective VASGRA cohort study. Overall, in patients with VASGRA definite VGEI, the MAGIC criteria offered a good sensitivity but a reduced specificity for diagnosis. Using the MAGIC criteria led to an overestimation of suspected VGEI.
      Comparison with other studies and the currently scarce literature is difficult, as neither suitable reference cohorts nor other validation studies in which the initially proposed Fitzgerald criteria,
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      the modified Duke criteria,
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      ,
      • Durack D.T.
      • Lukes A.S.
      • Bright D.K.
      New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Endocarditis Service.
      or the MAGIC criteria were applied to VGEI were found. One cohort study applied the Fitzgerald criteria as VGEI case definition for a risk factor analysis.
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      ,
      • Legout L.
      • Sarraz-Bournet B.
      • D’Elia P.V.
      • Devos P.
      • Pasquet A.
      • Caillaux M.
      • et al.
      Characteristics and prognosis in patients with prosthetic vascular graft infection: a prospective observational cohort study.
      However, the authors concluded that there were shortcomings challenging the introduction of the criteria as standard in VGEI case definition: two of three criteria were subject to personal judgement and inconsistent interpersonal rating or relied on arbitrarily chosen and not yet validated cutoff values. Moreover, the three rather vague criteria were all weighed equally and thus challenged a firm diagnosis of VGEI.
      Neither the Fitzgerald nor the modified Duke criteria were applied to the VASGRA participants and therefore the validation of their performance in VGEI diagnosis cannot be commented upon.
      • FitzGerald S.F.
      • Kelly C.
      • Humphreys H.
      Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
      ,
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      However, it is worth noting that there were similarities in the development and modification process of the MAGIC criteria and the Duke criteria.
      • Durack D.T.
      • Lukes A.S.
      • Bright D.K.
      New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Endocarditis Service.
      The Duke criteria were widely accepted and validated, and showed a high sensitivity and specificity in the diagnosis of infective endocarditis.
      • Durack D.T.
      • Lukes A.S.
      • Bright D.K.
      New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Endocarditis Service.
      ,
      • Hoen B.
      • Beguinot I.
      • Rabaud C.
      • Jaussaud R.
      • Selton-Suty C.
      • May T.
      • et al.
      The Duke criteria for diagnosing infective endocarditis are specific: analysis of 100 patients with acute fever or fever of unknown origin.
      Yet, they were criticised for the “overly broad categorisation of the suspected infective endocarditis group”.
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      In fact, the inclusion of additional criteria was suggested to achieve firm definite diagnoses and increase the diagnostic yield. Taking this into account, the modified Duke criteria added more refined imaging techniques (transoesophageal echocardiography in patients with suspected infective endocarditis) and serology methods in culture negative infective endocarditis to the scheme.
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      In 2015, the modified Duke criteria were redefined again, with the inclusion of new imaging modalities like PET/CT, cardiac CT, and magnetic resonance imaging.
      • Li J.S.
      • Sexton D.J.
      • Mick N.
      • Nettles R.
      • Fowler Jr., V.G.
      • Ryan T.
      • et al.
      Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
      ,
      • Habib G.
      • Lancellotti P.
      • Antunes M.J.
      • Bongiorni M.G.
      • Casalta J.P.
      • Del Zotti F.
      • et al.
      2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM).
      When new diagnostic criteria are assessed, it is not only the accuracy, but also the direct or indirect effects on patient management that should be considered. In the present study, the diagnostic accuracy of the MAGIC criteria in VGEI assessment was very good in the case of definite diagnoses. In suspected VGEI, there should be a balance between suspected overtreatment with unnecessary surgery and/or antimicrobial therapy, potential patient discomfort, and the undesirable consequence of delayed surgery in the case of a missed diagnosis. Counting suspected VGEI as “diseased” substantially reduced the specificity of the MAGIC criteria vs. the assessment of the multidisciplinary team in this study. This was especially true for VGEIs of the thoracic and abdominal aorta. Therefore, the analyses demonstrated the need for additional assessment tools leading to a more tailored application of a diagnostic algorithm for suspected VGEI and different graft locations. With the intention of obtaining an overarching case definition of VGEI, the MAGIC criteria were also assessed for abdominal, thoracic, and peripheral locations separately.
      • Chakfe N.
      • Diener H.
      • Lejay A.
      • Assadian O.
      • Berard X.
      • Caillon J.
      • et al.
      Editor’s Choice – European Society for Vascular Surgery (ESVS) 2020 Clinical Practice Guidelines on the Management of Vascular Graft and Endograft Infections.
      Hence, the overall test performance could be less accurate. However, sensitivity analyses were performed and the accuracy for each location calculated separately. Moreover, the likelihood ratios for putting equal weights to sensitivity and specificity were considered and thus accounted for the proportion of individuals with and without the disease. Finally, suitable cutoff values for inflammatory markers were not further elaborated and the availability of PET CT might be reduced or non-existent in some healthcare settings.
      After having applied and validated the proposed MAGIC categories among VASGRA patients, modifications to the criteria are suggested. Radiological findings are one of the three groups of the MAGIC criteria on which a VGEI diagnosis is reliant. CECT is the most commonly used for the diagnosis of VGEI, reaching high sensitivity and specificity, especially in acute infections.
      • Hasse B.
      • Husmann L.
      • Zinkernagel A.
      • Weber R.
      • Lachat M.
      • Mayer D.
      Vascular graft infections.
      ,
      • Wilson W.R.
      • Bower T.C.
      • Creager M.A.
      • Amin-Hanjani S.
      • O’Gara P.T.
      • Lockhart P.B.
      • et al.
      Vascular graft infections, mycotic aneurysms, and endovascular infections: a scientific statement from the American Heart Association.
      Suspicious findings on CECT are therefore clearly considered major MAGIC criteria. More ambiguous and undetermined results are found in suspected VGEI. More recently, PET CT and 99-mTC white blood cell scintigraphy with single photon emission CT (SPECT)/CT have been studied as imaging modalities in VGEI.
      • Kim S.J.
      • Lee S.W.
      • Jeong S.Y.
      • Pak K.
      • Kim K.
      A systematic review and meta-analysis of (18)F-fluorodeoxyglucose positron emission tomography or positron emission tomography/computed tomography for detection of infected prosthetic vascular grafts.
      ,
      • Reinders Folmer E.I.
      • Von Meijenfeldt G.C.I.
      • Van der Laan M.J.
      • Glaudemans A.
      • Slart R.
      • Saleem B.R.
      • et al.
      Diagnostic imaging in vascular graft infection: a systematic review and meta-analysis.
      Among patients with suspected VGEIs, the combination of the high sensitivity of PET/CT with the high specificity of CECT outperformed standalone imaging and reached very high diagnostic accuracy.
      • Husmann L.
      • Huellner M.W.
      • Ledergerber B.
      • Anagnostopoulos A.
      • Stolzmann P.
      • Sah B.R.
      • et al.
      Comparing diagnostic accuracy of (18)F-FDG-PET/CT, contrast enhanced CT and combined imaging in patients with suspected vascular graft infections.
      In a 2018 meta-analysis, SPECT/CT showed the highest accuracy in VGEI diagnosis.
      • Reinders Folmer E.I.
      • Von Meijenfeldt G.C.I.
      • Van der Laan M.J.
      • Glaudemans A.
      • Slart R.
      • Saleem B.R.
      • et al.
      Diagnostic imaging in vascular graft infection: a systematic review and meta-analysis.
      Thus, it is suggested that CECT and PET/CT and/or SPECT are equally important major criteria.
      Another part of the MAGIC criteria is the laboratory element. The major criteria include evidence of isolated microorganisms, either from explanted graft material, biopsy, or aspiration of perigraft fluid. In order for a patient to have a major criterion, an intervention must have been performed, which is not necessarily the case, depending on the patient’s situation. There is lack of consideration of histopathology and serology in this context, especially in the setting of culture negative VGEI. It is proposed that “microbiological findings” and “laboratory/ histopathology” are categorised separately.
      This study has several strengths. It was possible to validate the MAGIC criteria
      • Lyons O.T.
      • Baguneid M.
      • Barwick T.D.
      • Bell R.E.
      • Foster N.
      • Homer-Vanniasinkam S.
      • et al.
      Diagnosis of aortic graft infection: a case definition by the Management of Aortic Graft Infection Collaboration (MAGIC).
      in a well established cohort study by comparing the diagnostic criteria with an appropriate reference standard. Moreover, there was a standardised adjudication process based on clear criteria and a final multidisciplinary consensus.
      However, the study also has limitations. Firstly, an appropriate standard for comparison is difficult and this cohort study was not designed to validate the diagnostic criteria prospectively. A retrospective evaluation of prospectively collected information from the patients was carried out, and the original MAGIC criteria were published 2016, which may have changed the diagnostic assessment of VASGRA patients in the second part of the study period. However, when comparing the time period before and after 2016, the accuracy did not differ (data not shown). There is no record linkage to other hospitals and hence patients with a rejected VGEI or controls may have presented elsewhere with a subsequent diagnosis of VGEI. However, in VASGRA, there is a median follow up of 4.5 years for both VGEI patients and controls.
      • Anagnostopoulos A.
      • Ledergerber B.
      • Kuster S.P.
      • Scherrer A.U.
      • Naf B.
      • Greiner M.A.
      • et al.
      Inadequate perioperative prophylaxis and postsurgical complications after graft implantation are important risk factors for subsequent vascular graft infections - prospective results from the VASGRA Cohort Study.
      The long follow up, together with the high number of CEPET/CECT scans, may have contributed to the good agreement among the VASGRA team members on VGEI diagnosis (99%). However, the potential role of SPECT could not be assessed as this is not done routinely at the authors’ institution in patients with suspected VGEI. Secondly, a prospective unbiased validation of VGEI criteria is difficult as the evaluation is time sensitive and directly influences the decision process and the need for surgery. Moreover, even with a prospective validation, some criteria like “pus (definite by microscopy) around graft or aneurysm sac at surgery”, “microorganism recovered from explanted graft”, “microorganism recovered from an intra-operative specimen” or “microorganism from a percutaneous aspirate or perigraft fluid” are difficult to assess. Thirdly, a statement on whether using the MAGIC criteria would allow an earlier diagnosis in real life cannot be made. A retrospective review might lead to an overestimation of VGEI, as the first symptoms of the initial illness might already be interpreted as VGEI symptoms. Further, owing to the retrospective nature of the review, the evaluators were not blinded to results of alternative tests and reference standards. This was accounted for by the involvement of different medical specialties in the consensus review process. According to the Bayes theorem not only the accuracy, but also the pre-test probability of disease in a patient population affects the utility of diagnostic criteria.
      • Schunemann H.J.
      • Oxman A.D.
      • Brozek J.
      • Glasziou P.
      • Jaeschke R.
      • Vist G.E.
      • et al.
      Grading quality of evidence and strength of recommendations for diagnostic tests and strategies.
      To overcome this problem, one third of the study population included control patients undergoing routine vascular surgery. However, owing to the stringent criteria, controls could not be identified for all cases. Given the small number of patients, the subgroup analyses on anatomical location and graft material will have to be assessed in prospective multicentre studies.
      The MAGIC criteria may be used for a future VGEI case definition. As the MAGIC criteria put some rejected VGEI cases into the suspected or definite category, further modifications are suggested. Owing to the complexity of the disease and the heterogeneity of the affected population, criteria alone will still never suffice. Therefore, multidisciplinary management is a prerequisite for decision making in patients with VGEI.

      Conflicts of interest

      None.

      Funding

      This study was financed within the framework of the Vascular Graft Cohort Study (VASGRA), supported by the Swiss National Science Foundation (SNF) grant # 32473B_163132/1 , # 32473B_163132/2 and 320030_184918/1 (to B.H.). This work was also supported by the Clinical Research Priority Program (CRPP) of the University of Zurich for the CRPP “Precision Medicine for Bacterial Infections”. A.A. is supported by the academic career programme “Filling the Gap” of the Medical Faculty of the University of Zurich. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

      Acknowledgements

      The authors thank the participants of the Vascular Graft Cohort Study. The authors thank Nadia Eberhard, Ewerton Marques Maggio, Dieter Mayer, Rainer Weber, and Alexander Zimmermann for their contributions and critical review of the manuscript. The authors also thank the study nurses Caroline Mueller and Simone Buergin for their excellent work, and Christine Laich and Christine Voegtli for their administrative assistance. The members of the VASGRA Cohort Study are (in alphabetical order): Barbara Hasse (Principle Investigator), Nadia Eberhard, Michael Hoffmann, Lars Husmann, Reinhard Kopp, Bruno Ledergerber, Zoran Rancic, Carlos A. Mestres, Reinhard Zbinden, Annelies S. Zinkernagel, and Alexander Zimmermann.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Hasse B.
        • Husmann L.
        • Zinkernagel A.
        • Weber R.
        • Lachat M.
        • Mayer D.
        Vascular graft infections.
        Swiss Med Wkly. 2013; 143: w13754
        • Chakfe N.
        • Diener H.
        • Lejay A.
        • Assadian O.
        • Berard X.
        • Caillon J.
        • et al.
        Editor’s Choice – European Society for Vascular Surgery (ESVS) 2020 Clinical Practice Guidelines on the Management of Vascular Graft and Endograft Infections.
        Eur J Vasc Endovasc Surg. 2020; 59: 339-384
        • FitzGerald S.F.
        • Kelly C.
        • Humphreys H.
        Diagnosis and treatment of prosthetic aortic graft infections: confusion and inconsistency in the absence of evidence or consensus.
        J Antimicrob Chemother. 2005; 56: 996-999
        • Li J.S.
        • Sexton D.J.
        • Mick N.
        • Nettles R.
        • Fowler Jr., V.G.
        • Ryan T.
        • et al.
        Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.
        Clin Infect Dis. 2000; 30: 633-638
        • Lyons O.T.
        • Baguneid M.
        • Barwick T.D.
        • Bell R.E.
        • Foster N.
        • Homer-Vanniasinkam S.
        • et al.
        Diagnosis of aortic graft infection: a case definition by the Management of Aortic Graft Infection Collaboration (MAGIC).
        Eur J Vasc Endovasc Surg. 2016; 52: 758-763
        • Legout L.
        • Sarraz-Bournet B.
        • D’Elia P.V.
        • Devos P.
        • Pasquet A.
        • Caillaux M.
        • et al.
        Characteristics and prognosis in patients with prosthetic vascular graft infection: a prospective observational cohort study.
        Clin Microbiol Infect. 2012; 18: 352-358
        • Wilson W.R.
        • Bower T.C.
        • Creager M.A.
        • Amin-Hanjani S.
        • O’Gara P.T.
        • Lockhart P.B.
        • et al.
        Vascular graft infections, mycotic aneurysms, and endovascular infections: a scientific statement from the American Heart Association.
        Circulation. 2016; 134: e412-e460
        • Sah B.R.
        • Husmann L.
        • Mayer D.
        • Scherrer A.
        • Rancic Z.
        • Puippe G.
        • et al.
        Diagnostic performance of 18F-FDG-PET/CT in vascular graft infections.
        Eur J Vasc Endovasc Surg. 2015; 49: 455-464
        • Husmann L.
        • Huellner M.W.
        • Ledergerber B.
        • Anagnostopoulos A.
        • Stolzmann P.
        • Sah B.R.
        • et al.
        Comparing diagnostic accuracy of (18)F-FDG-PET/CT, contrast enhanced CT and combined imaging in patients with suspected vascular graft infections.
        Eur J Nucl Med Mol Imaging. 2019; 46: 1359-1368
        • Kim S.J.
        • Lee S.W.
        • Jeong S.Y.
        • Pak K.
        • Kim K.
        A systematic review and meta-analysis of (18)F-fluorodeoxyglucose positron emission tomography or positron emission tomography/computed tomography for detection of infected prosthetic vascular grafts.
        J Vasc Surg. 2019; 70: 307-313
        • Reinders Folmer E.I.
        • Von Meijenfeldt G.C.I.
        • Van der Laan M.J.
        • Glaudemans A.
        • Slart R.
        • Saleem B.R.
        • et al.
        Diagnostic imaging in vascular graft infection: a systematic review and meta-analysis.
        Eur J Vasc Endovasc Surg. 2018; 56: 719-729
        • Ajdler-Schaeffler E.
        • Scherrer A.U.
        • Keller P.M.
        • Anagnostopoulos A.
        • Hofmann M.
        • Rancic Z.
        • et al.
        Increased pathogen identification in vascular graft infections by the combined use of tissue cultures and 16S rRNA gene polymerase chain reaction.
        Front Med. 2018; 5: 169
        • Bisharat N.
        • Minuhin I.
        Prosthetic vascular graft infections between blood and concordance of graft culture pathogen.
        Am J Med Sci. 2012; 344: 431-435
        • Hagenaars J.C.
        • Wever P.C.
        • van Petersen A.S.
        • Lestrade P.J.
        • de Jager-Leclercq M.G.
        • Hermans M.H.
        • et al.
        Estimated prevalence of chronic Q fever among Coxiella burnetii seropositive patients with an abdominal aortic/iliac aneurysm or aorto-iliac reconstruction after a large Dutch Q fever outbreak.
        J Infect. 2014; 69: 154-160
        • Million M.
        • Raoult D.
        Recent advances in the study of Q fever epidemiology, diagnosis and management.
        J Infect. 2015; 71: 2-9
        • Mehta R.I.
        • Mukherjee A.K.
        • Patterson T.D.
        • Fishbein M.C.
        Pathology of explanted polytetrafluoroethylene vascular grafts.
        Cardiovasc Pathol. 2011; 20: 213-221
        • Van Damme H.
        • Deprez M.
        • Creemers E.
        • Limet R.
        Intrinsic structural failure of polyester (Dacron) vascular grafts. A general review.
        Acta Chir Belg. 2005; 105: 249-255
        • Samson R.H.
        • Veith F.J.
        • Janko G.S.
        • Gupta S.K.
        • Scher L.A.
        A modified classification and approach to the management of infections involving peripheral arterial prosthetic grafts.
        J Vasc Surg. 1988; 8: 147-153
        • Durack D.T.
        • Lukes A.S.
        • Bright D.K.
        New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Endocarditis Service.
        Am J Med. 1994; 96: 200-209
        • Hoen B.
        • Beguinot I.
        • Rabaud C.
        • Jaussaud R.
        • Selton-Suty C.
        • May T.
        • et al.
        The Duke criteria for diagnosing infective endocarditis are specific: analysis of 100 patients with acute fever or fever of unknown origin.
        Clin Infect Dis. 1996; 23: 298-302
        • Habib G.
        • Lancellotti P.
        • Antunes M.J.
        • Bongiorni M.G.
        • Casalta J.P.
        • Del Zotti F.
        • et al.
        2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM).
        Eur Heart J. 2015; 36: 3075-3128
        • Anagnostopoulos A.
        • Ledergerber B.
        • Kuster S.P.
        • Scherrer A.U.
        • Naf B.
        • Greiner M.A.
        • et al.
        Inadequate perioperative prophylaxis and postsurgical complications after graft implantation are important risk factors for subsequent vascular graft infections - prospective results from the VASGRA Cohort Study.
        Clin Infect Dis. 2019; 69: 621-630
        • Schunemann H.J.
        • Oxman A.D.
        • Brozek J.
        • Glasziou P.
        • Jaeschke R.
        • Vist G.E.
        • et al.
        Grading quality of evidence and strength of recommendations for diagnostic tests and strategies.
        BMJ. 2008; 336: 1106-1110

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.