Systematic review and/or Meta analysis

Background
Part I:
General introduction
Initial evidence base search
Part II
Guideline review
Part III
Systematic review
Evidence base search
 Systematic review and/or Meta analysis
 Vascular access population
 Preop, postop, surveillance
 Type of access – AVF, AVG, PD, CVC
 Care/ management
 Imaging, intervention, medication
 Guidelines
Methodology
 Search - MEDLINE, PubMed, Cochrane Database of Systematic Reviews
(CDSR), Database of Abstracts of Reviews of Effectiveness (DARE),
PROSPERO (CRD University of York) , Science direct
 Relevant key words
 Limited to English*
 Excluding grey literature
Results
 Up to February 2014
 20 systematic review and/meta-analysis
Topic
Systematic reviews /
Meta analysis
Management / Outcomes
6
Surveillance / Patency /
Followup
4
Technique / imaging
7
Medication
3
ERBP Working Groups
Topic
Nr
ERBP
Group
Central venous
catheter
3
1
Preoperative
2
2
Perioperative
5
3
Follow up
10
4
Follow up /
Perioperative
3
3&4
Part I:
General introduction
Initial evidence base search
Part II
Guideline review
Part III
Systematic review
AGREE II: Guideline Appraisal
Role of Imaging in Vascular Access
 Key areas for imaging in vascular access (VA):
1.
2.
3.
4.
Preoperative evaluation
Surveillance of VA
Diagnosis and treatment of stenosis and thrombosis
Central venous access
 What do the clinical practice guidelines recommend?
 Compare the studies used to make the recommendation
 Variation/similarities between different guidelines
Summary of Pre-op Guidelines
Criteria
Guideline recommendations
Venography (suspected cases of CVS)
EBPG
¤
¤
¤
KDOQI
¤
¤
¤
Magnetic resonance angiography (MRA)
¤
¤
Physical examination
Pre-Op assessment
Other (Suggestions to
care)
Ultrasound mapping
CARI
¤
¤
CNS
¤
Total
4
3
2
2
Preliminary analysis: Pre-op studies
 Pubmed 312 articles,
 after exclusion = 56* (30% )
 36 US, 12 Venography, 5 MRA and 1 CTA
 Only 3 RCT’s
 Reminder lower level evidence
Preliminary analysis: Pre-op studies
 Consensus with 3 RCT in guidelines
 Disagreement between lower levels of evidence studies
Quantitative Appraisal of the Guidelines
 The Appraisal of Guidelines for REsearch & Evaluation
(AGREE) Instrument II
“a framework for assessing variability & quality”
Ref: Brouwers M C., et al 2010, CMAJ 10.1503,
http://www.agreetrust.org
13
 AGREE II consists of 23 key questions organised within
6 domains followed and overall assessment
Domain 1. Scope and Purpose (items 1-3)
Domain 2. Stakeholder Involvement (items 4-6)
Domain 3. Rigor of Development (items 7-14)
Domain 4. Clarity of Presentation (items 15-17)
Domain 5. Applicability (items 18-21)
Domain 6. Editorial Independence (items 22-23)
Overall assessment
Methods
 Pilot study applied to international VA guidelines
 Four guidelines identified (EBPG, KDOQI, CARI, CNS)
 Three observers, categorical data scored (1-7)
 23 questions, under 7 topics (AGREE II)
 Overall scores
 Domain scores for each section
 Inter-observer repeatability
Results: Scaled Domain Scores
Scaled domain% = (obtained score-min possible score)/ (max possible score- min possible
score)
AGREE II
Domain
Section
ERBP
KDOQI
CNS
CARI
Domain 1
Scope and purpose
61%
81%
69%
52%
Domain 2
Stakeholder Involvement
56%
70%
43%
13%
Domain 3
Rigor of Development
68%
76%
60%
44%
Domain 4
Clarity of Presentation
98%
81%
72%
72%
Domain 5
Applicability
40%
74%
11%
38%
Domain 6
Editorial Independence
39%
89%
42%
78%
Domain 7
Overall assessment
56%
72%
50%
50%
Overall
Guideline ranking
59% (2nd)
78% (1st )
49% (4th)
50% (3rd)
*Analysis of variance: p=0.02
Results: Scaled Domain Scores
Scaled domain% = (obtained score-min possible score)/ (max possible score- min possible
score)
AGREE II
Domain
Section
ERBP
KDOQI
CNS
CARI
Domain 1
Scope and purpose
61%
81%
69%
52%
Domain 2
Stakeholder Involvement
56%
70%
43%
13%
Domain 3
Rigor of Development
68%
76%
60%
44%
Domain 4
Clarity of Presentation
98%
81%
72%
72%
Domain 5
Applicability
40%
74%
11%
38%
Domain 6
Editorial Independence
39%
89%
42%
78%
Domain 7
Overall assessment
56%
72%
50%
50%
Overall
Guideline ranking
59% (2nd)
78% (1st )
49% (4th)
50% (3rd)
*Analysis of variance: p=0.02
A G R E E II: F r e q u e n c y d i s t r ib u t io n ( A P 1 , A P 2 a n d A P 3 ) f o r e a c h q u e s t io n
EUAP1
EUAP2
6
EU
AP3
AP1
U AP2
SAP1
4
U AP3
SAP2
USAP3
2
C anAp1
C anAp2
C anAp3
3
2
O ZAP1
A
Z
O
O
Z
A
P
P
1
A
Z
O
n
a
C
G u id e lin e s
P
3
p
A
p
A
C
a
n
A
C
a
n
A
S
U
2
1
p
3
P
U
S
A
A
S
U
P
2
1
P
3
P
A
E
U
A
U
E
U
A
P
P
2
1
0
E
R a t in g ( M in 1 t o M a x 7 )
8
O ZAP2
O ZAP3
*Where error bars indicates median and
inter quartile ranges of distribution
Inter-operator agreement
Vascular Access
Guidelines
Average
Pairwise
Cohen’s
Kappa (κ)
Fleiss’
Kappa
Intra-class
Correlation
Coefficient
(ICC) (r)
Significance
for ICC
(p-value)
1
EU: ERPB
0.133
0.119
0.662
0.001*
2
US: KDOQI
0.013
0.636
0.002*
3
Can: CNS
0.067
0.016
0.708
0.000*
4
Oz: CARI
0.131
0.121
0.561
0.009*
0.084
Discussion
 Guideline methodology rigor was variable between the guidelines
(p<0.02)
 This may be due to different search strategies and different methods
of updating the guidelines
 Overall pair wise agreement and inter-operator agreement (Cohens’
kappa) was poor (κc: 0.3- 0.01)
 Weighted intraclass coefficient was significant (p<0.05) as it takes
into account for a consistently high/low observer
Conclusion
 Guidelines agrees with high evidence based studies
 Limitation: Not many RCT studies in VA
 Agree II was able to provide qualitative analysis
 Limitation: low agreement was found between
appraisers
 VA is an area that may need experts to grade guidelines
Methods
 CKD/ESRF - on or pre-HD patients
 All utilizing MRI
 Imaging in AV vascular access
Methods
 PubMed, Medline, Ovid SP, Science Direct,
PROSPERO, Cochrane
 Key words - Renal failure, Renal Disease,
Haemodialysis, vascular access, arteriorvenous
fistula, MRI, MRA, MRV
 Limited to English*
1996+ to 03/2014
Methods
Database screening
n = 598
Duplications removed
&
Exclusion criteria
applied
Prisma checklist*
Methods
Full text articles
reviewed
n = 37
Data synthesis
n = 10
(still to be
assessed n = 7)
Prisma checklist*
Software
 Reference manager – Mendeley
 MS Excel, MedCalc & MS Word
 RevMan
Publications included
n = 19
Nr. Analysed
n = 10
Author
Year
Type
Nr pts
AVF
AVG
Preop
Contrast Enhanced
Comparator
Not specified
intraoperative
Not specified
intraoperative
Waldman
1996
singe centre case series
13
8
7
Menegazzo
1998
single centre prospective case series
25
18
3
Laissy
1999
single centre prospective case series
19
17
2
Not specified
DSA
Cavagna
2000
single centre case series
13
13
0
gadodiamide, Omniscan, Nycomed, Oslo,
Norway)
DSA
Smits
2002
single centre prospective case series
18
8
14
gadopentetate
DSA
Planken
2003
single centre case series
15
7
8
gadolinium-DTPA ??
DSA
Han
2003
single centre prospective case series
10
2
8
gadoteridol (Prohance; Bracco, Milan, Italy)
DSA
Takahashi N
2004
single centre prospective case series
15
15
0
gadodiamide(Omniscan; Daiichi
Pharma,Tokyo,Japan)
DSA
Froger
2005
single centre prospective case series
48
51
0
gadoteridol (Prohance; Bracco, Milan, Italy)
DSA
Doelman
2005
single centre prospective case series
81
49
32
gadoteridol (Prohance; Bracco, Milan, Italy)
DSA
Zhang
2006
single retrospective centre case
series
9
8
1
gadopentetate (Magnevist, Schering)
DSA
Pinto
2006
single centre prospective case series
11
0
0
Gadolinium based
DSA
Duijm LE
2006
single centre prospective case series
101
66
35
gadoteridol (Prohance; Bracco, Milan, Italy)
DSA
Planken1
2008
2 large centre prospective study
73
49
24
gadopentetate (Magnevist, Schering)
Maturation
(along side US)
Planken
2008
case series/single centre
25
19
6
gadopentetate (Magnevist, Schering)
intraoperative
Sigovan,
2012
case series/single centre
10
10
0
ferumoxytol
Qualitative comparison
with NCE TOF
Bode
2012
case series/single centre
15
0
0
15
Gadovist, BayerSchering Pharma, Berlin,
Germany
compared to CE MR in
healthy
Gao
2012
case series/single centre
14
0
0
14
gadopentatate, Beilu Pharma,Beijing, China
DSA
Sigovan
2013
case series/single centre
3
3
0
NE
CFD analysis - fistula
maturation
3
Demographics
 518 Patients /
19 studies
 Mean
=
28 (SD 29 +/-13 CI 95% )
 F:M
=
1:1.4
 AVF
=
345
 AVG
=
140
 32 pre-op & 12 un-specified type
Imaging levels
Level of imaging Nr. of studies
Total Nr.
AVF
AVG
unspecified/ preop
Entire limb
9
167
126
46
4
Forearm
3
72
31
28
13
Arm
4
56
50
6
0
Limb & central
6
223
123
60
40
*518 patients , 527 segments
Characteristics
 1 Multi centre, 18 single centre case series
 10 prospective data collection
 14 studies - Blinded observers
Field type
Scanner make
TR/TE/FA
Voxel/FOV
0.5
Philips Gyroscan, Best, Netherlands
20/6.9,FA50
240 X 240 mm2
Menegazzo
1
Magnetom Siemens, Erlangen, Germany
27/9,FA60,
200mm
Laissy
1
Magnetom Siemens, Erlangen, Germany
3/10,FA60
200mm
Cavagna
0.5
Philips Gyroscan ACS-NT, Best, Netherlands
8.7/3.3, FA 60°
450 mm
Smits
1.5
Gyroscan; Philips Best, the Netherlands
16/9.5,7.5°
256
Planken
1.5
Philips Intera Best, Netherlands
5.4/1.6/40
3.1mm(3)
Han
1.5
Philips Gyroscan Intera, Best, Netherlands
4.4/1.5,20°
1.0x1.2x1.1 mm
Takahashi N 1
Signa Horizon, GE, Milwaukee, Wisc., USA
6.1-6.2/1.4FA30
24X14.4-30X18CM
Froger
1.5
Philips Gyroscan Intera, Best, Netherlands
4.1/1.3/20°
1.00x1.00x1.10 mm
Doelman
1.5
Philips Gyroscan Intera, Best, Netherlands
4.1/1.34/20,
1.00x1.00x1.10 mm
Zhang
1.5
Symphony/Avanto, Siemens, Erlangen,Germany
3.4/1.3; flip angle, 25°; 450/500mm
Pinto
1.5
Sonata/Avanto, Siemens, Erlangen,Germany
2.8/1.2,FA20
1.2 0.7 1.3 mm3
Duijm LE
1.5
Philips Gyroscan ACS-NT, Best, Netherlands
4.1/1.34/20
1.00x1.00x1.10 mm
Planken1
1.5
Philips Intera Best, Netherlands
4.4/2,FA50
voxel0.97x0.52x2
Planken
1.5
Philips Intera Best, Netherlands
4.4/2,FA50
voxel0.97x0.52x2
Sigovan,
1.5
Avanto; Siemens, Erlangen, Germany
30/7,FA25
0.25x0.25x0.6mm3
Bode
1.5
Philips Gyroscan Intera, Best, Netherlands
4.5/2.2/90
0.78x0.78x0.78
Gao
3
Magnetom Trio aTim Siemens, Munich, Germany
2.3/1/25
450mm
Sigovan
1.5
Siemens Avanto, Erlangen Germany
30/7/25°
0.25 × 0.25 × 0.6 mm3
Study
Waldman
Contrast Usage
1
Cavagna
0.1mmol/kg
gadodiamide, Omniscan, Nycomed, Oslo, Norway)
2
Smits
20-30mls
gadopentetate
3
Planken
35mls
gadolinium-DTPA
4
Han
39mls
gadoteridol (Prohance; Bracco, Milan, Italy)
5
Takahashi N
10mls
gadodiamide(Omniscan; Daiichi Pharma,Tokyo,Japan)
6
Froger
39mls
gadoteridol (Prohance; Bracco, Milan, Italy)
7
Doelman
39mls
gadoteridol (Prohance; Bracco, Milan, Italy)
8
Zhang
20mls
gadopentetate (Magnevist, Schering)
9
Pinto
20mls
Gadolinium based
10
Duijm LE
41mls
gadoteridol (Prohance; Bracco, Milan, Italy)
11
Planken1
45mls
gadopentetate (Magnevist, Schering)
12
Planken
45mls
gadopentetate (Magnevist, Schering)
13
Sigovan,
60mls
ferumoxytol
14
Bode
20mls/NCE
Gadovist, BayerSchering Pharma, Berlin, Germany
15
Gao
20-40mls (0.2mmol/kg)
gadopentatate, Beilu Pharma,Beijing, China
Contrast Usage
 Contrast used in 15 studies
 No reported adverse reactions
 11 series pre 2006 (NSF)
 Mean volume 34mls (SD 13, CI +/-7)
Results – analysed until present
 12 studies in 10 articles compared to DSA
 Data extractable & relevant
 Synthesizable for 2x2 table
Results
Results
Summary ROC Plot
Remaining
 Sub group analysis
 Completion of analysis & Weight summary
 Risk of Bias assessment
 Researcher 2 part deux