DVT Prophylaxis in Neurosurgery - Medical School · DVT and PE may be undetected In some studies as...
Transcript of DVT Prophylaxis in Neurosurgery - Medical School · DVT and PE may be undetected In some studies as...
DVT Prophylaxis in
Neurosurgery
“Seek simplicity and distrust it”
Alfred North Whitehead
54
VTE Prophylaxis in Critically Ill Patients with:
Intracranial Hemorrhage
Aneurysmal Subarachnoid Hemorrhage
Traumatic Brain Injury (TBI)
Brain Tumors
Spinal Cord Injury
Recommendations for Prevention of VTE
in Patients Undergoing Neurosurgical and
Neurovascular Interventions in:
Elective Spine Surgery
Complicated Spinal Surgery
Elective Craniotomy
Elective Intracranial/Intra-arterial
Procedures
Intracranial Endovascular Procedures
DVT, PE or VTE?
DVT is the most common, but is not
always serious
PE is the most serious, but is much less
common
If PE were always preceded by DVT, the
DVT would be a reasonable surrogate
measure, but it is not
Therefore most researchers use VTE as
the measure of choice for analysis
Is VTE a Reasonable Choice ?
It lumps minor superficial DVT with death
due to massive pulmonary embolism
DVT may be independent of PE, precede
PE, not be identified in patients with PE,
be disabling without leading to PE
DVT and PE may be undetected
In some studies as many as 50% of PE
have no associated DVT
Therefore, VTE probably
overestimates the risk of DVT and/or
PE with serious consequences
Soooo Much Data
By my count, there at least 20
randomized trials involving more than
8,000 patients assessing some
component of VTE prophylaxis
predominantly in intracranial surgery
There are also at least 4 systematic
reviews with meta analysis of the
subject(s)
REF 18
RCT
refs refer to "Beyond…" paper
Ns
Pooled N
4UFH vs LMWH
23, 24 247 2Craniotomy23UFH + ICD 75LMWH + ICD 75 24UFH + ICD 48LMWH + ICD 49
25, 26 379 2Spinal surgery25200UFH + ICD LMWH + ICD 26UFH + CS 92LMWH + CS 87
19, 20 792 2CS to LMWH Mostly craniotomy19LMWH + CS 153placebo + CS 154 20LMWH + CS 241placebo + CS 244
5, 27 188 2ICD to LMWH Mostly craniotomy27ICD 60LMWH 60 5ICD 22LMWH 23ICD+LMWH 23
21, 28 203 3UFH to placebo 2Craniotomy21UFH 50placebo 50 28UFH 55placebo 48
29 50 1Spinal surgery29N=50 UFH placebo
30-32 309 4CS to ICD 3Craniotomy Can’t confirm 3rd ref Wautrecht
30N=70 ICD CS 31CS + ICD 78CS 80placebo 81
33 117 1Spinal surgery33ICD 57CS 60
34 95 1ICD to placebo Mixed craniotomy and spine
34N=95 ICD placebo
35
104 1ICD to electrical stim Mixed craniotomy and spine 35UFH 58
Calf stim +
dextran 46
Used as cohort
36 134 1Thigh ICD to foot ICD Spinal surgery Used as cohort36
ICD foot75ICD thigh 59
12Cohort
37, 38 247 7Pharmacologic 2LMWH alone Craniotomy37
LMWH150 38
LMWH97
39 2823 1LMWH and CS Craniotomy39LMWH + CS 2823
40 150 1UFH alone Craniotomy40
UFH150
41-43
1116 3UFH and ICD
Craniotomy43UFH + ICD 872 42UFH + ICD 106
Frim is used in both RCT and
cohort
41
UFH + ICD
138ICD 473
44-46 370 5Mechanical 3ICD alone Spinal surgery 44ICD 31 45ICD 200 46 139
47, 48 392 2CS and ICD Spinal Surgery 47ICD + CS 75 48ICD + CS 317
7716
Meta AnalysisShould simplify by pooling data
Cannot make all the included trials have
the same:
Eligibility and exclusion criteria
Choice of prophylactic method
Choice of outcome measure
and yet, pooling the data assumes that all
of these things are so similar that the
differences among the studies don’t matter
Khan, et al. J Neurosurg 129:906–915, 2018
Assumptions Underlying Meta Analyses
The risk of VTE and Intracranial
Hemorrhage is the same in cranial and
spinal patients
Comparing Meta Analyses
Author Number of studies RCT Cohort Spine Included DVT PE VTE
Iorio 4 no yes no yes
Collen 12 RCT, 18 cohort yes yes yes no
Hamilton 6 no no no yes
Khan 9 yes yes no no
Outcomes
Separately
Analyzed
Author Number of studies RCT Cohort Spine Included DVT PE VTE
Iorio 4 no yes no yes
Collen 12 RCT, 18 cohort yes yes yes no
Hamilton 6 no no no yes
Khan 9 yes yes no no
Outcomes
Separately
Analyzed
Assumptions Underlying Meta Analyses
Is the risk of VTE and Intracranial
Hemorrhage is the same in cranial and
spinal patients ?
Is it OK to include non-randomized cohorts ?
Is it OK to assume that DVT and VTE
outcomes are similar enough to pool?
Comparing Meta Analyses
All methods used inconsistently across studies
Author venography echo fibrinogen angio CT V/Q
Iorio
Collen
Hamilton
Khan
Methods of Diagnosis for VTE
Assumptions Underlying Meta Analyses
Is the risk of VTE and Intracranial
Hemorrhage is the same in cranial and
spinal patients ?
Is it OK to include non-randomized cohorts ?
Is it OK to assume that DVT and VTE
outcomes are similar enough to pool?
Is it OK to assume that using different
diagnostic methods doesn’t affect outcome?
Comparing Meta Analyses
All methods used inconsistently across studies
Author placebo
compression
stockings
intermittent
compression
devices
unfractionated
heparin enoxaparin nadroparin
Iorio
Collen
Hamilton
Khan
Methods of Prophylaxis
Assumptions Underlying Meta Analyses
Is it OK to assume that using different
methods and combinations of prophylaxis
doesn’t affect outcome?
In addition, many studies mix populations of
patients known to have different risk of VTE
(brain tumor patients subarachnoid
hemorrhage patients and other elective
intracranial surgery patients)
Well, let’s just assume that all those assumptions are OK
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
Iorio Collen Hamilton Khan
Odds Ratio for DVT
lower 95% ci
Odds Ratio
upper 95% ci
Odds Ratio for DVT
Iorio Collen Hamilton Khan0.00
0.90
0.50
Odds Ratio for ICH or Maj. Hem.
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Iorio Collen Hamilton Khan
lower 95% ci
Odds Ratio
upper 95% ci
Iorio Collen Hamilton Khan0.00
3.00
1.00
2.00
4.00
NNT
Author
Number of
studiesnon-pharm
prophylaxis
chemo-
prophylaxis
absolute
risk
reduction
Number
Needed
to Treat
Agnelli 4 RCT 0.290 0.161 0.129 7.74
Collen 12 RCT, 18 cohort 0.061 0.033 0.029 34.59
Hamilton 6 RCT 0.260 0.148 0.112 8.93
Khan 9 RCT 0.215 0.126 0.089 11.24
proportion with VTE
Another View
NNH
Author
Number of
studiesnon-pharm
prophylaxis
chemo-
prophylaxis
absolute
risk
increase
Number
Needed to
Harm
Agnelli 4 RCT 0.0137 0.0235 0.0098 102.20
Collen 12 RCT, 18 cohort 0.0630 0.1244 0.0888 11.26
Hamilton 6 RCT 0.0400 0.0500 0.0100 100.00
Khan 9 RCT 0.0240 0.0330 0.0090 111.11
proportion with ICH or
major hemorrhage
Another View
NNT NNH
Author
absolute risk
reduction
Number
Needed to
Treat
absolute
risk increase
Number
Needed to
Harm
Collen 0.03 35 0.09 11
Iorio 0.13 8 0.01 102
Hamilton 0.11 9 0.01 100
Khan 0.09 11 0.01 111
trying to simplify the comparison
Agnelli
Nurmohamed
Cerrato
Collen
Hamilton
KhanIorio
Constantini
Goldhauber
MacDonald
Bostrom
Bucci
Nelson
Prestar
Skillman
Turpie
Voth
Wautrecht
Wood
Dickinson
MelonGruber
Kurtoglu
Hamidi
Conclusions
Meta-analysis does not fix flaws in the
included studies
Different approaches to inclusion and
exclusion lead to wildly different results
It isn’t clear which is right
More than 3 decades of uncoordinated
underpowered randomized trials have
failed to provide clear guidance on this
issue