STICH II

Title: Early Surgery Versus Initial Conservative Treatment in Patients with Spontaneous Lobar Intracerebral Hematomas (STICH II): a randomized trial

Authors: Mendelow et al
Lancet 2013

Background/Importance: Undetermined benefit of early surgical intervention for intracerebral hematoma

Trial design: international multicenter randomized trial
Duration: 67 months
Objectives: Determine if early surgical intervention for intracerebral hemorrhage improves outcomes
Primary Outcome/Endpoint: prognosis based favorable or unfavorable outcome at 6 months (what?)

Prognosis dichotomized based on equation:
P= 10 x GCS – age – 0.64(volume)
P>27.672 = good prognosis
P<27.672 = poor prognosis
No explanation in main text on how formula derived. Looks similar to ICH score in that it incorporates GCS, age, and volume. However this equation for prognosis further stratifies volume as opposed to a cutoff of 30cc in ICH score. Also the equation does not incorporate infratentorial or intraventricular status however these are not included in the study due to inclusion and exclusion criteria.

Outcome
Based on Extended Glasgow Outcome Scale (GOSE)
GOSE score
1 = dead
2 = vegetative state (absence of awareness of self or environment)
3 = lower severe disability (needs assistance with all ADLs)
4 = upper severe disability (needs partial assistance with ADLs)
5 = lower moderate disability (can perform ADLs but cannot resume work/school)
6 = upper moderate disbaility (some disability but able to resume partial work/previous activities)
7 = lower good recovery (minor physical/mental deficits affecting daily life)
8 = upper good recovery (full recovery or minor symptoms that do not affect daily life)

Prognosis based outcome:
Good prognosis group (P>27.672)
GOSE 5-8 = favorable
GOSE 1-4 = unfavorable

Poor prognosis group (P<27.672)
GOSE 4-8 = favorable
GOSE 1-3 = unfavorable

Inclusion criteria:
-spontaneous ICH
-superficial </= 1cm from cortical surface
-volume = 10-100cc
-<48 from ictus
-motor score on GCS greater than or equal to 5
-eye opening score on GCS greater than or equal to 2

Exclusion Criteria
-ICH associated with aneurysm, AVM, tumor, trauma
-location of bleed in basal ganglia, thalamus, cerebellum, brainstem
-intraventricular hemorrhage/extension
-pre existing conditions (not specified in main article)

Statistical Analysis: chi square test

Results:
n=601 (597 after removing protocol violations)
early surgical group n=305
initial conservative group n=292
62 (21%) crossover from initial medical management/conservative to surgery
Favorable outcome (based on intention to treat):
early surgery 123 (41%)
conservative 108 (38%)
OR = 0.86[95%: 0.62-1.2]p=0.357

Authors’ conclusion
“early surgery does not increase rate of death or disability at 6 months”

Limitations/weaknesses:
21% crossover from conservative to surgical group (may have artificially improved conservative group’s outcomes)
Strengths:
randomized controlled trial

Discussion:
This study population was specific to patients who could follow commands or were localizing and could at least open eyes to noxious stimulus. On the spectrum of GCS this is the better end were some of the patient would be able to speak and give their own consents. Therefore conclusions would be limited to those who many physicians might be undecided on whether to take to surgery or not. Therefore this study is an attempt to clarify/guide physicians in this gray area. The results do not reach statistical significance to guide decision making one way or another. However the wording of authors “early surgery does not increase rate of death or disability at 6 months” seems to show a surgical bias. Based on the same data another could write initial conservative management does not increase the rate of death or disability at 6 months. The important word in previous statement is initial, since the study had a 21% crossover rate. Which means if the patient clinically declines and is taken to surgery later, then delayed surgery due to decline when grouped with initial conservative management did not result in statistically significant differences in outcome at 6 months. Interestingly upon discussion with colleagues the same data led to different utilization in that some use this study to justify early surgery while other use this study to justify initial conservative management. As a resident I enjoy surgery, but with an unclear benefit I think when I get to make the decisions on my own I would give the patient a chance to whether the storm and ask the patient and family time to consider goals of care reserving surgery for clinical decline after patient and family are clearer on what they want and what would be an acceptable quality of life.