Low Molecular-weight Heparin for the Treatment of Acute Ischemic Stroke
Clinical Question
Is low-molecular weight heparin efficacious in reducing morbidity and
mortality in patients with acute ischemic strokes?
Clinical Bottom Line
- There is a small, but statistically significant, decrease in the risk of
death at 3 and 6 months if treatment with LMWH is used for ischemic stroke
started within the first 48 hours.
- There is a larger, statistically significant decrease in the degree of
physical dependency and increase in the degree of physical independence at 3 and
6 months.
The Evidence
Risk of Death, Relative Risk Reduction, and Number Needed to Treat at 3
months follow-up
|
Risk of Death |
RRR |
ARR |
NNT |
| Placebo |
14.3% |
|
|
|
| Low-dose LMWH |
14.9% |
-0.06 |
-0.006 |
NA |
| High-dose LMWH |
12.0% |
0.16 |
0.023 |
43 |
Risk of Death, Relative Risk Reduction, and Number Needed to Treat at 6
months follow-up
|
Risk of Death |
RRR |
ARR |
NNT |
| Placebo |
19% |
|
|
|
| Low-dose LMWH |
16.8% |
0.12 |
0.022 |
45 |
| High-dose LMWH |
13.0% |
0.32 |
0.06 |
17 |
Total Risk of Hemorrhagic Transformation (on follow-up CT)
|
Hemorrhage Risk |
Risk Difference |
Number needed to Treat |
| Placebo |
12.0% |
|
|
| Low-dose LMWH |
9.0% |
-0.03 |
33 |
| High-dose LMWH |
6.0% |
-0.06 |
17 |
Poor versus Good Outcomes at 3 and 6 months
| 3 months |
Good Outcome (%) |
Poor Outcome (%) |
RR (95%CI) for Poor Outcome |
| Placebo |
36 |
64 |
|
| Low-dose |
40 |
60 |
0.95 (0.76-1.17) |
| High-dose |
47 |
53 |
0.83 (0.66-1.05) |
| 6 months
|
|
|
|
| Placebo |
35 |
65 |
|
| Low-dose |
48 |
52 |
0.81 (0.64-1.02) |
| High-dose |
55 |
45 |
0.69 (0.54-0.90) |
Comments
- This was a well-designed study and is valid according to the JAMA
guidelines for articles on therapy.
- Data analysis was fairly rigorous.
- The results and conclusions of this study may not be applicable to all
patients as the study patients were not representative of the population in
general (they were all Chinese).
- Although the authors internally validate their assessment method, and
compare it to the Barthel Index (calculating their sensitivity and specificity),
this was not a diagnostic study. The mode of assessment is subject to both
patient/caregiver recall bias and interviewer bias.
- The only group that shows a benefit that is statistically powerful is the
group treated with high-dose LMWH at 6 months; this groups RR was 0.69, and the
CI excluded a RR of 1. An ARR of 0.10 is calculated for this group with a NNT
of 10 to prevent one poor outcome. However, this is only true if you combine,
as the authors do, alive (but dependent) patients and death. Excluding
dependency, and looking only at risk of death, 43 patients need to be treated
in the high dose group to prevent one death at 3 months and 17 to prevent one
death at 6 months (45 and 33, respectively for the low dose group at 3 and 6
months). The risk of hemorrhagic transformation (in terms of number needed to
harm) is far from substantial. In fact, these results indicate a protective
effect on hemorrhagic transformation -- 17 for the high-dose group and 33 for
the low dose group need to be treated to prevent a hemorrhagic transformation.
Taking this into consideration increases the overall qualitative treatment
effect.
- Over 2,400 patients were excluded from this study for various reasons,
leaving just over 10% included in the study. This raises concern about the
clinical significance of this study and the application of its results.
- No information is given about whether all three groups had equal
percentages of patients on secondary prophylaxis (aspirin).
- Despite the statistically significant decrease in poor outcomes reported by
the authors, this reviewer cannot recommend this therapy as routine treatment
for all patients with acute ischemic stroke in light of the above comments.
Appraised by: Patrick J. Sousa, MD
DATE: April 5, 1996
Kay R, et al. Low-molecular Weight Heparin for the Treatment of
Acute Ischemic Stroke. NEJM 1995;333:1588-1593.
