Aspirin has clinically significant benefit in high risk groups - Summary NNT can mislead clinicians
Editor – The systematic review(1;2) of antiplatelet drugs for
prevention of pre-eclampsia found statistically significant reduction in
pre-eclampsia and other outcomes such as fetal or neonatal death. The
authors concluded that the benefit was ‘small to moderate’ and the
implication for practice was that ‘relatively large numbers of women will
need to be treated to prevent a single adverse outcome’. With the numbers
of women needed to be treated to prevent one case of pre-eclampsia
reported as 100 (95% CI 59 to 167), clinicians (and women) might not think
treatment worthwhile.
However, calculating numbers needed to treat from pooled meta-
analysis data may be inappropriate, if it is possible to identify
subgroups of patients with substantially differing baseline risks(3). In
women with high levels of baseline risk, and assuming constant relative
risk from treatment, numbers needed to treat are smaller(4), and both
clinicians and women may be much more likely to wish to use aspirin to
prevent pre-eclampsia. It has been suggested(1;2) that meta-analysis of
individual patient data would be useful both in identifying high-risk
subgroups, and estimating the benefit they derive from antiplatelet
treatment. However, such meta-analyses generally take a long time to
complete(5). What should clinicians do in the mean time?
We can see no reason for thinking that the reduction in relative risk
for various risk levels will be substantially different. If high-risk (or
low risk) women can be identified, by any means, specific numbers needed
to treat can then be generated by using pooled relative risk estimates
from reviews of effectiveness(4), making the decision to treat (or not)
more appropriate and, in this particular case, probably more clear-cut for
most women.
We systematically reviewed the accuracy of uterine artery Doppler in
early pregnancy for predicting pre-eclampsia(6). In clinically high-risk
women, a positive Doppler result (abnormal flow velocimetry ratio or the
presence diastolic notch) meant a 23.5% (95% CI 18.6 to 29.2) risk of
developing pre-eclampsia. With baseline risk elevated to this level and
assuming the global estimated relative risk of 0.85(1), we estimate that
31 (95% CI 18 to 55) patients will be needed to be treated with aspirin to
prevent one case of pre-eclampsia. We would thus expect most women with
abnormal uterine artery Dopplers, when advised by their clinicians, to
request antiplatelet treatment.
A Coomarasamy
Research Fellow in Obstetrics
Education Resource Centre,
Birmingham Women's Hospital,
Metchley Park Road, Birmingham B15 2TG
Department of Public Health & Epidemiology,
Public Health Building,
University of Birmingham, B15 2TT
1. Duley L, Henderson-Smart DJ, Knight M, King JF. Anteplatelet
drugs for prevention of pre-eclampsia and its consequences: systematic
review. BMJ 2001;322: 329-333..
2. Knight M, Duley L, Henderson-Smart DJ, King JF. Antiplatelet
agents for preventing and treating pre-eclampsia. Cochrane
Database.Syst.Rev. 2000;CD000492.
3. Smeeth L, Haines A, Ebrahim S. Numbers needed to treat derived
from meta-analyses--sometimes informative, usually misleading. BMJ
1999;318:1548-51.
4. Guyatt GH, Sackett DL, Sinclair JC, Hayward R, Cook DJ, Cook RJ.
Users' guides to the medical literature. IX. A method for grading health
care recommendations. Evidence-Based Medicine Working Group. JAMA
1995;274:1800-4.
5. Stewart LA,.Clarke MJ. Practical methodology of meta-analyses
(overviews) using updated individual patient data. Cochrane Working Group.
Stat.Med. 1995;14:2057-79.
6. Chien PF, Arnott N, Gordon A, Owen P, Khan KS. How useful is
uterine artery Doppler flow velocimetry in the prediction of pre-
eclampsia, intrauterine growth retardation and perinatal death? An
overview. BJOG. 2000;107:196-208.
Rapid Response:
Aspirin has clinically significant benefit in high risk groups - Summary NNT can mislead clinicians
Editor – The systematic review(1;2) of antiplatelet drugs for
prevention of pre-eclampsia found statistically significant reduction in
pre-eclampsia and other outcomes such as fetal or neonatal death. The
authors concluded that the benefit was ‘small to moderate’ and the
implication for practice was that ‘relatively large numbers of women will
need to be treated to prevent a single adverse outcome’. With the numbers
of women needed to be treated to prevent one case of pre-eclampsia
reported as 100 (95% CI 59 to 167), clinicians (and women) might not think
treatment worthwhile.
However, calculating numbers needed to treat from pooled meta-
analysis data may be inappropriate, if it is possible to identify
subgroups of patients with substantially differing baseline risks(3). In
women with high levels of baseline risk, and assuming constant relative
risk from treatment, numbers needed to treat are smaller(4), and both
clinicians and women may be much more likely to wish to use aspirin to
prevent pre-eclampsia. It has been suggested(1;2) that meta-analysis of
individual patient data would be useful both in identifying high-risk
subgroups, and estimating the benefit they derive from antiplatelet
treatment. However, such meta-analyses generally take a long time to
complete(5). What should clinicians do in the mean time?
We can see no reason for thinking that the reduction in relative risk
for various risk levels will be substantially different. If high-risk (or
low risk) women can be identified, by any means, specific numbers needed
to treat can then be generated by using pooled relative risk estimates
from reviews of effectiveness(4), making the decision to treat (or not)
more appropriate and, in this particular case, probably more clear-cut for
most women.
We systematically reviewed the accuracy of uterine artery Doppler in
early pregnancy for predicting pre-eclampsia(6). In clinically high-risk
women, a positive Doppler result (abnormal flow velocimetry ratio or the
presence diastolic notch) meant a 23.5% (95% CI 18.6 to 29.2) risk of
developing pre-eclampsia. With baseline risk elevated to this level and
assuming the global estimated relative risk of 0.85(1), we estimate that
31 (95% CI 18 to 55) patients will be needed to be treated with aspirin to
prevent one case of pre-eclampsia. We would thus expect most women with
abnormal uterine artery Dopplers, when advised by their clinicians, to
request antiplatelet treatment.
A Coomarasamy
Research Fellow in Obstetrics
Education Resource Centre,
Birmingham Women's Hospital,
Metchley Park Road, Birmingham B15 2TG
arricoomar@hotmail.com
Harry Gee
Consultant Obstetrician
Birmingham Women's Hospital, B15 2TG
Khalid S Khan
Consultant Obstetrician and Gynaecologist
Birmingham Women's Hospital, B15 2TG
David Braunholtz
Senior Research Fellow
Department of Public Health & Epidemiology,
Public Health Building,
University of Birmingham, B15 2TT
1. Duley L, Henderson-Smart DJ, Knight M, King JF. Anteplatelet
drugs for prevention of pre-eclampsia and its consequences: systematic
review. BMJ 2001;322: 329-333..
2. Knight M, Duley L, Henderson-Smart DJ, King JF. Antiplatelet
agents for preventing and treating pre-eclampsia. Cochrane
Database.Syst.Rev. 2000;CD000492.
3. Smeeth L, Haines A, Ebrahim S. Numbers needed to treat derived
from meta-analyses--sometimes informative, usually misleading. BMJ
1999;318:1548-51.
4. Guyatt GH, Sackett DL, Sinclair JC, Hayward R, Cook DJ, Cook RJ.
Users' guides to the medical literature. IX. A method for grading health
care recommendations. Evidence-Based Medicine Working Group. JAMA
1995;274:1800-4.
5. Stewart LA,.Clarke MJ. Practical methodology of meta-analyses
(overviews) using updated individual patient data. Cochrane Working Group.
Stat.Med. 1995;14:2057-79.
6. Chien PF, Arnott N, Gordon A, Owen P, Khan KS. How useful is
uterine artery Doppler flow velocimetry in the prediction of pre-
eclampsia, intrauterine growth retardation and perinatal death? An
overview. BJOG. 2000;107:196-208.
Competing interests: No competing interests