Atenolol after myocardial infarction - absence of evidence is a failure of research policy
As I read through Freemantle et al's excellent systematic review of
beta blockade after myocardial infarction 1, I looked particularly for
evidence about the effectiveness of atenolol. In my practice, atenolol is
our first choice of beta blocker for hypertension, angina and secondary
prevention because it is convenient (once daily dose), cheap, and
relatively free from side effects in most patients.
After the promising, if inconclusive, results of early short term
trials of atenolol (pooled OR 0.93, 95%CI 0.85 to 1.02) it was frustrating
to find that there were so few long term trials that the results (OR 1.02,
95%C 0.52 to 1.99) were meaningless. The wide confidence intervals of the
long term trials reflect the weight of 1.6% for long term trials of
atenolol compared to 74.2% for the short term trials, and the authors
rightly conclude that atenolol has been inadequately evaluated for long
term use.
This gives me and thousands of other GPs a dilemma. Should we switch
patients from atenolol to propranolol (inconvenient dosage or expensive
sustained release preparations, plus more side effects) or timolol, which
is also considerably more expensive than atenolol? If we are to follow the
available evidence, then we should. On the other hand, absence of evidence
of effectiveness is not the same as evidence of absence of effectiveness,
and it is possible that atenolol is, in reality, as effective as
propranolol or timolol. That we lack evidence one way or the other for the
effectiveness of so ubiquitous a drug as atenolol in so important a
clinical area as secondary prophylaxis of myocardial infarction represents
a failure in research and development planning and policy.
Why was work on atenolol virtually abandoned after the promising
early results? Was it just that newer compounds seemed more exciting than
older ones? This seems unlikely, given that there is adequate evidence
about propranolol, an even older drug. It is more likely that the research
agenda was driven not only by the clinical need for evidence, but by the
need for pharmaceutical companies to obtain evidence supportive of their
products in order to increase sales. Perhaps atenolol, no longer under
patent, was just too unprofitable to justify adequate research funding?
I am worried that the principles of evidence-based medicine can be
manipulated by pharmaceutical companies. They have the financial resources
to carry out large randomised controlled trials, seeking outcomes which
show their products in the most favorable light, while research on equally
important questions remains underfunded.
1. Freemantle N, Cleland J, Young P, Mason J, Harrison J. Beta
blockade after myocardial infarction: systematic review and meta
regression analysis. BMJ 1999;318:1730-7
Competing interests:
No competing interests
26 June 1999
Toby Lipman
General practitioner, Northern and Yorkshire Research Training Fellow
Rapid Response:
Atenolol after myocardial infarction - absence of evidence is a failure of research policy
As I read through Freemantle et al's excellent systematic review of
beta blockade after myocardial infarction 1, I looked particularly for
evidence about the effectiveness of atenolol. In my practice, atenolol is
our first choice of beta blocker for hypertension, angina and secondary
prevention because it is convenient (once daily dose), cheap, and
relatively free from side effects in most patients.
After the promising, if inconclusive, results of early short term
trials of atenolol (pooled OR 0.93, 95%CI 0.85 to 1.02) it was frustrating
to find that there were so few long term trials that the results (OR 1.02,
95%C 0.52 to 1.99) were meaningless. The wide confidence intervals of the
long term trials reflect the weight of 1.6% for long term trials of
atenolol compared to 74.2% for the short term trials, and the authors
rightly conclude that atenolol has been inadequately evaluated for long
term use.
This gives me and thousands of other GPs a dilemma. Should we switch
patients from atenolol to propranolol (inconvenient dosage or expensive
sustained release preparations, plus more side effects) or timolol, which
is also considerably more expensive than atenolol? If we are to follow the
available evidence, then we should. On the other hand, absence of evidence
of effectiveness is not the same as evidence of absence of effectiveness,
and it is possible that atenolol is, in reality, as effective as
propranolol or timolol. That we lack evidence one way or the other for the
effectiveness of so ubiquitous a drug as atenolol in so important a
clinical area as secondary prophylaxis of myocardial infarction represents
a failure in research and development planning and policy.
Why was work on atenolol virtually abandoned after the promising
early results? Was it just that newer compounds seemed more exciting than
older ones? This seems unlikely, given that there is adequate evidence
about propranolol, an even older drug. It is more likely that the research
agenda was driven not only by the clinical need for evidence, but by the
need for pharmaceutical companies to obtain evidence supportive of their
products in order to increase sales. Perhaps atenolol, no longer under
patent, was just too unprofitable to justify adequate research funding?
I am worried that the principles of evidence-based medicine can be
manipulated by pharmaceutical companies. They have the financial resources
to carry out large randomised controlled trials, seeking outcomes which
show their products in the most favorable light, while research on equally
important questions remains underfunded.
1. Freemantle N, Cleland J, Young P, Mason J, Harrison J. Beta
blockade after myocardial infarction: systematic review and meta
regression analysis. BMJ 1999;318:1730-7
Competing interests: No competing interests