This note is based on the article ‘The Best Medicine’ by Sharon Begley in the July 2011 edition of Scientific American. The article discusses an alternative to the randomised control trial, the expensive gold-standard methodology that we use to establish facts.
This is about ALLHAT. ALLHAT is the Antihepertensive and Lipid-Lowering Treatment to prevent Heart Attack trial. It is the largest and most important investigation of treatments for high blood pressure that has ever been conducted. Doctors in the USA recruited 42,429 patients and then treated them with 4 commonly prescribed drugs for hypertension and they were followed for 5 years to see how well the medication controlled their blood pressure. The trial met the highest standards of medical research: neither the physicisans nor their patients knew who was being placed in which treatment group and patients had an equal chance of being assigned to any group. The study took 8 years to complete and cost $120 million.
The study produced clear results. They were announced in December 2002. The oldest and the cheapest of the drugs was more effective at reducing hypertension than the newer more expensive drugs.
The value of this finding to the USA in drug costs alone is estimated at $310 million per year.
In spite of ALLHAT, in 2009, 36% of first prescriptions for hypertension were for the cheapest and most effective medication. The results of ALLHAT have been ignored by the medical profession.
The question that the trial asks is very straightforward. Which of 4 drugs is most effective at reducing blood pressure? The answer is also likely to be very clear to understand. It is going to be drug A or B or C or D.
In order to answer this question, the US government took 8 years and spent $120 million. And the answer was very easy to understand. When you find that you have elevated blood pressure then, if you wish to use medication, then the best drug to start with is A.
These results were announced in December 2002. And yet in 2009, we find that only 36% of first time prescriptions for high blood pressure are for drug A.
It is important to note that in the USA any attempt to restrict physician autonomy or patient choice will face fierce resistance. In the political environment of the USA, you cannot tell doctors or patients which treatments they can and cannot have. What we are seeing here is a preference by doctors and patients.
We can assume that the results of the study would have been known to virutally all of the doctors that prescribed the medication. So the big question is why do we ignore the findings that our best efforts show will save money and improve patient outcomes.
Here are some ideas as to what is going on:
The question was simple. The methodology was the best we can devise. The cost was monumental. The time taken was near to eternal. The result was clear cut. And yet, the study did not lead to a large scale change in behaviour.
For the Constellation (and for most NGOs), the question that we are asked to answer in order to demonstrate the effectiveness of our ‘intervention’ is far from simple. We have neither the time nor the money to produce results that will have anything like the credibility of those that came from ALLHAT. In the light of the behaviour that we see from ALLHAT, what is the likelihood that any result that we produce will CHANGE the actions of ANYBODY?
We can argue that we use our personal experience to reach a conclusion as to the effectiveness or otherwise of the Community Life Competence process. If an external study supports our model of reality, then we are likely to embrace it (or at least to make use of it to support our position). If the external study does not support our model of reality then we can reject or ignore its results on the basis of flawed methodology or inapplicability to the case under consideration (or both).
The world of development is obsessed with measurement. The study discussed here shows that measurement in almost unimaginably ideal circumstances is not particularly effective in changing our actions.
Measurement of effectiveness (in some statistical sense) may be necessary in order to play the game. But we should recognise that it is nothing more than a charade. At best, we provide evidence to reinforce the prejudice of believers.
When I first began to discuss this subject in the Constellation, I held exactly the opposite point of view. I thought that if we could systematically collect information on our work, then we would be able to present a statistically valid justification of our work. This article continues my conversion to the view that we cannot collect valid statistical data to demonstrate the effectiveness of what we do AND that, even if we could provide such a demonstration, it would not be effective in changing people’s views.