University of Minnesota

Hugh Tunstall-Pedoe on WHO Myocardial Infarction Registers

I was recruited by Jerry Morris to run a Heart Attack Register in London’s Tower Hamlets. There had been two other English studies preceding it, one in Oxford and one in Edinburgh. This was a collaborative, multi-center study and it was really an attempt to put coronary care as it was practiced then in perspective.

There was this paradox at the time that the coronary care enthusiasts were saying they were halving mortality from disease. And that was at the time that in Britain at least, population CHD mortality rates were still increasing. And unless you put what happened in the coronary care units into a community perspective; you had to have the same denominator for both. This was very educational for me and for others because so many clinical statements are based on nebulous denominators.

The surveillance study taught me an enormous amount. I think it taught British clinicians things that many people in other countries took much longer to realize, which is that the hospital case fatalities were a small proportion of the total burden of deaths from coronary disease. I had never been taught as a medical student about sudden death from coronary disease. And then 70-80% of coronary deaths occurred out of hospital and there was a large amount of missed coronary disease and misdiagnosed coronary disease and people treating themselves for indigestion and so on when they had in fact myocardial infarction.

Jerry Morris, in fact, in his very nice little book, Uses of Epidemiology, actually quoted what he called a “register of coronary disease.” This had always been one of his ambitions, but he hadn’t been able to carry it out. In fact, he used insurance data as a substitute for community study in Uses of Epidemiology. And one of his dreams coming true was having this Tower Hamlet Study take place. I then met a much younger Zdenek Pisa in Jerry Morris’ office for the first time in 1969. He was then heading the European Office of Chronic Diseases and he had been getting these European register studies coordinated. Of course, another major input apart from the existing Edinburgh Study was the one in Gothenburg.

That was an interesting exercise, but I became very heavily involved not only in Tower Hamlets, but because at the very first meeting I was told, “You’re the Englishman so you’re going to write the report.” Because they usually had as the rapporteur someone who was fluent in English. And I’ve been rapporteur of WHO studies for 30 years now. Almost without exception every WHO meeting I go to. Not completely, but almost.

I then got involved in definitions and diagnostic criteria and realized that definitions, so called, were usually descriptions. They were not water-tight; they were descriptive. And so I then suggested circulating test cases and when we did that, the results were infinitely worse than anyone had realized. And that is the big contrast between MONICA and the original Heart Attack Registers is the quality control, standardization, measurement, and performance. Industry has been doing this for some time, but the medical world has only recently gotten into quality control. And we published all ours out on the web and there’s nothing that’s concealed. It’s all there.

Of course, you’d like to feel you’re practicing quality control but when you find there’s a problem it’s not so easy to know what to do about it. But the Heart Attack Registers got people together from different countries and it initiated people who hadn’t really thought about things into cardiovascular epidemiology and into studying what was happening. It put death outside hospital into perspective. And I suspect, at that time, in fact that we were really witnessing the unchanged natural history of the disease. I don’t think that any of the medical interventions that went on 30 years ago had much of an impact on disease. I now think, and I think a lot of epidemiologists haven’t caught up with this, that medical interventions, drugs and things really do change the natural history of a disease.

[ed. The Minnesota Heart Survey was also late to come to that view. It demonstrated that compared to very early in MHS the slope of the last 15 years has been much steeper for in-hospital survival. Whereas it had been much steeper for out-of-hospital in the 1980s. We finally came around, having made earlier calculations that coronary care simply couldn’t make an important impact on trends and rates.]

In MONICA, we found a stronger association with introduction of new therapies than we had expected. We’re still a bit baffled by it. But it’s interesting. Having gotten involved in epidemiology and prevention I find there are colleagues on the other side of me in epidemiology who still believe in the noble savage and that medical treatment and care and diagnosis is unimportant in the grand scheme. It’s a rather dangerous tendency, but it still remains. Of course, it makes it much more complicated if treatments do change what the underlying disease is doing or what it’s capable of. (Hugh Tunstall-Pedoe)

[ed. The frustration in the Minnesota Heart Survey is because the disease is changing and the treatment is changing and the diagnosis is changing and the reimbursement for different diagnoses is changing. And they are all changing at once. The Minnesota Heart Survey started in 1979 and is still being renewed because we’re getting information on how bad our information is. (Henry Blackburn)]