Gary Friedman

Quotes

First publications

And the other sub-study I did that I think is really interesting from a methodologic point of view is one Bill Kannel suggested, “This interest in case-control studies, why don’t you reanalyze some of the Framingham data as a case-control study?” And I did that and I presented it at this cardiovascular epidemiology meeting and I remember Warren Winklestein was chairman and he came up to me after and said, “Oh, you’ve got to write that up and submit that to theAmerican Journal of Epidemiology.” And it was published there and it was really hot methodologic stuff then. I was using the Cornfield formula for the odds ratio. But now when you look back on it with all the developments and analysis of case-control studies it seems really very primitive and naïve. (5)

Talkers vs. doers

I sort of view that there’s two kinds of epidemiologists. There’s the talking epidemiologist and the writing epidemiologist. I always viewed him [href=”http://www.epi.umn.edu/cvdepi/bio.asp?id=79″>Stallones] as the talking epidemiologist. You know, brilliant guy, but I could never really relate to him. (9)

I remember, the first paper I wrote in the Field Station [San Francisco USPHS] I somehow got a hold of the tobacco tax data and I did a study on the correlation of tobacco sales in each state with the coronary disease death rate. That had to be approved by the Public Health Service and Len Syme told me that first, ‘the guy’ in Washington wasn’t going to approve it, but then they finally did. I was really annoyed by their infringing on my academic freedom. But it was published in the Journal of Chronic Diseases. It was a nice correlation between cigarette sales in each state and the coronary disease mortality rate in each. (9)

Atrial Fibrillation and Stroke

Then I was trying to get studies out of the Seal Beach [population]…… You know Framingham in the US News and World Report and one of their claims was that they were the first ones to quantitate the risk of stroke with atrial fibrillation. (10)

Representativeness of an HMO population

The other thing I’ve had to deal with [in Kaiser reports], people sort of accept it as, “Well, it’s a community study.” [So] I always quote the characteristics of Framingham. You know, at Kaiser it’s an HMO population, but it’s huge, it’s a third of the people who live here, it’s very heterogeneous in terms of race, socioeconomic status, and I say, “We’ve got all these black people, Asian people….” And in Framingham. . . people are always questioning if you have a grant application from Kaiser, “Is this sort of a special population that can’t be generalized to….” . . . But no one questions Framingham. Framingham was all white, there were six African-Americans in the town, no Asians, they were almost all of Irish or Italian descent. It was so much less heterogeneous and generalizable than what we have for this HMO population. (11)

Tobacco and Carl Selzer

…we were approached by Carl Seltzer about studying the consequences of smoking and he had money we could get from the Council for Tobacco Research. So we went ahead with that. It wasn’t that controversial in those days and I was convinced that we could publish whatever we wanted without them telling us what to publish. Carl’s emphasis was always on looking at differences between smokers and non-smokers with the underlying idea that since they are different in all these ways, we can’t say that smoking is what’s causing all the problems. So we had differences in pulmonary function and ….., you know, all kinds of things in papers came out. Then we did one paper in which we were looking at differences in coronary risk factors and we found a lot of …… We looked at one where it was non-smokers versus smokers and another where it was quitters versus smokers and it always looked like the people who had quit or who didn’t smoke had lower risk by other means: their cholesterols were lower…. I forget all the details. We did publish this in the Journal of Chronic Disease and Carl, I think, was going around the world giving talks and saying that cigarettes don’t cause coronary disease. And I felt the obligation to pursue this further.

Well, by then multivariate analysis had come in . . . and I said, “Let’s see if we control for all these things in which smokers and non-smokers are different whether that explains away the risk of coronary disease.” So we went ahead and did that and found that it didn’t explain it away and Carl withdrew from that. He didn’t want to be associated with it. [He’d say]“Well, that multivariate analysis… You didn’t analyze the data properly.” Before he had always said, “Permanente data are the most wonderful thing in the world.” But now when we took it that far…

Those two papers got published in the NEJM and he wrote letters critical of them and so did Phillip Burch from England who was one of the other pro-smoking guys. . . he had these mathematical models where he tried to show that smoking really wasn’t causing heart disease, which were totally incomprehensible. So, anyway, Carl and I sort of broke apart and our funding from the Council on Tobacco Research dried up at that point. (14)

Multiphasic Screening

It’s been dropped. There were these two settings, Oakland and San Francisco where Maury Cullin had set up this multiphasic; it was a major screening program, each clinic examining about 25,000 people per year. It was mainly done as an efficient way to satisfy the demand for check-ups. You know in those days, people wanted annual check-ups and rather than having them go to the doctor and having an exam, the doctor orders lab tests and then they come back. They get all the lab tests, do a lot of screening, then it could be handled in a 15-minute doctor visit. And that was the main reason for doing it.

It was a great research tool and Maury had put in a lot of things that weren’t that clinically important like pain tolerance tests and huge questionnaires about all kinds of things. The clinicians weren’t that thrilled with it and finally in 1980 it was dropped from San Francisco and over the years it was reduced in terms of the number of tests, questionnaire items and it sort of petered out in Oakland in the 1990s. There are places that are sort of doing some kind of health appraisal, but it’s sort of at a low level type of thing within Kaiser. But it’s still, I think, pretty well established in the San Diego Kaiser . . . And I think they set up multiple centers in Japan as a result of that and there are others around the world, but not in the original place.

And the other thing that Maury Cullin was way ahead of his time with was that study for the FDA which I was brought on to be an epidemiologist. He had a system going for recording all the prescriptions dispensed in the pharmacy and all the doctors’ diagnoses in the clinic and when the outside funding for that ended in 1973 Kaiser dropped it. . . Now 20-30 years later that’s been re-established. All the pharmacies are computerized in Kaiser; all the docs are marking diagnoses on forms because they need it for quality of care and to demonstrate what’s going on. (18)