University of Minnesota
http://www.umn.edu/
612-625-5000
Menu

David Jacobs

Year: December 17th, 2001
Location: Minneapolis, Minnesota
Interviewed by: Blackburn, Henry

Abstract

INTERVIEW WITH DR. DAVID JACOBS

INTERVIEWED BY DR. HENRY BLACKBURN

Minneapolis, MN Sept 9, 2002

This is a flavorful talk about the design, implementation and interpretation of the UGDP trial on diet and oral hypoglycemic medications in adult diabetes. It is full of large personalities, of Chris Klimt, Alvan Feinstein, Jerry Cornfield, and of pioneering analysis. It touches on the battle for ‘ownership’ of research data in UGDP when the drug companies lost at the Supreme Court. The industry and the diabetologists were infuriated by the results that Tobutamide was ineffective and launched an attack felt by many. It treats the brilliance and contributions of Cornfield, Canner, and Feinstein and the education of the young Dr. Jacobs.


 

HB: Continuation of chats about UGDP.  This is David Jacobs who was in the UGDP Coordinating Center at Maryland. 

DJ: I joined the UGDP in August 1971. 

HB: That was right after the Tolbutamide report and all its controversy.

DJ: In looking here at Curt’s summary, he says,  “August 9, 1971 UGDP preliminary report on Phenformin published.”  So I joined in an environment, as a statistician with no medical or public health experience whatsoever, of this very exciting thing that had happened since 1962, when it was found that Tolbutamide, much to the amazement of the physicians who were prescribing it for many years, and to the dismay of Upjohn who was making a lot of money on it, was actually killing people if used in the dose prescribed.  Now, in retrospect, I wonder if they had had good glucose control at the same time, if Tolbutamide might not have done better.  And the same thing with Phenformin.  Metformin is a derivative of it and is back.  So questions do get revisited.  But at the time, that was the environment and Chris Klimt was there and Curt Meinert was there, Genelle Knatterud was a prominent person who was later to become head of the Maryland Medical Foundation that did all these clinical trials after it left the medical school over disagreements……

HB: Over disagreements…. You didn’t finish your sentence.

DG: Over disagreements with the University of Maryland Medical School about exactly who could do what when.  They were in the Department of Preventive Medicine but they weren’t really “doing” preventive medicine.  They were under Maureen Henderson, you might remember who came in at that time. 

HB: That’s right.  Battle of the giants maybe – Maureen and Chris?

DG: There were arguments about who would have what power.  Chris Klimt was an interesting man and Curt too was interesting.  All very powerful personalities.  Hard to get along with for junior people.  That’s not necessarily a bad statement about them, but they were interesting guys and very opinionated.  Chris was Teutonic.  He believed in total control.  So I never got along with him very well.  But again, I don’t think that’s such a bad statement about him. 

So anyway, I came into the UGDP and started working with it and the whole thing was about the diet, the people had been told to adhere to a good diet.  But nobody knew what that was and nobody had the behavioral skills to teach the patients to actually do it.  So they didn’t do it. 

HB: The diet intervention was weak in that study.

DJ: Very weak. 

HB: Hospital dietitians …..

DJ: Yes.  A number of years later we went through conversations with people who were skilled in interviewing and in exchange of information who were starting to point out that patients were not hearing what doctors were saying.  That was a very interesting thing for me to learn.  And I think it was part of why in the 60s and 70s patients were not getting the physician’s message.  Hopefully we’re starting to do better.  But it’s still a problem with a busy physician,  a high level of knowledge, and a very uneducated and uncertain patient.  It’s still an issue.

HB: I understand from Fred Goetz that the whole matter of oral hypoglycemic medications is still an issue.  They’re still putting them out and not doing adequate clinical trials.

DJ: Right.  And now as I look back, the understanding of beta cell exhaustion, for example, I was just reviewing  with Fred today.  We just got a paper accepted today speaking to secretory ability of the pancreas.  We didn’t have any measure in the Wadena Study of insulin resistance, but we did have secretory ability and we see a pattern in which it was going up in IGT people over five years and it was going down in the IDDMs who were not on insulin.  So they weren’t becoming deficient, but you could see a process which looked like beta cell exhaustion occurring. 

I just had another call today about triglycerides and heart disease.  I said, “The new scoop, 30 years later, is that triglycerides are responsive to insulin.  And if you’re insulin resistant you don’t pack away the fatty acids into your cells as well and so your triglycerides go up.”  And maybe that has a completely different meaning than we were thinking about as triglycerides as a kind of  companion with cholesterol.  So there are a lot of new clues about etiology that might lead you to have different ideas about the way the UGDP is interpreted and the way it was run.  But at the time, ……

HB: Could you go back again to how that study was run and the controversies as you remember?

DJ: There was a huge fuss by two groups.  Now, I was again hearing this as a new quasi-medical person in a new environment with a new PhD, and a young man, too.  I was only 26 in 1971.  So, the scuttlebutt in the Coordinating Center was that Upjohn had a conspiracy going because they only wanted to make money and the physicians who had been prescribing Tolbutamide for 15 years to their patients did not want to hear that they were wrong.  So it was a mixture of pride and power!  And they wanted access to the data and they wanted to show that we were all wrong.  There was a guy named Stanley Schorr, a statistician, and on our side was Jerry Cornfield, one of the world’s most impressive and the best statistician I’ve met.  A very important person in my training.  Stanley Schorr wrote an article which appeared, ……

HB: I’ve got the reference.  I think it was in JAMA, 1971. 

DJ: And I made the comment reading this, “This looks like he was, you know, just sitting on the pot and kind of scribbling something as he went along.”  It was very poorly written, just kind of a laundry list of things that might be wrong without any attention to whether they were wrong.  And Jerry Cornfield came back and answered the points one by one by one. 

HB: In the same journal in a letter or something?

DJ: Yeah, it was probably a letter responding to the Schorr article or it might have been a counterpoint article.  But, he responded in a very beautiful and well written and carefully thought out thing.  For example, it turns out that there was evidence of a little bit more congestive heart failure or digitalis use, something like that, in the Tolbutamide group.  So Stanley Schorr, sort of forgetting about confounding, just said, “If there were more of those, you can’t sort it out.”  And Jerry Cornfield said, “Well, here’s how you do sort it out.  You’d expect three more cases.  That wouldn’t change anything.”  So there was that kind of reply. You had excellent statisticians and thinkers and we had Max Miller who was from the Cleveland Clinic, who was a very excellent clinician, very interested in research.  He wasn’t as good in the research side of it, but we had a wonderful group of people who were thinking hard about it.  We had, in fact, done a randomized trial.  And, in fact, some of the randomized groups came out with a 30% higher cardiovascular disease incidence, I think.  So then there was an audit.

HB: Could you speak of the issue of access to the UGDP data.

DJ: My opinion was that we had an excellent group and that the study was well done.  Certainly standard for the time.  And one group came out worse than the other.  It was one of the very first big clinical trials and the community did not know how to receive it.  They were not used to it. 

HB: Chris interpreted it as the volunteers did not understand the intent to treat phenomenon.

DJ: Right, all of those things.  And they loom much larger now that many more clinical trials have been done and there are coordinating centers.  I was relating to you earlier that one observation was that in the group at Washington U., one of 12 centers actually had a finding in the wrong direction.  In other words, the right direction.  They found that Tolbutamide was protective.  But they only had 100 out of a thousand or 1200 people in the study.  And they had 12 cases of something and it went one or two cases in the hoped for direction.  So……

HB: Like a the doctor seeing 12 cases?

DJ: Yeah, so they were wondering about the integrity of their own clinic data and why shouldn’t their own clinical experience mean something, and wasn’t this completely contrary.  Another person was Bill Parsons who might or might not have been part of that [anomalous]group.  But, he also had a feeling that there was something wrong.  And there were numerous little things you could point at.  But the diet thing, I think now, is the most important thing.  People were not really talking about that.  I don’t think they could at the time.  They had no skills to have done a better job on diet. 

HB: Everybody wanted the data and Chris’ attitude didn’t encourage their confidence in getting them.

DJ: Right.  Chris Klimt was an autocrat who felt that he owned the data and the devil be damned, nobody else was going to get it.  There was an auditing team that came in that I mentioned before.  They looked it over and they couldn’t find anything wrong, you know.  It was a much more professional, academic audit than Stanley Schorr’s article.  And sure enough, what we were saying was true and right.  But the other groups didn’t want to back down and so there was a fight for the data.  As I understand it, the fight went to the Supreme Court and I guess resolved……

HB: Yes, 6 to 2.

DJ: The Supreme Court decided that the outsiders did not have the right to the data.  Now this is real interesting as we look at what’s going on today with all these data sets being made public.  And yet we have at the same time more privacy and more public regulations.  But, I was in that environment, discussing how we could maintain control of our own data and looking at it from the coordinating center’s point of view.  I must say, subsequently, especially working with the Lipid Research Clinic’s coordinating center, that the experience of being a statistician who could not get access to study data was a very bad experience.  It was also very inefficient way to look at the data, instead of putting it out and letting people present their interpretations.  But the clinical trials world is quite different from the observational studies world because they want an answer as opposed to things might exist in various ways. 

HB: Any other personal reactions of the people in the study, with Max, or other critics of the study?  Feinstein, did you ever go to any of his critiques of the study?

DJ: I know Alvin Feinstein.  Of all the people, I know Alvin Feinstein, and I’ve sat in a couple of meetings or committees with him.  I have chatted with him, had dinner with him.  I think his view is kind of extreme.  It’s a little bit like if everything isn’t right, then everything must be wrong.  But, he certainly, also, in the past 20 years has presented a lot of interesting methodology.  When he presents his ideas in pure form as just ideas, rather than applying them, perhaps incorrectly……

HB: Or get paid to apply them ….

DJ: Right, I think they are quite valuable.  So there certainly was a tone in the UGDP of “our side and their side.”  There was undoubtedly money being exchanged one way or the other and building on the sort of the current idea that we should not work with industry because you’ll be bought.  As opposed to the idea that I got out of MHHP that we should work with industry because it’s the only way to affect the public health.  We just need safeguards so that we’re not bought, you know.  They can do what they need to do, but we should do what we need to do.  I guess I didn’t know the opponents otherwise very well.  There were certainly a lot of great people working in that study whose careers began there or who made excellent contributions. 

HB: A great quote of Jerry Cornfield’s on being asked to talk on the principles of research, my first thought was to rise after the chairman’s introduction and say, ‘be careful’ and then to sit down. 

DJ: Right.  My view which is somewhat like that is that my textbook on statistics for epidemiology has four words in it, “Think about what you’re doing.”  Maybe that’s five.  But think about what you’re doing is very similar to Jerry’s and I learned a lot from working with him.  He died of pancreatic cancer.  I wonder why.

HB: Well, he smoked cigars constantly.  That’s been incriminated in some studies.  Now, Chris was depressed wasn’t he?

DJ: Chris ended life very depressed.  I understood that he committed suicide.  People weren’t talking about how he actually died.  But I think he had contracted some illness and it was probably secondary to being sick all the time.  Paul Canner was another really fine person.  Was he working in the UGDP?  Maybe he was just ancillary to it.  He was more in the CDP.  But he was a wonderful, one of the most ….

HB: Is he still there?

DJ: No, he’s not.  Martha Canner is still there last I heard, his wife.  I actually took over here from his sister, Jean.

HB: Who disappeared into some religious commune.

DJ: Went to Iowa.

HB; June, Jane.  Jean.

DJ: Jean, that’s who it was.  Jean Canner.  And the Canners were a family of extremely fundamentalist religious people.  Paul, at the same time, was a fantastic, immaculate statistician.  Every time we would have a meeting with him he would present one of his coordinating center reports and he would have done exactly the right and elegant thing, totally innovative, to answer whatever the question.  But he, at the same time, believed strongly in Jesus and he believed in Creationism.  He did not believe in Evolution.  I said, “How is this possible?  How can you reconcile these things?  You know, I’m a liberal Jewish guy and just want to understand.”  I was interested in his religious concepts.  He tried to explain to me, “Well, whether I’m fixing a door handle or about to undertake a difficult statistical problem, I say a prayer, Jesus helps me, and I’m fine.”  And he’s completely convinced about Creationism and didn’t really want to address that.  There was a very funny incident with Tominaga who was there in the CDP coordinating center and who was my office mate and was the one who actually brought me to Minnesota in a way.  Tominaga was teaching his class in 1973 on ECG interpretation and Paul Canner and I were both in the class and Tominaga was extraordinarily clear, with wonderful, beautiful diagrams and everything.  And Paul raised his had at one point when Tominaga was talking about “action potential.”  So Paul said, “Could you explain in more detail was action potential is?”  And Tominago said, “Well, action potential, you know, is when you have a differential of the electrolytes one side or the other side of the cell membrane and it creates a voltage.”  So Paul raises his hand again and says, “You know, I still don’t quite get it.  Why is there that voltage?”  He replied, “Well, because there are these enzymes or whatever, they move the electrolytes back and forth.”  And Paul kept persisting, “Why, why is this?”  He wanted Tominaga to say that this was evidence of God.  And Tomi finally looked at him and he said, “It just is.  That’s all.  It just is.” 

HB: Good for Tomi.  He got his Japanese back up. 

DJ: Paul, in subsequent years, much to his credit no matter what, went on a ministry and so, last I heard, he’s 80% time editing a religious newspaper and still doing statistics 20% of the time.  And his wife, Martha, is in the clinic.  His sister Jean was an excellent young statistician.

HB: Yes, we hired her here. 

DJ: Right.  Then she moved to Iowa to be a farm wife because that’s what the group believed was the meaning of life.  So I’m not going to argue with that.  But I’m going to do it my way. 

HB: Thank you, David, it’s been fun. 

 (Henry Blackburn)

Quotes

DJ: So I joined in an environment, as a statistician with no medical or public health experience whatsoever, in this very exciting thing that had happened since 1962, when it was found that Tolbutamide, much to the amazement of the physicians who were prescribing it for many years, and to the dismay of Upjohn who was making a lot of money on it, was actually killing people if used in the dose prescribed.

HB: A great quote of Jerry Cornfield’s: “On being asked to give a talk on the principles of research, my first thought was to rise after the chairman’s introduction and say [only], ‘be careful’ and then to sit down.

DJ: Right. My view which is somewhat like that is that my textbook on statistics for epidemiology has four words in it, “Think about what you’re doing.” Maybe that’s five.

Full Transcript Access

Full transcripts of interviews may be made available to those engaged with original materials for scholarly studies by contacting us.