David Himmelberger (M.S. 1972)
As an applied statistician working in the health care field, David Himmelberger’s job demands that he master both precision and nuance. The precision comes in designing studies that quantitatively assess the quality of life in patients suffering from chronic diseases like asthma and arthritis. The nuance comes in accounting for the cultural and linguistic differences in patients from as many as 45 countries.
"In English one of the very standard types of questions is, ‘How much of the time did you feel depressed or tired?’" explained Himmelberger. "Most of the time it’s asked as a categorical question: none of the time, a little of the time, some of the time, most of the time, all of the time. In other languages, that’s an unnatural way to ask the question. In another culture, it would be asked as ‘How often did you feel depressed?’ Then the answers are: one time, two times, three times. The whole nature of the data has changed from a continuous variable to a frequency. So now you’re mixing apples and oranges together. This is something that one has to think about for multinational clinical trials. What’s the natural way to ask the question in different parts of the world, and will that result in equivalent data?"
Once Himmelberger has designed a questionnaire, he travels to the country to pilot test it with a small group of patients. "Rather than going to the expense of running the study for the first time, we take a small sample," he said. "If you’ve made a mistake, you haven’t collected a lot of worthless data." In the various countries, he works collaboratively with linguists, physicians, and the sponsoring organizations to fine tune the surveys. He learns how to account for cultural differences by listening to people. "To me, that’s been one of the most interesting aspects of the work, getting a deeper first-hand understanding of different cultures," Himmelberger said. "What’s been the process before is the research and questionnaires have been developed mostly in the U.S. and the U.K. People thought, just be careful of how you translate. But you have to be aware of the structure of the data that comes from the questionnaire."
Himmelberger acknowledges that international quality of life evaluation is a very specialized field. He had the idea to found Health Outcomes Group after working at Stanford University as a biostatistician for 15 years. "At the medical school at Stanford where I consulted with faculty and researchers on the design of their experiments and the analysis of their data, I could clearly see that virtually all of the proposed treatments for chronic diseases were unlikely to be cures, at least during the course of my lifetime," Himmelberger said. "With the absence of premature death in these diseases, it was clear to me that measurement of meaningful quality of life outcomes would be the critical challenge in evaluating the benefits of new intervention." Himmelberger combined this vision of a future need with his love of traveling. "Part of the motivation when I left Stanford was to make the business multinational," he said.
Himmelberger’s time at Purdue was also instrumental in preparing him for his future career. "My experiences at Purdue in the consulting lab with Virgil Anderson and George McCabe were the basis for my interest in working in a consulting capacity in applied statistics," he said. One of the problems he worked on as a graduate student was helping a group in Industrial Park analyze remote sensing data they had collected from flying over large portions of the United States. "They wanted to know, Could you identify crops based on this data? Could you quantitate that? Many times when you asked farmers how much of a crop they were growing, either they didn’t know, or they didn’t want to tell you." Himmelberger said, "What I learned at the lab at Purdue is that underneath all of those problems, there’s a certain structure. As a consultant, my job is to strip away the layers and figure out what the problem is."
To current statistics students Himmelberger advised, "If you choose to do something in an applied area of statistics, choose something that you love doing and that you believe is meaningful." He also pointed out that contract research organizations are an option for those who do not want to work in a big industry. "What I appreciate most is that statisticians are often asked to be objective," he said. "That’s why different companies come to me, because I have some expertise, but I have no opinion one way or the other. In a big company, there’s a lot of pressure many times to get a certain result. However, if you’re not giving good advice, somewhere down the line the fact that the treatment is no good is going to catch up with the company. That is something that can be tough for young people to do. Contract research organizations are an alternative that students don’t often consider."
To read more Alumni Profiles, please visit our Alumni Profiles archive.
March 2006
