Abstract
Dr. Wilson, you are one of the pioneers in the area of asthma education research and have developed and evaluated several major educational programs during the past 20+ years. How did your interests as a student, your educational background, and your early training lead you to pursue a field focused on many different aspects of healthcare, health education, disease management, and prevention?
Like many professionals, when they reflect on their career they may realize that the path from point A to point B was not as direct as one might think it should be. Even when I was in elementary school I was very interested in medicine, psychology, and research. I can recall many humorous examples of how that interest manifested itself and caused my mother some distress; for example, when I used her cooking equipment to sterilize agar and Petri dishes to grow various fungi in my closet. I also finished dissecting my dogfish shark for Comparative Anatomy class on one of her cookie sheets. Fortunately, we both survived those experiences! My interest in medicine, and my other passion, music, have continued throughout my life.
In college I double-majored in psychology and zoology, and in graduate school I was in the Experimental Psychology program, specifically, physiological psychology. I was a research assistant as an undergraduate at the University of Minnesota, doing behavioral studies of learning in rats, and in the Student Counseling Bureau analyzing predictors of academic success in medical school. As a junior, I also taught a laboratory section of the undergraduate Introduction to Biology course.
In graduate school I had a brief foray into mathematical psychology, developing and testing models of learning in Rhesus monkeys. I did my research at the Stanford University School of Medicine, rather than in the Stanford Psychology Department where I was pursuing my degree. My dissertation focused on cortical function in primates and, specifically, on visual processing and pattern recognition. I had a foundation in basic biomedical research with an emphasis on behavioral sciences.
After my postdoctoral training at Stanford I took a position in an independent research institute, the American Institutes for Research, largely because by then I had a young family, felt I needed to stay in the Bay Area, and there were no academic positions available at the time in my field, either at Stanford or other nearby universities. I began as a Research Assistant to John C. Flanagan, a very well-known psychologist who had founded AIR and whose field of study was educational measurement. In the mid-1970s an opportunity arose to develop a proposal to revise the test for medical school admissions, the MCAT, for the Association of American Medical Colleges. We put together a team that consisted of educational measurement specialists, medical educators, and subject matter experts. I applied my background in the sciences, with some experience in measurement and statistics, and together we defined a process for determining the detailed specifications for what would and would not be tested in the areas of biology, chemistry, physics, and measurement of reading and quantitative skills. We then developed the item pool and the first forms of the new MCAT, which went into use in 1976. I headed the MCAT development contract until 1991.
Meanwhile, I was getting involved with other studies that focused on the career choices of physicians, because one motivating force for the revision of the MCAT was the need to encourage more physicians to pursue careers in primary care and to work in rural and underserved urban areas. I designed and conducted studies of 2 large student cohorts that aimed at identifying predictors of entry into medicine and of specialty choices and practice locations. This process gave me additional experience, not only with obtaining contracts for research, but also in obtaining grants to support my own ideas of what needed to be investigated. At this time, I also conducted research and development to provide selection tools for use in awarding National Health Service Corps (NHSC) scholarships.
By the 1980s, I had developed a research group that included other investigators interested in health-related and health professions research, and we began to look at ways in which behavioral science approaches and educational methodology could be applied to study the behaviors of health professionals and patients with chronic health problems and to develop educational tools to help them improve clinical outcomes.
As I mentioned, I was always interested in both medicine and psychology, and at the core of that was an interest in the clinical problems of patients and how the behaviors of patients and healthcare professionals determine outcomes. In addition to factors such as the state of knowledge about a disease at a given point in time, and the care and medical treatments available for that condition, it is critical that patients do what they need to do in order to prevent or to treat a disease. I focused on the “how to” component: if we know about disease risk factors, how do we minimize those, how do we intervene to change risk behaviors? If we know how to treat a disease, how do we help patients implement the treatment so they derive the intended benefits? That has essentially been the focus of my research for the last 30 years. I went from the MCATs, to health professionals’ career choices, to the behaviors of patients with chronic health problems and the ways in which health professionals provide care, which has really been my interest all along—what is going on in encounters between the healthcare system and patients that influences whether medical knowledge and therapy gets translated into better health outcomes.
For example, it is widely known that only about half of the medications prescribed for asthma patients are actually used. That includes a substantial proportion of instances in which patients never even pick up from the pharmacy the medications prescribed for them. Even if they do, they might try them for a little while and then stop using them, or they might not use them correctly and not get the benefits they should be getting. For example, an inhaled asthma medication may be used in such a way that it does not get distributed down into the lungs properly. In our and other studies it has been found that, when you determine how much medication a person has actually used over a period of time, maybe half or less of what has been prescribed for them is even taken out of the pharmacy, let alone used properly. Very effective medications can be available, but they will not provide any benefit if they are not used, or not used properly.
Please briefly describe some of the early projects you helped develop and guide, such as the AIR/Kaiser asthma education program, one of the first validated asthma education programs, the Wee Wheezers program for toddlers, and the Fresno Asthma Project. How did you and your colleagues develop an evidence base for asthma education? What were some of the main difficulties and limitations in implementing these programs?
You have listed here the 3 main programs we developed. In all of these studies, we were aiming to find out whether a certain theoretically based approach did or did not achieve the benefits we hoped it would, or was better than another approach; we were not simply developing a program and assuming it was effective. We were particularly careful to document exactly what the educators did in the form of a program manual with standardized handouts and visual aids. It was only later that various federal agencies and health associations began to pick up the programs, repackage them for clinical use, and distribute them on a broad scale. Our focus was mainly on the research—does an approach work or not? Is it better than an alternative approach? Is it cost-effective in terms of reducing healthcare costs?
One of the studies targeted adults with asthma and was conducted in conjunction with Kaiser. We used Flanagan’s Critical Incident Technique to collect and analyze data from asthma patients, parents, and clinicians in order to develop several of the asthma education programs. We gathered information about instances in which a patient did something that was especially effective or ineffective in managing his or her asthma. When you have gathered enough such real-life examples you can develop a taxonomy of effective and ineffective behaviors. That is what we used as the basis for these programs. I think the use of critical incident studies is unique to our research and proved to be very valuable. Once you identify what a patient should do or should not do, you can then work backwards from that and identify the things a patient needs to know and is willing (motivated) and able (have the skills) to do to control the symptoms of a disease. We first did this type of needs assessment, then engaged in instructional development and pilot testing, and then carried out controlled clinical trials of the programs we had developed.
For the original AIR/Kaiser program, we developed three versions of, each of which had a different format. One involved small groups led by a nurse educator, one was designed for a nurse educator to use with individual patients, and the third was packaged as a manual for use by patients at home with no direct professional involvement. A fourth group simply received their usual care or “none of the above.” We found the group program to be more effective and more cost-effective on certain dimensions. When people work in small groups they can challenge each other, support each other, share experiences, role play, and engage in other types of group processes that tend to encourage the intended behavior changes.
My group does not conduct educational programs. We do research, and out of this have come many tools that healthcare systems and others adopt. I have observed the extent to which demonstrably effective programs are or are not implemented in various settings. We recognize that it is difficult to sustain group programs; there has to be careful thought to advertising these programs and getting the physicians who care for patients with asthma to recommend and support the participation of the patients and their families. If this can be achieved, it is clear that such programs can be effective and that the healthcare system will be motivated to continue the program. A good example of this is the Darnell Community Hospital, in Texas, which is linked to a local military base. The hospital implemented our Wee Wheezers program for children under the age of 6 years, and also required that physicians refer pediatric patients diagnosed with asthma to the educational program. The hospital also made other changes in the care of patients with asthma. Together, these changes resulted in about a quarter of a million dollars in cost savings in the first year, which caused the base commander to fund the asthma educator to continue the program. This example was written up as a case study by the Centers for Disease Control and Prevention (CDC).
You also mentioned the Fresno Asthma Project in your question. That was a very different kind of program. An opportunity arose for support from the National Heart, Lung, and Blood Institute (NHLBI) to develop a community-wide approach to improving the care and education of people with asthma. Together with a long-time colleague from Kaiser who had moved from the Bay Area to Fresno, we put together a proposal for the Fresno Asthma Project, which involved multiple healthcare systems in that area, including Valley Children’s Hospital, the American Lung Association of Central California, and the San Joaquin Valley Health Consortium. This was the first model of an asthma coalition in the United States. Subsequently, with the encouragement of NHLBI, The Robert Wood Johnson Foundation, the California Endowment, and other agencies, asthma coalitions developed all over the country, and the Fresno Asthma Project spread out over multiple counties and eventually became the Central California Asthma Project, which is still very active.
Recently, you developed and evaluated the Lowering Environmental Tobacco Smoke: LETS Manage Asthma program, a behaviorally based cotinine feedback intervention for reducing tobacco smoke exposure in children with asthma. What major conclusions can be drawn from this and previous educational programs aimed at intervening in asthma?
Although smoking rates in California have gone down dramatically, many children are still exposed to secondhand smoke. This is harmful enough in general, but especially so if a child has asthma. There is a known relationship both between maternal smoking and the development of asthma and between maternal smoking and the amount of morbidity a child suffers.
This study grew out of our work in Fresno, where we developed an intervention for parents who were smokers and had children with asthma who were receiving care from the pediatric pulmonary service at Valley Children’s Hospital. We gave the parents feedback on their child’s exposure, using urinary cotinine levels as a marker of exposure. Cotinine is a metabolic breakdown product of nicotine that is present in all body fluids including urine. We repeatedly tested the children as the intervention proceeded and worked with the families over time to reduce exposure, demonstrating some decrease in the cotinine level compared with children whose parents did not receive this intervention, and a significant decrease in the children’s need for medical care for asthma.
That initial study was not as successful as we would have liked, however. Although we demonstrated a strong trend toward reduced cotinine levels, it was not statistically significant. We planned a larger study together with Dr. Harold Farber and Kaiser in Northern California to try to enhance the previous intervention. In this study we carefully interviewed the parents to determine where the child was being exposed to tobacco smoke and by whom. We wanted to identify the most critical sources of exposure; for example, was it a daycare provider either inside or outside the home; did more than one person in the house smoke and how often were those people present. We developed a 4-day timeline recall procedure that identified very specifically where the child was throughout each day, who they were with, and whether those people were smoking. Any exposure to tobacco smoke within that period would show up in the urinary cotinine levels. We then focused on what had to change in order to eliminate the child’s exposure. We prioritized the sources of exposure and determined how ready the parents were to make the necessary changes. If they were not ready to make those changes, then we implemented various strategies to encourage them to start thinking about how to implement the changes. If the parents were more ready to make changes, then the focus shifted to providing the skills they would need to make those changes. This might include a range of behaviors such as getting a prescription for a smoking cessation aid, enrolling in a smoking cessation program, or role-playing how they would talk to their daycare provider about the need to eliminate their child’s tobacco smoke exposure.
The program was very tailored to the individual parents’ needs and readiness to change, and we continually measured whether progress was being made. All of the families in this study received basic education about asthma, including making sure the children were on a correct medication regimen. We wanted to ensure that it was not a lack of asthma education or treatment, but rather the smoking cessation messages and skills we provided, that really made the difference.
We are still preparing the results of this study for publication so I cannot talk about all of the evidence. In general, however, we found the program to be effective at reducing biologically verified exposure, but not in all children. We saw substantial reductions in exposure and in asthma symptoms, but only in those children whose asthma history indicated that they were at risk for asthma exacerbations. This research helped to identify children most in need of intervention and most likely to benefit, and identified at least one approach that can be successful.
Why is it important that health education programs be grounded in behavior change theory?
Parents (or adult patients) can have an understanding of what asthma is, how asthma controller medications work, what rescue medications are, and so on. This understanding of the disease and its treatment is necessary, but it is not sufficient. You can probably see from the types of behaviors I described in connection with the smoking exposure program why that would not be sufficient. Parents have to do many things differently—that is, change their behavior—to reduce a child’s exposure to environmental tobacco smoke. This requires motivation and, in many cases, skills they do not currently have. There is a lot of evidence to support the critical importance of focusing on behavior change, rather than just information transmission, behavior change in health education, rather if one wants to improve the outcomes of treatment of a medical condition like asthma or diabetes, or in order to reduce a health risk such as smoking or smoke exposure, risky sexual behavior, obesity, and many other health risks.
Why is shared decision-making important in asthma management in adults? How did the Better Outcomes of Asthma Treatment (BOAT) study encourage shared decision-making between health professionals and patients? What is the status of this study and what do you hope to learn from it?
There have been a lot of efforts to improve patient adherence to treatment regimens, specifically to medication, but those efforts have had limited success. There is also limited evidence of success in improving health outcomes as a consequence of improved treatment adherence. For >30 years, there have been observational studies showing an association between whether patients take their medications and various aspects of physician communication behavior—whether the physician tailors the regimen to the patient, for example. However, the fact that we know about these associations does not mean that we know how to intervene, change the communication between patients and physicians, and improve adherence or health outcomes. The problems that have been seen as a result of at least some types of hormone replacement therapy in women are a clear example of why it is not wise to promote broad implementation of a regimen based on observational, cross-sectional studies, rather than on experimental evidence.
In our previous research, I had noted that the educational approaches we were using did not seem to be having a big effect on patient use of medication. A colleague of mine, Dr. Sonia Buist who was then the head of Pulmonary Medicine at the Oregon Health Sciences University, and I decided that, with all the talk about shared decision-making and patients being more involved in treatment decisions, we would find out whether changing clinician–patient communication would actually affect adherence and outcomes. Our target was adult patients with poorly controlled asthma. The underlying hypothesis was that adherence to asthma treatment would increase if the patients were involved in choosing which medications they would use. That shifts the concept from adherence—in which the physician prescribes and the patient adheres—to looking at patients’ preferences and what a patient is trying to achieve as a result of asthma treatment, involving the patient in the choice of treatment in order to accomplish the goal, developing a plan with the patient for evaluating whether the treatment is working, and adjusting treatment, with the patient, until the desired outcome is obtained. This study is in the process of being published. We found that a shared decision-making approach, compared with either usual care or an alternative approach in which patients received education but did not share in treatment decisions, was more effective in terms of both adherence and health outcomes. The alternative model, which included education but no involvement in treatment choice, was better than not intervening at all, but not as effective as involving the patient in choosing treatment. The patients who participated in the treatment choice became much more committed to implementing the treatment chosen.
What are the key attributes/components of a health education and awareness program that contribute to its success or failure? How has that changed with increasing use of technology and Web-based strategies?
In my research group, Dr. Jun Ma, who is especially interested in using technology to improve the care of patients with diabetes and metabolic syndrome, is looking at Web-based programs that help patients and the professionals who care for them to monitor their disease and to set goals for changes that will improve control of the disease—including medication, exercise, and diet. Technology can be quite powerful for helping patients, although some people are less eager to use technology to monitor their disease than others.
Technology will continue to move forward and will be utilized in more and different ways. At the Palo Alto Medical Foundation, where my research program is located, the electronic health record has an online communication portal between patients and health professionals. This portal is being used to enhance data collection from electronic monitoring devices using automatic uploads and to communicate with patients to help them better monitor their condition. Web-based approaches are also being tested.
What types of asthma-focused studies and educational programs would you like to see funded in the near future?
I would like to see more work done in adults with asthma, and especially in those 65 years of age and older. There is a lot of emphasis on pediatric asthma, and I do not mean to diminish the importance of that in any way, but, in point of fact, there are many more adults than children with asthma. We have studied some programs for the elderly, such as the Peak Flow study, which looked at whether peak flow monitors are beneficial for elderly patients, but that is one of only a few programs that have targeted the Medicare-age asthma population. I would also like to see a real expansion and testing of the limits and best practices for implementing a shared decision-making approach to treatment choices in chronic diseases and for conditions that involve single decisions, such as the decision between surgery and other treatment approaches for cancer and other conditions.
What is the scope of your research and activities at the Palo Alto Medical Foundation Research Institute?
Within the PAMF Research Institute, I chair the Department of Health Services. We currently have two investigators in addition to myself, 2 postdoctoral fellows, as well as research associates with advanced degrees in epidemiology and biostatistics, and data analysts. Our focus is on studies of disease prevention and the management of chronic diseases, and we are doing research on a variety of topics. The most recent study, led by Dr. Ma, concerns the association that has been observed between obesity and asthma, and especially between obesity and poorly controlled asthma. This study, referred to as BE WELL, is a controlled trial of a lifestyle modification and weight loss intervention to determine whether weight loss results in improved asthma control. Dr. Ma is also conducting a study in patients with diabetes testing a Web-based intervention to help patients achieve better diabetes control, as mentioned above. Dr. Dominic Frosh, another investigator, is looking at issues such as the impact of direct-to-consumer advertising of medications and shared decision-making between physicians and patients concerning prostate specific antigen (PSA)-testing and prostate cancer. We are also planning other studies, including a study of the use of alcohol by elderly patients and the particular risks of alcohol in this population, which has a high rate of multiple medication use. There’s no chance we will run out of interesting, important scientific questions to ask!
—Interview by Vicki Glaser