|
Abstract
IntroductionNon-insulin-dependent diabetes mellitus (NIDDM) has emerged as a leading cause of morbidity and mortality in aboriginal communities throughout North America.1–4 It is thought that the roots of the high diabetes prevalence in First Nations populations lie in the profound social, environmental and lifestyle changes of the past 50 years, including dietary changes and reduced activity levels, acting on a susceptible genotype.5–9 To date, a great deal of scientific effort has focused on explaining the pathogenesis of diabetes in aboriginal populations and on the medical management of diabetes in individuals. However, effective strategies for primary prevention in communities are only beginning to be developed.The Sandy Lake Health and Diabetes Project was initiated in 1992 to address this issue. The study’s objectives have already been discussed by Hanley et al. in the preceding article in this issue.10 This paper will focus on the third objective, the development of community-based intervention strategies to prevent diabetes and its complications, based on the formative qualitative and quantitative research.
BackgroundFormative assessment methods are used mainly by social scientists to assess people’s beliefs, perceptions and behaviours using a combination of qualitative and quantitative approaches. In addition, they seek to describe the context in which these behaviours take place and to understand why people do what they do. Thus, planners can anticipate reactions to health programs and better adapt programs to local conditions.11–16 Such adapted programs are therefore believed to be more effective in changing human behaviour and improving health status.The linkage of qualitative and quantitative approaches has received increased attention over the past 20 years. It has been observed that the strengths and weaknesses of qualitative research complement those of quantitative research, and the reverse is also true.17 To date, the primary link has been the use of qualitative information to improve quantitative research. This has included the use of qualitative data to generate testable hypotheses and to design better quantitative instruments;18 the conduct of parallel streams of information-gathering that, when combined, yield increased confidence about research findings;19–21 and the use of qualitative data to assist in interpreting quantitative findings.22 Examples of the linkage of qualitative and quantitative methods for the design and implementation of health interventions include developing approaches for smoking prevention,23 developing effective AIDS risk reduction messages for use with commercial sex workers 24 and improving workplace conditions.20 Probably the most concerted set of efforts linking the two approaches has been in the development of a series of problem-focused "rapid assessment" manuals over the past eight years. Starting with the original Rapid Assessment Procedures manual developed by Scrimshaw and Hurtado in 1987,25 there have now been over 10 manuals produced on such topics as acute respiratory infections, women’s health, vitamin A, sexually transmitted diseases, malaria and water sanitation.26 In this paper, we present a model for linking qualitative and quantitative information to develop interventions targeted to prevent obesity and diabetes in an Ojibwa-Cree reserve. We will emphasize the contributions of formative data only. Certainly, there are other elements that enter into intervention design, such as review of the literature and our own personal experiences, but these will not be considered here.
SettingThe study was conducted in the First Nations Ojibwa-Cree community of Sandy Lake, Ontario, a description of which has been provided previously by Hanley et al.10The major supplier of food in the community is the Northern, a branch of retail chain that services many remote communities throughout Canada. The federal government’s Northern Food Mail Program partially subsidizes the transportation of perishable foods to these communities. In addition, there are a few small convenience stores that sell a wide range of canned foods, candy, snacks and fast foods (e.g. pizza, fried chicken). The community is considered a "dry reserve"; no alcohol is permitted on reserve land. The area has substantial wild food resources including moose, duck, geese, rabbit, beaver and fish, although hunting, fishing and consumption of these foods is seasonal and limited. Although the community is remote, it is not lacking in modern technology. Most households have TVs and VCRs, many have satellite dishes and snowmobiles. The community has its own radio and community television stations. The two local schools currently have over 500 students enrolled from kindergarten to grade 12. The prevalence of NIDDM and impaired glucose tolerance in Sandy Lake is high; based on our study, roughly 16% of the study population over the age of 10 have diabetes, and another 10% have impaired glucose tolerance.
MethodsAn integral part of the research project and its design was the development of a partnership with the community of Sandy Lake. This occurred at a variety of levels including local government (Band Chief and Council), elders, other community leaders and the community at large. This was facilitated by regular meetings with the Chief and Council (five to six times per year), community-wide radio phone-in talk shows and the sharing of results with the community at community feasts and other events. A permanent project co-ordinator was hired to live in the project house, which was strategically located within the community.A wide variety of both qualitative and quantitative data was gathered through the course of the project. Table 1 gives an idea of the range of methods used in conducting the formative qualitative research. These methods consisted of key informant interviews, systematic interview techniques such as free listing and pile sorts, direct observations and review of existing written materials. Themes explored included commonly eaten foods, beliefs about foods, leisure and other activities, and health and illness beliefs. Both children and adults participated as informants and respondents. Results of this data collection and specific details of the methodology have been presented in detail elsewhere.27 Quantitative data were gathered at both the individual and the household level over 20 months (July 1993–March 1995). These data were collected to obtain basic community information on obesity and diabetes prevalence as well as associated risk factors in this setting. For individuals aged 10 or older, a structured survey was performed that included the following components: sociodemographics, health beliefs and knowledge, food frequency (usual food intake over the past three months), 24-hour dietary recall, substance use, activity recall, concepts of body image and a family tree for history of diabetes. Physical examination included anthropometry and assessment of body composition by bioelectrical impedance assessment (BIA). Serum was drawn for laboratory investigations of fasting glucose, oral glucose tolerance test, lipids, creatinine, urea, etc. A 73% participation rate was achieved by testing 728 (out of 1018 eligible) individuals above the age of 10 years. At the household level, information was collected about household demographics, economic status and food preparation practices from a sample of 250 households. A detailed description of these methods is presented earlier in this issue.10
|
||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||
Model For Developing Interventions Using Qualitative and Quantitative InformationDevelopmental SequenceSelected interventions to improve the diet will be used as examples in the sections that follow. Figure 1 outlines the overall process of intervention development. After the initial formative qualitative research was analyzed, important findings were presented to the community and group sessions were held to generate ideas for appropriate interventions. A list of potential interventions was developed, and the feasibility of each of these was qualitatively assessed. This information, in combination with preliminary analysis of the quantitative research, reduced the list of potential interventions even further and refined the specific strategies in terms of the medium to be used, the details of the message and the target group. These three items (in bold on Figure 1) are the stages we emphasize in this paper. The final steps in intervention development are pre-testing the interventions and obtaining community feedback about their appropriateness and effectiveness. This process is currently under way in the community. Following further refinement, a community-wide strategy will be implemented. Making a Worksheet
|
|||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||
Applications of the ModelWe will present two examples to highlight how the model is used for incorporating qualitative and quantitative information into intervention development.Example 1: Children’s Consumption of Fat a) Problem definition Parents frequently emphasized the poor eating habits of their children, and their own healthier preference for food from the bush, like fish and moose meat. The quantitative data supports and complements the findings of the qualitative data analysis. Based on the results of the food frequency survey, children were significantly more likely than adults to report frequent (three or more times per week) consumption of Klik (a canned meat product with 82% of total calories from fat), chips, (whole) milk, pop, chocolate and cookies. In addition, body composition assessment by BIA supports the finding that obesity prevention in children may be important. The BIA results showed average percent body fat in 10–19-year-old males to be 21.0 (SD = 1.0); and in 10–19-year-old females, 36.7 (SD = 1.1). Both of these averages are well above mean reference values for this age group, especially for females. b) Intervention strategy Where did these ideas come from? Portions of the qualitative data suggested working on the school curriculum. Pile sort exercises found that, as early as the second and third grade, children categorize some foods as "junk foods" and feel they are unhealthy. Children also recognize sweet foods as unhealthy. Thus, children in Sandy Lake already have a sense of the healthiness and unhealthiness of food. Currently there is no health or nutrition curriculum until ninth grade. However, there is strong teacher interest to begin one in earlier grades. Changes in the foods available at the Northern’s delicatessen will likely have some impact as well. Parents report that children prefer "junk foods" to bush foods: "They used to eat Indian foods when they were small, but they don’t really like it now. Now they only like French fries, Klik, beef stew, ... canned food." Thus, children do exercise choice in the selection of food in their environment, particularly when they are on their own, getting food during the school lunch period. Differences in the eating habits of children and adults in the community will help us target interventions to children. While children eat significantly more snack foods, adults report eating eggs, lard, hot cereal, soup, potatoes (not French fries or chips) and tea significantly more often than children. We will use this information to develop age-specific, food-specific messages. Community input and feedback about these interventions will be sought in the summer and fall of 1995. Example 2: Added Fat Consumption a) Problem definition The quantitative food frequency data were helpful in confirming the importance of the issue of added fat consumption. Most of the frequently (3+ times/week) consumed foods were store-bought foods such as tea (94%), white bread (90%), butter/lard (80%), evaporated milk (73%) and pop (64%). The majority of the 10 most frequently consumed foods on the list are either high in fat or are commonly consumed with added fat. The least frequently consumed foods tended to be wild foods, available only during specific seasons. For example, only 16% of the population reported eating moose three or more times a week. The following added fats were consumed daily by the given percentages of the population: evaporated milk, 67%; lard or butter, 66%; margarine, 48%. The corresponding figures for foods to which fat is added are bannock, 43%; white bread, 79%; tea, 90%. b) Intervention strategy Our health education messages will focus on those specific foods that contribute the most added fat to the diet. Out of over 300 different foods consumed in the community, only 20 foods accounted for 77.1% of total fat consumption (grams). Four of these foods consumed as added fat—lard, evaporated milk, butter and margarine— account for 14% of the total amount of fat consumed. The single highest source of fat in the diet, bannock (11.6% of the total amount of fat consumed), is also commonly eaten with added fat. Efforts to reduce the amount of fat in the diet will focus on reducing the amount of lard and butter eaten on bannock, replacing whole evaporated milk with 2% evaporated milk and encouraging people to try bannock recipes with reduced fat. We anticipate a reduction of total fat intake by 5–10% in individuals by intervening with these foods alone. The qualitative information suggesting a community-wide mass media intervention strategy included multiple observations that radios are commonly left on all day long, tuned to the community station. Most community announcements are made on radio, which is seen as a source of "service, information and help." Local television is another means of communication for special community events, such as funerals and fund-raising activities. These community media are effective ways of reaching a broad spectrum of the population on a regular basis. The quantitative data support the use of these media, as our questionnaire of material possessions revealed that 100% of homes have working radios, 99.2% have televisions and 78.7% have VCRs. Of course, health education messages to reduce the amount of added fat will also be communicated and reinforced in many other ways through other components of the intervention plan, including activities such as home visiting and health fairs. Community input and feedback concerning these strategies will be sought in the summer and fall of 1995.
Overall Intervention StrategyThe above examples illustrate a few specific elements of the intervention plan. The overall intervention strategy designed for Sandy Lake will include two major components: a community-based strategy and a randomized clinical trial, featuring an intensive set of lifestyle interventions for high-risk individuals identified by the prevalence screening. Community-based Strategy
The primary objective of this component will be to evaluate whether an intensive lifestyle intervention strategy in individuals at high risk for developing diabetes will demonstrate health benefits, such as weight loss, self-motivation and knowledge, that are significantly greater than matched controls randomly assigned to receive the community-level intervention alone. The high-risk individuals are defined as those at risk for developing diabetes, i.e. individuals with impaired glucose tolerance and/or significant obesity (body mass index >27) and hyperinsulinemia (highest 20th percentile) who are between 20 and 50 years of age. This group at risk was identified during the previous survey. The intensive lifestyle intervention group will receive enhanced and individualized diet and exercise education, motivational support and increased exercise training from qualified community-based physical activity leaders (PALS) recruited and trained locally. The PALS will have regular and frequent contact with these individuals in order to optimize lifestyle changes. Overall, this strategy will help to determine if a community approach to lifestyle intervention can be successfully implemented in an aboriginal setting.
ConclusionsIn this paper, we have sought to provide a model with examples of the ways that qualitative and quantitative information can be combined to develop culturally appropriate, community-based diabetes prevention intervention strategies. The quantitative information helped us to target interventions to individuals most at risk for obesity and diabetes and to identify specific foods on which educational efforts will be focused. Qualitative information helped us to identify specific risk behaviours, to select appropriate language and phrasing of messages and to choose the correct media for communication. Qualitative methods also helped and continue to help to build rapport with community members, and they will be used to obtain feedback about specific interventions on a continuing basis.
AcknowledgementsThe invaluable partnership and support of the Chief and Council and the people of Sandy Lake, Ontario, are gratefully acknowledged.This project was funded by grants from the National Institute of Health (91-DK-01) and the Ontario Ministry of Health (#04307).
References1. West KM. Diabetes in American Indians and other native populations of the new world. Diabetes 1974;23:841–55.2. Long TP. The prevalence of clinically treated diabetes among Zuni Reservation residents. Am J Public Health 1978;68:901–2. 3. Knowler WC, Pettitt DJ, Savage PJ, Bennett PH. Diabetes in Pima Indians: contributions of obesity and parental diabetes. Am J Epidemiol 1981;113:144–56. 4. Szathmary EJE. Diabetes in arctic and subarctic populations undergoing acculturation. Collegium Anthropologicum (Zagreb) 1986;10:145–58. 5. Neel JV. Diabetes mellitus: a thrifty genotype rendered detrimental by ‘progress’? Am J Hum Genet 1962;14:353–62. 6. Knowler WC, Pettitt D, Bennett PH, Williams RC. Diabetes mellitus in the Pima Indians: genetic and evolutionary considerations. Am J Phys Anthropol 1983;62:107–14. 7. Ritenbaugh C, Goodby C. Beyond the thrifty gene: metabolic implications of prehistoric migration into the new world. Medical Anthropology: Cross Cultural Studies in Health and Illness 1989;11(3):227–36. 8. Dowse GK, Zimmet PZ, Gareeboo H, Alberti KGMM, Tuomilchto J, Finch CF, et al. Abdominal obesity and physical inactivity are risk factors for NIDDM and impaired glucose tolerance in Indian, Creole, and Chinese Mauritians. Diabetes Care 1991;14:271–82. 9. Kriska AM, LaPorte R, Pettitt D, Charles M, Nelson R, Kuller L, et al. The association of physical activity with obesity, fat distribution and glucose intolerance in Pima Indians. Diabetologia 1993;36:863–9. 10. Hanley AJG, Harris SB, Barnie A, Gittlesohn J, Wolever TMS, Logan A, et al. The Sandy Lake Health and Diabetes Project: design, methods and lessons learned. Chronic Dis Can 1995;16(4):149–56. 11. Atkin CK, Freimuth V. Formative evaluation research in campaign design. In: Rice RE, Atkin CK, eds. Public communication campaigns. 2nd ed. Newbury Park (CA): Sage, 1991. 12. Gittelsohn J, Bentley ME, Pelto PJ, Nag M, Harrison A, Landman LT, eds. Listening to women talk about their health: issues and evidence from India. New Delhi: Har-Anand Publications, 1994. 13. Helitzer-Allen D, Kendall C, Wirima J. The role of ethnographic research in malaria control: an example from Malawi. Res Sociol Health Care 1993;10:269–86. 14. Helitzer-Allen D, Zukoski A, Newes G, Doniger A. Applying an ecological framework to understand factors contributing to early initiation of sexual activity and social desirability of teen pregnancy: a case study in Monroe County, New York. Health Education Research: Theory and Practice. In press. 15. Lloyd LS, Winch P, Ortega-Canto J, Kendall C. The design of a community-based health education intervention for the control of aedes aegypti. Am J Trop Med Hyg 1994;50(4):401–11. 16. Young TK. The health of native Americans: towards a biocultural epidemiology. New York: Oxford University Press, 1994. 17. Steckler A, McLeroy K, Goodman R, Bird S, McCormick L. Toward integrating qualitative and quantitative methods: an introduction. Health Educ Q 1992;19(1):1–9. 18. Coreil J, Augustin A, Holt E, Halsey NA. Use of ethnographic research for instrument development in a case control study of immunization use in Haiti. Int J Epidemiol 1989;18 Suppl 2:S33–7. 19. Helitzer-Allen D, Kendall C. Explaining differences between qualitative and quantitative data: a study of chemoprophylaxis during pregnancy. Health Educ Q 1992;19(1):41–54. 20. Hugentobler MK, Israel BA, Schurman SJ. An action research approach to workplace health: integrating methods. Health Educ Q 1992;19(1):55–76. 21. Jick T. Mixing qualitative and quantitative methods: triangulation in action. Adm Sci Q 1979;24(4):602–11. 22. Buchanan DR. An uneasy alliance: combining qualitative and quantitative research methods. Health Educ Q 1992;19(1):117–35. 23. de Vries H, Weijts W, Dijkstra M, Kok G. The utilization of qualitative and quantitative data for health education program planning, implementation and evaluation: a spiral approach. Health Educ Q 1992;19(1):101–15. 24. Dorfman LE, Derish PA, Cohen JB. Hey girlfriend: an evaluation of AIDS prevention among women in the sex industry. Health Educ Q 1992;19(1):25–40. 25. Scrimshaw SCM, Hurtado E. Rapid assessment procedures for nutrition and primary health care. Los Angeles: University of California at Los Angeles, Latin American Center Publications; 1987. 26. Manderson L, Aaby P. An epidemic in the field? Rapid assessment procedures and health research. Soc Sci Med 1992;35(7):839–50. 27. Gittelsohn J, Harris SB, Burris K, Kakegamic L, Landman L, Sharma A, et al. Use of ethnographic methods for applied research on diabetes among Ojibway-Cree Indians in Northern Ontario. Health Educ Q. In press. Author ReferencesJoel Gittelsohn, Division of Human Nutrition, Department of International Health, School of Hygiene and Public Health, The Johns Hopkins University, 615 N Wolfe Street, Baltimore, Maryland, USA 21205-2179; E-mail: JGITTEL@PHNET.SPH.JHU.EDUStewart B Harris, Thames Valley Family Practice Research Unit, University of Western Ontario, London, Ontario (formerly Sioux Lookout Programme, University of Toronto) Sara Whitehead, Medical Services Branch, Health Canada, Sioux Lookout, Ontario Thomas MS Wolever, Department of Nutritional Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario Anthony JG Hanley, Annette Barnie and Bernard Zinman, Diabetes Clinical Research Unit, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario Louisa Kakegamic, Sandy Lake, Ontario Alexander Logan, Department of Clinical Epidemiology, Mount Sinai Hospital, Toronto, Ontario Based on a paper presented at the 3rd International Conference on Diabetes and Indigenous Peoples: "Theory, Reality and Hope," held in Winnipeg, Manitoba, May 26–30, 1995. |
Last Updated: 2002-10-29 |