Unformatted Attachment Preview
BMC Medical Research Methodology BioMed Central Open Access Study protocol Methodology of a diabetes prevention translational research project utilizing a community-academic partnership for implementation in an underserved Latino community Philip A Merriam*1, Trinidad L Tellez2, Milagros C Rosal1, Barbara C Olendzki1, Yunsheng Ma1, Sherry L Pagoto1 and Ira S Ockene3 Address: 1Division of Preventive and Behavioral Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester, MA, USA, 2Greater Lawrence Family Health Center, Lawrence, MA and Department of Family Medicine and Community Health, University of Massachusetts Medical School, Worcester, MA, USA and 3Division of Cardiovascular Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester, MA, USA Email: Philip A Merriam* – Philip.Merriam@umassmed.edu; Trinidad L Tellez – TTellez@glfhc.org; Milagros C Rosal – Milagros.Rosal@umassmed.edu; Barbara C Olendzki – Barbara.Olendzki@umassmed.edu; Yunsheng Ma – Yunsheng.Ma@umassmed.edu; Sherry L Pagoto – Sherry.Pagoto@umassmed.edu; Ira S Ockene – Ira.Ockene@umassmed.edu
* Corresponding author Published: 13 March 2009 BMC Medical Research Methodology 2009, 9:20 doi:10.1186/1471-2288-9-20 Received: 4 November 2008 Accepted: 13 March 2009 This article is available from: http://www.biomedcentral.com/1471-2288/9/20 © 2009 Merriam et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: Latinos comprise the largest racial/ethnic group in the United States and have 2–3 times the prevalence of type 2 diabetes mellitus as Caucasians. Methods and design: The Lawrence Latino Diabetes Prevention Project (LLDPP) is a community-based translational research study which aims to reduce the risk of diabetes among Latinos who have a ≥ 30% probability of developing diabetes in the next 7.5 years per a predictive equation. The project was conducted in Lawrence, Massachusetts, a predominantly Caribbean-origin urban Latino community. Individuals were identified primarily from a community health center’s patient panel, screened for study eligibility, randomized to either a usual care or a lifestyle intervention condition, and followed for one year. Like the efficacious Diabetes Prevention Program (DPP), the LLDPP intervention targeted weight loss through dietary change and increased physical activity.
However, unlike the DPP, the LLDPP intervention was less intensive, tailored to literacy needs and cultural preferences, and delivered in Spanish. The group format of the intervention (13 group sessions over 1 year) was complemented by 3 individual home visits and was implemented by individuals from the community with training and supervision by a clinical research nutritionist and a behavioral psychologist. Study measures included demographics, Stern predictive equation components (age, gender, ethnicity, fasting glucose, systolic blood pressure, HDLcholesterol, body mass index, and family history of diabetes), glycosylated hemoglobin, dietary intake, physical activity, depressive symptoms, social support, quality of life, and medication use. Body weight was measured at baseline, 6-months, and one-year; all other measures were assessed at baseline and one-year. All surveys were orally administered in Spanish. Results: A community-academic partnership enabled the successful recruitment, intervention, and assessment of Latinos at risk of diabetes with a one-year study retention rate of 93%. Trial registration: NCT00810290 Page 1 of 9 (page number not for citation purposes) BMC Medical Research Methodology 2009, 9:20 Background Latinos are the largest minority group in the United States representing 13.7% of the total population [1]. The Centers for Disease Control analyzed data from the Behavioral Risk Factor Surveillance System (BRFSS) and found that Hispanics continue to have a higher prevalence of diabetes than non-Hispanic whites [2]. Cultural Health Promotion Plan in Latino Hispanic Community
Overall, 7.4% of Hispanics in the BRFSS had been told by a doctor that they had diabetes. Given the increasing prevalence of sedentary lifestyle and obesity and their correlation with diabetes and heart disease [3,4], it is likely that the number of individuals with type 2 diabetes mellitus will continue to increase, and that this will be an especially significant burden among Latino communities. Latinos have a very high risk of developing diabetes in their lifetime – a 50% probability for Hispanic women, compared to the approximately 1 in 3 chance for the average American, making primary prevention of type 2 diabetes an important priority in this population [5]. The Diabetes Prevention Program (DPP) was a randomized clinical trial that successfully demonstrated that modest weight loss and increased physical activity could reduce the incidence of diabetes in a group of pre-diabetic patients by 58% [6]. However, both the recruitment methodology and the intervention were very costly. The trial included 27 centers recruiting approximately 1 participant/center/week over a 3 year period at a cost (excluding staff) of approximately $1075 per randomized participant [7]. The DPP intervention began with 16 weekly one-hour individual intervention sessions carried out over 24-weeks and continued with monthly individual and group sessions [7]. Cultural Health Promotion Plan in Latino Hispanic Community
The methodology used in the DPP may be too expensive to implement in real world settings, requiring less expensive methods to be developed and tested [8,9]. The primary objective of the Lawrence Latino Diabetes Prevention Project (LLDPP) is to design and test a less intensive intervention that, like the DPP, targets weight loss through dietary change and increased physical activity, in order to reduce the risk of type 2 diabetes in a lowincome Latino community. The LLDPP study methodology was designed to decrease the high cost of screening and recruitment seen in the DPP, in part by using an accurate but inexpensive screening procedure based on a predictive equation that weights age, gender, ethnicity, fasting blood glucose (FBG), systolic blood pressure, high density lipoprotein (HDL-C), body mass index (BMI), and natal family history of diabetes to estimate relative risk of developing diabetes in the subsequent 7.5 years [10]. See Figure 1 for Stern formula. The purpose of this paper is to present the methodology used to translate the DPP program to a diabetes-prevention research project in a high-risk Latino community, http://www.biomedcentral.com/1471-2288/9/20 and to highlight the community-academic partnership formed to carry this out. Methods and design Setting This project was carried out in the city of Lawrence, Massachusetts, a primarily Latino community characterized by families struggling with high levels of poverty, limited access to jobs, and limited access to resources families need to prosper [11]. Community and academic collaborators Study planning and implementation involved collaboration among the Greater Lawrence Family Health Center (GLFHC), the Lawrence Council on Aging (LCOA)/Senior Center, the YWCA of Greater Lawrence, and investigators from 2 campuses of the University of Massachusetts (UMass). As the study progressed, the Mayor’s Health Task Force joined the partnership. The study principal investigator (PI) and a co-PI are UMass Medical School (UMMS) faculty, and the community-PI is a Greater Lawrence Family Health Center (GLFHC) physician and UMMS faculty member. The PI is also a UMass Memorial Medical Center (UMMMC) physician. The GLFHC provides healthcare to approximately 80% of the Lawrence Latino population. The health center houses a UMMS-affiliated Family Medicine Residency program and a CDC-funded Racial and Ethnic Approaches to Community Health (REACH) diabetes disparities reduction project. Cultural Health Promotion Plan in Latino Hispanic Community
The study administrative support for this project was based at GLFHC under the direction of the community-PI, who bridged the partnership among the 6 community and academic collaborators. The Lawrence Council on Aging (LCOA)/Senior Center, a conveniently located and well-respected social service facility, housed all study screening, recruitment, and assessment appointments. The LCOA and YWCA each provided a community coordinator to staff the study through subcontract arrangements. The community coordinators were chosen for being well-known, respected, and having longstanding relationships and community work experience within the community of Lawrence. Figure 2 presents the study partnership diagram. Study management The partners in this collaboration maintained formal communication through thrice-yearly meetings held at the Lawrence Senior Center, and also informally through regular telephone contacts and e-mail. The study staff was divided into a recruitment/retention team and an intervention team; both teams met monthly. The project director, community-PI, and lead nutritionist shared responsibility for team oversight. The project director, a Page 2 of 9 (page number not for citation purposes) BMC Medical Research Methodology 2009, 9:20 Ͳ • • • • • • • • • Ͳ Ͳ Ͳ http://www.biomedcentral.com/1471-2288/9/20 Ͳ Ͳ Figure Stern formula 1 Stern formula. UMMS faculty member with a background in research management and social work, had primary responsibility for the smooth operation of the study on a day-to-day basis, visiting the clinic at least weekly to meet with study staff, and coordinating the two monthly team meetings. The community-PI, a practicing Latina physicianresearcher who was responsible for bridging the community-academic partnership, oversaw the local administrative operations and met regularly with the recruitment/ retention team. The lead clinical research nutritionist, who had worked with the co-PI Latina behavioral psychologist to design and develop the intervention materials and train the community intervention team, also met regularly with them [12,13]. Daily management of study activities was facilitated by the use of Lotus Notes/IBM tracking system software (Lotus Notes R5.0.11 ®). Cultural Health Promotion Plan in Latino Hispanic Community
The database was kept on a server at UMMS which could be accessed easily by study personnel in Lawrence and at UMass. Multiple levels of password protection were used to ensure data security. Recruitment and community outreach The primary recruitment outreach method entailed drawing from the GLFHC patient panel by identifying potentially eligible patients who received a mailed letter of invitation, and then telephone recruitment calls from the study community coordinators. A mailing list was generated every 6 to 8 weeks by running a screening query of the current GLFHC database to identify potentially eligible Latino patients with an age ≥ 25 years who had a high likelihood of meeting eligibility criteria (e.g., overweight, history of hypertension, low HDL-C, or FBG 100–125 mg/ dl, and not diabetic). A 2nd query was run to remove patients who had already been approached to be in the study; with approximately 250 names randomly selected for each mailing. Personalized patient screening invitation letters were created, signed by the patient’s primary care physician (PCP) and the community-PI, and then mailed. Patients were eliminated by their PCPs if deemed ineligible or a poor study candidate (such as having severe psychiatric illness, etc.). PCPs were kept engaged through regular updates at provider meetings and through communications by the community-PI. The selected GLFHC patient names were downloaded into the Lotus Notes tracking database, and divided equally between the two community coordinators for subsequent telephone outreach. Additional outreach methods included public service announcements on public access television, guest spots on local Spanish radio programs, advertisements in the local Spanish and bilingual newspapers, flier inserts in the Senior Center newsletter, and mailings to non-GLFHC physicians with the purpose of creating awareness of the study. Telephone pre-screening Pre-screening activities were conducted by each community coordinator who followed up on the mailed invitation letters with telephone calls two weeks after the mailing, unless an individual had already responded and declined further contact. The latter occurred rarely. The initial telephone contact included an assessment of preliminary eligibility, an invitation to schedule a fasting screening appointment, and if scheduled, instruction to bring in all current medications. Reminder calls were made to patients the day before, and/or the morning of, all scheduled appointments. Cultural Health Promotion Plan in Latino Hispanic Community
Screening appointment In addition to those patients scheduled for screening appointments as described above, patients were also screened as walk-ins if they learned of a screening event via word-of-mouth or community outreach. Following a screening protocol, a community coordinator would explain the study, highlight what would occur at the visit, and obtain a signed screening consent form which was available in English and Spanish. Individuals then were administered a one-page survey which assessed diabetes risk perceptions, and underwent the study screening. All current medications were recorded, and anthropometric measures were taken by the clinic assistant. These included height and weight (without shoes and outerwear), blood pressure (after sitting quietly for 2 minutes) using the Dinamap XL® automated BP monitor, and a fasting fingerstick lipid profile and glucose measure (Cholestech LDX System®). The Stern predictive formula value and BMI were calculated using a Microsoft Excel® program. The community Page 3 of 9 (page number not for citation purposes) BMC Medical Research Methodology 2009, 9:20 http://www.biomedcentral.com/1471-2288/9/20 agents (not at stable dose for 6 months or more), or prescription weight loss medications. Nine hundred and forty-nine individuals had screening appointments during the screening phase which spanned 34 months, beginning on October 10, 2004. Figurepartners Study 2 Study partners. coordinator informed the screened individual of the results of their blood pressure, FBG, weight and BMI, total cholesterol and HDL-C, both verbally and in writing (written into a brief educational brochure), and their potential study eligibility. Patients were informed that their PCP would be receiving the results directly, and were encouraged to review their results with their PCP. Those Latino individuals who were ≥ 25 years of age, had a BMI ≥ 24 kg/m2and a ≥ 30% likelihood of developing diabetes in 7.5 years as predicted by the Stern equation were determined to be pre-eligible and invited to schedule a baseline appointment in 3–4 weeks. Each pre-eligible individual’s PCP was mailed a medical clearance form that reviewed the eligibility criteria, and asked for the PCP’s permission for the individual to participate. Through screening or the PCP medical clearance form, the following criteria defined an ineligible state: a fasting glucose of 126 mg/dL or greater, inability or unwillingness to give informed consent, clinically diagnosed diabetes, a plan to move out of the area within the study period, presence of a psychiatric illness which limits ability to participate, no telephone, inability to walk unaided or walk five city blocks (1/4 mile) without stopping, having a medical condition likely to limit lifespan, taking a medication or having a medical condition that interfered with the assessment for diabetes, or having an endocrine disorder that alters blood sugar. In addition to asking these questions on the medical clearance form, all pre-eligible individuals’ medication lists and screening flow-sheets were reviewed by the community-PI for study contraindications which included beta-blocking agents (not at stable dose for 3 months or more), thiazide diuretics at doses higher than 25 mg/day, niacin in pharmacologic doses, systemic glucocorticoids, protease inhibitors, atypical antipsychotic Baseline and follow-up assessments Recruitment into the study occurred at the baseline appointment. Cultural Health Promotion Plan in Latino Hispanic Community
Each individual’s community coordinator explained the study again, this time in more detail and highlighting the commitment to three visits over the oneyear study period; and a second study consent form was reviewed and signed. The participant was given a copy of the informed consent, a study brochure, a flier reminding them of the 3 pending telephone assessment calls, and the expected dates of their 6-month and 1 year follow-up appointments. Each participant was also given a food portion visual handout for reference during the telephone assessments and community coordinator contact information. All study participants completed intervieweradministered assessments which included demographic questions (age, gender, education, occupation, and household data), as well as social support (the Medical Outcomes Study scale) [14], depressive symptoms (the Center for Epidemiological Studies Depression Scale (CES-D) [15], and quality of life (SF-12) questions [16]. The baseline assessment visit also included anthropometric measures (weight, height and waist circumference); two blood pressure readings (ten minutes apart); and a fasting venous blood sample for lipid, glucose, and HbA1c assays. Serum and plasma aliquots were prepared and the buffy coat layer saved. A serum aliquot was sent to the University of Massachusetts Lowell for the lipid profile and a plasma aliquot for the glucose assay; and a frozen whole blood sample (with EDTA) was sent to the Diabetes Diagnostic Laboratory at the University of Missouri for analysis of HbA1c measures. Extra serum and plasma was saved from those who provided informed consent for additional studies. Three randomly selected 24-hour dietary and physical activity assessments (24 HR) [17] (NDSR-2007®) were conducted by trained bilingual Spanish-speaking registered dietitians not involved in the intervention and blinded to participant’s condition, via unannounced telephone interviews (on two weekdays and one weekend day) within the following 2 weeks of the assessment visit. Study participants were asked to refer to a food portion visuals booklet they had received at the baseline assessment to facilitate portion size estimation. At 6 months post-baseline, a measure of weight was scheduled. At one year, the measures collected at baseline were repeated and included demographics, one-year quesPage 4 of 9 (page number not for citation purposes) BMC Medical Research Methodology 2009, 9:20 tionnaire, Stern predictive equation variables, weight, laboratory measures, and 24 HR. Cash incentives of $25 were given at the baseline visit and the 6-month assessment; and $50 was given for study completion at the one-year assessment. Cultural Health Promotion Plan in Latino Hispanic Community
The Institutional Review Boards of the University of Massachusetts Medical School and Greater Lawrence Family Health Center approved the subject recruitment and data collection procedures. The complete process is outlined in Figure 3. Randomization and study conditions After recruitment and completion of baseline data collection, individuals were randomized to receive the usual care (UC) or to a lifestyle intervention (LI) condition. Randomization occurred at the household level. If any new participant was from the same household as someone already in the study, they were assigned to the same condition already assigned to that household. Lifestyle intervention A group-based intervention (13 group sessions) complemented by 3 individual home visits was develope … Cultural Health Promotion Plan in Latino Hispanic Community