Study setting and population
South Asian people (defined as anyone of Indian subcontinent origin, whether from Pakistan, India, Bangladesh or Sri Lanka) with Type 2 diabetes over the age of 30 years were identified from day care centres catering for significant numbers of ethnic minorities, and three general practices which had more than 70% of their patients from different ethnic minority groups. Letters were sent to all individuals via their general practitioner and/or day care staff inviting them to participate in this study. If there was no response, patients were telephoned by the relevant care staff. One hundred and forty five patients, [Pakistani (n = 85), Indian (n = 33), White (n = 27)] were available and willing to participate. Before collecting the baseline data written informed consent was obtained from all patients. For those who were unable to read any language the consent form and details of intervention were read and explained by bilingual staff, and then their agreement was obtained. The study was approved by the Greater Glasgow Community/Primary Care local Research Ethics Committee.
Study design
We conducted a controlled trial in South Asians, and also studied a white group for comparative purposes. The South Asian group was divided based on gender, then each stratum was further divided based on their reading ability in any language. Finally minimization was performed for the resulting four strata to allocate individuals into intervention (n = 59) and control groups (n = 59). The White comparison group (n = 27) was separate (Figure 1). Before randomisation a questionnaire at the start of the study (baseline) assessed the patients' knowledge, attitudes, and practice, about diabetes. The development, content and structure of the questionnaire have been fully described elsewhere [10]. The outline of the questions used is shown in figure 2. After randomisation, baseline results were used to develop an appropriate, culturally competent, educational intervention about diabetes which was carried out by a bilingual health educator team (one podiatrist and one dietician). The baseline questionnaire was then re-administered at the end of the study to all subjects (control and intervention) with change in score as the primary outcome, and difference in changes in score as the secondary outcome.
Intervention
Despite the existence of several types of educational programmes for Type 2 diabetes, there was no specific standard for diabetes education, particularly for ethnic minority groups in developed countries available during this study. Therefore the questionnaire responses were analyzed and used to help develop the educational tool for sessions. In addition, recommendations of national and international diabetes organisations, such as Diabetes UK, and also information from previous studies were utilised for this purpose [11]. The educational intervention was carried out in day care centres and in GP's surgeries. It consisted of three sessions, one dietician-led sessions of about one hour's duration, and one podiatrist-led session of about one and a half hours' duration. They were carried out and completed within three months. The format of the educational programme was based on group education; the size of each group being between 6 and 12. The learning process in the groups was a combination of didactic elements (lecture) and interactive group discussion. During classes patients were asked to discuss some of their experiences and problems. Then the educator explained issues relevant to diabetes such as pathophysiology and cause of diabetes, short and long-term complications of diabetes, blood glucose control, recommendations for appropriate lifestyle changes (e.g., exercise, smoking cessation), nutrition recommendations (culturally appropriate), foot care, instruction about when and how to contact the physician or other members of the health care team when the patients were unable to solve acute problems themselves. The educators also used some simple support material including visual aids, food examples (real, models or packages, as applicable). Additionally some booklets and leaflets about diabetes, diet and foot care, translated into Urdu, Punjabi and Hindi, provided by Diabetes UK, were given to each patient after each session. Furthermore the educational team used an information video (25 minutes) in the second session which was recorded in their language.
Both team members were able to speak Punjabi, the preferred language that most patients speak at home. The Punjabi language is a common language even for those who also speak Hindi or Urdu. A total of 18 educational sessions in 6 separate programmes were carried out. Since patients had different religions and came from different cultures three programmes were implemented only for women, two programmes for men, and one programme for mixed women and men. In day care centres suitable rooms properly furnished and equipped with audio-visual instruments, a white board, and posters were used for this programme. In the GP's surgery one room was allocated for the diabetes educational intervention. This room was comfortable for patients and had similar equipment.
Analysis
To analyse data both descriptive and analytic statistics were applied. For descriptive analysis results are expressed as numbers, percentages, mean (± SD and 95% CI). Differences between the characteristics of the patient groups were compared, using Chi-Squared (for categorical variables) and t-test (for continuous variables). One-way analysis of variance (ANOVA) was applied to examine differences in variables between the three groups at baseline. Paired t-test was used to examine outcome by comparing baseline and follow-up assessments. The two-sample t-test was used to examine the changes in scores when comparing the intervention group with the ethnic control group and the changes in the two control groups. SPSS, version 11.5 for Windows (SPSS, Chicago, IL) was used.