A cross sectional survey was distributed over twelve weeks between June and August 2008. A two stage sampling design was conducted which consisted of a pilot test followed by a cross-sectional multicenter study using self-completed anonymous questionnaires. The questionnaire was designed by the researchers following a literature review to identify the type of CAM previously used by people living with diabetes. The Sydney West Area Health Service (SWAHS) Scientific Advisory and the Human Research Ethics Committee (HREC), Nepean Campus approved the project (HREC Reference Number: 08/027/08/NEPEAN/42).
The pilot test involved the researcher distributing the questionnaire to five people with diabetes at a tertiary teaching hospital to establish face validity and identify any questions that needed to be altered prior to using the questionnaire in the main study. The pilot test was important to ensure the target audience understood the questions; the wording was appropriate and would yield the required data. Content validity was established by one researcher with expertise in CAM use and questionnaire design through a comprehensive literature review of similar studies and surveys. The pilot study was successful and no questions were modified prior to the main study.
The questionnaire comprised of 32 questions using a combination of open and closed question response formats [See Additional File 1
]. Questions were divided into three domains related to demographic data, diabetes-related information and CAM usage. Socioeconomic information sought included gender, age group, educational, marital and employment status and country of birth. Diabetes-specific information included type and duration of diabetes, frequency of blood glucose monitoring, last glycosylated haemoglobin A1c level, and presence of any diabetes complications. CAM-related information included type of CAM used and the reason for use.
Following a literature review, the researchers grouped the CAM into three broad categories to account for the potential variety of therapies used, which included: vitamins and minerals, herbal medicines and relaxation and other therapies. Information was collected about the following CAM therapies:
Vitamins/minerals: chromium, vitamin E, Co-enzyme Q10, L-carnitine, selenium, vitamin C, Vanadium, and magnesium.
Herbal products: ginseng (Panax quinquefolium), garlic (Allium sativum), onion (Allium cepa), ivy gourd (Coccinia grandis), holy basil (Ocimum tenuiflorum), cinnamon (Cinnamomum verrum), fenugreek (Trigonella foenum-graecum), milk thistle (Silybum mariunum), bitter melon (Momordica charantia) and Gymnena sylvestre.
Relaxation and other therapies: acupuncture, tai chi, massage, yoga, prayer, essential oils, and reflexology.
Although botanical names are cited here, the common names were used in the questionnaire because most respondents were more likely to recognise them.
The inclusion criteria included
• Those aged 18 yrs and over
• Known T1DM or T2DM and able to give informed consent to participate.
The exclusion criteria included
• Inability to communicate in spoken or written English
• Age under 18
• Those having participated in the pilot test.
• Those with intellectual disability or active psychiatric disease that prevented them from giving informed consent.
Following ethics approval, the main study involved distributing the questionnaire to individuals with diabetes in two separate private endocrine clinics in Parramatta and public outpatient clinics and medical wards of the Nepean Hospital in Western Sydney over a 12 week period. These centers were selected because the researcher was able to access the relevant clinics to recruit participants.
The researchers estimated that there were approximately 8 patients per week in both of the endocrine clinics and overall 24 patients from the public outpatient clinics and wards of Nepean hospital during the 12 week collection period, thus a total estimated sampling population of 120 individuals in Western Sydney. From this estimated sample, there were a total of 69 respondents.
In each setting, the treating endocrinologist or clinic health staff asked people with DM consecutively whether they would be willing to complete the anonymous questionnaire. A participant information statement that explained the aims of the project and indicated that the return of the questionnaire would be regarded as consent to use the anonymous information for the research purposes outlined was attached to each questionnaire.
Qualitative and quantitative data analysis processes were used to analyse the data. The data analysis was conducted using Statistical Package for the Social Sciences (SPSS; SPSS Inc, Chicago, IL) software 9 version 12) for Windows. Results for CAM and non-CAM users were recorded as mean ± standard error and, where appropriate, analysed using standard non-parametric methods including Chi-squared test and Fisher Exact analysis. A p
value of 0.05 was used to determine any statistically significant differences between CAM and non-CAM users. The qualitative content analysis of open-ended questions was undertaken using the framework method [10
]. The framework method consists of a five-step process that involves becoming familiar with the data, identifying a thematic framework, indexing and charting key themes and mapping and interpreting the findings.