Study design and subjects
This study utilized a non-experimental ‘modified’ cross-sectional survey design, whereby data collection took place at two time points, three to four weeks apart. The data collected at each time point was cross-sectional in nature. The study was conducted between October 2013 and August 2014. Subjects were recruited from two major health care facilities in Beirut: a philanthropic general hospital and a private academic medical center. The philanthropic hospital and the medical center are both accredited by the Lebanese Ministry of Health and attract a large proportion of the patient population in Lebanon. The philanthropic general hospital generally serves patients belonging to a lower socioeconomic class as compared to the private academic medical center. Ethical approval for this study was obtained from both sites where the study took place; the private academic center and the philanthropic hospital. At the academic medical center, approval was obtained from the Institutional Review Board (IRB), the division of the Social and Behavioral Sciences, under the following protocol ID: NUT FN.11. At the philanthropic hospital, the hospital’s ethics committee approved the study protocol.
Based on sample size calculations, a sample of 190 patients was needed to estimate CAM use prevalence among breast cancer patients, assuming a 95 % confidence interval, a 5 % margin of error, and an assumed prevalence of CAM use of 15 %. The latter prevalence was based on the findings of a previous investigation on CAM use prevalence among leukemia patients in Lebanon [
19].
Data collection
Recruitment of breast cancer patients took place at the waiting room in the clinics of the academic medical center and the philanthropic hospital. To ensure a representative cross-sectional sample of patients from the two recruitment sites, interviews were conducted on different days of the week and at varying times of the day.
As indicated earlier in this section, data collection took place at two separate time points. Through face-to-face interviews at the recruitment site, study participants completed a CAM-related questionnaire. Three to four weeks later, QOL assessment was carried by telephone. There were two main reasons for carrying out the QOL assessment by telephone and separate from the face-to-face interview at the hospital: First, research approvals obtained from the two participating health care facilities advised against long questionnaires, especially with cancer patients, and required the research team to carry the QOL questioning only for patients who, during the hospitals face-to-face interview, consented to being contacted later for future research. Second, the QOL questionnaire included a few sensitive questions about sexuality and intimate relationships which were deemed more appropriate to be asked over the phone as compared to face-to-face interviews.
At the hospitals, trained research assistants approached patients and obtained written consent from those agreeing to participate. Patients were reassured that their answers were confidential and would not be shared with their health care providers. The signed consent forms were kept separate from questionnaires in order to ensure that none of the consent forms can be linked to their corresponding questionnaires, thus protecting the anonymity of each patient. Completed questionnaires were kept in locked cabinets and electronic data was saved in password-protected computers with access available only to the investigators.
During the face-to-face interviews at the hospitals, patients completed the CAM questionnaire used in this study and which consisted of three sections; the first section included questions assessing socio-demographic and lifestyle characteristics of the study participants such as age, marital status, educational level, employment status, and health insurance. The second section included questions specific to breast cancer and general health, such as the duration of breast cancer, history of breast cancer in the family, and stage of the disease (early, locally advanced, or metastatic). The last section of the questionnaire included questions assessing the frequency and types of CAM used as well as the characteristics of the CAM use such as the factors influencing CAM choice, reasons for using CAM, rate of disclosure to treating physicians, and CAM-related side effects. CAM use was defined as using CAM at least once after breast cancer diagnosis. The types of CAM used were assessed using the following question: ‘What type of CAM product have you used?’ Patients were given seven choices to answer this question (more than one choice could be indicated whenever applicable). The choices were: vitamins and minerals; dietary supplements; special foods; herbal remedies/herbal preparations; spiritual healing; folk medicine (bloodletting/cupping); and ‘other’ (specify). These choices for the types of CAM included in this question were based on prevalent CAM modalities reported in other investigations of CAM use in Lebanon [
19,
21‐
23]. The content validity of the questionnaire was confirmed by a panel of experts consisting of an oncologist, nutrition epidemiologist and health policy expert. The questionnaire was originally written in English and then translated to Arabic by a professional translator. The translated Arabic version was back-translated by another professional translator to ensure the parallel-form reliability of the questionnaire. The original and the back-translated versions were reviewed for consistency in meaning by two bilingual experts. A copy of this questionnaire is found as Additional file
1.
Three to four weeks following the face-to-face interviews, the QOL assessment was carried out via telephone using a second questionnaire: the Functional Assessment of Cancer Therapy-Breast Symptom Index (FACT-B) questionnaire, the Arabic version (Additional file
2). FACT-B comprised of 37 items which measure multidimensional QOL in patients with breast cancer. It includes five subscales (Physical Well-Being (PWB), Social Well-Being (SWB), Emotional Well-Being (EWB), Functional Well-Being (FWB), and Breast Cancer Scale (BCS)). The face validity of this version of the FACT-B was recently examined among breast cancer patients in Lebanon, and involved both quantitative (face-to-face interviews with breast cancer patients (
n = 33)) and qualitative assessments (two focus groups (4 women per group)). Results indicated that for most of its subcategories, the instrument adequately tackled the different aspects that could possibly affect the QOL of women with breast cancer. In both the qualitative and quantitative assessments, the instrument was considered easy to follow, short, simple, culturally appropriate and pertinent to the women’s experience with the disease [
24].
For the QOL assessment, the research assistant was extensively trained to collect data over the phone, deal with sensitive topics, and address questions in a considerate manner with no judgment, tone, or attitude.
Statistical analysis
The questionnaires were checked for completeness, and responses were coded and entered into the Statistical Package for the Social Sciences (SPSS) software version 21.0 for Windows. Frequencies and percentages as well as means and standard deviations were used to describe categorical and continuous variables, respectively. CAM use, the main outcome in this study, was dichotomous and defined as either using CAM at least once after breast cancer diagnosis or not. Bivariate and multivariate logistic regression analyses were applied to determine the correlates of CAM use. Odds ratios and their respective 95 % confidence intervals were computed. The characteristics of CAM use, including factors influencing CAM choice, reasons for using CAM, and rate of disclosure to treating physicians, were presented as frequencies and proportions [n (%)]. The FACT-B QOL total score and the scores on each of its five subscales were calculated as means and standard deviations. The difference in QOL scores between CAM users and non-users was assessed using an independent sample t-test. A p-value of 0.05 was used to determine statistical significance.