Study setting
A cross-sectional study was conducted using a self-administered questionnaire among primary health care (PHC) physicians working at primary health care centres (PHCCs) under the Ministry of Health (MOH) in Riyadh, Saudi Arabia, during the period from September 2018 to March 2019.
According to the database of the General Directorate of Health Affairs in Riyadh, approximately 656 physicians work in 135 PHCCs of the MOH in the city of Riyadh. All PCPs (general practitioners or family physicians) working in PHCCs under the MOH in Riyadh were eligible for inclusion. All specialists who were not in direct contact with patients with diabetes, physicians who were on an extended leave of duty, and physicians who declined to participate were excluded.
A cluster multistage random sampling technique was employed. For the purpose of the study, PHCCs were clustered according to the city’s geographic divisions into five region (middle, southern, northern, eastern, and western), with 18–34 PHCCs in each region. Of these PHCCs, 10 PHCCs in each region were randomly chosen. Therefore, 50 PHCCs were included in the study.
Sample size
The sample size was calculated using a standard sample size equation "n=z2p(1-p)/e2" and an assumed proportion of 50% (proportion of medical physicians who had correct knowledge regarding pre-travel counselling for patients with diabetes). Using a 95% confidence interval and a 5% margin of error, the sample size was estimated to be 385 and was adjusted to 410 to compensate for the non-response rate.
Participants and survey instrument
All physicians present at the time of data collection in the selected PHCCs were included; hard copies of the questionnaires were delivered to the available physicians. Without seeing it first, the physicians were asked to complete the anonymous self-administered survey in English in order to assess their basic background knowledge.
The self-administered questionnaire was developed by the principal investigator (RD) based on the study objectives and after a literature review of similar studies.8-10 A panel of two diabetologists and one family physician, all of whom provide clinical care for patients with diabetes and are familiar with diabetes guidelines and the survey’s development, assessed the questionnaire for appropriateness, accuracy, and relevance and were asked to critique the questionnaire’s content. To ensure the face validity of the questionnaire, it was presented to a sample of 20 participants in a pilot study and then finalized. The results of the piloted questionnaires were not included in the analysis.
The questionnaire is divided into four sections with a total of 24 questions. The first section pertains to demographic characteristics, including age, gender, level of education, and nationality.
The second section assesses the physicians’ knowledge regarding diabetes and travel. The response choices for knowledge items include “yes”, “no” and “do not know”. Correct answers were scored as 1, while incorrect answers and “do not know” were scored as 0. The total knowledge score ranges from 0 to 10 (10 items). A higher score indicates more knowledge of the subject.
The third section assesses the attitudes of physicians towards pre-travel counselling for patients with diabetes. Five-point Likert scale items were used; strongly agree responses were scored as 5, agree as 4, uncertain as 3, disagree as 2, and strongly disagree as 1. The total attitudes score ranges from 6 to 30 (6 items), with higher scores indicating a higher degree of agreement. The fourth section assesses the practices of physicians towards pre-travel counselling.
Physicians at or above the mean score were considered to present good knowledge, high attitudes, or optimal practices, while those under the mean score were categorized as showing poor knowledge, low attitudes, or poor practices.
Data management and analysis plan
Data were coded and entered using the Statistical Package for Social Sciences Version 22 (IBM Corporation, Armonk, NY, USA). Categorical data were presented as numbers and percentages and were analysed using the chi-square test. KAP scores were calculated; scores ≥80% were considered good, scores of 60-<80% were considered moderate, and scores <60% were considered poor. Continuous data were tested for normality by using the Shapiro-Wilk test. Data that were not normally distributed were expressed as medians and interquartile ranges (expressed as 25th-75th percentiles). Continuous, non-normally distributed independent data were analysed using the Mann-Whitney U test, whereas continuous, non-normally distributed paired data were analysed by using Friedman’s two-way analysis of variance by ranks. P≤0.05 was considered statistically significant.
Ethical considerations
Approval for the study was obtained from the Institutional Review Board, College of Medicine, King Saud University (no. E-18-0488), Riyadh, Saudi Arabia. Official approval letters were obtained from the Directorate of Health Affairs in Riyadh. Each participant received the questionnaire and was informed about the objective of the present study. The Institutional Review Board has agreed that completing the questionnaire will imply consent.