Health Professionals’ Knowledge and Attitude toward Patient Confidentiality and Associated Factors in a Resource-Limited Setting: A Cross-Sectional Study | BMC Medical Ethics
Design and implementation of the study
An institutional cross-sectional study was conducted among health professionals from August to September 2021. Gondar is a historic city located in the northwestern part of Ethiopia, 772 km from the capital Addis Ababa and 168 km by Bahir Dar. . The University of Gondar Specialty Hospital is one of the largest teaching hospitals in the Amhara region, providing tertiary level care to over seven million people in the northwestern part of the country in from the regions of Amhara, Tigray and Benishangul Gumuz. . It has 960 health professionals in 30 service units responsible for delivering health services to an average of 800 patients per day.
Study Subjects and Eligibility Criteria
All healthcare professionals working at the University of Gondar Specialist Hospital and those available during the study period were the sources and the study population. The study excluded healthcare professionals with less than six months of experience, those who had not been found in hospital for various reasons and those on annual leave during the data collection period.
Sample size determination and sampling procedure
The sample size was calculated using the single population proportion formula, n = Z(α/2)2 pq/d2 . We assumed: n = the required sample size, Z = the value of the standard normal distribution corresponding to α/2, 1.96, p = proportion of healthcare professionals who had a good knowledge and attitude towards confidentiality, q = proportion of health professionals who had knowledge and attitude towards confidentiality, and d = precision assumed equal to 0.05. To our knowledge, no studies have been conducted in Ethiopia to determine healthcare professionals’ knowledge and attitude towards patient confidentiality. Therefore, we assumed that p (proportion of health professionals who had good knowledge and a good attitude towards confidentiality) was 0.5. Therefore, the required sample size was calculated at 384. After adding a non-response rate of 10%, 423 healthcare professionals were enrolled in the study.
Stratified with a simple random sampling method was used to select the 423 participants. First, the sample was stratified according to their department. Then, the selection was distributed proportionally in each stratum based on the number of health care providers in each stratum or department to assess their knowledge, attitude and associated factors related to patient confidentiality. After proportionally distributing the samples in each stratum, a simple computer-generated random sampling technique was used to select the study subjects in each department (Fig. 1).
The main outcome variable of this study was knowledge and attitude towards patient confidentiality. The questionnaires used in this study were developed based on a review of related literature [20, 21]. Sociodemographic and work-related characteristics were used as independent variables in this study.
Knowledge of patient confidentiality was assessed using seven items with ‘yes’ and ‘no’ responses. Each correct answer was worth one point, while each incorrect answer was worth zero points, with a possible height score of 7 for the knowledge part. An average of 7 questions regarding knowledge about patient confidentiality was calculated. And those above the average score were classified as “good” knowledge, and those below were classified as “poor” knowledge. .
Attitudes toward patient confidentiality were assessed using 14 questions with a 5-point Likert scale ranging from “strongly disagree” (score 1) to “strongly agree” (score 5) . The final score in the attitude section ranges from 14 to 70. An average of the 14 attitude toward patient confidentiality questions was calculated. Those above the average value were classified as “favorable” attitude, and those below the average value were classified as “unfavorable” attitude. .
Data collection and quality control tool
A self-administered, organized and pre-tested questionnaire was created in English. The data collection process included two supervisors and ten data collectors. A one-day training was given to data collectors to remove ambiguities. A pre-test was carried out outside the study area, in the health centers of the city of Gondar, with 10% of the population studied. The validity and reliability of the data collection instrument were assessed using the results of the pre-test. Cronbach’s alpha for the attitude questions was 0.82, while Cronbach’s alpha for the knowledge questions was 0.76. These figures show that the questionnaire is very reliable.
Data processing and analysis
Data entry was performed using Epi Data version 4.6 software packages and analyzed using the Statistical Package for the Social Sciences (SPSS) version 25. Descriptive statistics were calculated to describe socio-demographic variables and knowledge. and attitudes of healthcare professionals regarding patient confidentiality and data sharing. . Bivariate and multivariate binary logistic regression analyzes were performed to measure the association between dependent and independent variables. In bivariate regression analysis, variables with a p less than 0.2 were included in the multivariate regression analysis to assess their adjusted impacts on the dependent variables. Odds ratio with a confidence level of 95% and P value were calculated to determine the strength of association and to decide on statistical significance. For all significantly associated variables, the threshold value was p