3.1. Study Population
We circulated questionnaires among participants in two viral hepatitis congresses: in Tehran (capital city of Iran, a national congress) and in Zanjan (in western Iran, a regional congress) and also among physicians of two university hospitals in Ahvaz (located in southern part of Iran). Participants returned 370 (82%) of the 450 distributed questionnaires. We excluded the results of one questionnaire due to the absence of more than 50% of the data. In this cross sectional study, we considered dentists, general practitioners (GPs), paraclinicians (laboratory specialists, pathologists, anesthesiologists, radiologists, and parasitologists), surgeons (gynecologists, general surgeons, and orthopedic surgeons) and internists (cardiologists, pediatricians, dermatologists, specialists in internal medicine, infectious disease, and emergency medicine) as different groups. This project consisted of different steps for producing and standardizing a questionnaire about KAP of physicians as it relates to hepatitis B and C. Further to designing the questionnaire as a first step, we conducted a cross sectional study of 60 doctors with different specialties, and had them answer the questions in a self-administered manner. Second step was followed by a reliability analysis and factor analysis for reducing the items as much as possible. As a third step, we had 10 experts ensure the validity of the questionnaire.
3.2. Designing and Standardizing the Questionnaire
A 29-item self-administered questionnaire assessing risk of transmission; seroconversion rates; the actual prevalence of HBV and HCV in Iran; vaccination against HBV; use of double gloves and protective eyewear; rate of needle stick injury and its reporting; checking the status of viral hepatitis; the use of disposable syringes and how they are discarded; and post-exposure prophylaxis had been designed. The questionnaire used the Likert scale; yes/no and a few open-ended questions, in addition to some demographics (age, gender, specialty, number of hours spent at work per week).
3.2.1. Item Generation
After a thorough search through the available literature, we planned two focus group discussions and one expert panel in order to design flowchart (Appendix 1
) and for characterizing the main domains of our KAP survey. Then, we detailed our main domains to some questions. Experts in the field of methodology (epidemiologists; specifically one who was the specialist in designing the questionnaire), a psychiatrist, a specialist in Community Medicine, an infectious diseases expert and gastroenterologists were most important part of our expert panel and focus group discussions.
3.2.2. Item Modification
According to the expert methodologists’ opinions, the structure and the content of some questions had been altered. A pilot study on 60 Iranian physicians with various specialties (participating in the scientific national congresses) was performed to assist in order to modify the structure and content of the primary questionnaire. Answers to each question were also revised according to the different answers to questions in the pilot study.
3.2.3. Item Reduction (Factor Analysis)
Factor analysis provides an enhanced understanding of which variables forms a “relatively coherent subset, independent of others” (18
). We performed this analysis on 370 subjects to realize if our most important domains (knowledge, attitude and practice) were categorized by this analysis in the same pattern we first categorized them. We planned to confirm the primary flowchart (Appendix 1
) of the questionnaire in this approach.
3.2.4. Item Standardization 126.96.36.199. Reliability
Internal consistency reliability (Cronbach’s Alpha), measuring the extent that the questions in each domain and all three main parts (knowledge, attitude and practice) tap a particular concept (19
) was determined according to the pilot study on 60 physicians with various specialties.
188.8.131.52. Face Validity
A separate sample of ten experts in the field of liver diseases and/or designing the questionnaire reviewed the questionnaire and answered the question: “How well do you think the questionnaire measures knowledge, attitude and practice of a physician about hepatitis B and C”? They responded using a 5-point Likert scale from 1 (not at all) to 5 (very well).
184.108.40.206. Content Validity
The content validity of the final questionnaire was determined according to the clarity, relevancy, simplicity, and consistency of each question with the questions set from 10 experts in the field of liver diseases (5 persons) and methodologists (5 persons). They examined the questionnaire for important omissions or inappropriate choice of items. To decrease the prestige bias, demographic variables were inserted at the end of questionnaire. For quality assurance, there was also guidance provided for some questions so the recipient would know how some specific questions should be answered.
3.3. Demographic Variables
Age, gender, specialty, place of work, and the quantity of activity in medical practice were among our demographic variables.
3.4. Data Analysis
We calculated the internal consistency of the questionnaire using Cronbach’s Alpha coefficient. Factor analysis was done for data reduction and grouping the related variables in conceptually similar and statistically related groups. The extraction method was aprincipal components analysis, the varimax rotation method, and we extracted factors based on an Eigenvalue larger than 1. Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett’s test of sphericity were used and the cut off point for loading on each factor was 0.3. We used mean ± SD for expressing quantitative variables and correlation test with Pearson coefficient for assessing the relationship between these variables. The analysis was done by SPSS 13 (SPSS Inc. Chicago, Illinois, USA) and Excel software. The authors considered differences and correlations, with P < 0.05 being statistically significant. For calculating the Item Content Validity Index (I-CVI), the average of our experts which believed that item is desirable/completely desirable was calculated and expressed as percent. The Scale Content Validity Index (S-CVI) for clarity, relevancy, simplicity, and consistency was calculated as the average of items which our experts believed that were desirable/completely desirable.
All subjects signed an informed written consent before participating in the study. The ethics committee of the Baqyiatallah Research Center for Gastroenterology and Liver Disease from Baqyiatallah University of Medical Sciences, Tehran, Iran approved the study proposal.