We conducted a cross-sectional survey by means of a self-administered questionnaire. The survey was voluntary, with no identifiable information such as participants’ or centers’ name collected. The Ethics Committee of Tehran University of Medical Sciences approved the survey. The target population comprised all 241 physicians working in the public health centers of Tehran city in April 2011 who were eligible for the study. District Health Centers (DHC) agreed to distribute self-administered questionnaires to physicians, along with a pack containing a toothbrush and dental floss as a gift. Responses were anonymous. In total, 220 physicians returned completed questionnaires, for a response rate of 92% (Table ).
Distribution of physicians (n = 220) working in public health centers according to background characteristics in Tehran, Iran
Tehran has seven District Health Centers (DHC) with various functions including family health, environmental health, and dental health, and a medical education unit which took the responsibility of distributing the questionnaires in this study. Each DHC supervises 15 to 30 public health centers with a various number of physicians in each unit.
Questionnaire and variables
The questionnaire of the study was developed based on previous validated surveys [16
] with minor modifications. First, the questionnaire was assessed for content validity by experts in dental public health for relevancy and clearness. Then, a pilot study was conducted among physicians (n=30) working in the public health centers of Qazvin, a city near Tehran. Based on the pilot study (test/re-test at a two-week interval) and the participants’ opinion on the clarity of the questions, the questionnaire was slightly revised, and we omitted two questions with an actual agreement score under 70%.
Socio-demographics included the respondents' age, gender, their working profile (public only/public and private practice) and region (affluent/non-affluent).
The questionnaire contained statements on knowledge, attitude, and willingness for OHC education. The knowledge category included three domains: a) pediatric dentistry knowledge (12 questions), b) general dental knowledge (9 questions), and c) dentistry-related medical knowledge (13 questions). In the pediatric dentistry domain, questions included the timing of tooth eruption, the time/age to begin tooth cleaning and brushing for children and usage of fluoride for them, transmission of the bacteria that causes dental decay, the effects of pacifier sucking and mouth breathing, and the advantages of sealant therapy. The dental domain included questions on the first signs of dental decay and its etiology, the effects of fluoride and xylitol, the best time to refer a pregnant woman for a dental procedure, and the main causes of periodontal diseases. The medical domain consisted of questions on relations between systemic and periodontal diseases, drugs increasing risk for dental caries, and lesions in the oral cavity with need for biopsy. The responses were on the 5-point Likert scale with response alternatives ranging from "strongly agree" to ''strongly disagree,” including “don’t know.” In addition, we included four knowledge questions with multiple-choice answers.
Similarly, attitudes were measured with two statements on their opinion about the necessity of having OHC information and a request for the physicians to assess their competence in oral health. Their willingness was measured similarly by two statements on readiness to obtain more education and willingness to implement OHC activities.
We dichotomized answers to the knowledge and attitude questions with a score of one for correct/willing/positive answers, and 0 for false/unwilling/negative and don’t-know answers. Thereafter, we calculated a total score for total knowledge (theoretical sum score: 34) and for its three domains: pediatric dental (theoretical sum score: 12) general dental (theoretical sum score: 9) and medical (theoretical sum score: 13) according to socio-demographic background factors.
Evaluation of statistical significance of the differences between subgroups included the independent samples t-test for comparison of mean values and the Chi-square test for frequencies. To compare the domains, the scores were standardized by calculating the percentage of correct answers inside domains. Statistical significances between the standardized mean scores were evaluated using analysis of variance with repeated measurements and paired samples t-test.
Linear Regression Models were used to investigate functional relationships associated with physicians’ knowledge scores, controlling for background characteristics. Logistic regression models served for the multivariable assessment of factors related to the willingness. The corresponding odds ratios (OR) and their 95% confidence intervals (95% CI) were determined. Goodness of fit was assessed by means of the R square (0.1 < R2 < 0.2) and Hosmer and Lemeshow tests (p = 0.67).