Assessment of preventive practices towards hepatitis B infection among nursing students in Bangladesh: role of knowledge, attitudes and sociodemographic factors

This study was carried out among nursing students of the Barishal and Khulna divisions, specifically Khulna, Barishal, Jashore and Patuakhali districts in Bangladesh. These two divisions were purposefully chosen since, in Bangladesh, Barishal ranked second and Khulna ranked third, with 27% and 23% of cases of acute HBI, respectively [24]. Additionally, the incidences of nonalcoholic fatty liver disease and hepatic encephalopathy among chronic liver disease were also high [24]. Data collectors visited randomly selected nursing colleges. Participants were selected using a stratified random sampling technique. Stratification was performed considering various factors such as nursing program, year of study, and gender. The data was collected from the participants in their classroom during the break of their classes.

All the participants were in the diploma and midwifery nursing program. Male and female students were included within the age range between 18 and 26 because that is the satisfactory age range for nursing students. Second and final year students were chosen as participants for this research. The 1st year students were excluded from this study because of their less interaction with the patients. Moreover, nursing students undergoing internships were excluded due to their inaccessibility.

The Cochran formula was used to calculate the sample size [25]. The sample size was estimated based on the following parameters: (i) we used 50% as a predictive prevalence (i.e., p = 0.5) due to the lack of data on preventive practices towards hepatitis B among nursing students in Bangladesh, (ii) 95% confidence level (Z = 1.96), and (iii) 5% margin of error (e = 0.05).

Thus, minimum sample size for this study was, \( n=\frac{{z}^{2}pq}{{d}^{2}}= \frac{{\left(1.960\right)}^{2} \left(0.5\right) (1-0.5)}{{\left(0.05\right)}^{2}}\) = 384.16 ≈ 384

Using a 20% non-response rate, the optimum sample size was 461. Instead of 461 samples, we took 737 because larger samples better reflect the population and provide more accuracy.

A cross-sectional study was conducted between January and April 2023 among students at nursing colleges. A structured questionnaire was adopted from a previous study and modified to make it appropriate for our demographic [23]. For instance, we changed the demographic questions based on our country’s perspective (e.g., family income categories, academic year, etc.) and translated it to our native language for better understanding. Additionally, we changed some question categories to make it convenient for the students. During pilot testing, we asked the students about their understanding of the questionnaire and noted their feedback. Based on their feedback, we made some modifications without changing the actual meaning of the questions. The questionnaire was converted into Bengali from English and then back translated into English from Bengali by a translator to ensure that the questionnaire’s uniformity was maintained. An independent research assistant double-checked the procedure. Prior to data collection, a pilot test was done among 5% of our sample size. Uncertain questions were modified based on the results of the pilot testing. These 5% of participants were excluded in our main study population. A questionnaire-based data collection method was used to collect information. Prior to data collection, everyone was asked if he/she wanted to participate or not. We only collected data from those who willingly participated. If anyone said that they did not want to participate, we thanked them and went forward. Information was provided about the study’s aims and verbal consent was obtained from the participants. During the interview, individuals were also informed that their participation in this survey was entirely voluntary and they were under no obligation to continue.

The survey questionnaire consisted of a total of 43 questions, which included 9 demographic variables (Part A), 18 questions to determine knowledge (Part B), 8 attitude-related questions to determine attitude (Part C), and 8 questions related to practice to determine practice (Part D). All participants were required to answer all the questions from Part A to Part D.

The knowledge questions contained three options: Yes, No, and Not sure. The scoring was 1 for each correct answer and 0 for each wrong answer. The total score for knowledge was 18 and was categorized as low (0–9), moderate (10–14), and high (15–18). The attitude section was coded on five-point Likert scale, with responses of “strongly agree,” “agree,” “uncertain,” “disagree,” and “strongly disagree.” The scoring for the attitude section was done in the same way as the knowledge section, with 1 for each correct answer and 0 for each wrong answer. The total score for attitudes was 8 and was categorized as poor (0–4), moderate (5–6), and good (7–8). Practice scoring and categories were the same as attitude [23].

The quantitative data was analyzed using SPSS for Windows Version 25.0. Descriptive statistics, such as frequency, percentages, mean, median and standard deviation were used to analyze the demographic details of the students. Chi-square test was done to see significant difference among the distribution. Logistic regression was performed to estimate the relationship between a dependent variable and one or more independent variable. Multicollinearity was checked before establishing logistic regression. In multinomial logistic regression analysis, the significant variables and those with p-value < 0.25 in the chi-square function were entered into adjusted regression model. The reliability test of the instrument was performed and the Cronbach’s alpha of knowledge was 0.712, attitude was 0.653 and practice was 0.678.

Of the total 737 participants, the majority were female (88.3%). The participants were between 18 and 24 years old. Two-thirds of participants (65.4%) were studying in the third year of their study. The majority of the respondents (95.5%) had no family history of hepatitis B. Table 1 presents participants’ socio-demographic data.

The overview of knowledge levels about HBI prevention among nursing students are highlighted in Table 2. The majority of the participants (90.5%) knew that the dsDNA virus is responsible for HBI. The study participants showed varying levels of knowledge about the clinical manifestations of HBI, with 74.6% identifying jaundice as a symptom, more than a half (55.1%) acknowledging that not all infected individuals display symptoms and three quarters (75.6%) acknowledging the infectiousness of carriers (Table 2).

The summary of responses of attitudes towards HBI prevention is presented in Table 3. The majority of respondents (87.4%) believed that they were not at risk of getting a HBI, but 53.9% of them recognized that any blood exposure carries inherent risks (Table 3).

Table 4 illustrates the practices taken by the nursing students to prevent HBI. The frequency of the practices was reported as; hepatitis B screening (65.4%), hepatitis B vaccination (68.4%), vaccination ≥ 3 doses (52.6%), post-hepatitis B vaccination antibody testing (8.7%), changed gloves per client (67.6%), never recapped needles (28.4%), never splashed blood on the body (52.9%) and sometimes needle stick injuries in the past (50.6%).

The mean (SD) knowledge score among study participants was 11.42 (± 2.88) on a scale of 17.00. The mean (SD) attitude score was 4.33 (± 1.91) out of 8.00. The mean practice score for preventing HBI was 4.27 (± 2.056) out of 8.00.

Overall, two-thirds of the respondents (65.8%) knowledge was moderate, and one-fifth (21.3%) was low. Nearly half of the respondents (49.3%) had poor attitudes toward HBI prevention. 53.2% participants had poor practice of HBI prevention and quarter of participants (25.1%) had moderate practices (Fig. 1).

The chi-square test shows that participants’ marital status (chi-square, X² = 11.176, p = 0.004), family history with HBI (X² = 23.809, p = 0.000), knowledge (X² = 33.702, p = 0.000) and attitude (X² = 63.630, p = 0.000) were significantly correlated with the HBI prevention practice (Table 5). In addition, the findings of unadjusted multinomial regression analysis are shown in the Supplementary Table 1.

Table 6 represents the results of adjusted estimated effects concerning factors related to hepatitis B prevention practice. Nursing students with low knowledge regarding HBI prevention were found to be approximately three times more likely to engage in poor prevention practices when compared to those with a higher level of knowledge (aOR = 2.562; 95% CI = 1.29–5.07; p = 0.007). Nursing students who had poor attitudes towards preventing HBI were approximately seven times more likely to demonstrate poor HBI prevention practice in contrast to those with good attitudes (aOR = 5.730; 95% CI = 3.19–10.28; p = 0.000). Also, in contrast to unmarried students, married students were found to have a reduced likelihood of exhibiting poor practices for preventing HBI (aOR = 0.320; 95% CI = 0.13–0.82; p = 0.017). Nursing students in the 19–20 age group exhibited a threefold higher likelihood of showing poor practices for HBI prevention compared to those in the 23–24 age group (aOR = 3.038; 95% CI = 1.30–7.09; p = 0.010). In contrast to 3rd year students, 2nd year students were about twofold more likely to execute poor practices of HBI prevention than good practices (aOR = 2.147; 95% CI = 1.19–3.86; p = 0.011). However, nursing students with a family history of HBI were less likely to practice poor HBI prevention than those without one (aOR = 0.134; 95% CI = 0.05–0.36; p = 0.000) in contrast to those without.

This is one of the first studies to look into the factors that affect the practice towards the prevention of HBI among nursing students in the southern region of Bangladesh. Thus, the study’s findings give baseline data that will aid policymakers and public health specialists in developing preventive strategies to reduce HBI in Bangladesh. Another significant aspect of this study is its large sample size, as well as its appropriate methodological and statistical procedures. However, some limitations of our study should be considered. For example, we cannot determine the causal associations due to the cross-sectional study design. During the pre-testing phase of the questionnaire a small group of students were involved, and the same school was used for data collection. It is understandable that some questions may have been discussed by these students with their classmates who participated in the study later. Since the data was collected by the self-administered questionnaire, self-reported bias may be present. Furthermore, some environmental impacts may influence the students during data collection. Such as, the classroom used for data collection was designed for examinations which may put pressure on the students as they are attending the real exam. Despite all these probable influences, it is noteworthy that the data collection went smoothly without any external pressure. The participants completed the questionnaire independently. A clear clarification was given to all the students that their participation was entirely voluntary and that it was not an examination.