Abstract
Infection with the liver fluke Opisthorchis viverrini is the major causative factor inducing cholangiocarcinoma in the Mekong region of Southeast Asia. Northeast Thailand has the highest incidence of this cancer worldwide leading to about 20,000 deaths every year. Infection with the liver fluke comes from eating raw or undercooked fish, a tradition in this area that can potentially be countered by education programs at school level. Here we develop a school-based health education model, based on protection motivation theory (PMT), including module design, learning materials, student activities, and capacity building amongst teachers. This education program was applied and tested in primary school to pupils (9–13 years) in Khon Kaen province, northeast Thailand. Using a randomized control trial, four schools served as intervention groups (n = 118 pupils) and another four acted as controls (n = 113 pupils). Based on PMT constructs, we found that the pupils in the intervention group had significantly greater knowledge and perceived the severity, vulnerability, response efficacy, and self-efficacy parameters concerning the dangers of eating raw fish and of developing cholangiocarcinoma than those in the control schools (p < 0.05). All of the PMT constructs measured were significantly intercorrelated with each other (p < 0.001). At the same time, some background knowledge, from community-based education programs, was present in the control schools. The result from this initial study suggests that PMT can be used to predict protective attitude as well as behavior changes in evaluating the consequence of school health intervention programs.
Introduction
Cholangiocarcinoma (CCA) is a significant health problem in continental East and Southeast Asia, particularly in the countries bordering the Mekong River. The incidence of CCA in Khon Kaen province, located in northeast Thailand, is the highest in the world, ranging from 93.8 to 317.6 per 100,000 persons/year between 1990 and 2001 (1,2). The National Cancer Registry of Thailand also shows that liver and bile duct cancer are the most common cancers in both males and females in northeast Thailand, unlike the rest of the country (3).
The major risk factor for developing CCA in Southeast Asia is infection with the liver fluke Opisthorchis viverrini sensu lato (4). This liver fluke is widespread in the Lower Mekong Basin including Thailand, the Lao People’s Democratic Republic (Lao PDR), Cambodia, and south and central Vietnam (2,5). Approximately eight million people in Thailand are infected, of whom more than two thirds live in the northeast of the country (1). The national helminth survey showed that the overall prevalence of O. viverrini was 8.7%, being highest in the northeast (16.6%), followed by the north (10.0%), the central region (1.3%), and the south (0.01%) (1,6). A recent mass survey in 2017 in northeast Thailand (n = 18,436) revealed that the prevalence of O. viverrini was 11.3% (Sithithaworn, unpublished data). The age group with the highest prevalence were 40–60+ with the highest district prevalence being between 70% and 84% (7). This is less than the prevalence found previously, which averaged over 60% between 1984 and 1987 (8), but approached 100% in surveys in the 1980s (9).
The liver fluke infection occurs with the consumption of raw or undercooked freshwater, cyprinid fish carrying the parasite (10). The consumption of raw freshwater fish along the Lower Mekong Basin, including northeast Thailand, is an embedded cultural tradition (10–14). The favorite dishes are called koi-pla, pla-som, and a fermented fish call pla-ra. The northeastern Thai population is reliant on freshwater fish as a major source of dietary protein and these foods are eaten regularly (10). To date, health education programs to avoid raw and undercooked fish have proven ineffective in altering the eating behavior predominantly in the older villagers (15).
Control programs for O. viverrini infection started in the 1950s with measures against helminths in Thailand. Since early 1980s, there have been sporadic and local health education programs and the use of the anthelminthic drug praziquantel (16).
The significance of health education programs in primary schools has been studied in various global contexts, including Thailand (17–21). The essential activities in these schools are health education and promotion, hygiene awareness, and improved sanitation. Providing scientifically accurate information in a manner that is accessible to children and young people helps improve their awareness of public health problems and stimulates precautionary behavior (22). In the past (prior to 1983) the prevalence of O. viverrini in school age children was high (14–37%) in Thailand and Lao PDR (9,23,24). However, in Thailand the prevalence of infection has dropped substantially over the last three decades (6,9,25). For long-term and sustainable control in many infectious diseases, school-based health education is strongly recommended by the World Health Organization (26). Thus, in Thailand school children, who will eventually enter the adult population, should be the target group for the liver fluke and CCA control program (27).
Generally, theory-based interventions are more successful in achieving stable and permanent behavioral changes than are non-theory-based interventions. Protection motivation theory (PMT) is one model that explains why people engage in unhealthy practices and offers suggestions for changing such behaviors via public health campaigns (28,29). Being both educational and motivational (30), PMT proposes that we protect ourselves based on four factors: the perceived severity of a threatening event (in this case opisthorchiasis and CCA), the perceived probability of the occurrence, or vulnerability (consumption of koi-pla, pla-som, pla-ra), the efficacy of the recommended preventive behavior (consuming of cooked fish only), and the perceived self-efficacy (cooking freshwater fish).
Presently, no formal school-based health education program to prevent and control opisthorchiasis and CCA is available in Thailand. Therefore, our aims were to assess the efficacy of a school-based health education model, based on PMT, in school children, including module design, learning materials, student activities, and capacity building amongst teachers. The outcome from this study is expected to provide solid foundation for further improvement and modification for future expansion to more primary schools in northeast Thailand as part of the liver fluke control and elimination program. Here we present data on an initial trial of the program used.
Methods
Study design and sampling method
A cluster randomized control trial (RCT) following the CONSORT guideline (updated 2010: http://www.consort-statement.org/) was conducted in 2012–2013 in Khon Kaen Province, northeast Thailand. The target population was drawn from 210 primary schools under the responsibility of Khon Kaen Primary Educational Service Area Office 2. For inclusion criteria, the primary schools were stratified into two groups, those using the health education package and those not using the package. The 25 schools where health education was implemented are grouped in the ‘intervention strata’, and the 185 schools without health education implementation were grouped into ‘other strata’. The next step was to randomly select four schools as the ‘intervention group’ and another four schools as the ‘control group’. The key participants were primary school children aged 10–12 years old (level 4–6). The eight primary schools chosen were at least 5 km apart from each other (see Supplementary Figure 1 online). The total number of students in levels 4–6 was more than 30 for each school, and they all had reasonable literacy standards. Schools were chosen according to the inclusion criteria (see below) with the use of RCT. In each chosen school, 30 students were recruited randomly for the study.
The sample size was calculated based on the estimation for comparing two independent means. Pool variance between the mean of knowledge of two groups was calculated according to Sota et al. (22). The significance level was set at α < 0.05 and the power of the analysis at 1 – β = 0.90. The sample size required for this study was calculated to be 105 per group and the number of clusters per group was 4.
Ethical approval for this study was obtained from the Human Research Ethics Committee covering clinical trials, at Khon Kaen University (HE57IO7). All participants and relevant persons provided written consent, including school principals, teachers, parents and students. At the end of the study the same teaching package were supplied to the control schools for their own future use.
The teaching program
The program was developed based on PMT in 2012–2013 by a teacher–researcher team. It consisted of seven units: orientation, CCA, liver fluke, risky food of liver fluke infection and liver fluke free Isan (the regional name for the northeast of Thailand), recipe inspector and evaluation (see Supplementary Table 1). The program activities were divided into three types: counseling activities, student activities, and activities for social and public interest, all provided by the teacher. Then the students were enrolled as members of the ‘Junior Food and Drug Administration Club’. They were asked to learn about opisthorchiasis and CCA by self-learning and showed their understanding using drawings, pictures, pop-ups, mini pocket books, and mind maps. The outputs included both individual and group exercises. Since they gained knowledge and perception of the severity of opisthorchiasis and CCA, they were assigned to inspect risky dishes eaten by their families. They would discuss the problems with their parents and take note of what kinds of risky foods were found in their own homes and those of their relatives and neighbors. This enabled the children to realize that risk factors related to opisthorchiasis and CCA were found at home and in the community.
The effect of this program was evaluated once at the end of the intervention. The knowledge and perception of school children related to the prevention and control of opisthorchiasis and CCA were assessed by a self-administered questionnaire. There were 10 questions on the child’s knowledge of liver flukes, the consequences of liver fluke infection, CCA and prevention measures, such as: ‘Do you get opisthorchiasis by eating raw freshwater fish?’, ‘Can you get opisthorchiasis by skin penetration of liver fluke when you swim in canal or swamp?’ The internal consistency coefficient was good (α = 0.84).
The four components of PMT were measured. Perceived severity was measured using five items, for example, ‘Opisthorchiasis can cause CCA’. Again, the item responses were measured on a three-point Likert scale ranging from 1 (agree), 2 (not sure), 3 (disagree). The internal consistency coefficient was well above acceptable level (α = 0.71).
Perceived vulnerability was assessed by five items, for example, ‘The one who most often eats raw fish will always get opisthorchiasis’ and ‘If we continue to eat raw fish, we will get repeated liver fluke infection’. Item responses were again measured on the three-point Likert scale. An acceptable internal consistency of α = 0.75 was found.
Perceived response efficacy was assessed using five items, for example, ‘Eating cooked fresh water fish helps prevent liver fluke infection’ and ‘Using toilets can prevent liver fluke dispersal into the environment’. Again, the responses ranged from 1 (agree), 2 (not sure), 3 (disagree). An acceptable internal consistency of α = 0.72 was found.
Perceived self-efficacy was measured using five items. For example, ‘My health status will be better if I do not get liver fluke infection’ and ‘Pla-som is sold in my community so I must eat it’. An acceptable internal consistency of α = 0.73 was found.
Data processing and statistical analysis
Data collected were manually entered twice into a computer using Excel and after quality check they were transferred to the STATA program for analyses. The comparative mean difference for knowledge, perception of vulnerability, severity, response efficacy, and self-efficacy were analyzed by analysis of covariance.
Results
A total of 118 school children participated in the intervention group and 113 in the control school. The demographic characteristics of each group were similar, with no significant differences between them in terms of age, gender mix, average grades, number of family members, occupational mix, and toilet availability (Table 1).
The intervention group of schools showed significantly higher knowledge of the health problems caused by liver flukes. We found that the intervention schools scored better than the control group in each prevention motivation theory variable (Table 2). A total of 66.9% of intervention schools showed a high level score for knowledge of liver flukes, such as how infection occurs, the sources of infection and how to prevent opisthorchiasis. The scores for the control group of schools were generally lower. The students in the control group showed a poorer knowledge of risky foods and did not necessarily attribute this to raw fish consumption. For instance, a number of students replied that eating raw beef and raw snails could cause liver fluke infection. Very few students could identify freshwater fish as a transmission agent for liver fluke. Most of them identified the wrong prevention measures from negative questions. For example, to the question ‘Can hand washing can protect you from liver fluke infection?’ 62% of students wrongly replied in the affirmative, and 64% of students believed that swimming in local canals or swamps could cause liver fluke infection. The mean scores (standard deviation (SD)) of the intervention group and control group were 7.6 (2.7) and 3.6 (1.9), respectively, and the adjusted mean difference of 3.35 (95% confidence interval (CI) = 2.67–4.02) in the intervention group was significantly higher than the control group (p < 0.001) (Table 3).
Demographic characteristics of the school children.
Knowledge, perception of severity, perception of vulnerability, perception of response efficacy, and perception of self-efficacy between intervention and control groups.
Comparison of knowledge and perception among school children between intervention schools (n = 118) and control schools (n = 113) at the end of the second semester, 2012.
Mean difference was adjusted for age, sex, family member, and parent occupation.
The threat or risk appraisal of prevention motivation theory among intervention school children showed high scores in terms of perception of severity: 60.2% for the intervention group and 24.8% for the control group. There was little consensus about perceived severity amongst the control group, for instance some of the children did not perceive CCA as being caused by liver fluke. The students were also not sure that opisthorchiasis could be cured and they believed that if they seldom eat raw fish they will not get infected. Thus the mean score of perception of severity in intervention group and control group were 12.3 (2.5) and 10.0 (0.2), respectively, with an adjusted mean difference of 1.72 (95% CI = 0.97–2.47) (p < 0.001) (Table 3).
In terms of their perception of vulnerability, 63.6% of the intervention group showed high scores compared with only 37.2% of the control group. Many students in the control group indicated that they ate raw fish because pla-ra and pla-som was sold in their village. The mean scores relating to perception of vulnerability in the intervention and control groups were 13.0 (2.4) and 11.2 (2.5), respectively, with an adjusted mean difference of 1.34 (95% CI = 0.62–2.08) (p < 0.001) (Table 3).
For coping appraisal, 57.6% of the school children showed a high score for perception of response efficacy after the intervention, while this was only 35.4% for the control group. Control group pupils showed poor agreement of perception of response efficacy that toilet use could prevent liver fluke dispersion but not liver fluke infection that toilet use was not a risk factor to CCA, and that taking anti-liver fluke tablets could eradicate liver fluke in humans. These correct perceptions were found more commonly among the intervention group. The mean score of perception of response efficacy in the intervention and control groups were 12.0 (2.5) and 10.5 (2.8), respectively, with an adjusted mean different of 1.1 (95% CI = 0.35–1.94) (p = 0.005) (Table 3).
Perception of self-efficacy scored 79.7% for the intervention group, while the control group had a moderate score of 57.5%. The control group showed poorer agreement of perception of self-efficacy, such as avoidance som-tam whenever it was mixed with raw pla-ra. The mean score of perception of self-efficacy was somewhat lower than other perception variables, the intervention group was 9.7 (1.7) and control group was 8.3 (1.6), with an adjusted mean difference of 1.67 (95% CI = 1.18–2.15) (p < 0.001) (Table 3). All of the PMT constructs, including knowledge and perceived severity, vulnerability, response efficacy and self-efficacy parameters concerning of the dangers of eating raw fish and of developing CCA, were significantly intercorrelated with each other (p < 0.001).
Our results showed that school children in the intervention groups ate raw fish less often than those in the control group. Pupils in the intervention group preferred to eat dishes mixed with raw pla-ra such as som-tam (52%) and jaewbong (30%), but they very rarely ate koi-pla, som-pla, etc. Pupils in the control group ate many kinds of raw fish dishes, often as som-tam mixed with raw pla-ra (66%), as well as jaewbong mixed with raw pla-ra (47%), som-pla (29.2%), pla-som (25%), and koi-pla (20.1%).
Discussion
Our data show that substantial knowledge of the life cycle and dangers of opisthorchiasis remain within the school population in northeast Thailand, but that this can be significantly improved using the PMT teaching module designed for this purpose. Our study will also act as a baseline to determine the knowledge base in school children in the future.
The intervention program based on PMT was originally introduced by Rogers in order to understand mechanisms by which people adopt protective behaviors to reduce a perceived threat of disease (31). The PMT, which has been used in HIV prevention research (19,32), can be an alternative to the knowledge, attitude, and practices approach because it integrates cognitive process with information, knowledge, attitudes to behavioral intentions, and further behaviors (33), all of which are highly relevant to the case of opisthorchiasis and CCA.
Our goal was to assess efficacy of a primary school-based health education program built on PMT. The intervention program was designed to build specific positive beliefs and health behaviors related to opisthorchiasis prevention and control by enhancing knowledge, perception of severity and vulnerability, and focusing on self-response and self-efficacy of the school children.
Our results indicate the success of this health educational intervention in enhancing the knowledge and perception of O. viverrini infection behavior as the intervention group consistently scored better that the control group. The intervention group showed an improvement in knowledge of how O. viverrini infection occurs, the type of freshwater fish that can be infected and risky dishes. The perception of severity and vulnerability to opisthorchiasis and CCA were also increased. Finally this program increased the children’s response efficacy and self-efficacy in avoiding raw fish consumption. Nevertheless, data from the control group also showed significant knowledge of the causes and consequences of O. viverrini infection, suggesting that previous community-based control and prevention programs have had a lingering effect (34).
Nevertheless, certain misconceptions were more common in the control schools, including how infection occurred. Almost 90% of the children thought that the consumption of raw snails and raw beef could also lead to opisthorchiasis. A total of 62% replied that infection occurred by parasites penetrate their skin while swimming in canals or swamps, which is true for Schistosoma mekongi infections but not from liver flukes (35). Our findings in the control schools were similar to those found among older persons in the Isan community where there are common misconceptions on the route of transmission. For example, in addition to the idea that eating any raw food, drinking contaminated water or skin penetration can lead to infection, there is a common belief drinking rice whisky (40% alcohol) or mixing lime juice with raw fish can kill the flukes (13). The self-response and self-efficacy behaviors are more complicated, involving critical thinking to promote desirable health behavior. Our findings reflected self-confident in avoiding raw fish dishes. However, as the children are dependent on parents or guardian in their daily routine to promote this health behavior, the family must accept child’s activities.
The advantages of this study are: it directly compared the intervention and control groups using a clustered, randomization control design; it was carried out in an area where school children are at high risk due to the raw food culture present and it was carried out in a real-life setting with simultaneous assessment, making its results generalizable. In addition, it used learner development activities and a club atmosphere which emphasized to teamwork appropriate and consistent with learners’ maturity the school setting and local context.
Thus our health education program succeeded in improving knowledge on infectious fish, cyprinid fishes, i.e. the type of foods that harbor O. viverrini and leads to opisthorchiasis and potentially CCA. However, the stability of the protection effect could be increased by combining our school-based learning with other strategies, such as social support to enhance the participation of family members, and community education as currently carried out by the Cholangiocarcinoma Screening and Care Program (CASCAP) (34).
Conclusion
The current low prevalence of opisthorchiasis among children in Thailand suggests the success of previous community-based control and prevention programs. This is reflected in the background knowledge found in the control group children. However, the ‘in school’ learning program tested here clearly increased the knowledge and perception of the public health problems associated with O. viverrini infection. This program provides opportunities to enhance health behavior by promoting ‘anti-raw fish dishes behavior’. This school based health education, focusing on learner activities development and providing a club atmosphere, indicates how educational interventions that are based upon ideas drawn from PMT can yield impressive improvements in terms of knowledge of public health problems and self-prevention measures. The success of such public health education suggests that it should become school policy in effected areas, and could be modified to include other potential diseases when appropriate.
Supplemental Material
Supplement – Supplemental material for Analysis of a school-based health education model to prevent opisthorchiasis and cholangiocarcinoma in primary school children in northeast Thailand
Supplemental material, Supplement for Analysis of a school-based health education model to prevent opisthorchiasis and cholangiocarcinoma in primary school children in northeast Thailand by Luxana Laithavewat, Carl Grundy-Warr, Narong Khuntikeo, Ross H. Andrews, Trevor N. Petney, Puangrat Yongvanit, Pannee Banchonhattakit and Paiboon Sithithaworn in Global Health Promotion
Footnotes
Acknowledgements
We would like to thank the International Bureau of the German Federal Ministry of Education and Research (IB-BMBF) National Science and Technology Development Agency (NSTDA) and Khon Kaen University 2013 for providing funding for cooperative workshops.
Conflict of interest
The authors declare that there are no conflicts of interest.
Funding
This work was supported by the Higher Education Research Promotion and office of the Higher Education Commission, through health cluster (SHeP-GMS), Khon Kaen University, Thailand and Cholangiocarcinoma Screening and Care Program (CASCAP), Khon Kaen University.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
