Study site and participants

This study was conducted in Macanip, a local government unit or “barangay” in the municipality of Jaro, Leyte. Leyte is an Eastern Visayas province of the Philippines, and is one of the major endemic foci of schistosomiasis [5]. According to the 2010 census by the Philippine National Statistics Office, the population of Macanip was 1,755 [15]. The RITM has conducted other schistosomiasis research in the area, including bovine vaccine trials. Macanip is loosely organized into three zones located in the town center, and eight sitios or territories distanced from it. All three zones and all eight sitios were included in the study, and each zone or sitio was sampled proportional to its number of households.

The minimum sample size calculated based on the primary study objective to assess knowledge, attitudes, and practices in the community was 96 participants (95% CI = 95.98–96.02, ME = 10%), based on the 2010 population size, a prevalence of 50%, and 80% power. No data for schistosomiasis prevalence in Macanip within the last 10 years were published at the time of the study. In each zone or sitio, participants were recruited by randomly selecting households on a predetermined walking route, and then knocking on doors. Questionnaires were administered via oral interview to one adult (18 years of age or older) in each selected household. In each household only the first participant to consent to the study was interviewed.

This study was approved by the Institutional Review Board (IRB) of the Tufts Medical Center / Tufts University Health Sciences in Boston, MA (IRB#11699) and the Research Institute of Tropical Medicine (RITM) in Manila, Philippines (IRB #2015–11). The study qualified for exemption status with the Tufts IRB. All relevant research procedures adhered to the guidelines of both institutions, including the use of oral consent for data collection. Village officers were asked for permission prior to initiation of the study. All participants were orally informed of the purpose and procedures of the study, with emphasis on voluntary participation and confidentiality. All participants were also provided with an information sheet satisfying the basic elements of informed consent, translated into Waray-waray and verbally communicated by community health workers. Only participants who verbally consented to the study after considering this sheet were included.

Study design

This study was analytic and cross-sectional in design and employed the use of a standardized questionnaire. The questionnaire was designed based on consultation with researchers from both the Cummings School of Veterinary Medicine at Tufts University and the RITM. The questionnaire included both open- and closed-ended questions and was designed to take no more than 30 minutes to administer. The questionnaire was written in English, translated to Waray-waray by RITM staff who had experience working in the area, and back-translated to English for IRB approval. Questions covered 1) knowledge and attitudes related to schistosomiasis, its symptoms, and its transmission; 2) attitudes and practices in relation to schistosomiasis prevention; 3) willingness to comply with public health control programs such as mass treatment and carabao vaccination; and 4) whether the respondent had contracted schistosomiasis in the past. Additional questions covered socioeconomic characteristics, medical history and relevant household conditions such as access to different local water sources and sanitary facilities.

Questionnaires were administered verbally by six RITM staff members who had previous experience assisting with survey-based schistosomiasis research in Macanip. Due to other RITM studies, all interviewers had previously received training on maintaining confidentiality and administering questionnaires in a standardized manner. During the study, each interviewer was accompanied by one Barangay Health Worker (BHW), a community health worker trained by the Philippine Department of Health. BHWs did not participate in the interviews, but assisted in navigating the area, locating households and greeting participants. All six interviewers spoke Waray-waray and Tagalog fluently, as well as basic English. Interviews primarily followed the written protocol in Waray-waray, with some flexibility allowed in clarifying participants’ answers with additional questions. For open-ended questions, answers were written down word-for-word in Waray-waray and then translated to English by RITM staff on the same day. Participants were not paid monetarily but were informed that the results of the study would be made available to them via local control programs as well as RITM-led meetings.

Data processing and statistical analysis

‘Knowledge scores’ were calculated from questions on knowledge of schistosomiasis symptoms, transmission and prevention using the sum of total correct responses. ‘Prevention scores’ were similarly calculated from questions related to engagement in prevention strategies that are recognized to be effective (scored as correct or 1) and in strategies that are recognized as ineffective (scored as incorrect or 0). For example, a participant who answers yes to whether she avoids water contact and also answers yes to whether she visits a traditional healer is awarded one point to her prevention score, for avoiding water contact. Answers to open-ended questions were qualitatively grouped into common themes and counted by theme. These included questions such as where the participant had first heard about schistosomiasis.

Data analysis was conducted using R v 3.3.1. The data were found to be non-normally distributed, and the following analyses were conducted to determine significant associations between outcomes of interest and predictor variables: 1) Spearman’s correlation was performed to determine the relationship between knowledge score and prevention score; 2) Mann-Whitney U test was used to assess significant differences in knowledge scores between those willing or unwilling to participate in mass treatment in the future; 3) Mann-Whitney U test was also used to assess differences in knowledge scores between carabao owners willing or unwilling to vaccinate their carabaos; and 4) multiple logistic regressions were used to identify factors predictive of 3 different outcome variables. These outcome variables included a) willingness to participate in mass treatment in the future; b) willingness to vaccinate carabaos, with only carabao owners included; and c) having contracted schistosomiasis in the past. A p-value of ≤0.05 was accepted as statistically significant for all analyses.

Multiple logistic regression

Predictive models for the three outcome variables described above (a, b, and c) were developed using multiple logistic regression. For each regression, predictor variables were selected by performing a univariate analysis of each variable with the outcome of interest. These variables included demographic data, knowledge score, prevention score, household conditions such as presence of a working toilet, and attitudes around schistosomiasis such as belief that the disease can be prevented. Variables also included whether participants had heard of schistosomiasis, whether they had learned about it in school, and whether they described schistosomiasis as common to the area. A p-value of ≤0.10 was used to initially determine inclusion in the first model. Next, a second model was developed, including all predictor variables that had a p-value of ≤0.05 in the first model. Each of the remaining variables that had been included in the first model but not the second (i.e. those with a p-value between 0.05–0.10) were then added back individually and tested for significance. The final model included all variables that remained significant in this second model, as well as the demographic variables of age, sex, and education level. A p-value of ≤0.05 was used for statistical significance for the final model.

Knowledge

All participants had heard of schistosomiasis, and almost all participants knew it was common to the area. These findings suggest high awareness of the disease in Macanip. As well as being endemic to Leyte, schistosomiasis has necessitated multiple research projects by RITM staff, which may account for such high awareness. Additionally, 72.6% of participants had previously contracted schistosomiasis, further confirming endemicity of the disease. Knowledge scores were fairly high, and participants demonstrated accurate knowledge of transmission, clinical signs and proper prevention techniques (Table 2). However, most participants were unaware of a microscopic agent causing the disease (Table 2). Additionally, almost half the participants did not remember learning about schistosomiasis in school, with most citing existing RITM programs or previous research as the source of their knowledge (Table 2). The high prevalence of disease and long history of control efforts in Macanip could have necessitated basic knowledge of schistosomiasis. However, our findings suggest limited understanding of the biology of the disease. Improving the quality of disseminated information might encourage compliance with disease control efforts, as well as preventing misconceptions about mass treatment or other interventions.

Attitudes

Our study also reveals important attitudes surrounding schistosomiasis control and the effects of these attitudes on receptiveness to control programs. Most participants considered schistosomiasis to be a very serious disease, and expressed willingness to participate in mass treatment, suggesting high compliance in future mass treatment programs (Table 2). However, research in Samar, the endemic province closest to Leyte, has showed far lower compliance on the day of mass treatment than predicted from surveys [13]. Further study is therefore needed on how willingness to participate can be translated into actual compliance in mass treatment.

Additionally, only half the carabao owners in this study expressed willingness to have their carabaos vaccinated for schistosomiasis (Table 2). The most frequently cited reason for being unwilling was fear for the carabao’s health, suggesting misconceptions and doubts surrounding the idea of livestock vaccination. Knowledge scores did not vary significantly between participants who were willing and unwilling to participate in mass treatment or between owners willing and unwilling to vaccinate their carabaos. There is therefore no evidence that current knowledge of the disease affects willingness to comply with these control programs. Knowledge alone may therefore be an insufficient motivator for compliance in this community.

Instead, belief in preventability of the disease was revealed to be a key attitude in this study. About 60% of participants believed it was possible to prevent schistosomiasis, while 40% of participants did not (Table 2). These findings suggest that a significant proportion of Macanip residents do not view schistosomiasis as a solvable problem for their community. This most likely stems from the persistence of schistosomiasis in Macanip despite decades of control measures. However, carabao owners that did believe in prevention were over five times more likely to be willing to vaccinate their carabaos (Table 4). This finding suggests that those who believe in disease prevention are more open to new strategies for disease control. The success of future interventions in Macanip may therefore rely on preserving this attitude within the community via health promotion. Efforts to educate residents on the importance of water contact, for example, should emphasize that this practice makes it possible to be free of schistosomiasis.

We also found that age had a significant effect on attitudes about disease control, where older participants were less likely to be willing to participate in future mass treatment programs (Table 3). Additionally, carabao owners who were willing to participate in future mass treatment were much more willing to have their carabaos vaccinated (Table 4). These findings once again highlight the influence of sociocultural context on the impact of public health interventions. If compliance with disease control tends to decrease with age, targeting an older cohort may improve the impact of future control efforts. Mass treatment campaigns, for example, should be expanded toward older members of the community in messaging and accessibility. Additionally, if receptiveness to different aspects of disease control are interrelated, and if belief in prevention plays a role, then interventions must adequately address these attitudes in order to succeed. To this end, regular recruiting efforts for mass treatment could also disseminate media aiming to positively influence participant attitudes, such as belief in prevention.

Practices

Prevention scores were high among participants, but our findings indicate inconsistent behavior around disease prevention. Belief in the preventability of schistosomiasis was not found to significantly impact prevention score. However, around 40% of participants said they did not actively avoid water contact, and more than 30% said they did not use protective gear while contacting water, despite high awareness of both these prevention strategies (Table 2). Additionally, a large majority of participants (85.8%) reported regular water contact, despite about 60% saying they actively tried to avoid it (Tables 1 and 2). This discrepancy suggests that Macanip residents are aware of disease prevention or might believe in its efficacy, but still encounter barriers to it in practice. During interviews, participants mentioned neglecting these strategies when they were not practical, such as declining to wear heavy boots while working in a rice field for many hours. Some participants mentioned using herbal treatments and visiting the traditional healer as health-seeking behaviors, but spoke in the context of maintaining wellness rather than treating schistosomiasis. These anecdotes suggest further obstacles to disease prevention that Macanip residents may encounter, and reinforce the need for behavioral and sociocultural research related to this disease. Identifying the community’s unique barriers to disease prevention via qualitative research would allow for more effective design of public health interventions or recommendations.

Previous disease

Our study found a high proportion of participants (72.6%) who reported that they had previously contracted schistosomiasis (Table 1). Neither knowledge score nor prevention score was found to have a significant effect on previous disease, implying that current levels of awareness and disease-preventing behavior may not be related to previous experience with schistosomiasis. Additionally, respondents who had access to a working toilet at home were more likely to report previous disease (Table 5). Although this finding is unexpected, participants were not asked when they had acquired their toilet in relation to contracting schistosomiasis. Appropriate waste disposal may therefore be insufficiently protective of disease in this setting, and those who experience the disease may then become motivated to acquire a toilet in the home.

Once again, belief in the preventability of schistosomiasis was a key attitude, with participants who did not believe in prevention being more likely to report previous disease (Table 5). Although this attitude did not have a significant effect on prevention score, belief in prevention might impact behavior in other ways, such as decreasing the consistency or effort behind prevention practices. Conversely, participants might cease to believe that schistosomiasis is preventable after experiencing the disease themselves, in a community where it is highly prevalent. Further research is needed to investigate the relationship of this attitude to disease status and prevention behavior.

Finally, attitudes and practices around mass treatment programs were also associated with contracting schistosomiasis. Those who reported willingness to participate in future mass treatment were more likely to report previous disease. Additionally, those who reported previous participation in mass treatment were much more likely to report previous disease (Table 5). This association is consistent with our finding that over half the participants who sought schistosomiasis treatment in the past did so through mass treatment programs (Table 1). Those who contract schistosomiasis might therefore pursue mass treatment, and then continue to view it favorably once treated successfully. Ultimately, mass treatment appears to be a well-established and popular form of health care in Macanip, but fails to fully control schistosomiasis. This status quo of well-known but insufficient care might have shaped community attitudes on current disease control as well as receptiveness to new interventions.

Limitations

This study was most significantly limited by the format of the KAP questionnaire. Although attempts were made to ask open-ended questions, most of our data relied upon dichotomous and numerical observations. This was especially relevant to prevention practices, where many participants with high prevention scores still reported frequent water contact. Further studies should aim for more open-ended assessment of disease prevention, including more qualitative data on barriers to behavior change, and incorporate an observational component to validate questionnaires. Additionally, our sampling strategy may have created bias by selecting households upon a predetermined walking route during the work day. Macanip residents who lived farther from main pathways or worked away from home during the day may have been excluded by convenience sampling. These limitations in study design stem from conducting the first attempt at KAP research in this area on a limited timeline and budget. We invite further studies to spend more time recruiting participants and using more open-ended methods to assess knowledge, attitudes, and practices.

Our study aimed to quantify self-reported disease alongside attitudes on disease prevention. However, future research in Macanip should continue to monitor adult human prevalence when possible, using screening tests such as Kato-Katz stool examination. Because schistosomiasis has been prevalent for so many years, future studies should also obtain temporal data on disease prevalence, as much of our data may have been shaped by previous medical histories.

Conclusion

Our study revealed fairly high levels of awareness and good baseline knowledge of schistosomiasis in Macanip, as well as frequent attempts at disease-preventing behavior. These findings serve as a testament to the many years of research, training, and disease control programs in Macanip. However, we found inconsistencies in prevention practices that suggest barriers to lasting behavior change. Future interventions in Macanip should address specific obstacles to disease-preventing behavior via further qualitative research.

Our results highlight belief in the preventability of schistosomiasis as a pivotal attitude, influencing participant willingness to vaccinate livestock as well as their likelihood of reporting previous disease. Our findings imply that belief in prevention may be associated with more receptiveness to new control strategies such as carabao vaccination. Future public health interventions could therefore include efforts to bolster this belief, such as providing examples of other communities that have successfully lowered their burden of disease.

Recent studies from the People’s Republic of China are often referenced as success stories to this end. In 2004, China implemented an integrated strategy for schistosomiasis control that includes mass treatment and education but also snail control, improved sanitation, and replacement of working bovines with agricultural machinery [16, 17]. This shift has resulted in the elimination of human schistosomiasis cases in formerly endemic communities, as well as the elimination of infected bovines and snails in selected areas [16, 17]. However, it is important to note that most rural endemic communities such as Macanip suffer from long-term poverty as well as disease. Local poverty fosters disease transmission by increasing exposure to contaminated water as well as limiting the means to treat or prevent disease [3]. Interventions that target the parasite’s life cycle (e.g. bovine vaccination) should therefore be employed alongside strategies for poverty reduction [18]. Ultimately, effective disease control requires a long-term multi-sectoral commitment toward breaking the cycle of S. japonicum transmission as well as mitigating the effects of poverty on public health outcomes.

Our findings emphasize the importance of cross-disciplinary work to successful schistosomiasis control in Macanip and in similar endemic communities. Studies that provide sociocultural and behavioral data should continue to supplement biomedical research in shaping future interventions. This focus on a comprehensive, integrated approach is relevant to similar endemic communities currently designing and implementing programs for the control and elimination of schistosomiasis.