Study design and setting
The cross-sectional survey was conducted using a structured questionnaire from May to July 2022. All methods followed relevant guidelines or regulations66, including compliance with ethical requirements and specifically sampling without biological specimens. The interviews were conducted by obtaining informed consent; the respondents signed the consent form after understanding the aims and objectives of the research and voluntarily providing data to the researcher without coercion, and the identity of the respondents was made anonymous by the researcher to maintain confidentiality.
The study was approved by the Health Research Ethics Committee of the Faculty of Public Health, Universitas Sriwijaya, with ethical approval No: 313/UN9.FKM/TU.KKE/2022. Participation in this study was voluntary in the field. All analyses were performed using participant identification codes to ensure maximum confidentiality.
Using a purposive method, the investigator chose three villages, namely Tanjung Lalang, Tanjung Agung, and Penyandingan, in Tanjung Agung Sub-District, Muara Enim District. The study areas were selected based on information about the sites with the highest mining activities. The study areas are shown in Fig. 1
Sample size
This study uses a hypothesis test for two population proportions (two-sided test) with a level of significance of 95%, a type I error (α) of 5%, and a test power of 80%67.
It utilised P1 and P2 from previous studies for sample size determination and selected variables. Based on the literature review, the anticipated population proportion 1 (P1) for patients who do not use mosquito nets and have malaria is 0.79, while the anticipated population proportion 2 (P2) for patients who use mosquito nets and have malaria is 0.50. Therefore, the minimum sample size for each group in this study is 42 people. The results are multiplied by two because a two-proportion sample formula is used, resulting in an estimated total minimum sample size of 84 people. To account for potential dropouts and other factors, the sample size was increased by 10% to 92.4 and rounded to 92 respondents. Thus, the total sample calculation yielded 92 people. This recent study used a cross-sectional study design and revealed that 78 (84.78%) respondents had no malaria, whereas 14 (15.22%) had malaria.
Sampling technique
This study employed a purposive sampling technique to select clusters or villages. The village with the highest number of miners was chosen as the criteria for selection. The sampling method was done randomly in each village using the existing sampling frame in each chosen village. The number of samples taken in each village was determined proportionally, where the minimum sample obtained represents the population of artisanal small-scale mining (ASM) in these three villages. Due to social and environmental factors affecting malaria in small-scale mining areas, there were malaria-positive cases in this population, making the respondents at risk for malaria transmission. An illustration of the sampling technique is shown in Fig. 2
Flowchart of a sampling frame of the study.
Using Muara Enim District’s Central Bureau of Statistics data, the investigator made a cluster fraction sample and grouped people in each village. Finally, the respondents were chosen using two-stage cluster sampling in the study area. The current study has improved Internal and external validity. The population’s livelihoods in the villages are heterogeneous; apart from labourers and employees, they also contribute to plantations, livestock, fisheries, farmers, and ASM. More mining locations exist in these three selected villages than in other areas.
Data collection
After random sampling, inclusion/exclusion criteria are based on the at-risk population. Thus, the eligibility criteria or inclusion criteria of this current study were that the respondent has lived in Muara Enim Regency for more than six months, is an illegal miner, can communicate properly and is ready to be a respondent. While the exclusion criteria are if the respondent does not finish the questionnaire and the sample respondents cannot speak effectively.
Epidemiological, socio-demographic characteristics, behavioural risk factors and observational environmental data were collected using a structured questionnaire. This questionnaire’s validity and reliability were tested on 30 respondents who live in Darmo village, which has mining activities. The respondents have the same characteristics as the sample study. After getting a valid and reliable questionnaire, the questionnaire was used for the respondents to collect information on risk factors of malaria occurrence in ASM in the study area. Eligible participants were informed about the general objectives. Oral consent was obtained before the interview, and information was recorded on paper case record forms. Investigators interviewed 92 miners who met the requirements to participate in the study. The interview, a structured questionnaire with open-ended and closed items, was conducted face-to-face.
Scope of variables
The data was processed using Stata software: on the dependent variable (malaria occurrence), respondents who did not experience malaria were coded 1, and respondents who had or were experiencing malaria were given code 2 in this study, conducted in 2022. Malaria infection (positive malaria) are patients with symptoms of malaria who have been declared positive based on rapid diagnostic tests (RDTs) test and microscopy. In this study, the respondents who did not experience positive malaria were those who had negative malaria and were also healthy. In the questionnaire, for the dependent variable, the investigator asked respondents, “Had the respondents ever had blood drawn for malaria examination by a health worker?”. The answer was binary “Yes or Not”. If the answer was “Yes”, the next question was: had the respondents tested positive for malaria after examination by a health worker?
Likewise, the independent variable, code “small,” is given to describe the variable in the “bad/low” condition or the “at-risk” group. This category is grouped under code 1. At the same time, the code “large” is given to describe variables in “good/high” conditions or groups with the category “no risk” with code 2.
Socio-demographic characteristics of participants
All independent variables were coded as one risk category, with code 2 indicating no risk. The respondent’s age is recorded in years 35 (code 1) and 35 (code 2), then the years of service are divided using years five years and < 5 years and the division of working hours for respondents is calculated hourly 8 h and < 8 h in a day. The gender differences were divided into male and female and taken from the questionnaire. In this study, education is the highest level of education achieved by participants. Participants were considered highly educated after graduating high school and coded = 2. Participants who did not complete high school were classified as uneducated and assigned the code = 1.
Behavioural risk factors
From the questionnaire, the use of mosquito nets is categorised as follows: the respondent who does not use mosquito nets at night when sleeping is given a code of 1. If the respondent uses a mosquito net, it is coded as 2. The habit of using mosquito repellent is then divided; if not or only occasionally using mosquito repellent, code 1 was assigned, and code 2 was assigned to respondents who used mosquito repellent. The habit of going out at night and taking self-medication uses code 1 if yes, or sometimes code 2 if you never go out at night and take self-medication. The respondents’ knowledge is then divided into lack of knowledge and high knowledge and code 1 and code 2 for those with high knowledge. Meanwhile, attitude and practice are categorised with code 1 if they are not good and code 2 if they are good.
Environmental risk factors
Environmental factors are divided into two categories: outside and inside the home. In the outdoor environment, whether there is a breeding place for mosquitoes and a resting place for mosquitoes. The presence of a breeding and resting place is coded 1 if it is at risk or if the distance is 100 m from the respondent’s home location, and code 2 is not at risk if it is > 100 m. Regarding environmental elements, the condition of the house’s walls and the existence of the ceiling are significant. Additionally, the state of the house’s flooring Researchers assigned code 1 for a housing condition that does not match the requirements and is not an eligible house and code 2 for a housing condition that meets the requirements.
Factors associated with malaria occurrence using a structured questionnaire.
Information was primarily collected from respondents who were mining workers. The dependent data from a structured questionnaire included epidemiological data on malaria occurrence; At the same time, the independent variable included respondents’ socio-demographic characteristics, including age, sex, years of service, length of work, and education level. In addition, behavioural risk factors included using a mosquito net, mosquito repellent, being out-of-the-house at night, self-medication, and KAP (knowledge, attitude and practice) also investigated. Furthermore, environmental risk factors included mosquito breeding, resting place, house wall condition, the existence of the house ceiling, and house floor condition. Another study found that bamboo/wood house walls, no insecticide, and a distance < 100 m from the mosquito breeding site were malaria risk factors68. The previous study showed environmental housing variables affect transmission. Mosquito breeding places near dwellings influence malaria transmission69. Household and neighbourhood-focused malaria interventions may assist high-burden nations. Targeted vector control methods like indoor residual pesticide spraying can help reduce malaria, address age-related differences in malaria outcomes, and ITN use70.
Anopheles larvae breed in former mining excavations, lagoons, swamps, buffalo puddles, ponds, ditches, rice and fields as breeding sites. A questionnaire and GPS measurements were used to determine whether each dwelling had a breeding site. Previously, Anopheles data were collected by the South Sumatra Provincial Health Office, Baturaja Litbangkes, and Muara Enim District Health Office.
Data analysis
Data were analysed using descriptive statistics, chi-square and logistic regression analysis using STATA statistical software. Other data extracted were the availability of association between these variables and the occurrence of malaria in miners, provided in Tables 1 and 2, respectively.
Descriptive analysis
Individual characteristics of respondents include age, sex, years of service, length of work, and education level. Furthermore, behavioural risk factors include a mosquito net, mosquito repellent, out-of-the-house at night, and self-medication, then KAP (knowledge, attitude and practice) are the goals of descriptive analysis.
Bivariate analysis
The significance of the statistical relationship between each independent and dependent variable was analysed using the Chi-Square test by comparing the probability value (p value) with the alpha (α) = 0.05. If the p value is < 0.05, Ho (the null hypothesis) is rejected. It means there is a significant statistical relationship between the independent and dependent variables; if the p value > (0.05), Ho is accepted or failed to be rejected, meaning there is no significant statistical relationship between the independent and dependent variables.
Multivariable analysis
The malaria risk severity was evaluated using crude and adjusted prevalence ratio (aPR)†. If the prevalence ratio (PR) is more than one, the probability of developing malaria increases.
Limitations of research
In the study area, malaria transmission is influenced by various risk factors, although not all of them have been considered as research variables. This aspect presents an opportunity for further investigation. However, the current variables under study primarily pertain to social and environmental factors that impact malaria in small-scale mining areas. It is important to note that the study’s sample size is relatively small, which may introduce some bias in the results, particularly in assessing the proportion of malaria-positive cases. The sample size was determined using the sample formula for testing the two-proportion hypothesis, using data from previous studies regarding P1 and P2, as well as selected variables. This approach has improved the internal and external validity of the study, making it representative of artisanal mining in the three villages. In these villages, the residents’ livelihoods are diverse, involving activities such as smallholder plantations, livestock keeping, fisheries, farming, and mining. Interestingly, the study did not specify the respondents’ working hours, which could be associated with the risk of Anopheles mosquito bites. Despite this, it was found that respondents worked varying hours, with some working less than 8 h and others working 8 h or more. The majority of respondents (56.5%) opted for longer working hours to increase their daily income, as more hours of work resulted in higher wages. Hence, economic considerations drove many respondents to extend their working hours.The working hours typically span from morning until evening, although some individuals work until late evening. It is worth noting that An. sinensis tends to bite outdoors, while An. vagus prefers indoor biting. Both species are active from 9 p.m. to 4 a.m. These findings may have broader implications and can potentially be applied to other small-scale mining areas.
In summary, while additional risk factors may affect malaria transmission in the study area, the current research is primarily centered on the social and environmental factors impacting malaria in small-scale mining regions. Although small, the study’s sample size was carefully determined, enhancing the validity of the findings for artisanal mining in the three villages. Working hours play a role in malaria risk, with economic factors driving some respondents to work longer. Different mosquito species’ time preferences and biting behaviours shed light on the potential applicability of the study’s concept to other similar mining locations.
Informed consent
I undersign a certificate that I have the written consent of the identifiable person or their legal guardian to present the cases in this scientific paper.