Evaluation of IR3535 as a Spatial Repellent for Malaria Control. (REPELMALARIA)

Evaluation of IR3535 as a Spatial Repellent for Malaria Control. (Avaliação da eficácia do Repelente Ambiental 3-(N-acetyl-n-butyl) Aminopropionic Acid Ethyl Ester (IR3535) Como Ferramenta Para o Controlo da malária).

Mozambique contributes with 5% of global malaria cases, and despite control efforts the Sofala province continues to experience a high burden of malaria. The resistance to insecticides and changes in vector habits can compromise the use of common vector control tools. The use of spatial repellents is thus an interesting alternative, as it does not exert selective pressure on resistance genes or eliminate other insects with impact on biodiversity. IR3535 is a non-toxic repellent and if used at community level can extend protection to outdoor biting. Hypothesis: Using the IR3535 repellent for indoor and outdoor spraying will reduce the prevalence of malaria and reduce vector density and infection. An experimental Before-After-Control-Intervention will be carried out with two groups: a) Intervention (Tambai Q2 and Q6): with intra and extra-household spraying with IR3535 and b) Control (Tambai Q3 and Q4): without spraying. Tambai is acommunity of Bebedo, Nhamatanda, Sofala, Mozambique. The mosquito distribution, diversity, density and sporozoite rate will be monitored indoors and outdoors in both communities for 2 years. The prevalence of malaria will be determined in under five years old children at the beginning, the end of the 1st year and at the end of the study. Additionally, cross-sectional studies with a mixed approach assessing the community knowledge, attitudes and practices (KAP) will be carried out to detect gaps that persist at the community level. Heads of households and health professionals will be interviewed at the beginning of the study, the end of 1st year and at the end of the study. The data will be analyzed using SPSS and R software packages. For matching situations (before and after), the McNemar test will be used to ascertain statistical significance. Generalized Linear Models (GLM) will be used to jointly analyze several explanatory variables. Linear Mixed Models (LMM) and Generalized Estimation Equation (GEE) models will be used to compare longitudinal data. The prevalence of malaria and entomological indices relevant for transmission are expected to decrease with the intervention while community knowledge on malaria and its control are expected to increase.

Study area and population The study will be carried out in the communities of Tambai, locality of Bebedo, district of Nhamatanda, Sofala province, Mozambique. The community is covered by the Mutondo Health Center. Study design A prospective experimental Before-After-Control-Intervention test will be carried out with two groups: a) Intervention (community of Tambai quarters 2 (Q2) and 6 (Q6)): with indoors and outdoors spraying with IR3535 (3-(N-acetyl-N-butyl) aminopropionic acid ethyl ester) and b) Control (community of Tambai quarters 3 (Q3) and 4 (Q4)): without spraying. This longitudinal study will last for 2 years. Mosquitoes will be captured indoors and outdoors to determine entomological indicators relevant to transmission. Mosquito breeding sites in the communities will be mapped using "Participatory Mapping". The prevalence of malaria in children under five will be determined in the control and in the intervention communities, at the beginning, middle and end of study. A KAP survey on malaria to Tambai householders (Q2, Q3, Q4; Q6) and health professionals and key community actors. Population interventions and procedures Initially, the Beira Operational Research Center (CIOB) and the Institute of Hygiene and Tropical Medicine of the Nova University Lisbon (IHMT-NOVA), in collaboration with the Mutondo health center, Provincial Directorate of Health of Sofala and the Directorate district health district of Nhamatanda and the Gorongosa National Park, will present the intervention in a participatory manner and explain the approach directed to the community in order to guarantee its appropriation by the community from the beginning of the study. Before conducting the study, communities will be informed, and the local structures will be briefed in order for the community to understand the objectives of the study. Surveys on community knowledge, attitudes and practices (KAP). Three KAP surveys will be carried out, at the start (baseline), at the end of the first year and at the end of the second year. The study population will comprise all households from Q2, Q3, Q4 and Q6 of Tambai. Household data will be collected using a questionnaire designed for tablets, with Android system, through the ODK application. Interviews to the head of household will take place at their homes or at a location proposed by them. The KAP about malaria and the different aspects related to the acceptability of the use of repellents will be complemented by focus groups and interviews to local key actors. The collection of the data will be carried out based on two techniques: a) semi-structured interviews (Flick, 2006) to health professionals at the Vinho Health Center (a Maternal and Child Health Nurse and a preventive medicine professional), two activists and local leaders; b) focus groups (Flick, 2006) with matrons and practitioners of traditional medicine from the two communities under study. All participants in this component of the study will be selected intentionally (Hancock, Ockleford and Windridge, 2009). Participation will be made by signing the informed consent. Malaria prevalence The malaria prevalence will be determined in children under 5 years of age in both locations Tambai Q2 and Q6 (intervention) and Tambai Q3 and Q4 (control) at the beginning of the study, at the beginning of the second year and at the end of the second year (end of the project). Consent will be requested to the legal guardian. Capillary blood will be collected using a sterile pipette and individual lancet from all children in the study aged 6-59 months of age for: a) Rapid Diagnostic Test (RDT) for Plasmodium falciparum; b) blood drop smears and c) filter paper for nucleic acids extraction (FTA Cards ThermoFisher). The samples will be coded, and information about the coordinates of the associated household will be recorded. The initial sample processing will be carried out at the laboratory of the Health Center of Ponta Gea - Beira, and the molecular analyzes at IHMT-NOVA - Lisbon, as defined in the MTA. Identification of Plasmodium will be done by PCR and genotyping by whole genome sequencing (Ocholla et al. 2014). Children who are diagnosed with malaria by RDT during the study will be referred to the Mutondo Health Center for treatment. House Spraying with repellent The spraying with IR3535 will start at the beginning of the study and will be performed monthly during the first 3 months, bimonthly during the next 6 months and then quarterly thereafter. The spraying will be carried out inside and outside the houses during the morning. Spraying will only be carried out in intervention quarters, while there will be no spraying in control quarters. Entomology For entomological study households will be randomly selected. In case of failure of satellite imagery and geographical analysis (Wampler et al. 2013, Escamilla et al. 2014), a classic approach based on household inventories in each community will be used manually. Using local mapping and numbering, two lists of numbered households will be created. After a random number generator (e.g., Random.org) will be used to randomly select the households from each list (Wampler et al. 2013). Mosquitoes will be collected twice a month inside and outside the houses at the selected households for 2 years in the intervention and the control quarters. Inside each household, resting mosquitoes that fed the previous night will be collected using a Prokopak aspirator. Outside the residences, CDC- light traps will be placed in the early evening and collected at the following morning. The traps will be placed outdoors, in places equidistant from the buildings (houses) that compose the household. The physiological state of the mosquitoes will be determined, and the females classified as unfed, partially fed, fed, semi-gravid or gravid (Detinova, 1962). Mosquito females will be identified morphologically and confirmed by the PCR (Abilio et al. 2015). The density of mosquitoes caught in the control and intervention communities will be estimated using Williams' geometric mean. GEE and LMM will be explored to compare the two groups along time. The rate of infection will be determined by detecting and quantifying sporozoites in mosquitoes. The mosquitoes' heads and thorax will be removed and prepared to isolate DNA (Abilio et al. 2011). Sporozoites will be detected and quantified by quantitative PCR (qPCR) as described in (Marie et al. 2013). Breeding sites mapping The mosquito breeding sites will be mapped using the "participatory mapping" approach (Dickin et al. 2014). The method will be used to identify the spatial perceptions of malaria risk in the community, based on the spatial identification of existing breeding sites. Interview participants and focus groups will be invited to draw a map of the areas they associate with breeding in communities and describe the important features of the map. These maps can then be digitized on a geographic information system (GIS) to create a georeferenced map of community knowledge. Malaria cases in the community During the study, cases of malaria in Tambai will be monitored through patients who visit the Mutondo Health Center. Patients of all ages from the communities with suspected malaria will be recruited. A questionnaire will be administered. A RDT, a thick drop and blood smear will be performed, and a drop of blood collected for subsequent molecular diagnosis by PCR (as described above). Enrolment and consent will be asked after the patients have received all the care they need in the health unit. Written informed consent will be obtained from the participants or their legal representatives for patients under 18 years old. Study variables The study variables can be divided into variables from the survey (questionnaire) and semi-structured interviews. Variables can be grouped into the following categories: Socio-demographic variables: Age; Sex; Marital status; Profession; Education level; Degree of kinship to the householder; Household size (nº of houses, area); Number of rooms in the house; nº of people per division; Type of construction. Ownership and use of mosquito nets: Ownership of mosquito nets; Use of mosquito nets; Mobility and activities of household members. Malaria knowledge, attitudes and practices: Malaria knowledge level; Malaria attitudes; Mosquito control and bites; Ownership, conservation and use of mosquito nets; Knowledge about repellents; After analyzing the questionnaire data, the variables are deduced; Level of knowledge, attitude and practices about malaria (score categorized as: bad (> 50%), medium (50% -74%) and high (≥75%)). Variables of the study will also be those resulting from the collection of biological material (peripheral blood and mosquitoes): Malaria cases in children under five (RDT, optical microscopy and PCR); Malaria cases attending the Vinho Health Center; Mosquito density inside and outside homes; Percentage of infected mosquitoes; Distribution of mosquito species inside and outside homes; Breeding site (participatory mapping) Sample size for KAP Approximately 126 householders will be surveyed (according to the National Institute of Statistics of Moçambique (INE) report, 2017) corresponding to the total number of households in the communities under study. The chosen locations are comprised of only 126 households, therefore everyone contacted will be included to participate in order to guarantee a maximum power for the study. Prevalence of malaria in the intervention and control community According to the INE report, 2017, the total number of the population in the two study areas is 630 inhabitants. To calculate the number of children under 5 years of age, the total population was multiplied by 17.5%, so the sample will consist of about 110 children. The attempt to include all children aims to try to obtain a sample size that guarantees us the highest possible power. Sample size and statistical power for entomological indicators Different scenarios and programs were explored to calculate the sample size. Initially, a scenario corresponding to a sample size of 40 households (20 in each community) followed for 48 moments was considered as being feasible. The statistical power was calculated using the G * Power v.3.2 program (Faul et al. 2007, 2009), with a correlation between measurements made at different times 0.2; with effect of size f = 0.25; a type I error probability equal to 0.05. A statistical power of 91.1% was estimated. Conversely, for the same parameter values, fixing the power at 95%, in the perspective of using ANOVA for repeated measurements, 24 aggregates were obtained in each community. However, if the assumptions of ANOVA for repeated measurements fail, LMM or GEE models will be used, so it is important to have an idea of the statistical power for these models. Using the Longpower package (https://www.rdocumentation.org/packages/longpower/versions/1.0-19/topics/longpower-package) of the R software, the power associated with the GEE and LMM models, for the previous scenario (24 households), would not be satisfactory. Considering the possibility of missing values (up to 20% in the last visit), a more realistic autoregressive correlation matrix for measurements in each visit, several scenarios were explored in Longpower to guarantee at least powers of 80%, for the various types of models. Thus, also keeping the budget and human resources available, we have 40 households in each location (80 households in total), with a single monthly visit to each household. Data analysis The data will be analyzed using SPSS version 25.0 and R package. Initially, descriptive statistics and exploratory analysis will be performed. Parametric and non-parametric tests will be used, either for independent samples (for example, comparing the two communities at a given time), or for paired samples (comparing each community at the beginning, middle and end of the study). The Chi-square or Fisher's test will be used to test the association between qualitative variables or to compare proportions between independent groups. For paired qualitative data, the McNemar test and Cochran's Q test (for binary data) will be applied. For quantitative variables, the comparison of two independent groups will be performed using the T-Student test, in case of compliance with the assumptions of normality and homogeneity of variances, or alternatively the Mann-Whitney-Wilcoxon test. ANOVA or Kruskal-Wallis test will be used for more than two groups. Signs, Wilcoxon and Friedman tests will be used to compare two or more paired samples. If the sample sizes allow, generalized linear models can be used to jointly analyze several explanatory variables for some events of interest (for example, use of repellent). For the longitudinal comparison between the intervention and control groups, as for entomological indices, Mixed Linear Models and Generalized Estimation Equation Model or a Nonparametric Analysis of Longitudinal Data in Factorial Experiments (nparLD) approach will be applied, using several packages of the R. software. The (CONsolidated Standards of Reporting Trials) CONSORT2010 standards, an intention-to-treat (IT) analysis will be used, and there may be a need to proceed previously with a treatment for missing values. Differences and associations will be considered statistically significant when p <0.05. In order to estimate the prevalence or other proportions of interest, 95% confidence intervals will be obtained by Wald method or alternatively the Wilson or Jeffreys methods. The sporozoite rate (% of mosquitoes with sporozoites in the salivary glands) will be compared between the control and intervention community and the percentage of reduction determined. The temporal and spatial distribution of the different mosquito species and their densities will be explored using ArcGIS 10 with Spatial Analysis and Statistical Tools) and compared between the 2 locations (control x intervention). In the qualitative component, after recording the interviews, their full transcription will be carried out. Data processing and analysis will be manual, using the content analysis technique. Data management - Questionnaires The Open Data Kit (ODK) will be used for data collection and its quality control. The supervisor will review the interviews on a daily basis, before sending them to the secure server located at the CIOB for storing. The server will be accessible only to the data manager and the team of researchers. The transferred data does not include the names of participants, only their codes. The data manager will clean and validate the data daily and provide feedback to the field staff. All physical documents such as informed consents, data collection forms and electronic devices (GPS and tablets) will be stored and only accessible to the CIOB Principal Investigator. The paper documents will be kept for 5 years after the study is completed and then destroyed. Data Management - Prevalence and entomology The data will be collected through a paper data collection form. Every day, each data collector must enter all data in a database available on a computer and send it to the data manager, after a careful revision by responsible of the team. After the data manager receives the data, she/he can create a password protected folder to archive the data on a daily basis. Quality control and monitoring - Training of staff Study personnel will be properly trained by the study's research team. The field team will be subdivided into three (3) groups: Collecting blood samples and administering the survey to householders and conducting interviews with health professionals. The team will consist of 3 health technicians with experience in research. Mosquitoes collection and breeding sites mapping. The team will consist of 2 entomology technicians. Sprayers that will be spraying houses with repellent. The team will consist of 4 people. The training will be focused on ethical aspects, selection of households, children aged 6-59 months, conducting interviews, blood sampling procedures, good clinical and laboratory practices, mosquito collection, manipulation, identification, processing and storing, breeding mapping and spraying techniques and safety. Quality control and monitoring - Supervision of field work The teams will have a field coordinator who will be responsible for monitoring the field work and ensuring data quality. During the study, a team of supervisors will periodically visit the field teams to ensure proper practices. The quality control (QC) of the data will start in the pilot phase, with the identification and replacement of unclear terms and study questions. The field coordinator will also continuously assess compliance with standard operating procedures (SOP) and ethical principles. Data quality and quality assurance The baseline study will include a pilot that will be conducted during training in a community that is not part of the study. The pilot will have the following objectives (i) to evaluate and validate the contents of the data collection instruments, (ii) to ensure compliance with the ethical aspects related to the study. The pilot will help to identify the difficulties and omissions during the questionnaire. The pilot will analyze the fieldwork strategy, team composition, staff responsibilities and time required to complete all activities. The field coordinator will do the quality control of the inquiries made before they are sent to the CIOB and in addition to the quality control done by the field coordinator, the main investigators will also do the supervision to guarantee the quality of the data. Storage of Data The data will be stored at the CIOB. Study documents and data collection instruments, including informed consent in physical format, that will be placed in file folders and kept locked, with restricted access to researchers. All documents related to the research will be filed according to the rules of the legislation and will be kept at CIOB for a period of 5 years. Storage of Biological samples Part of the blood and mosquito sample will be processed in the laboratory of IHMT-NOVA, Portugal, for that purpose a material transfer agreement (MTA) has been signed between CIOB and IHMT-NOVA, Portugal. Part of the blood sample will be kept in the CIOB for future analysis in the scope of PhDs and/or Master programs, explicitly referring to the present project. The samples will be kept in a freezer at CIOB at -80 ° C, for 5 years after the end of the study. Biological samples that will not be used for further analysis will be destroyed at the end of the analysis within the study. Ethical review This study protocol has been submitted for ethical approval to the Institutional Committee on Bioethics for Health of the National Institute of Health (CIBS-INS) and to the National Committee on Bioethics for Health (CNBS) - Clearance Ref. 63/CNBS/20 and Addendum Ref. 11/CNBS/21. The study will be carried out in accordance with the rules of "good clinical practice (GCP)", "good clinical and laboratory practice (GCLP)" and all applicable regulatory requirements, such as the Declaration of Helsinki 2013, at its 64th General Assembly. Informed consent The consent from the heads of households, parents or legal representatives of the children and health professionals selected for the study will be obtained by trained interviewers. They will be fluent in the local languages (Sena and Ndau) or accompanied by an interpreter during the request for informed consent and interviews. All participants will be informed about the objectives of the study, procedures to be performed during the study and potential associated inconveniences, as well as the possibility of leaving the study even after signing the informed consent without any implications for them. The participant will have the right to privacy during the moment of the interview, it will be held in a place where the participant feels comfortable, explaining that all the information provided by him will be confidential, authorization for audio recording of the interview will be specifically requested. Permission from the householder will be ask, for the collection of a finger prick blood sample from the children in the household aged 6-59 months. If the child's legal representative is different from the head of household, informed consent will also be administered to her/him. If the child's parent/legal guardian is under 18, informed consent will be administered to authorize the child to participate. Authorizations for periodic visits throughout the study will be requested while explaining the importance of the procedures for: spraying of their homes (indoor and outdoor) with mosquito repellent (monthly). Only in Tambai Q2 and Q6; collection of mosquitoes from the interior and exterior of their houses (twice a month) in two localities. The inquirers and each participant will sign 2 copies of the informed consent form, in which one copy will remain with the participant and the other will remain with the inquirer for later filing in the study folder. The contact of the principal investigator of the study is an integral part of the document. The informed consent form will serve as proof that the participant accepted to be part of the study by their own free will. In order to guarantee confidentiality, each participant will be assigned an identification code for the questionnaire. Each questionnaire made will have a code corresponding to consent. Participation risks The risks of participation are minimal, but in the event that a participant feels uncomfortable during the interview with any question, they can ask to pass the question on. During the blood sample collection, there may be same risk of bleeding or inflammation from the collection site. To minimize that adequate material will be provided to collect capillary blood (sterilized lancets and disposable gloves) through health professionals trained for that specific activity. Regarding the spraying of houses with repellent, participants will be informed about potential rare adverse effects. If someone in the residence, where the spraying was carried out, has any reaction, which is suspected to be associated with the product, the spraying will be interrupted in that residence and the individual referred to the health service. Benefits of participation Frequently, after carrying out interviews, an increase of knowledge in the communities is noted. This is due to the fact that often those who apply the questionnaire can alert to best practices, which can contribute to improvements in the health of the community. One of the benefits will be knowing whether or not you have malaria, plus children diagnosed with malaria will be able to benefit from treatment.

A prospective experimental Before-After-Control-Intervention test will be carried out with two groups: a) Intervention (community of Tambai): with indoors and outdoors spraying with IR3535 (3-(N-acetyl-N-butyl) aminopropionic acid ethyl ester) and b) Control (community of Micheu 1): without spraying.

Community of Tambai: Indoors and outdoors spraying with IR3535 (3-(N-acetyl-N-butyl) aminopropionic acid ethyl ester)

The spraying will be performed monthly during the first 3 months, bimonthly during the next 6 months and then quarterly thereafter. The spraying will be carried out inside and outside the houses during the morning.

An evaluation scale will be defined with the attribution of points for correct answers, and whoever has <50% will be considered a participant with a bad level of knowledge, attitudes and practices and whoever has 50% -74% and ≥75% will be considered medium and high level of knowledge, attitudes or practices respectively

An evaluation scale will be defined with the attribution of points for correct answers, and whoever has <50% will be considered a participant with a bad level of knowledge, attitudes and practices and whoever has 50% -74% and ≥75% will be considered medium and high level of knowledge, attitudes or practices respectively

The mosquito breeding sites will be mapped using the "participatory mapping" approach (Dickin et al. 2014). Community generated maps will be digitized on a GIS to create a georeferenced map of community knowledge.

Malaria cases in the locality of Bebedo will be monitored through patients of all ages with suspected malaria, attending the Vinho Health Center. Blood will be collected, for blood-smears and PCR.

Inclusion Criteria for children: aged 6 to 59 months resident in the communities under study to give informed consent for participation by parents or legal representatives availability on the day of the study. Exclusion Criteria for children: aged below 6 months and older than 59 months being infected with or having undergone malaria treatment for up to one month lack of consent from parents or legal representatives. Inclusion Criteria for heads of household and health professionals: Be the head of household or someone appointed by her/him and a resident at the place of study Be a health professional in the area of Vinho health center Be available on the day of the interview Have consented to participate in the study Exclusion Criteria for heads of household and health professionals: • Having a mental illness or having a change in behavior

Centro de Investigação Operacional da Beira, Instituto Nacional de Saúde, Ministério de Saúde, Mozambique

The data collected from this study will be disseminated to different audiences at different times. A final report will be presented to the Sofala Provincial Health Directorate, Nhamatanda District Health Directorate and the Ministry of Health trough the National Institute of Health. In order to return the results to the communities, a meeting with the local structures of the communities will be held. The session will open to the general population using a language that will be accessible to the participants in this study, with emphasis on implementing and maintaining practices that protect them from malaria over time. The results will also be disseminated to the scientific community through publications and presentations at national and international scientific conferences. Data and specimens will be available for related studies to MSc and PhD theses at CIOB, Mozambique.

Abilio AP, Kleinschmidt I, Rehman AM, Cuamba N, Ramdeen V, Mthembu DS, Coetzer S, Maharaj R, Wilding CS, Steven A, Coleman M, Hemingway J, Coleman M. The emergence of insecticide resistance in central Mozambique and potential threat to the successful indoor residual spraying malaria control programme. Malar J. 2011 May 2;10:110. doi: 10.1186/1475-2875-10-110.

Abilio AP, Marrune P, de Deus N, Mbofana F, Muianga P, Kampango A. Bio-efficacy of new long-lasting insecticide-treated bed nets against Anopheles funestus and Anopheles gambiae from central and northern Mozambique. Malar J. 2015 Sep 17;14:352. doi: 10.1186/s12936-015-0885-y.

DETINOVA TS. Age-grouping methods in Diptera of medical importance with special reference to some vectors of malaria. Monogr Ser World Health Organ. 1962;47:13-191. No abstract available.

Dickin SK, Schuster-Wallaceab CJ, Elliottc SJ. Mosquitoes & vulnerable spaces: Mapping local knowledge of sites for dengue control in Seremban and Putrajaya Malaysia. Applied Geography. 2014. 46: 71-79. DOI: 10.1016/j.apgeog.2013.11.003

Escamilla V, Emch M, Dandalo L, Miller WC, Martinson F, Hoffman I. Sampling at community level by using satellite imagery and geographical analysis. Bull World Health Organ. 2014 Sep 1;92(9):690-4. doi: 10.2471/BLT.14.140756. Epub 2014 Jun 17.

Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods. 2009 Nov;41(4):1149-60. doi: 10.3758/BRM.41.4.1149.

Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007 May;39(2):175-91. doi: 10.3758/bf03193146.

Marie A, Boissiere A, Tsapi MT, Poinsignon A, Awono-Ambene PH, Morlais I, Remoue F, Cornelie S. Evaluation of a real-time quantitative PCR to measure the wild Plasmodium falciparum infectivity rate in salivary glands of Anopheles gambiae. Malar J. 2013 Jul 2;12:224. doi: 10.1186/1475-2875-12-224.

Morrison AC, Astete H, Chapilliquen F, Ramirez-Prada C, Diaz G, Getis A, Gray K, Scott TW. Evaluation of a sampling methodology for rapid assessment of Aedes aegypti infestation levels in Iquitos, Peru. J Med Entomol. 2004 May;41(3):502-10. doi: 10.1603/0022-2585-41.3.502.

Ocholla H, Preston MD, Mipando M, Jensen AT, Campino S, MacInnis B, Alcock D, Terlouw A, Zongo I, Oudraogo JB, Djimde AA, Assefa S, Doumbo OK, Borrmann S, Nzila A, Marsh K, Fairhurst RM, Nosten F, Anderson TJ, Kwiatkowski DP, Craig A, Clark TG, Montgomery J. Whole-genome scans provide evidence of adaptive evolution in Malawian Plasmodium falciparum isolates. J Infect Dis. 2014 Dec 15;210(12):1991-2000. doi: 10.1093/infdis/jiu349. Epub 2014 Jun 19.

Wampler PJ, Rediske RR, Molla AR. Using ArcMap, Google Earth, and Global Positioning Systems to select and locate random households in rural Haiti. Int J Health Geogr. 2013 Jan 18;12:3. doi: 10.1186/1476-072X-12-3.