BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on October 1, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from

Demographic and health characteristics of Magude district, Southern

Journal: BMJ Open ManuscriptFor ID peerbmjopen-2019-033985 review only Article Type: Original research

Date Submitted by the 04-Sep-2019 Author:

Complete List of Authors: Galatas, Beatriz; Instituto de Salud Global Barcelona, Malaria Elimination Initiative; Centro de Investigacao em Saude de Manhica, ISGLobal Nhacolo, Ariel; Centro de Investigacao em Saude de Manhica Munguambe, Humberto; Centro de Investigacao em Saude de Manhica Marti, Helena; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Jamise, Edgar; Centro de Investigaçao em Saúde de Manhiça Guinovart, Caterina; Barcelona Institute for Global Health Cirera, Laia; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Amone, Felimone; Centro de Investigacao em Saude de Manhica Macete, Eusebio; Centro de Investigação em Saúde de Manhiça Bassat, Quique; BARCELONA INSTITUTE FOR GLOBAL HEALTH

Rabinovich, Regina; Instituto de Salud Global Barcelona, Malaria http://bmjopen.bmj.com/ Elimination Initiative Alonso, Pedro; WHO, GMP Aide, Pedro; Manhiça Health Research Centre Saute, Francisco; Centro de Investigação em Saúde de Manhiça Sacoor, Charfudin; Centro de Investigacao em Saude de Manhica, Department of Demography

Public health < INFECTIOUS DISEASES, Demography < TROPICAL Keywords: MEDICINE, Magude project, Malaria elimination, Population and health on October 1, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 40 BMJ Open

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 40

1 2 3 1 Demographic and health characteristics of Magude district, Southern 4

5 2 Mozambique BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 3 8 4 Authors 9 1, 2 1 1,2 10 5 Beatriz Galatas *¶, Ariel Nhacolo¶ , Helena Martí-Soler , Humberto 11 6 Munguambe1, Edgar Jamise1, Caterina Guinovart2, Laia Cirera1,2, Felimone 12 13 7 Amone1, Eusébio Macete1,3, Quique Bassat1,2,5,6, N. Regina Rabinovich2,7, Pedro 14 8 L. Alonso1,2., Pedro Aide1,4, Francisco Saute1, Charfudin Sacoor1 15 16 9 17 10 Affiliations 18 For peer review only 19 11 1. Centro de Investigação em Saúde da Manhiça (CISM), Manhiça, Mozambique; 20 21 12 2. Barcelona Institute for Global Health (ISGlobal), Hospital Clínic- 22 13 Universitat de Barcelona, Barcelona, Spain. 23 24 14 3. National Directorate of Health, Ministry of Health, Mozambique 25 15 4. National Institute of Health, Ministry of Health, , Mozambique 26 27 16 5. ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain. 28 17 6. Pediatric Infectious Diseases Unit, Pediatrics Department, Hospital Sant 29 30 18 Joan de Déu (University of Barcelona), Barcelona, Spain 31 32 19 7. Harvard T.H. Chan School of Public Health, Boston, MA, USA. 33 34 20 *Corresponding Author 35 36 21 E-mail: [email protected] 37 22 ¶ These authors contributed equally to this work http://bmjopen.bmj.com/ 38 39 23 40 24 Emails of Authors 41 42 25 [email protected] 43 26 [email protected] 44 45 27 [email protected] on October 1, 2021 by guest. Protected copyright. 46 47 28 [email protected] 48 29 [email protected] 49 50 30 [email protected] 51 31 [email protected] 52 53 32 [email protected] 54 33 [email protected] 55 56 34 [email protected] 57 35 [email protected] 58 59 36 [email protected] 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 40 BMJ Open

1 2 3 37 [email protected] 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 38 Key words 8 9 39 10 40 Magude district, population, health, malaria control tools, malaria 11 12 41 elimination 13 42 14 15 16 17 43 Word count 18 For peer review only 19 44 5082 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 40

1 2 3 45 Abstract 4 5 46 (Word count: 289) BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 47 Objectives: A Demographic and Health Platform (DHP) was established in 8 48 Magude in 2015, prior to the deployment of the project’s interventions. 9 10 49 This platform aimed to inform the design, implementation and evaluation of 11 50 a malaria elimination project planned for the district of Magude, through 12 13 51 the identification of households, and population; and the collection of 14 15 52 demographic, health and malaria information. 16 53 Setting: Magude is a rural district of Southern Mozambique which borders 17 18 54 . It Forhas nine peer rural healthreview facilities only and one referral health 19 55 center with an inpatient ward. 20 21 56 Intervention: A baseline census geolocated and enumerated all the 22 57 households, residents and non-residents, collected demographic and socio- 23 24 58 economic information at the household and individual levels, as well as 25 59 information on the coverage and usage of malaria control tools. Inpatient 26 27 60 and outpatient data were obtained from the district health authorities. The 28 29 61 demographic platform was updated in 2016. 30 62 Results: The baseline census conducted in 2015 reported 48,448 (92.1%) 31 32 63 residents and 4,133 (7.9%) non-residents, and 10,965 households in Magude. 33 64 Magude’s population is predominantly young, half of the population is 34 35 65 illiterate and the main economic activities are agriculture and fishing. 36 66 Houses are mainly built with traditional non-durable materials and have 37 http://bmjopen.bmj.com/ 38 67 poor sanitation facilities. Between 2010 and 2014, malaria was the most 39 68 common cause of inpatient admissions (representing up to 20-40% of all 40 41 69 inpatient admissions), followed by HIV (12-22%), and anemia (12-15%). In 42 43 70 early 2015, all-age bed-net usage was between 21.8%-27.1% and the reported 44 71 coverage of indoor residual spraying varied across the district between 45 on October 1, 2021 by guest. Protected copyright. 46 72 30.7% and 79%. 47 73 Conclusion: This study revealed that Magude has limited socio-economic 48 49 74 conditions, poor access to health care services and low coverage of malaria 50 75 vector control interventions. Thus, Magude represented an area where it is 51 52 76 most pressing to demonstrate the feasibility of malaria elimination 53 77 strategies. 54 55 56 57 58 59 60 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 40 BMJ Open

1 2 3 78 Strengths and limitations of this study 4 5 79  Prior to the implementation of the malaria elimination project named BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 80 the Magude project, the most recent census data available had been 8 81 collected during the national census that took place in 2007. Thus, a 9 10 82 Demographic and Health Platform (DHP) was established in Magude in 11 12 83 2015 to inform and accurately plan the interventions contemplated 13 84 under the elimination project. The DHP geolocated and enumerated all 14 15 85 the households, residents and non-residents of Magude, collected 16 86 demographic and socio-economic information at the household and 17 18 87 individual levels,For peer as well reviewas information only on the coverage and usage 19 88 of malaria control tools. 20 21 89  A second census round was conducted one year later acknowledging that 22 23 90 detailed population data collected at individual level at one point 24 91 in time may miss individuals or households that can be captured after 25 26 92 a later update. As a result, this DHP found 48,448 residents and 27 93 10,965 households in Magude in 2015, and 56,943 residents and 11,960 28 29 94 households in 2016. Socio-economic and population mobility patterns 30 95 were depicted by location, sex or age. In 2015, all-age bed-net usage 31 32 96 was between 21.8%-27.1% and the reported coverage of indoor residual 33 34 97 spraying (IRS) varied across the district between 30.7% and 79%. 35 98  The census rounds were planned in close collaboration with the 36 37 99 community leaders and district authorities, and counted with http://bmjopen.bmj.com/ 38100 intensified training of field workers, and a strong component of field 39 40101 and data supervision by experienced demographers. Nevertheless, the 41 102 data collected through the DHP could have been affected by 42 43103 inaccuracies during data collection or entry, or by recall or 44 45104 desirability bias of census participants. on October 1, 2021 by guest. Protected copyright. 46105  To gather baseline data for the Magude project, and understand the 47 48106 health profile of the district, inpatient and outpatient data were 49107 obtained from the Weekly epidemiological bulletin (BES) reported from 50 51108 nine active rural health facilities; and monthly inpatient data was 52 53109 obtained from the health center with an inpatient ward. Between 2010 54110 and 2014, malaria was the most common cause of inpatient admissions 55 56111 (representing up to 20-40% of all inpatient admissions), followed by 57112 HIV (12-22%), and anemia (12-15%). 58 59 60 4

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 40

1 2 3 113  Inpatient information was limited by the quality and accuracy of the 4

5 114 data at the time it was collected, by disease-specific interventions BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 115 or changes in diagnostics, referral or reporting practices. Cases 7 8 116 reported by BES are subject to RDTs stock-outs or reporting 9 10117 inaccuracies, and usually exclude cases identified by community health 11118 workers. This supported the plans to establish a stronger surveillance 12 13119 system in Magude to fully capture changes in malaria during the malaria 14120 elimination project. 15 16121 17 18 For peer review only 19 20122 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 5

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 40 BMJ Open

1 2 3 123 Background 4 5 124 Mozambique is one of the countries with the highest malaria burden in the BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 125 world [1]. Malaria prevalence is heterogeneous within the country ranging 8 126 from high transmission intensity in the North (>50%) to less than 3% in the 9 10127 South [2], which has experienced great progress against malaria in the past 11128 decades, partly as a result of the regional initiatives aiming for malaria 12 13129 elimination in the area. In line with the vision of a malaria free world 14 15130 established by the World Health Organization in its Global Technical 16131 Strategy for 2016-2030 [3], the National Malaria Control Program of 17 18132 Mozambique (NMCP)For decided peer to redefine review its strategic only objectives in order to 19133 include the implementation of malaria elimination activities in the south. 20 21134 In this context, a malaria elimination project named the Magude Project was 22135 designed and evaluated in Magude district, , by the Manhiça 23 24136 Health Research Center (CISM) and the Barcelona Institute of Global Health 25137 (ISGlobal), to assist the NMCP in adopting a malaria elimination strategy 26 27138 based on local evidence [4]. 28 29139 30140 Prior to the initiation of the Magude project in 2015, there was limited 31 32141 and outdated information with regards to the number of individuals living 33142 in Magude, as well as to their demographic and socio-economic 34 35143 characteristics [5]. Thus, detailed information from the whole district was 36144 deemed crucial to inform the elimination strategies that had been planned 37 http://bmjopen.bmj.com/ 38145 for the following years in the district. The process of filling in this 39 146 knowledge gap also aimed to identify and contact key leaders at provincial, 40 41147 district, and at community level, to inform and engage them in the 42 43148 activities prior to their deployment. In this context, a Demographic and 44149 Health Platform adapted from the Health and Demographic Surveillance System 45 on October 1, 2021 by guest. Protected copyright. 46150 (HDSS) method was established in the district of Magude, in February of 47151 2015 with the objective of providing reliable and updated demographic data 48 49152 to inform the project. This platform allowed to plan the activities and to 50153 provide a sampling frame to measure indicators in the community such as 51 52154 malaria prevalence, and to estimate the coverage of indoor residual spraying 53 155 (IRS), coverage and usage of long-lasting insecticide treated nets (LLIN) 54 55156 and mass drug administration (MDA) campaigns. The DHP’s permanent 56 57157 identification numbers were used to track individual’s participation in 58158 each intervention of the Magude project longitudinally, thus allowing to 59 60159 identify and quantify potential challenges to the project, such as reasons 6

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 40

1 2 3 160 for non-participation, or probable sources of imported infections. These 4

5 161 data was also used to accurately measure prevalence at the community BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 162 stratified by age groups and areas of residence. Overall, these findings 7 8 163 were crucial to the design of the Magude project, and offered robust 9 164 evidence to guide malaria elimination strategies in southern Mozambique. 10 11165 12 13166 This article presents the demographic, socio-economic and health 14167 characteristics of the population of Magude, as well as the coverage of 15 16168 malaria control interventions estimated through the baseline census 17169 conducted in early 2015, prior to the deployment of interventions. It also 18 For peer review only 19170 provides a summary of the demographic profile of Magude after updating the 20171 census in 2016. 21 22 23 24172 Methods 25 26 27173 Study area 28174 The district of Magude is located in the North-Western part of Maputo 29 30175 Province and borders with the districts of Massingir, Chókwe and Bilene, 31 32176 from on the North and North-East; with the districts of 33177 Manhiça and Moamba, of Maputo Province in the East and South; and with the 34 35178 South African National Kruger Park, in the West (Figure 1A). 36179 37 http://bmjopen.bmj.com/ 38180 Magude was selected as an appropriate demonstration area for the malaria 39181 elimination project as it was expected to pose the types of challenges that 40 41182 the NMCP would face when implementing a malaria elimination campaign in the 42183 south of the country. First, there were more than 13,000 malaria cases 43 44184 reported in the district in 2014, with the majority of cases observed 45 on October 1, 2021 by guest. Protected copyright. 46185 between January and May, suggesting that the epidemiology of malaria in the 47186 district was representative of most endemic areas in the country with the 48 49187 typical seasonal pattern coinciding with the rainy season [6,7]. Second, 50188 the socio-economic and infrastructural limitations reported for Magude made 51 52189 it sufficiently representative of a rural district of Mozambique, while 53190 still at reach of CISM’s facilities (located in Manhiça district), which 54 55191 facilitated logistics, supervision and quality control processes. Finally, 56 192 the population of Magude had had very limited exposure to research projects 57 58193 or targeted innovative malaria interventions prior to 2015, having only 59 60194 received the programmatic IRS, LLINs and child immunization campaigns 7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 40 BMJ Open

1 2 3 195 conducted by the government. This allowed facing the challenges of working 4

5 196 in an unexposed population that was not biased by previous activities. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 197 Data collection procedures 8 9 198 A Demographic and Health Platform (DHP) was established in Magude in 10199 February 2015, by CISM, which was adapted from the HDSS previously 11 12200 established in Manhiça district by CISM [8]. A baseline census was conducted 13 14201 at that time (February to June) to identify and enumerate all neighborhoods, 15202 households and resident and non-resident individuals in the district. GPS 16 17203 coordinates were also captured for every household. Each individual was 18204 assigned a permanentFor andpeer unique review identification only number, linked to the 19 20205 household number where they were first enumerated. A second census round 21206 was conducted between August and September of 2016, to review and update 22 23207 all information collected during the baseline census and record live births, 24208 deaths, and migrations. New households and new members were enumerated, and 25 26209 their information was collected to update the baseline databases. 27 28210 29211 This DHP defined a household as a structure or set of constructions where 30 31212 an individual or group of individuals live and share domestic activities 32213 and costs (such as eating and sleeping) and recognize one of them as their 33 34214 superior or chief, regardless of their kinship ties. Individuals were 35215 defined as residents if they had lived and slept in a household within the 36 37216 study area for a period of three months or more or intended to do so. Non- http://bmjopen.bmj.com/ 38 217 resident members of a household were defined as those who left the district 39 40218 or who had never lived in the study area for a period of three or more 41 42219 months, but who kept their roles in the households as head of household, 43220 husband, or other important breadwinners, and paid regular visits to their 44 45221 households in Magude. These individuals leave their households for specific on October 1, 2021 by guest. Protected copyright. 46222 reasons such work, studies, or imprisonment; and would otherwise reside in 47 48223 the referred household. This category excluded offspring or other members 49224 who had left the household to build their new residencies outside the study 50 51225 area, irrespective of whether they kept visiting the reference household 52 226 or not. 53 54227 55 56228 Baseline and updated data from the census rounds were collected through 57229 standardized questionnaires using Open Data Kit (https://opendatakit.org/) 58 59230 installed in Android tablets. Data were sent to a secure server at CISM 60231 using Wi-Fi. Information collected on households’ socio-economic 8

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 40

1 2 3 232 characteristics included building materials used for the main house, source 4

5 233 of water and electricity, and household assets and livestock. At the BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 234 individual level, the information collected included the name, sex, date 7 8 235 of birth, relation to the head of household, education, and occupation of 9 236 all residents and non-residents of Magude. Information was also collected 10 11237 on malaria prevention tools, including the possession and use of LLINs, and 12 13238 whether the house had received IRS in the preceding year. Individuals were 14239 also asked about history of fever during the preceding 30 days, and whether 15 16240 they had sought care at a health facility, community health worker, 17241 traditional healer, or none. The specific health facility where the 18 For peer review only 19242 individual sought care was also specified to delineate health facility 20243 catchment areas (HFCA) according to the community. Finally, information was 21 22244 collected on the migration and mobility patterns of all participants, to 23 245 better characterize the mobility profile of the district, and offer 24 25246 potential information on the sources of new infections if transmission was 26 27247 eventually significantly reduced. 28248 29 30249 Administrative and geographical information was obtained directly from the 31250 district authorities and from the most recent district profile of Magude 32 33251 published in 2014 [7] and district statistics performed by the National 34252 Institute of Statistics [9,10]. Data on inpatient discharges and weekly 35 36253 outpatient malaria cases (Boletim Epidemiológico Semanal or “BES”) reported 37 http://bmjopen.bmj.com/ 254 through the national health information system between 2010-2014 were 38 39255 obtained from the District Health authorities. 40 41 256 Patient and Public Involvement 42 43257 This study counted with the support and involvement of the community of 44 45258 Magude in all of its stages. During its design and planning, meetings on October 1, 2021 by guest. Protected copyright. 46259 were held with the district authorities and community leaders in order to 47 48260 inform them of the purpose of the DHP under the scope of the Magude 49261 project, as well as to landscape the number of villages in Magude that 50 51262 had to be covered by the DHP. The study team worked closely with village 52 263 chiefs when performing the household visits, which ensured that all 53 54264 district households were covered by the DHP. Reports were submitted to 55 56265 the administrative authorities to support vector control activities, 57266 among others. Study findings were also communicated in meetings held with 58 59267 the community prior to the deployment of MDAs, where the findings 60 9

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 40 BMJ Open

1 2 3 268 obtained from the study – particularly focusing on malaria – where 4

5 269 discussed in detail [11]. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 270 Data analysis 8 9 271 Data management and descriptive analyses were performed using R Software 10272 [12]. No inferential statistics (p-values or 95% confidence intervals) were 11 12273 calculated as the DHP covers the entire district, and therefore offers 13 14274 direct population parameters. Spatial visualization of the data collected 15275 as well as the open source administrative data bases from DIVA-GIS [13] 16 17276 were performed using QGIS Software [14]. 18 For peer review only 19 20277 Results 21 22 23278 Geography of Magude district 24 2 25279 The district of Magude has an area of 6,961 km and is divided in five 26280 administrative posts, namely: Magude-sede, Motaze, Panjane, Mahele and 27 28281 Mapulanguene (Figure 1A). The vegetation of Magude is dominated by open 29282 forests and savannahs hosting animals such as impalas, warthogs, lions, 30 31283 buffalos and elephants. There is one permanent river (Incomati), which 32 284 flows through the south-western region of the district and constitutes the 33 34285 main source of water in the area, and three intermittent rivers dependent 35 36286 on rainfall, called Massitonto, Uanétse and Mazimuchopes [7]. 37 http://bmjopen.bmj.com/ 38287 Demographic and socio-economic characteristics 39 40288 The baseline census of 2015 registered 10,965 households and 52,802 41289 individuals, of whom 48,448 (91.8%) were residents and 4,133 (7.8%) were 42 43290 non-residents. The district had a population density of 6.9 inhabitants per 44291 km2. The census update conducted between August and October of 2016 45 on October 1, 2021 by guest. Protected copyright. 46292 registered a population of 61,868 individuals. Of the population censed in 47 293 2015, 42,792 (81%) were found during the update round, 6,099 (11.5%) were 48 49294 reported to still be in Magude by a family member although field workers 50 51295 did not find them in other households and therefore were not able to confirm 52296 and complete the registration of migration. There were an additional 3,244 53 54297 (6.1%) individuals censed in 2015 who were not found in 2016 and for whom 55298 there were no informants, but were still considered to be in Magude for 56 57299 this analysis. This update also recorded the death of 670 individuals censed 58300 in 2015, 1,687 live births and 721 immigrations since the baseline census. 59 60 10

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 40

1 2 3 301 An additional 7,325 individuals who were missed in 2015 were censed during 4

5 302 this update (Table 1). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 303 7 8 304 The population in Magude is irregularly distributed among the five 9 305 administrative posts. The majority of the population (73%) lives in the 10 11306 Magude-Sede administrative post (the capital of the district); 14.1% in 12307 Motaze, 5.9% in Panjane, 3.7% in Mapulanguene, and 3.4% in Mahele (Table 13 14308 1). The age and sex structure of the population is dominated by a young 15309 population, and a sharp reduction in the adult population, particularly 16 17310 among males (Figure 1B). There is a steady reduction in the male to female 18 311 ratio from the ageFor of 15 peeronwards, reviewwith a notable only reduction in the proportion 19 20312 of males older than 30. 21 22313 23314 Sixty-two percent of males and 63.1% of females above the age of 14 reported 24 25315 being married or in de-facto union; and 3% of married men practice polygamy. 26316 The illiteracy rate among those aged 6 and older in Magude is 51.9% among 27 28317 females and 47.4% among males. Approximately 37.2% of individuals reached 29318 5th to 9th grade, and 9.5% have completed between 10th and 12th grade 30 31319 (Sup.Table 1). In 2015 Magude had 31 primary schools offering 1st to 5th 32 320 grades; 33 offering 1st-7th grades; one secondary school offering 8th-12th 33 34321 grades; and one private higher-education training center. Xichangana is the 35 36322 main local language of the district and is the mother tongue of 92% of the 37323 population of Magude. However, 51% of the population also reports being http://bmjopen.bmj.com/ 38 39324 able to speak Portuguese, which is the mother tongue of only 3.2% of the 40325 population [7]. 41 42326 43327 Occupations are unevenly distributed between males and females older than 44 45328 18 years of age. The majority of the population (26.1% of males and 70.7% on October 1, 2021 by guest. Protected copyright. 46 47329 of females) relies on subsistence agriculture, fishing or work as cane 48330 cutters in the sugar plantations within Magude, or in Xinavane, in the 49 50331 nearby district of Manhiça. Other occupations include being a salesperson 51332 (8.9% of men and 11% of women), doing construction work (22.1% of men, 0.7% 52 53333 of women), or making coal (11.5% of men and 3.1% of women). A small 54334 proportion of the population are security guards, or work in the public 55 56335 sector, particularly as health professionals, teachers or in the military 57 336 (Sup.Table 1). 58 59337 60 11

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 40 BMJ Open

1 2 3 338 Men represent 54.6% of heads of households in Magude. Only 0.4% of household 4

5 339 heads are younger than 18 and 5.6% are between the ages of 18-24, while BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 340 74.7% are between 25-64 years old, and 19.2% are older than 65 years of age 7 8 341 (Table 2). Forty-three percent of heads of household are single, 35.4% are 9 342 married or in de-facto union, 16.8% are widows and 4.3% separated. The 10 11343 illiteracy rate among heads of households is 59%, while 28.2% have completed

12 th th th 13344 up to 5-7 grade, 8.7% have completed between 10 to 12 grade, and 0.4% 14345 have a university degree (Table 2). 15 16346 17 18347 Living conditionsFor in Magude’speer reviewhouseholds only 19 20348 Settlements in Magude are made of individual household compounds that are 21349 built in proximity to each other in the central areas of each administrative 22 23350 post, and more spread out in the rest of the district. The median household 24351 size is 5, although 1,171 households (10.7%) have only one resident, 40% 25 26352 of whom are older than 65 years old. 27 28353 29354 The majority of households are traditional round-shaped or rectangular- 30 31355 shaped huts constructed using cane (32.5%), cement (26%), mud bricks 32356 (21.6%), or reeds covered by adobe (15.6%) (Table 2). More than half of the 33 34357 households have traditional latrines (53.5%), 10.6% have improved latrines, 35358 while 33.8% of households do not have any form of sanitation facility. 36 37359 Flush toilets can only be found in 2.2% of the households. The primary http://bmjopen.bmj.com/ 38 360 lighting sources used in the households are paraffin (49.6%), electricity 39 40361 (30.8%), candles (9.9%) and solar panels (4.6%). The majority of households 41 42362 cook using wood logs (59.3%). Water is mainly obtained from pumps (34.3%) 43363 or piped water (20.5%), although a quarter of the population collects water 44 45364 directly from the river (24.9%) or from open wells near the river (10.5%) on October 1, 2021 by guest. Protected copyright. 46365 (Table 2). 47 48366 49367 With regards to household assets, 68.5% of households own at least one 50 51368 mobile phone, 36.6% own a radio, 32.7% have a television (TV), while 16.8% 52 369 and 12.2% of households have a fridge and a freezer, respectively. Almost 53 54370 a quarter of households reported having bicycles (23.7%), while 10.8% 55 56371 reported owning cars, 5.3% motorbikes, and 0.7% trucks (Sup.Figure 1). 57372 58 59 60 12

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 40

1 2 3 373 Mobility patterns 4

5 374 Five percent of residents reported having spent the night before the census BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 375 outside of Magude (6.4% of males and 5% of females). Of these, 43% were 7 8 376 younger than 14 years old, 25.1% were 15-29, 18.1% were 30-44, 10.3% were 9 10377 45-64 and 3.4% were older than 65 years old (Sup.Table 2). 11378 12 13379 Almost half of the population (43% of males and 46.9% of females) did not 14380 report frequently travelling outside of Magude. Among the 54.4% who did, 15 16381 the places most frequently visited included Maputo City (55.3% of males, 17382 58.3% of females who travel), South Africa (23.1% of males and 16.8% of 18 For peer review only 19383 females), Gaza Province (12.2% of males and 15.9% of females), Inhambane 20 384 Province (4.3% males and 2.6% females), and other Northern provinces within 21 22385 Mozambique (5.2% of males and 6.3% of females). Younger individuals (less 23 24386 than 15 years old) reported travelling the most to all provinces within the 25387 country, followed by those between the ages of 15-30; while 15 to 45-year- 26 27388 olds are more likely to travel to South Africa. Individuals older than 45 28389 reported travelling more to provinces northern of Maputo province, as well 29 30390 as abroad (Sup.Figure 2). 31 32391 Health and malaria information 33 34392 In 2015, the district health services had eight health facilities (HF), of 35 393 which only one was of Type I, and the other 7 were of Type II. This Type I 36 37394 HF is located in the Magude-Sede administrative post and had a maternity http://bmjopen.bmj.com/ 38 39395 and inpatient department, with 57 beds and three active medical doctors 40396 (0.06 doctors per 1,000 inhabitants). The other 7 health facilities without 41 42397 an inpatient unit but with maternity wards are of Type II, and offer 29 43398 beds altogether but no medical doctors. The Type II HFs are distributed 44 45399 among all administrative posts of the district: two in Magude Sede, one in on October 1, 2021 by guest. Protected copyright. 46400 Motaze, one in Panjane, two in Mahele and one in Mapulanguene. An additional 47 48401 Type II health facility with 3 beds was introduced in 2016 in Magude Sede 49 402 and another one with 5 beds in 2017 in Mapulanguene, thus altogether 50 51403 contributing to 9 Type II health facilities with 37 beds and one Type I 52 53404 health facility for the entire district, with three medical doctors, 15 54405 general nurses and 19 maternal and child health nurses (Figure 1A). Of 55 56406 these, only 3 have access to piped water, 4 have access to public network 57407 electricity, and 5 rely only on solar panels. All health facilities are 58 59408 located on a main road to facilitate access, and the median Euclidian 60 13

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 40 BMJ Open

1 2 3 409 distance from households to the nearest health facility is 2.7 km 4

5 410 (interquartile range [IQR] 1.4–7.9 km), although households are as close BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 411 as 15 m or as far as 38.8 km. There is only one ambulance in Magude, which 7 8 412 is used for transferring patients within the district, but also to the 9 413 hospitals in Xinavane, Manhiça, and Maputo. 10 11414 12 13415 There are 27 Community Health Workers (or APEs, from its acronym in 14416 Portuguese) in Magude distributed throughout the district (Figure 1A). The 15 16417 average distance between households and the assigned APE is approximately 17418 6.3 km (median of 4.4 km, IQR 2.8–5.4 km). APEs are trained by the Ministry 18 For peer review only 19419 of Health to offer primary health services in areas with poor access to 20420 HFs. They provide diagnosis and treatment for malaria, diarrhea, and 21 22421 pneumonia, and refer patients with signs of sickness requiring higher 23 422 medical attention [15]. 24 25423 26 27424 All HFs and APEs in Magude are equipped to diagnose malaria through Rapid 28425 Diagnostic Tests (RDTs) –light microscopy is only available in the Magude 29 30426 Sede Type I HF. They also offer treatment to all positive cases with 31427 Artemether-Lumefantrine, the first line treatment in Mozambique. 32 33428 Intermittent Preventive Treatment in pregnancy (IPTp) using Sulfadoxine- 34429 pyrimethamine is also offered in all HFs of the district. In May of 2014 35 36430 the NMCP conducted an LLIN universal distribution campaign, with 35,432 bed 37 http://bmjopen.bmj.com/ 431 nets distributed in Magude district. This was followed by a focal IRS 38 39432 campaign between October and December of 2014 using the insecticides 40 41433 deltamethrine and DDT, which was only deployed in the Motaze administrative 42434 post, where the malaria case-load was highest. 43 44435 45436 Malaria has traditionally been the first cause of disease in the district, on October 1, 2021 by guest. Protected copyright. 46 47437 responsible for approximately 53% of all consultations reported in 2003 48438 [7]. According to the inpatient discharge data available from January of 49 50439 2011 to December of 2014, between 20-40% of all-age discharges in the 51 440 inpatient department were due to malaria, followed by HIV (12-22%), anemia 52 53441 (12-15%), pneumonia (6-12%), and diarrhea (3-8%), varying by month and year 54 55442 (Fig 2A). The weekly number of malaria cases reported through the BES in 56443 Magude between 2010 and 2014 follows a seasonal pattern with a peak between 57 58444 December and May and a reduction of cases during the dry season (Fig 2B). 59445 According to BES data, there were 293 cases per 1000 population at risk 60 14

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 40

1 2 3 446 reported during the transmission year of July 2011 - June 2012, 247 per 4

5 447 1000 between July 2012 and June 2013, 252 cases per 1000 between July 2013 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 448 – June 2014, and 110 cases per 1,000 between July 2014 and June 2015. [7]. 7 8 449 9 450 The prevalence of history of fever during the 30 days prior to the baseline 10 11451 census was 12.1% in children under the age of 5, 7.7% among 5-14 year-olds, 12 13452 and 11.6% in individuals older than 15. Almost all individuals reported 14453 going to a health facility as their primary source of health care (>99%), 15 16454 while a small proportion mentioned seeking care first from traditional 17455 healers or community health workers. The proportion of individuals who 18 For peer review only 19456 reported sleeping under a bed net the night before was 27.1% among children 20457 under the age of 5, 21.8% in 5-14 year-olds, and 27% among those > 15 years 21 22458 of age (Table 3). The universal bed-net coverage (i.e. one net for every 23 459 two individuals of a household) in 2015 was 52.7% in the administrative 24 25460 post of Magude-Sede, 59.9% in Motaze, 44.1% in Panjane, 52.7% in Mahele and 26 27461 52.1% in Mapulanguene. Finally, the reported coverage of IRS during the 28462 previous 12 months was 49% in Magude-Sede, 79% in Motaze, 41.1% in Panjane, 29 30463 61% in Mahele, and 30.7% in Mapulanguene (Table 2 and Sup.Table 3). 31464 32 33 34 35465 Discussion 36 37466 Overall, the Demographic and Health Platform established in Magude offered http://bmjopen.bmj.com/ 38467 detailed insight on the baseline socio-demographic profile of Magude 39 40468 district prior to the Magude project. It revealed that the number of 41 469 residents in Magude in 2015 (52,804 individuals and 10,965 households) was 42 43470 similar to the number reported by the 2007 national census conducted by the 44 45471 National Institute of Statistics (INE) (53,229 individuals and 11,408 on October 1, 2021 by guest. Protected copyright. 46472 households) [9] but did not coincide with the projected population for 47 48473 Magude in 2015 estimated by the INE (n=62,000) [10]. The census update 49474 conducted in 2016 identified individuals who had been missed by the baseline 50 51475 census, and recorded the births, deaths, and in-migrations since baseline. 52476 As a result, the population of Magude in 2016 assuming that those censed 53 54477 in 2015 who were not found in 2016 were still living in the district, was 55 478 estimated to be 61,868 individuals, which is similar to the projection by 56 57479 INE for this year (62,924). 58 59480 60 15

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 40 BMJ Open

1 2 3 481 The usefulness of a Demographic and Health Surveillance System in 4

5 482 demographic and bio-medical research in African countries has been BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 483 previously described [16]. Having a precise source of population data is 7 8 484 crucial for the correct estimation of disease risk and coverage of health 9 485 indicators in a population, as well as for planning purposes of public 10 11486 health interventions [17]. This platform has shown that detailed population 12 13487 data collected at individual level at one point in time still misses a 14488 number of individuals, who can be captured after a later update. However, 15 16489 all census rounds are also limited by the capacity to find individuals at 17490 home, or informants of unavailable individuals, to identify whether a member 18 For peer review only 19491 lives in the area despite not being found during a specific census round. 20492 To identify or collect information about absent individuals, field 21 22493 supervisors attempt to arrange interviews with them at their places of 23 494 work, or interview their family members at their households during weekends. 24 25495 The same phenomenon is expected to take place when implementing malaria 26 27496 interventions in the community (such as rounds of IRS or mass drug 28497 administrations), which generally challenges the estimation of intervention 29 30498 coverage. 31499 32 33500 Household-level data indicates that the households in Magude are typical 34501 of a rural area with suboptimal conditions regarding access to water, 35 36502 electricity, and sanitation. The majority of houses were relatively 37 http://bmjopen.bmj.com/ 503 homogeneous, built with non-durable materials, such as cane and adobe, and 38 39504 did not have access to clean water, conditions indicative of a low socio- 40 41505 economic status [18]. The age and sex composition of Magude’s population 42506 is typical of rural Mozambique and sub-Saharan Africa, composed primarily 43 44507 by young individuals and a decreasing proportion of adults as age increases 45508 [19]. Another important aspect of the structure of the population of Magude on October 1, 2021 by guest. Protected copyright. 46 47509 is the lower sex ratio (number of males in relation to that of females) 48510 among young and middle-aged adults, which has been reported, in the 49 50511 neighbouring district of Manhiça, to be related to male labour migration 51 512 to Maputo city or South Africa and higher male mortality [19]. 52 53513 54 55514 A large proportion of the heads of households and of the overall population 56515 of Magude reported being illiterate, a major risk factor for health 57 58516 outcomes. Additionally, most individuals reported having occupations 59517 related to agriculture or fishing, which are usually carried out early in 60 16

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 40

1 2 3 518 the morning until around noon. These aspects should be taken into 4

5 519 consideration when planning mobilization campaigns and community-based BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 520 interventions, as visits might have to be conducted at certain times of day 7 8 521 and/or venues outside the households to be able to find these individuals. 9 522 10 11523 Magude is also subject to significant mobility and migration, as more than 12 13524 half of the population reported travelling frequently outside of the 14525 district. The destinations most reported by travelers living in Magude were 15 16526 Maputo and South Africa, followed by Gaza and Inhambane provinces. This 17527 information is useful to evaluate the risk of malaria importation from 18 For peer review only 19528 other areas due to travel or migration [20]. Maputo city and South Africa 20529 are areas with lower malaria burden than Magude [21]; however, the districts 21 22530 surrounding Magude (Manhiça, Moamba), the provinces of Gaza, Inhambane and 23 531 the rest of Mozambique have higher malaria prevalence estimates than Magude 24 25532 district [2], and could be a source of imported infections [22]. These 26 27533 places are commonly visited by all age groups, whereas the lower-endemic 28534 areas are mainly visited by <30 year-olds. This age pattern of migration 29 30535 is typical of the rural areas of Southern Mozambique, where migration rates 31536 are higher among individuals aged 20-45 and their children, and decrease 32 33537 with increasing age when individuals establish as permanent residents in 34538 an area. 35 36539 37 http://bmjopen.bmj.com/ 540 The population of Magude has access to ten health facilities, one of them 38 39541 with an inpatient ward with 57 beds. Thus there are 11.8 beds per 10,000 40 41542 population, a ratio that is higher than the national level (7 beds per 42543 10,000 in Mozambique) [23]. While most respondents indicated seeking health 43 44544 care primarily from formal health facilities, the majority of the population 45545 is scattered throughout the district as far as 38 kms away from a HF and on October 1, 2021 by guest. Protected copyright. 46 47546 it is difficult to access health facilities or APEs due to limited roads 48547 and lack of transportation. This has implications for the delivery of public 49 50548 health interventions, especially for those that rely on the passive 51 549 detection of cases to control transmission in the community. 52 53550 54 55551 Data between 2011 and 2014 from the outpatient and inpatient department of 56552 the Magude Sede health center indicated that malaria was the main cause of 57 58553 hospitalization and later discharge with a clear seasonal pattern. The 59554 population of Magude is also burdened by HIV, anemia, pneumonia, and 60 17

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 40 BMJ Open

1 2 3 555 diarrhea, similar to its neighboring districts of Manhiça and Chokwe [24– 4

5 556 31]. Inpatient information was retrospectively collected, and thus, is BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 557 limited by the quality and accuracy of the data at the time at which it was 7 8 558 collected. It may also be biased by disease-specific interventions or 9 559 changes in diagnostics, referral protocols of severe cases to other 10 11560 districts, or reporting practices. Additionally, cases reported by BES are 12 13561 subject to RDTs stock-outs or reporting inaccuracies. Cases reported by the 14562 APEs are usually excluded, and neither the total number of fevers nor the 15 16563 number of individuals tested are reported. Thus, a stronger surveillance 17564 system than BES is required to fully capture the clinical malaria profile 18 For peer review only 19565 of a district aiming to eliminate malaria [32]. 20566 21 22567 The age-specific self-reported history of fever in the past 30 days (7.7%- 23 568 12.1%) were slightly lower than the average percentage of fevers reported 24 25569 in the past 2 weeks in Maputo Province, which was 15% according to the most 26 27570 recent malaria indicator survey [2]. Approximately half of the households 28571 in Magude reported owning one bed net for every two individuals (universal 29 30572 coverage). However, only a fourth of the population reported sleeping under 31573 a bed net the previous night. This information is indicative of an area 32 33574 where bed net owners do not necessarily use the net as a preventive tool 34575 against malaria during the rainy season, despite this being a prevalent 35 36576 disease in the district. Other demographic and socio-economic risk factors 37 http://bmjopen.bmj.com/ 577 that have been associated with LLIN use, such as households with more 38 39578 children, an undereducated household head, or households far away from the 40 41579 health facilities, might explain the low bed net usage in Magude [33–35]. 42580 43 44581 The reported IRS coverage in the area of Motaze was higher, which 45582 corresponds with the focal IRS campaign conducted by the NMCP that took on October 1, 2021 by guest. Protected copyright. 46 47583 place only there in September of 2014. The coverage reported in other areas 48584 was probably the consequence of recall bias, as a district-wide IRS campaign 49 50585 took place in 2013. Overall, the coverage reported for LLIN and IRS 51 586 throughout the district was below the World Health Organization recommended 52 53587 coverage of >80% [36], leaving a significant proportion of the population 54 55588 unprotected by any of the standard preventative measures. These findings 56589 mirror the estimates reported for Maputo Province in 2011 by the Demographic 57 58590 Health Survey, which found a reported bed net usage rate of 27% in Maputo 59591 Province [37]. The identification of such vector control coverage gaps 60 18

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 40

1 2 3 592 called for a strong community engagement campaign focused on the use of the 4

5 593 bed nets and participation in the yearly rounds of IRS. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 594 Conclusion 10595 Through the establishment of a Demographic and Health Platform in Magude 11 12596 in 2015, which was updated in 2016, and a baseline assessment of the 13597 relevant health indicators, it was possible to fully characterize a district 14 15598 where malaria elimination interventions were to be deployed and evaluated. 16 17599 Magude represents a typical rural district of Mozambique characterized by 18600 limited social andFor economic peer infrastructures, review which only had to be considered for 19 20601 the design and operationalization of the community-based interventions. 21602 This study also revealed low literacy rates among a large proportion of the 22 23603 population and identified agriculture as the main economic activity in the 24604 district. These socio-demographic characteristics suggest that a strong 25 26605 community engagement would be necessary for the implementation of a malaria 27606 elimination project. Half of the population of Magude reported travelling 28 29607 outside of the district to areas of high and low malaria transmission 30 31608 intensity, showing that malaria importation will likely be a source of 32609 continuous transmission even if interruption of local transmission is 33 34610 achieved. Also, the poor access to health care services, as well as to core 35611 malaria vector control interventions such as IRS and bed nets distributed 36 37612 by the NMCP, indicated that these aspects needed to be integrated within http://bmjopen.bmj.com/ 38613 the malaria elimination program planned for the district, in order to 39 40614 maximize the impact of the interventions aiming to interrupt transmission. 41 615 Overall, this district represents the reality of the majority of malaria 42 43616 endemic areas in Sub-Saharan Africa, where elimination is most needed, and 44 45617 where it is most pressing to demonstrate the feasibility of elimination on October 1, 2021 by guest. Protected copyright. 46618 strategies. 47 48619 49 50 51 620 52 53 54 55 56 57 58 59 60 19

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 40 BMJ Open

1 2 3 621 Abbreviations 4 5 622 APEs – Agentes Polivalentes Elementais (Community Health Workers) BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 623 BES – Boletim Epidemiológico Semanal (Weekly Epidemiological Bulletin) 8 624 CISM – Manhiça Health Research Center 9 10625 DHP – Demographic and Health Platform 11626 GPS – Global Positioning System 12 13627 HDSS – Health and Demographic Surveillance System 14 15628 HF – Health Facility 16629 HIV – Human Immunodeficiency virus 17 18630 INE – Instituto NacionalFor peer de Estatística review (National only Institute of Statistics) 19631 IPTp – Intermittent Preventive Treatment in pregnancy 20 21632 IRS – indoor residual spraying 22633 ISGlobal – Barcelona Institute of Global Health 23 24634 LLINs – Long-Lasting Insecticide treated Nets 25635 MDAs – Mass Drug Administration 26 27636 NMCP - National Malaria Control Program of Mozambique 28 29637 RDTs – Rapid Diagnostic Tests 30638 TV – Television 31 32 33 34639 Declarations 35 36 37640 Acknowledgements http://bmjopen.bmj.com/ 38 39641 We would like to thank the community of Magude for participating in this 40642 study, and the team of field workers who have collected the data presented 41 42643 here. We would also like to acknowledge the Administrative and District 43644 Health Authorities of Magude for their collaboration and for providing some 44 45645 of the information also included in this article. We thank everyone who on October 1, 2021 by guest. Protected copyright. 46 646 supported this study directly or indirectly through fieldwork, or analysis 47 48647 support. 49 50 648 51 Ethics approval and consent to participate 52649 The Demographic and Health Platform protocol, consent forms and 53 54650 questionnaires were approved by the CISM internal ethics committee and 55651 written consent from the health authorities was also sought prior to its 56 57652 implementation. Meetings were held with community leaders and with general 58653 members of the community of Magude to inform them about the DHP operations 59 60654 and the malaria elimination project as a whole. A written informed consent 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 40

1 2 3 655 was obtained from all household heads to record household-level information, 4

5 656 as well as from all individuals providing individual information. Informed BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 657 consents for children under the age of 18 years were sought from their 7 8 658 parents or primary caretakers. The collection of district health 9 659 surveillance data was conducted under the protocol that aimed to evaluate 10 11660 the impact of the Magude project on malaria transmission, which was approved 12 13661 by CISM’s internal ethical committee, the Ethics Committee of the Hospital 14662 Clínic of Barcelona, and the Ministry of Health National Bioethics Committee 15 16663 of Mozambique (IRB00002657). 17 18664 Competing InterestsFor peer review only 19 20665 The authors declare that they have no competing interests 21 22666 Funding 23 24667 Funding was provided by the Bill and Melinda Gates Foundation and the 25668 Fundación “la Caixa” Partnership for the Elimination of Malaria in Southern 26 27669 Mozambique (OPP1115265). Q. B. is an ICREA (Institut Catal. de la Recerca 28 670 i Estudis Avan.ats; Catalan Government) Research Professor. ISGlobal is a 29 30671 member of the CERCA Programme, Generalitat de Catalunya. 31 32 33672 Author’s Contributions 34673 BG: participated in the study design and fieldwork, supported in the data 35 36674 cleaning and data analysis process, and wrote the draft of this article 37675 AN: participated in the study design, in study analyses, in the http://bmjopen.bmj.com/ 38 39676 interpretation of results and writing of this article 40677 HMS: cleaned and analyzed the data, and contributed to the writing of this 41 42678 article 43 679 HM: Led the implementation of field activities and data collection process, 44 45680 and participated in the interpretation of results on October 1, 2021 by guest. Protected copyright. 46 47681 EJ: Participated in the study design and implementation of field activities 48682 and data collection. 49 50683 CG: Participated in the interpretation of results and writing of this 51684 article 52 53685 FA: Designed the data collection tools, supported the implementation of 54686 field activities and data cleaning. 55 56687 EM: participated in the study design, and interpretation of results 57 58688 QB: participated in the study design, interpretation of results and writing 59689 of this article 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 40 BMJ Open

1 2 3 690 NRR: participated in the interpretation of results and writing of this 4

5 691 article BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 692 PLA: participated in the study design, interpretation of results and writing 7 8 693 of this article 9 694 PA: participated in the study design, field implementation interpretation 10 11695 of results and writing of this article 12 13696 FS: participated in the study design, supervised field activities, 14697 interpretation of results and writing of this article 15 16698 CS: Conceived the study, participated in interpretation of results and 17699 writing of this article 18 For peer review only 19700 20 21701 Availability of data 22 23702 The datasets collected as part of the Demographic Health Platform and 24703 analysed during the current study are available from the corresponding 25 26704 author upon written request. 27 28705 The data presented in figures 2a and 2b belongs to the Ministry of Health 29706 of Mozambique and is considered as third-party data which can be accessed 30 31707 by contacting the following individuals appointed by the Ministry of health: 32708  To gain access to Inpatient Department data collected by the Ministry 33 34709 of health please contact Dr Baltazar Candrinho, Head of NMCP 35 710 Mozambique, phone number:+258828665730 or e- 36 37711 mail: [email protected] http://bmjopen.bmj.com/ 38 39712  To gain access to the data collected through the Boletim 40713 Epidemiologico Semanal (BES) please contact Dr Lorna Gujral, 41 42714 Department of Epidemiology/MOH/Mozambique, phone 43715 number: +25882 3250800 or e-mail: [email protected] 44 45716 on October 1, 2021 by guest. Protected copyright. 46 47717 48 49 50 51 52 53 54 55 56 57 58 59 60 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 40

1 2 3 718 References 4 5 719 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 720 1. World Health Organization. World malaria report 2018. :210. 8 721 2. Ministério da Saúde (MISAU), Instituto Nacional de Estatística (INE), 9 10722 ICF Internacional. Inquérito de Indicadores de Imunização, Malária e 11723 HIV/SIDA em Moçambique 2015. Maputo. Moçambique. Rockville, Maryland, EUA: 12 13724 INS, INE e ICF International; 2015. 14 15725 3. World Health Organization, others. A framework for malaria elimination. 16726 2017 [cited 2017 Mar 28]; Available from: 17 18727 http://apps.who.int/iris/handle/10665/254761For peer review only 19728 4. Aide P, Candrinho B, Galatas B, Munguambe K, Guinovart C, Luis F, et al. 20 21729 Setting the scene and generating evidence for malaria elimination in 22730 Southern Mozambique. Malar J. 2019;18:190. 23 24731 5. Instituto Nacional de Estatistica. Relatório Final do Inquérito ao 25732 Orçamento Familiar- IOF 2014/15 [Internet]. 2015. Available from: 26 27733 http://www.ine.gov.mz/operacoes-estatisticas/inqueritos/inquerito-sobre- 28 29734 orcamento-familiar 30735 6. Maharaj R, Moonasar D, Baltazar C, Kunene S, Morris N. Sustaining 31 32736 control: lessons from the Lubombo spatial development initiative in southern 33737 Africa. Malar J [Internet]. 2016 [cited 2016 Oct 17];15. Available from: 34 35738 http://malariajournal.biomedcentral.com/articles/10.1186/s12936-016-1453- 36739 9 37 http://bmjopen.bmj.com/ 38740 7. Ministerio de Administração Estatal. Perfil do distrito de Magude, 39 741 Provincia de Maputo (2014). Ministerio de Administração Estatal; 2014. 40 41742 8. Alonso P, Saúte F, Aponte J, Gómez-Olivé F, Nhacolo A, Thompson R, et 42 43743 al. Manhiça demographic surveillance system, Mozambique. Popul Health 44744 Surviv INDEPTH Sites. 2002;1:295–308. 45 on October 1, 2021 by guest. Protected copyright. 46745 9. Instituto Nacional de Estatística. Estatísticas do Distrito de Magude. 47746 2010. 48 49747 10. Instituto Nacional de Estatistica. População Projectada por distritos, 50748 Maputo Província 2007_2040.xls — Instituto Nacional de Estatistica 51 52749 [Internet]. [cited 2018 May 29]. Available from: 53 750 http://www.ine.gov.mz/estatisticas/estatisticas-demograficas-e- 54 55751 indicadores-sociais/projeccoes-da-populacao/populacao-projectada-por- 56 57752 distritos-maputo-provincia-2007_2040.xls/view 58753 11. Portugaliza HP, Galatas B, Nhantumbo H, Djive H, Murato I, Saúte F, et 59 60754 al. Examining community perceptions of malaria to inform elimination efforts 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 40 BMJ Open

1 2 3 755 in Southern Mozambique: a qualitative study. Malar J [Internet]. 2019 [cited 4

5 756 2019 Aug 21];18. Available from: BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 757 https://malariajournal.biomedcentral.com/articles/10.1186/s12936-019- 7 8 758 2867-y 9 759 12. R, Foundation for Statistical Computing. R: A language and environment 10 11760 for statistical computing [Internet]. Vienna, Austria; Available from: 12 13761 http://www.R-project.org/ 14762 13. Hijmans R, Cruz M, Rojas E, Guarino L. DIVA-GIS, version 1.4. A 15 16763 geographic information system for the management and analysis of genetic 17764 resources data. 2001; 18 For peer review only 19765 14. QGIS Development Team (2017) [Internet]. Open Source Geospatial 20766 Foundation Project; Available from: http://qgis.osgeo.org 21 22767 15. Ndima SD, Sidat M, Give C, Ormel H, Kok MC, Taegtmeyer M. Supervision 23 768 of community health workers in Mozambique: a qualitative study of factors 24 25769 influencing motivation and programme implementation. Hum Resour Health. 26 27770 2015;13:63. 28771 16. Bocquier P, Sankoh O, Byass P. Are health and demographic surveillance 29 30772 system estimates sufficiently generalisable? Glob Health Action. 31773 2017;10:1356621. 32 33774 17. Tatem AJ, Campiz N, Gething PW, Snow RW, Linard C. The effects of 34775 spatial population dataset choice on estimates of population at risk of 35 36776 disease. Popul Health Metr. 2011;9:4. 37 http://bmjopen.bmj.com/ 777 18. Pons-Duran C, González R, Quintó L, Munguambe K, Tallada J, Naniche D, 38 39778 et al. Association between HIV infection and socio-economic status: evidence 40 41779 from a semirural area of southern Mozambique. Trop Med Int Health. 42780 2016;21:1513–21. 43 44781 19. Nhacolo AQ, Nhalungo DA, Sacoor CN, Aponte JJ, Thompson R, Alonso P. 45782 Levels and trends of demographic indices in southern rural Mozambique: on October 1, 2021 by guest. Protected copyright. 46 47783 evidence from demographic surveillance in Manhiça district. BMC Public 48784 Health [Internet]. 2006 [cited 2017 Jun 6];6. Available from: 49 50785 http://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-6-291 51 786 20. Ruktanonchai NW, Bhavnani D, Sorichetta A, Bengtsson L, Carter KH, 52 53787 Córdoba RC, et al. Census-derived migration data as a tool for informing 54 55788 malaria elimination policy. Malar J. 2016;15:273. 56789 21. Moonasar D, Maharaj R, Kunene S, Candrinho B, Saute F, Ntshalintshali 57 58790 N, et al. Towards malaria elimination in the MOSASWA (Mozambique, South 59791 Africa and Swaziland) region. Malar J [Internet]. 2016 [cited 2016 Aug 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 40

1 2 3 792 25];15. Available from: 4

5 793 http://malariajournal.biomedcentral.com/articles/10.1186/s12936-016-1470- BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 794 8 7 8 795 22. Ruktanonchai NW, Smith DL, De Leenheer P. Parasite sources and sinks 9 796 in a patched Ross–Macdonald malaria model with human and mosquito movement: 10 11797 Implications for control. Math Biosci. 2016;279:90–101. 12 13798 23. World Health Organization. GHO | By country | Mozambique - statistics 14799 summary (2002 - present) [Internet]. WHO. [cited 2018 May 29]. Available 15 16800 from: http://apps.who.int/gho/data/node.country.country-MOZ?lang=en 17801 24. González R, Augusto OJ, Munguambe K, Pierrat C, Pedro EN, Sacoor C, et 18 For peer review only 19802 al. HIV Incidence and Spatial Clustering in a Rural Area of Southern 20803 Mozambique. PloS One. 2015;10:e0132053. 21 22804 25. Feldblum PJ, Enosse S, Dubé K, Arnaldo P, Muluana C, Banze R, et al. 23 805 HIV prevalence and incidence in a cohort of women at higher risk for HIV 24 25806 acquisition in Chókwè, southern Mozambique. PloS One. 2014;9:e97547. 26 27807 26. Deus N de, João E, Cuamba A, Cassocera M, Luís L, Acácio S, et al. 28808 Epidemiology of Rotavirus Infection in Children from a Rural and Urban 29 30809 Area, in Maputo, Southern Mozambique, before Vaccine Introduction. J Trop 31810 Pediatr. 2018;64:141–5. 32 33811 27. Nhampossa T, Mandomando I, Acacio S, Quintó L, Vubil D, Ruiz J, et al. 34812 Diarrheal Disease in Rural Mozambique: Burden, Risk Factors and Etiology 35 36813 of Diarrheal Disease among Children Aged 0-59 Months Seeking Care at Health 37 http://bmjopen.bmj.com/ 814 Facilities. PloS One. 2015;10:e0119824. 38 39815 28. Aguilar R, Moraleda C, Quintó L, Renom M, Mussacate L, Macete E, et al. 40 41816 Challenges in the diagnosis of iron deficiency in children exposed to high 42817 prevalence of infections. PloS One. 2012;7:e50584. 43 44818 29. Garcia-Basteiro AL, Miranda Ribeiro R, Brew J, Sacoor C, Valencia S, 45819 Bulo H, et al. Tuberculosis on the rise in southern Mozambique (1997-2012). on October 1, 2021 by guest. Protected copyright. 46 47820 Eur Respir J. 2017;49. 48821 30. Lanaspa M, O’Callaghan-Gordo C, Machevo S, Madrid L, Nhampossa T, Acácio 49 50822 S, et al. High prevalence of Pneumocystis jirovecii pneumonia among 51 823 Mozambican children <5 years of age admitted to hospital with clinical 52 53824 severe pneumonia. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol 54 55825 Infect Dis. 2015;21:1018.e9-1018.e15. 56826 31. Roca A, Sigauque B, Quinto L, Morais L, Berenguera A, Corachan M, et 57 58827 al. Estimating the vaccine-preventable burden of hospitalized pneumonia 59828 among young Mozambican children. Vaccine. 2010;28:4851–7. 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 40 BMJ Open

1 2 3 829 32. Barclay VC, Smith RA, Findeis JL. Surveillance considerations for 4

5 830 malaria elimination. Malar J. 2012;11:304. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 831 33. Pinchoff J, Hamapumbu H, Kobayashi T, Simubali L, Stevenson JC, Norris 7 8 832 DE, et al. Factors Associated with Sustained Use of Long-Lasting 9 833 Insecticide-Treated Nets Following a Reduction in Malaria Transmission in 10 11834 Southern Zambia. Am J Trop Med Hyg. 2015;93:954–60. 12 13835 34. García-Basteiro AL, Schwabe C, Aragon C, Baltazar G, Rehman AM, Matias 14836 A, et al. Determinants of bed net use in children under five and household 15 16837 bed net ownership on Bioko Island, Equatorial Guinea. Malar J. 2011;10:179. 17838 35. Atieli HE, Zhou G, Afrane Y, Lee M-C, Mwanzo I, Githeko AK, et al. 18 For peer review only 19839 Insecticide-treated net (ITN) ownership, usage, and malaria transmission 20840 in the highlands of western Kenya. Parasit Vectors. 2011;4:113. 21 22841 36. World Health Organization, World Health Organization, Global Malaria 23 842 Programme. Global technical strategy for malaria, 2016-2030 [Internet]. 24 25843 2015 [cited 2018 Mar 5]. Available from: 26 27844 http://apps.who.int/iris/bitstream/10665/176712/1/9789241564991_eng.pdf?u 28845 a=1 29 30846 37. Ministerio da Saude (MISAU), Instituto Nacional de Estatística (INE) e 31847 ICF International (ICFI). Moçambique Inquérito Demográfico e de Saúde 2011. 32 33848 Calverton Md USA MISAU INE E ICFI. 2011; 34849 35 36850 37 http://bmjopen.bmj.com/ 851 38 39852 40 41853 42854 43 44855 45856 on October 1, 2021 by guest. Protected copyright. 46 47857 48858 49 50859 51 860 52 53 54 55861 56 57 58 59 60 26

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 40

1 2 3 862 Tables 4 5 863 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 864 Table 1: Summary of Magude Population in 2015 and 2016. 8 District Magude Mapulan- Motaze Panjane Mahele 9 Level Sede guene 10 N % N % N % N % N % N % 11 Households 13 12 10965 8011 73.1 1471 627 5.7 377 3.4 479 4.4 13 2015 .4 Households 12 14 11960 8520 71.3 1441 645 5.4 803 6.7 547 4.6 15 2016 .1 Population 14 16 52802 38534 73.0 7421 3122 5.9 1784 3.4 1934 3.7 17 2015 .1 18 Residents For91.8 peer review13 only 48448 35346 73.0 6605 2930 6.0 1695 3.5 1868 3.9 19 * .6 20 Non- 19 4133 7.8* 2970 71.9 816 191 4.6 89 2.2 66 1.6 21 residents .7 22 Unclassif 23 221 0.4* 218 99.5 0 - 1 0.5 0 - 0 - ied 24 Population 13 25 61868 44203 71.4 8149 3576 5.8 3499 5.7 2404 3.8 26 2016 .2 Residents 56943 92.1 12 27 40623 71.4 7275 3361 5.9 3329 5.8 2322 4.1 28 ± * .8 29 Non- 4535 18 7.3* 3266 72.0 842 206 4.5 143 3.2 76 1.7 30 residents ø .6 31 Unclassif 8. 390 § 0.6* 314 80.9 32 9 2.3 27 7.0 6 1.5 32 ied 2 33 Demographic 12 34 Events 2015- 3078 2204 71.6 390 186 6.0 168 5.5 129 4.2 .7 35 16 36

Births 13 http://bmjopen.bmj.com/ 37 1687 54.8 1203 71.4 220 92 5.5 91 5.4 80 4.7 .0 38 Immigrati 8. 39 721 23.4 521 72.3 60 51 7.1 60 8.3 29 4.0 40 ons 3 Deaths 16 41 670 21.8 480 71.6 110 43 6.4 17 2.5 20 3.0 42 .4 43865 * Column % 44866 ± 56,943 = 48,448 (censed in 2015) – 616 (deaths) + 1,670 (births) + 1,235 (immigrations) 45867 + 6,206 (censed in 2016) on October 1, 2021 by guest. Protected copyright. 46868 ø 4,535 = 4,133– 42 (deaths) + 12 (births) + 64 (immigrations) + 368 (censed in 2016) 47869 § 390 = 221 – 8 (deaths) + 5 (births) + 21 (immigrations) + 151 (censed in 2016) 48870 49 50 51 52 53 54 55 56 57 58 59 60 27

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 40 BMJ Open

1 2 3 871 Table 2: Socio-demographic characteristics of heads of households, and 4

5 872 household characteristics (2015). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Characteristics of Heads of households* N % 7 Sex Males 5051 54.6 8 Females 4206 45.4 9 < 18 40 0.4 10 18 – 24 520 5.6 Age group (years) 11 25 – 64 6919 74.7 12 > = 65 1777 19.2 13 Residents of Magude 9257 84.4 14 Illiterate 5460 59.0 5-7th grade 2611 28.2 15 Education level 8 -12th grade 807 8.7 16 (>=6 years old) 17 University 37 0.4 18 For Missingpeer Information review only342 3.7 19 Single 4007 43.3 20 Married or de-facto 3275 35.4 Marital status union 21 (> 14 years old) Divorced 16 0.2 22 Separated 393 4.3 23 Widow 1554 16.8 24 Households 25 characteristics 26 Average household 5 [3-7] 27 size** 28 Lone-resident 1171 10.7 29 households 30 Cane 3506 32.5 31 Cement 2807 26.0 32 Wall material of the Mud bricks 2338 21.6 33 main building Adobe 1703 15.8 34 Zinc plates 319 3.0 35 Wood 128 1.2 36 Traditional latrine 5774 53.5 37 Improved latrine 1144 10.6 http://bmjopen.bmj.com/ 38 Type of toilet WC connected to septic 233 2.2 tank 39 No latrine 3650 33.8 40 Paraffin 5362 49.6 41 Electricity 3321 30.8 Main source of lighting 42 Candles 1074 9.9 energy 43 Solar panels 502 4.6 44 Other 539 5.0 45 Wood logs 6401 59.3 on October 1, 2021 by guest. Protected copyright. 46 Coal 740 6.9 47 Gas 46 0.4 Kitchen Fuel 48 Electricity 65 0.6 49 Paraffin 8 0.1 50 Missing Information 3541 32,8 51 Water pumps 3700 34.3 52 Directly from the 2691 24.9 river 53 Primary source of Piped water 2218 20.5 54 drinking water 55 Open well near the 1138 10.5 river 56 Other 1054 9.8 57 Vector control tools 58 59 60 28

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 40

1 2 3 Universal ITN coverage 5690 53.2 4 (one net per two 5 household members) BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Yes 5722 52.2 7 IRS in past 12 months No 4719 43.0 8 Unknown 524 4.8 9 873 * or subhead (if head does not live in the household) 10874 **Median [Interquartile Range] 11 12875 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 40 BMJ Open

1 2 3 876 Table 3: Individual-level health and malaria prevention indicators in Magude 4

5 877 district 2015). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Age group in Years < 5 5-14 >15 7 N % N % N % 8 Had fever in preceding 30 9 10 days 11 Yes 1043 12.1 1128 7.7 2931 11.6 12 No 7492 87.2 13325 91.2 21977 87.1 13 Unknown 55 0.6 154 1.1 319 1.3 14 Primary source of health 15 care 16 Health Facility 8552 99.6 14548 99.6 25075 99.4 17 Traditional Healer 13 0.2 27 0.2 61 0.2 18 Community HealthFor Worker peer22 review0.3 32 only0.2 66 0.3 19 Slept under a bed net the 20 preceding night 21 22 Yes 2330 27.1 3179 21.8 6799 27.0 23 No 6260 72.9 11428 78.2 18428 73.0 24878 25 879 26 27880 28 29881 30 31 32882 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 30

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 40

1 2 3 883 Figure Titles and Legends 4 5 884 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 885 Fig 1: Map of Magude district (2015). 8 886 A) Administrative and permanent river shape files were obtained from 9 10887 DIVA-GIS ((http://www.diva-gis.org/Data), and confirmed with Magude’s key 11888 informants. GPS positions of households, health facilities and community 12 13889 health workers obtained directly from the field and mapped using QGIS. B) 14 15890 Population pyramid of Magude district (2015). Proportions out of the 16891 total population were calculated for every five-year age and sex group 17 18892 For peer review only 19893 Fig 2: Health profile of Magude district prior to the baseline census 20 21894 (2010-2014). 22895 A) Most common diseases leading to all-age hospitalization and later 23 24896 discharge in Magude reported by the District Health Authorities between 25897 2010-2014. B) Weekly number of outpatient malaria cases observed in 26 27898 Magude between 2010 and 2014 reported through the Weekly Epidemiological 28 29899 Bulletin (BES). 30900 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 40 BMJ Open

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 40

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 40 BMJ Open

1 2 3 Supplementary Tables 4

BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 5 6 Sup.Table 1: Selected demographic and socioeconomic characteristics of Magude’s population 7 (2015). 8 9 Variable / by sex Males Females 10 N Col % N Col % 11 12 Age groups (years) 21775 26673 13 < 1 826 3.8 898 3.4 14 1-4 3391 15.6 3475 13.0 15 16 5-14 7345 33.8 7262 27.2 17 15-64 9334 42.9 13205 49.5 18 For peer review only 19 >=65 864 4.0 1824 6.8 20 Education Level (>= 6 years old) 16712 21461 21 22 Illiterate 7925 47.4 11136 51.9 23 5-9 grade 6221 37.2 7616 35.5 24 25 10-12 grade 1595 9.5 1825 8.5 26 University 43 0.3 20 0.1 27 Missing Information 928 5.6 864 4.0 28 29 Occupation (> 18 years old) 5243 5532 30 Farmer or Fisherman 1371 26.1 3909 70.7 31 32 Salesperson 466 8.9 609 11.0 33 Construction work occupations 1159 22.1 39 0.7 34 35 Coal maker / Lumberjack 603 11.5 173 3.1 36 Guards/Police/Military 539 10.3 45 0.8 37 http://bmjopen.bmj.com/ 38 Health professional 85 1.6 133 2.4 39 Teacher 263 5.0 180 3.3 40 Miner 106 2.0 9 0.2 41 42 Other 651 12.4 435 7.9 43 Marital Status (> 14 years old) 10157 14986 44 45 Single (never married) 6168 60.7 7772 51.9 on October 1, 2021 by guest. Protected copyright. 46 Married or de-facto union 3690 36.3 4702 31.4 47 48 In a polygamic relationship 202 2.0 - - 49 Divorced 7 0.1 13 0.1 50 51 Separated 138 1.4 541 3.6 52 Widow 154 1.5 1958 13.1 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 40

1 2 3 Sup.Table 2: Number of individuals by age group who reported spending one night outside of 4

5 Magude the day before the census visit was performed (2015). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Men Women All 7 Previous night spent Age group N (%) N (%) N (%)* 8 outside of Magude 9 592 5.1 580 5.0 1172 43,0 Yes 10 * 11 0 - 14 No 10874 94.0 10970 94.3 12 NA± 96 0.8 85 0.7 13 324 6.3 362 5.6 686 25.1 Yes 14 * 15 – 29 15 No 4774 92.7 6022 93.3 16 NA± 52 1.0 70 1.1 17 298 11.3 197 5.3 495 18.1 Yes 18 For peer review only * 30 – 44 19 No 2323 87.8 3497 93.9 20 NA± 24 0.9 30 0.8 21 133 8.6 149 4.9 282 10.3 Yes * 22 45 – 64 23 No 1393 90.5 2839 93.8 24 NA± 13 0.8 39 1.3 25 Yes 36 4.2 57 3.1 93 3.4* 26 > 65 No 818 94.7 1744 95.6 27 NA± 10 1.2 23 1.3 28 Yes 1383 6.4 1345 5.0 2728 5.6 29 All ages No 20182 92.7 25072 94,0 45254 93.5 30 NA± 195 0.9 247 1.0 442 0.9 31 * Percentage of those who reported travelling (2,728) 32 ± Missing Information 33 34 35 36 Sup.Table 3: Household-level health and malaria prevention indicators in Magude district (2015). 37 http://bmjopen.bmj.com/ Administrative Magude- 38 Motaze Panjane Mahele Mapulanguene 39 Post Sede 40 Household size 4 [3-6] 4 [3-7] 4 [2-6] 4 [2-6] 3 [2-5.5] 41 42 Members per net 2.0 [1.4-4.0] 2.0 [1.3-3.0] 2.0 [1.6-5.5] 2.0 [1.5-4.0] 2.0 [1.3-4.0] 43 Median [IQR] 44

Universal ITN on October 1, 2021 by guest. Protected copyright. 45 52.7% 59.9% 44.1% 52.7% 52.1% 46 coverage * 47 IRS in past 12 N % N % N % N % N % 48 49 months 50 Yes 3925 49.0 1162 79.0 258 41.1 230 61.0 147 30.7 51 52 No 3627 45.3 293 19.9 352 56.1 133 35.3 314 65.6 53 Unknown 459 5.7 16 1.1 17 2.7 14 3.7 18 3.8 54 55 ** One net for every 2 members of the households 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 40 BMJ Open

1 2 3 Supplementary Figures 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Sup.Figure 1: Household Assets in Magude district in 2015. 7 8 Percentage of households in Magude district with at least one item per asset, and median and 9 10 interquartile range of the assets for which the number of items available was higher than 1. 11 12 13 Sup.Figure 2: Mobility patterns of Magude’s population in 2015. 14 Main destinations per age group among those who reported travelling outside of Magude in 2015. 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 40

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 40 BMJ Open

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 40

1 2 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 3 4 Item Check

5 No Recommendation BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the Yes 7 abstract 8 (b) Provide in the abstract an informative and balanced summary of what was Yes 9 10 done and what was found 11 Introduction 12 Background/rationale 2 Explain the scientific background and rationale for the investigation being Yes 13 14 reported 15 Objectives 3 State specific objectives, including any prespecified hypotheses Yes 16 Methods 17 18 Study design For4 Present peer key elements review of study design earlyonly in the paper Yes 19 Setting 5 Describe the setting, locations, and relevant dates, including periods of Yes 20 recruitment, exposure, follow-up, and data collection 21 22 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of Yes 23 participants 24 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and NA 25 effect modifiers. Give diagnostic criteria, if applicable 26 27 Data sources/ 8* For each variable of interest, give sources of data and details of methods of Yes 28 measurement assessment (measurement). Describe comparability of assessment methods if 29 there is more than one group 30 Bias 9 Describe any efforts to address potential sources of bias Yes 31 32 Study size 10 Explain how the study size was arrived at NA 33 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, Yes 34 describe which groupings were chosen and why 35 36 Statistical methods 12 (a) Describe all statistical methods, including those used to control for Yes 37 confounding http://bmjopen.bmj.com/ 38 (b) Describe any methods used to examine subgroups and interactions NA 39 (c) Explain how missing data were addressed Yes 40 41 (d) If applicable, describe analytical methods taking account of sampling Yes 42 strategy 43 (e) Describe any sensitivity analyses NA 44 45 Results on October 1, 2021 by guest. Protected copyright. 46 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers Yes 47 potentially eligible, examined for eligibility, confirmed eligible, included in the 48 study, completing follow-up, and analysed 49 50 (b) Give reasons for non-participation at each stage Yes 51 (c) Consider use of a flow diagram Yes 52 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) Yes 53 54 and information on exposures and potential confounders 55 (b) Indicate number of participants with missing data for each variable of Yes 56 interest 57 Outcome data 15* Report numbers of outcome events or summary measures Yes 58 59 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates NA 60 and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 40 BMJ Open

1 2 (b) Report category boundaries when continuous variables were categorized Yes 3 (c) If relevant, consider translating estimates of relative risk into absolute risk NA 4

5 for a meaningful time period BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and Yes 7 sensitivity analyses 8 9 Discussion 10 Key results 18 Summarise key results with reference to study objectives Yes 11 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or Yes 12 imprecision. Discuss both direction and magnitude of any potential bias 13 14 Interpretation 20 Give a cautious overall interpretation of results considering objectives, Yes 15 limitations, multiplicity of analyses, results from similar studies, and other 16 relevant evidence 17 18 Generalisability For21 Discuss peer the generalisability review (external validity)only of the study results Yes 19 Other information 20 Funding 22 Give the source of funding and the role of the funders for the present study and, Yes 21 22 if applicable, for the original study on which the present article is based 23 24 *Give information separately for exposed and unexposed groups. 25 26 27 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 28 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 29 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 30 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 31 32 available at www.strobe-statement.org. 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from

A demographic and health platform to inform a malaria elimination project in Magude district, Southern Mozambique ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-033985.R1

Article Type: Original research

Date Submitted by the 18-Feb-2020 Author:

Complete List of Authors: Galatas, Beatriz; Instituto de Salud Global Barcelona, Malaria Elimination Initiative; Centro de Investigacao em Saude de Manhica, ISGLobal Nhacolo, Ariel; Centro de Investigacao em Saude de Manhica Marti, Helena; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Munguambe, Humberto; Centro de Investigacao em Saude de Manhica Jamise, Edgar; Centro de Investigaçao em Saúde de Manhiça Guinovart, Caterina; Barcelona Institute for Global Health Cirera, Laia; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Amone, Felimone; Centro de Investigacao em Saude de Manhica Macete, Eusebio; Centro de Investigação em Saúde de Manhiça Bassat, Quique; BARCELONA INSTITUTE FOR GLOBAL HEALTH http://bmjopen.bmj.com/ Rabinovich, Regina; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Alonso, Pedro; WHO, GMP Aide, Pedro; Manhiça Health Research Centre Saute, Francisco; Centro de Investigação em Saúde de Manhiça Sacoor, Charfudin; Centro de Investigacao em Saude de Manhica, Department of Demography

Primary Subject on October 1, 2021 by guest. Protected copyright. Global health Heading:

Secondary Subject Heading: Epidemiology, Global health, Public health

Public health < INFECTIOUS DISEASES, Demography < TROPICAL Keywords: MEDICINE, Magude project, Malaria elimination, Population and health

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 50 BMJ Open

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 50

1 2 3 4 1 A demographic and health platform to inform a malaria elimination project in Magude

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 2 district, Southern Mozambique 7 8 9 3 10 11 4 Authors 12 13 1, 2 1 1,2 1 1 14 5 Beatriz Galatas *¶, Ariel Nhacolo¶ , Helena Martí-Soler , Humberto Munguambe , Edgar Jamise , 15 16 6 Caterina Guinovart2, Laia Cirera1,2, Felimone Amone1, Eusébio Macete1,3, Quique Bassat1,2,5,6, N. 17 18 7 Regina Rabinovich2,7, PedroFor L. Alonsopeer1,2., Pedro review Aide1,4, Francisco only Saute1, Charfudin Sacoor1 19 20 21 8 22 23 9 Affiliations 24 25 26 10 1. Centro de Investigação em Saúde da Manhiça (CISM), Manhiça, Mozambique; 27 28 11 2. Barcelona Institute for Global Health (ISGlobal), Hospital Clínic-Universitat de Barcelona, 29 30 12 Barcelona, Spain. 31 32 33 13 3. National Directorate of Health, Ministry of Health, Mozambique 34 35 14 4. National Institute of Health, Ministry of Health, Maputo, Mozambique 36 37 http://bmjopen.bmj.com/ 38 15 5. ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain. 39 40 16 6. Pediatric Infectious Diseases Unit, Pediatrics Department, Hospital Sant Joan de Déu (University 41 42 17 of Barcelona), Barcelona, Spain 43 44 45 18 7. Harvard T.H. Chan School of Public Health, Boston, MA, USA. on October 1, 2021 by guest. Protected copyright. 46 47 48 19 *Corresponding Author 49 50 51 20 E-mail: [email protected] 52 53 21 ¶ These authors contributed equally to this work 54 55 22 56 57 58 23 Emails of Authors 59 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 50 BMJ Open

1 2 3 4 24 [email protected]

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 25 [email protected] 7 8 26 [email protected] 9 10 11 27 [email protected] 12 13 28 [email protected] 14 15 16 29 [email protected] 17 18 30 [email protected] peer review only 19 20 31 [email protected] 21 22 23 32 [email protected] 24 25 33 [email protected] 26 27 28 34 [email protected] 29 30 35 [email protected] 31 32 36 [email protected] 33 34 35 36 37 37 Key words http://bmjopen.bmj.com/ 38 39 40 38 41 42 39 Magude district, census, demographic surveillance site, population, health, malaria control tools, 43 44 45 40 malaria elimination on October 1, 2021 by guest. Protected copyright. 46 47 41 48 49 50 51 52 42 Word count 53 54 43 5467 55 56 57 58 59 60 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 50

1 2 3 4 44 Abstract

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 45 (Word count: 302) 8 9 46 Objectives: A Demographic and Health Platform (DHP) was established in Magude in 2015, prior to 10 11 47 the deployment of a project aiming to evaluate the feasibility of malaria elimination in southern 12 13 14 48 Mozambique, named the Magude project. This platform aimed to inform the design, implementation 15 16 49 and evaluation of the Magude project, through the identification of households, and population; and 17 18 For peer review only 19 50 the collection of demographic, health and malaria information. 20 21 51 Setting: Magude is a rural district of Southern Mozambique which borders South Africa. It has nine 22 23 52 peripheral health facilities and one referral health center with an inpatient ward. 24 25 26 53 Intervention: A baseline census enumerated and geolocated all the households, and their resident 27 28 54 and non-resident members, collecting demographic and socio-economic information, and data on the 29 30 31 55 coverage and usage of malaria control tools. Inpatient and outpatient data during the five years (2010- 32 33 56 2014) before the survey were obtained from the district health authorities. The demographic platform 34 35 57 was updated in 2016. 36 37 http://bmjopen.bmj.com/ 38 58 Results: The baseline census conducted in 2015 reported 48,448 (92.1%) residents and 4,133 (7.9%) 39 40 59 non-residents, and 10,965 households. Magude’s population is predominantly young, half of the 41 42 43 60 population has no formal education and the main economic activities are agriculture and fishing. 44 45 61 Houses are mainly built with traditional non-durable materials and have poor sanitation facilities. on October 1, 2021 by guest. Protected copyright. 46 47 62 Between 2010 and 2014, malaria was the most common cause of all-age inpatient discharges 48 49 50 63 (representing 20-40% of all discharges), followed by HIV (12-22%), and anemia (12-15%). In early 51 52 64 2015, all-age bed-net usage was between 21.8% and27.1% and the reported coverage of indoor 53 54 55 65 residual spraying varied across the district between 30.7% and 79%. 56 57 58 59 60 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 50 BMJ Open

1 2 3 4 66 Conclusion: This study revealed that Magude has limited socio-economic conditions, poor access to

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 67 health care services and low coverage of malaria vector control interventions. Thus, Magude 7 8 68 represented an area where it is most pressing to demonstrate the feasibility of malaria elimination. 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 50

1 2 3 4 69 Strengths and limitations of this study

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 70  Strengths: 8 9 71 o A Demographic and Health Platform (DHP) was established in Magude in 2015 to inform 10 11 72 and accurately plan the interventions contemplated under the elimination project, given 12 13 14 73 that the district’s demographic data had not been updated since the 2007 national census. 15 16 74 o A second census round was conducted one year later acknowledging that detailed 17 18 For peer review only 19 75 population data collected at individual level at one point in time may miss individuals or 20 21 76 households that can be captured after a later update. 22 23 77 o The census rounds were planned in close collaboration with the community leaders and 24 25 26 78 district authorities, and counted with intensified training of fieldworkers, and a strong 27 28 79 component of field and data supervision by experienced demographers. 29 30 31 80  Limitations: 32 33 81 o Data collected through the DHP could have been affected by inaccuracies during data 34 35 82 collection or entry, or by recall or desirability bias of census participants. Inpatient 36 37 http://bmjopen.bmj.com/ 38 83 information was limited by the quality and accuracy of the data at the time it was collected, 39 40 84 by disease-specific interventions or changes in diagnostics, referral or reporting practices. 41 42 43 85 The data reported in the weekly epidemiological bulletin (BES) do not distinguish 44 45 86 presumed from confirmed cases, neither by the diagnostic used (rapid diagnostic test or on October 1, 2021 by guest. Protected copyright. 46 47 87 microscopy) or the place of detection (health facility or community), and are subject to 48 49 50 88 RDTs stock-outs or reporting inaccuracies. 51 52 53 54 89 55 56 57 58 59 60 5

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 50 BMJ Open

1 2 3 4 90 Background

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 91 Mozambique is one of the countries with the highest malaria burden in the world [1]. Malaria 8 9 92 prevalence among children of 6 to 59 months of age is heterogeneous within the country ranging from 10 11 93 high transmission intensity in the North (>50%) to less than 3% in the South, as reported in the latest 12 13 14 94 malaria survey conducted in 2015 [2]. Lower prevalence in the Southern region may be related to the 15 16 95 great progress against malaria in the past decades, partly as a result of the regional initiatives aiming 17 18 For peer review only 19 96 for malaria elimination in the area. In line with the vision of a malaria free world established by the 20 21 97 World Health Organization in its Global Technical Strategy for 2016-2030 [3], the National Malaria 22 23 98 Control Program of Mozambique (NMCP) decided to redefine its strategic objectives in order to 24 25 26 99 include the implementation of malaria elimination activities in the South. In this context, a malaria 27 28100 elimination project named the Magude Project was designed and evaluated in Magude district, 29 30 31101 Maputo province, by the Manhiça Health Research Center (CISM) and the Barcelona Institute of 32 33102 Global Health (ISGlobal), to assist the NMCP in adopting a malaria elimination strategy based on 34 35103 local evidence [4]. 36 37 http://bmjopen.bmj.com/ 38104 39 40105 Prior to the initiation of the Magude project in 2015, there was limited and outdated information with 41 42 43106 regards to the number of individuals living in Magude, as well as to their demographic and socio- 44 45107 economic characteristics [5]. Thus, detailed information from the whole district was deemed crucial to on October 1, 2021 by guest. Protected copyright. 46 47108 inform the elimination strategies that had been planned for the following years in the district. The 48 49 50109 process of filling in this knowledge gap also aimed to identify and contact key leaders at provincial, 51 52110 district, and at community level, to inform and engage them in the activities prior to their deployment. 53 54 55111 In this context, a Demographic and Health Platform adapted from the Health and Demographic 56 57112 Surveillance System (HDSS) method was established in the district of Magude, in February of 2015 58 59 60 6

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 50

1 2 3 4 113 with the objective of providing reliable and updated demographic data to inform the project. This

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 114 platform allowed to plan the activities and to provide a sampling frame to measure indicators in the 7 8 115 community such as malaria prevalence, and to estimate the coverage of indoor residual spraying 9 10 11116 (IRS), coverage and usage of long-lasting insecticide treated nets (LLIN) and mass drug 12 13117 administration (MDA) campaigns. The DHP’s permanent identification numbers were used to track 14 15 16118 individual’s participation in each intervention of the Magude project longitudinally, thus allowing to 17 18119 identify and quantify potentialFor challengespeer toreview the project, such only as reasons for non-participation, or 19 20120 probable sources of imported infections. These data was also used to accurately measure prevalence 21 22 23121 at the community stratified by age groups and place of residence. Overall, these findings were crucial 24 25122 to the design of the Magude project, and offered robust evidence to guide malaria elimination 26 27 28123 strategies in Southern Mozambique. 29 30124 31 32125 This article presents the demographic, socio-economic and health characteristics of the population of 33 34 35126 Magude, as well as the coverage of malaria control interventions estimated through the baseline 36 37127 census conducted between February and June of 2015 It also presents the burden of disease in http://bmjopen.bmj.com/ 38 39 40128 Magude during the five years (2010-2014) before the DHP, using the inpatient and outpatient data 41 42129 obtained from the district health authorities. A summary of the demographic profile of Magude after 43 44130 updating the census between August and September of 2016 is also provided Methods 45 on October 1, 2021 by guest. Protected copyright. 46 47 48131 Study area 49 50132 The district of Magude is located in the North-Western part of Maputo Province and borders with the 51 52 53133 districts of Massingir, Chókwe and Bilene, from Gaza Province on the North and North-East; with the 54 55134 districts of Manhiça and Moamba, of Maputo Province in the East and South; and with the South 56 57135 African National Kruger Park, in the West (Figure 1A). 58 59 60 7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 50 BMJ Open

1 2 3 4 136

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 137 Magude was selected as an appropriate demonstration area for the malaria elimination project as it 7 8 138 was expected to pose the types of challenges that the NMCP would face when implementing a malaria 9 10 11139 elimination campaign in the south of the country. This is, there were more than 13,000 malaria cases 12 13140 reported in the district in 2014, with the majority of cases observed between January and May, 14 15 16141 suggesting that the epidemiology of malaria in the district was representative of most endemic areas 17 18142 in the country with the Fortypical seasonal peer pattern review coinciding with only the rainy season [6,7]. Additionally, 19 20143 the socio-economic and infrastructural limitations reported for Magude made it sufficiently 21 22 23144 representative of a rural district of Mozambique, while still at reach of CISM’s facilities (located in 24 25145 Manhiça district), which facilitated the logistics, supervision and quality control processes. Finally, the 26 27 28146 population of Magude had had very limited exposure to research projects or targeted innovative 29 30147 malaria interventions prior to 2015, having only received the programmatic IRS, LLINs and child 31 32148 immunization campaigns conducted by the government. This allowed facing the challenges of working 33 34 35149 in an unexposed population that was not biased by previous activities. 36 37150 http://bmjopen.bmj.com/ 38 39 40 151 Study design 41 42 43152 A Demographic and Health Platform (DHP) was established in Magude in February 2015, by CISM, 44 on October 1, 2021 by guest. Protected copyright. 45153 which was adapted from the health and demographic surveillance system (HDSS) previously 46 47 48154 established in Manhiça district by CISM [8]. A baseline census was conducted at that time (February 49 50155 to June) to identify and enumerate all neighborhoods, households and resident and non-resident 51 52 53156 individuals in these households. Geographic positioning system (GPS) coordinates were also 54 55157 captured for every household. was All individuals were assigned a permanent and unique 56 57158 identification number, linked to the household number where they were first enumerated. A second 58 59 60 8

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 50

1 2 3 4 159 census round was conducted between August and September of 2016, to review and update all

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 160 information collected during the baseline census and record live births, deaths, and migrations that 7 8 161 occurred between the two censuses. New households and new members were enumerated, and their 9 10 11162 information was collected to update the baseline databases. 12 13163 14 15 16164 This DHP defined a household as a structure or set of constructions where an individual or group of 17 18165 individuals live and shareFor domestic peer activities review and costs (such onlyas eating and sleeping) and recognize 19 20166 one of them as their superior or chief, regardless of their kinship ties. Individuals were defined as 21 22 23167 residents if they had lived and slept in a household within the study area for a period of three months 24 25168 or more or intended to do so. Non-resident members of a household were defined as those who left 26 27 28169 the district or who had never lived in the study area for a period of three or more months, but who 29 30170 kept their roles in the households as head of household, husband, or other important breadwinners, 31 32171 and paid regular visits to their households in Magude. These individuals leave their households for 33 34 35172 specific reasons such work, studies, or imprisonment; and would otherwise reside in the referred 36 37173 household. This category excluded offspring or other members who had left the household to live in http://bmjopen.bmj.com/ 38 39 40174 their own households outside the study area, irrespective of whether they kept visiting the reference 41 42175 household or not. 43 44 45176 Data collection procedures on October 1, 2021 by guest. Protected copyright. 46 47 48177 Baseline and updated data from the census rounds were collected through standardized 49 50178 questionnaires using Open Data Kit (https://opendatakit.org/) installed in Android tablets. Data were 51 52 53179 sent to a secure server at CISM using Wi-Fi. Information collected on households’ socio-economic 54 55180 characteristics included building materials used for the main house, source of water and electricity, 56 57181 and household assets and livestock. At the individual level, the information collected included the 58 59 60 9

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 50 BMJ Open

1 2 3 4 182 name, sex, date of birth, relation to the head of household, education, and occupation of all residents

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 183 and non-residents of Magude. Information was also collected on malaria prevention tools, including 7 8 184 the possession and use of LLINs, and whether the house had received IRS in the 12 months 9 10 11185 preceding the census. Individuals were also asked about history of fever during the preceding 30 12 13186 days, and whether they had sought care at a health facility, community health worker, traditional 14 15 16187 healer, or none. The specific health facility where the individual sought care was also specified to 17 18188 delineate health facility Forcatchment peer areas (HFCA) review according to theonly community. Finally, information was 19 20189 collected on the migration and mobility patterns of all participants, to better characterize the mobility 21 22 23190 profile of the district, and offer potential information on the sources of new infections if transmission 24 25191 was eventually significantly reduced. 26 27 28192 29 30193 Administrative and geographical information was obtained directly from the district authorities and 31 32194 from the most recent district profile of Magude published in 2014 by the central government authority 33 34 35195 [7] and district statistics performed by the National Institute of Statistics [9,10]. We retrospectively 36 37196 collected monthly data on inpatient discharges from the referral health center of Magude Sede, as http://bmjopen.bmj.com/ 38 39 40197 well as weekly outpatient malaria cases reported through the weekly epidemiological bulletin (Boletim 41 42198 Epidemiológico Semanal or “BES” in Portuguese) for the period of 2010-2014. The BES did not 43 44199 distinguish between presumed and confirmed cases, the diagnostic used (rapid diagnostic test or 45 on October 1, 2021 by guest. Protected copyright. 46 47200 microscopy) or the location of detection (health facility or in the community). 48 49 50201 Patient and Public Involvement 51 52 53202 This study counted with the support and involvement of the community of Magude in all of its 54 55203 stages. During its design and planning, meetings were held with the district authorities and 56 57204 community leaders in order to inform them of the purpose of the DHP under the scope of the 58 59 60 10

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 50

1 2 3 4 205 Magude project, as well as to landscape the number of villages in Magude that had to be covered

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 206 by the DHP. The study team worked closely with village chiefs when performing the household 7 8 207 visits, which ensured that all district households were covered by the DHP. Reports were submitted 9 10 11208 to the administrative authorities to support vector control activities, among others. Preliminary 12 13209 findings were also communicated in meetings held with the community prior to the deployment of 14 15 16210 MDAs, where the findings obtained from the study – particularly focusing on malaria – where 17 18211 discussed in detail [11].For peer review only 19 20 21212 Data analysis 22 23 24213 Data management and descriptive analyses were performed using R Software [12]. The percentage 25 26214 distribution of categories within any given variable were calculated considering the observations with 27 28 29215 missing values in the denominator. No inferential statistics (p-values or 95% confidence intervals) 30 31216 were calculated as the DHP covers the entire district, and therefore offers direct population 32 33217 parameters. Spatial visualization of the data collected on the field and obtained from web-based open 34 35 36218 source administrative data bases from DIVA-GIS [13] were performed using QGIS Software [14]. 37 http://bmjopen.bmj.com/ 38219 39 40 41220 Household’s socio-economic status was measured using a modified version of Oxford’s Poverty and 42 43221 Human Development Initiative Multidimensional Poverty Index (MPI) [15], following the same 44 on October 1, 2021 by guest. Protected copyright. 45222 methodology used by the INDEPTH group to estimate the poverty index of various HDSS sites 46 47 48223 throughout Africa [16]. In summary, this method categorizes households as “deprived” or “not 49 50224 deprived” based on six deprivation indicators: 1) lack of electricity, 2) lack or sharing of an improved 51 52 53225 sanitation facility; 3) lack of access to improved drinking water source, or source only available more 54 55226 than 30-minute walk, round trip; 4) sand floors in main houses; 5) dung, wood or charcoal used for 56 57227 cooking fuel; and 6) households that do not own a car nor truck and do not own more than one of 58 59 60 11

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 50 BMJ Open

1 2 3 4 228 the following: radio, TV, telephone, bike, motorbike, or refrigerator. A deprivation indicator is

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 229 generated as lowest depravation (0-2), moderate depravation (3-4), or highest depravation (5-6) 7 8 230 [16]. 9 10 11 12 13231 Results 14 15 16 17232 The geography of Magude district 18 For peer review only 19233 The district of Magude has an area of 6,961 km2 and is divided in five administrative posts, namely: 20 21 22234 Magude-sede, Motaze, Panjane, Mahele and Mapulanguene (Figure 1A). The vegetation of Magude 23 24235 is dominated by open forests and savannahs hosting animals such as impalas, warthogs, lions, 25 26 236 buffalos and elephants. There is one permanent river (Incomati), which flows through the south- 27 28 29237 western region of the district and constitutes the main source of water in the area, and three 30 31238 intermittent rivers dependent on rainfall, called Massitonto, Uanétse and Mazimuchopes [7]. 32 33 34 35239 Demographic and socio-economic characteristics 36 37240 The baseline census of 2015 registered 10,965 households and 52,802 individuals, of whom 48,448 http://bmjopen.bmj.com/ 38 39 241 (91.8%) were residents and 4,133 (7.8%) were non-residents. The district had a population density 40 41 42242 of 6.9 inhabitants per km2, which varies by administrative post. The majority of the population (73%) 43 44243 lives in the Magude-Sede administrative post (the capital of the district); 14.1% in Motaze, 5.9% in 45 on October 1, 2021 by guest. Protected copyright. 46 47244 Panjane, 3.7% in Mapulanguene, and 3.4% in Mahele (Table 1). The age and sex structure of the 48 49245 population is dominated by children and young people, and a sharp reduction of adults, particularly 50 51 246 among males (Figure 1B). 52 53 54247 55 56 57 58 59 60 12

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 50

1 2 3 4 248 The census update conducted between August and October of 2016 registered a population of 61,868

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 249 individuals. Of the population censed in 2015, 42,792 (81%) were found during the update round, 7 8 250 6,099 (11.5%) were reported to still be in Magude by a family member although field workers did not 9 10 11251 find them in other households and therefore were not able to confirm and complete the registration of 12 13252 migration. There were an additional 3,244 (6.1%) individuals censed in 2015 who were not found in 14 15 16253 2016 and for whom there were no informants, but were still considered to be in Magude for this 17 18254 analysis. This update alsoFor recorded peer the death review of 670 individuals only censed in 2015, 1,687 live births and 19 20255 721 immigrations since the baseline census. An additional 7,325 individuals who were missed in 2015 21 22 23256 were censed during this update (Table 1). 24 25257 The baseline census indicated that 62% of males and 63.1% of females above the age of 14 reported 26 27 28258 being married or in de-facto union; and 3% of married men practice polygamy. The illiteracy rate 29 30259 among those aged 6 and older in Magude is 51.9% among females and 47.4% among males. 31 32260 Approximately 37.2% of individuals reached 5th to 9th grade, and 9.5% have completed between 10th 33 34 th st th 35261 and 12 grade (Sup.Table 1). In 2015 Magude had 31 primary schools offering 1 to 5 grades; 33 36 37262 offering 1st-7th grades; one secondary school offering 8th-12th grades; and one private higher- http://bmjopen.bmj.com/ 38 39 40263 education training center. According to the national census performed in 2007, Xichangana is the 41 42264 main local language of the district and is the mother tongue of 92% of the population of Magude. 43 44265 However, 51% of the population also reports being able to speak Portuguese (the official language in 45 on October 1, 2021 by guest. Protected copyright. 46 47266 Mozambique), which is the mother tongue of only 3.2%[7]. 48 49267 50 51 268 Occupations are unevenly distributed between males and females older than 18 years of age. The 52 53 54269 majority of the population (26.1% of males and 70.7% of females) relies on subsistence agriculture, 55 56270 fishing or work as cane cutters in the sugar plantations within Magude, or in Xinavane, in the nearby 57 58 59 60 13

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 50 BMJ Open

1 2 3 4 271 district of Manhiça. Other occupations include being a salesperson (8.9% of men and 11% of women),

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 272 doing construction work (22.1% of men, 0.7% of women), or making coal (11.5% of men and 3.1% of 7 8 273 women). A small proportion of the population works as security guards, or work in the public sector, 9 10 11274 particularly as health professionals, teachers or in the army (Sup.Table 1). 12 13275 14 15 16276 Men represent 54.6% of heads of households in Magude. Only 0.4% of household heads are younger 17 18277 than 18 years of age andFor 5.6% arepeer between thereview ages of 18-24, only while 74.7% are between 25-64 years 19 20278 old, and 19.2% are older than 65 years of age (Table 2). Forty-three percent of heads of household 21 22 23279 are single, 35.4% are married or in de-facto union, 16.8% are widows and 4.3% separated. Fifty-nine 24 25280 percent of heads of households reported having no formal education, while 28.2% reported 26

27 th th th 28281 completing up to 5-9 grade, 8.7% up to 10 to 12 grade, and 0.4% reported having a university 29 30282 degree (Table 2). 31 32 33283 Living conditions in Magude’s households 34 35 36284 Settlements in Magude are made of individual household compounds that are built in proximity to 37 http://bmjopen.bmj.com/ 38285 each other in the central areas of each administrative post, and more spread out in the rest of the 39 40 41286 district. The median household size is 5, although 1,171 households (10.7%) have only one resident, 42 43287 40% of whom are older than 65 years old. 44 on October 1, 2021 by guest. Protected copyright. 45288 46 47 48289 The majority of households are traditional round-shaped or rectangular-shaped huts constructed 49 50290 using cane (32.5%), cement (26%), mud bricks (21.6%), or reeds covered by adobe (15.6%) (Table 51 52 53291 2). More than half of the households have traditional latrines (53.5%), 10.6% have improved latrines, 54 55292 while 33.8% of households do not have any form of sanitation facility. Flush toilets can only be found 56 57293 in 2.2% of the households. The primary lighting sources used in the households are paraffin (49.6%), 58 59 60 14

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 50

1 2 3 4 294 electricity (30.8%), candles (9.9%) and solar panels (4.6%). The majority of households cook using

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 295 wood logs (59.3%). Water is mainly obtained from pumps (34.3%) or piped water (20.5%), although 7 8 296 a quarter of the population collects water directly from the river (24.9%) or from open wells near the 9 10 11297 river (10.5%) (Table 2). 12 13298 14 15 16299 With regards to household assets, 68.5% of households own at least one mobile phone, 36.6% own 17 18300 a radio, 32.7% have aFor television peer (TV), while review 16.8% and 12.2% only of households have a fridge and a 19 20301 freezer, respectively. Almost a quarter of households reported having bicycles (23.7%), while 10.8% 21 22 23302 reported owning cars, 5.3% motorbikes, and 0.7% trucks (Sup.Figure 1). 24 25303 26 27 28304 Households in Magude had a median number of 4 depravations. Twenty-five percent of the 29 30305 households were in the lowest depravation category, with 0-2 deprivation indicators; while 55..3% fell 31 32306 in the 3-4 depravation group and 22.4% in the 5-6 depravation group. This distribution was unevenly 33 34 35307 observed among the different administrative posts. Magude Sede had the highest proportion of 36 37308 households in the lowest depravation group (32%), while Mapulanguene had the highest proportion http://bmjopen.bmj.com/ 38 39 40309 of households in the moderate depravation category (70.8%). Panjane and Mahele were the 41 42310 administrative posts with the highest proportion of highly deprived households (Sup. Table 2). 43 44311 45 on October 1, 2021 by guest. Protected copyright. 46 47 48312 Mobility patterns 49 50313 The baseline census shouwed that 5% of residents reported having spent the night before the census 51 52 53314 outside of Magude (6.4% of males and 5% of females). Of these, 43% were younger than 14 years 54 55315 old, 25.1% were 15-29, 18.1% were 30-44, 10.3% were 45-64 and 3.4% were older than 65 years old 56 57316 (Sup.Table 3). 58 59 60 15

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 50 BMJ Open

1 2 3 317 4 5 318 Almost half of the population (43% of males and 46.9% of females) did not report frequently travelling BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 319 outside of Magude during the baseline census. Among the 54.4% who did, the places most frequently 9 10320 visited included Maputo City (55.3% of males, 58.3% of females), South Africa (23.1% of males and 11 12321 16.8% of females), Gaza Province (12.2% of males and 15.9% of females), 13 14 15322 (4.3% males and 2.6% females), and other Northern provinces within Mozambique (5.2% of males 16 17323 and 6.3% of females). Younger individuals (less than 15 years old) reported travelling the most to all 18 For peer review only 19 20324 provinces within the country, followed by those between the ages of 15-30; while 15 to 45-year-olds 21 22325 are more likely to travel to South Africa. Individuals older than 45 reported travelling more to provinces 23 24326 northern of Maputo province, as well as abroad (Sup.Figure 2). 25 26 27 28327 Health and malaria information 29 30328 The district has 10 functional health facilities– one with an inpatient and maternity ward (Type I) and 31 32 33329 nine HFs only with maternity wards (Type II). Eight of the ten HFs were active in 2015, and 2 more 34 35330 Type II HFs were added in 2016 and 2017. The referral HF, located in the Magude-Sede 36 http://bmjopen.bmj.com/ 37331 administrative post, has 57 beds and three active medical doctors (0.06 doctors per 1,000 38 39 40332 inhabitants). The other 9 Type II health facilities have 37 beds in total. The whole district has only 41 42333 three medical doctors, 15 general nurses and 19 maternal and child health nurses (Figure 1A). Only 43 44 45334 3 HFs have access to piped water, 4 have access to public network electricity, and 5 rely only on on October 1, 2021 by guest. Protected copyright. 46 47335 solar panels. All health facilities are located on a local main road to facilitate access, and the median 48 49336 Euclidian distance from households to the nearest health facility is 2.7 km (interquartile range [IQR] 50 51 52337 1.4–7.9 km), although households are as close as 15 m or as far as 38.8 km. There is only one 53 54338 ambulance in Magude, which is used for transferring patients within the district, but also to the 55 56 57339 hospitals in Xinavane, Manhiça, and Maputo. 58 59 60 16

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 50

1 2 3 4 340

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 341 The health system in Mozambique includes Community Health Workers (or APEs, from its acronym 7 8 342 in Portuguese) who are trained by the Ministry of Health to offer primary health services in areas with 9 10 11343 poor access to HFs. They provide diagnosis and treatment for malaria, diarrhea, and pneumonia, and 12 13344 to refer patients with signs of sickness requiring higher medical attention [17]. In Magude there are 14 15 16345 27 APEs distributed throughout the district (Figure 1A). The average distance between households 17 18346 and the nearest APE isFor approximately peer 6.3 km review (median of 4.4 km,only IQR 2.8–5.4 km). 19 20347 All HFs and APEs in Magude are equipped to diagnose malaria through Rapid Diagnostic Tests 21 22 23348 (RDTs) –light microscopy is only available in the Magude Sede Type I HF. They also offer treatment 24 25349 to all positive cases with Artemether-Lumefantrine, the first line treatment in Mozambique. Intermittent 26 27 28350 Preventive Treatment in pregnancy (IPTp) using Sulfadoxine-pyrimethamine is also offered in all HFs 29 30351 of the district. In May of 2014 the NMCP conducted an LLIN universal distribution campaign, with 31 32352 35,432 bed nets distributed in Magude district. This was followed by a focal IRS campaign between 33 34 35353 October and December of 2014 using the insecticides deltamethrine and DDT, which was only 36 37354 deployed in the Motaze administrative post, where the malaria burden was highest. http://bmjopen.bmj.com/ 38 39 40355 41 42356 Malaria has traditionally been the first cause of disease in the district, responsible for approximately 43 44357 53% of all consultations reported in 2003 [7]. According to the inpatient discharge data available from 45 on October 1, 2021 by guest. Protected copyright. 46 47358 January of 2010 to December of 2014, between 20-40% of all-age discharges in the inpatient 48 49359 department were due to malaria, followed by HIV (12-22%), anemia (12-15%), pneumonia (6-12%), 50 51 360 and diarrhea (3-8%), varying by month and year (Fig 2A). The weekly number of malaria cases 52 53 54361 reported through the BES in Magude between 2010 and 2014 follows a seasonal pattern with a peak 55 56362 between December and May and a reduction of cases during the dry season (Fig 2B). According to 57 58 59 60 17

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 50 BMJ Open

1 2 3 4 363 BES data, there were 293 cases per 1000 populationreported during the transmission year of July

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 364 2011 to June 2012, 247 per 1000 between July 2012 and June 2013, 252 cases per 1000 between 7 8 365 July 2013 and June 2014, and 110 cases per 1,000 between July 2014 and June 2015. [7]. 9 10 11366 12 13367 The prevalence of history of fever during the 30 days prior to the baseline census was 12.1% in 14 15 16368 children under the age of 5, 7.7% among 5-14 year-olds, and 11.6% in individuals older than 15 years. 17 18369 Almost all individuals reportedFor going peer to a health review facility as their only primary source of health care (>99%), 19 20370 while a small proportion mentioned seeking care first from traditional healers or community health 21 22 23371 workers. The proportion of individuals who reported sleeping under a bed net the night before was 24 25372 27.1% among children under the age of 5, 21.8% in 5-14 year-olds, and 27% among those > 15 years 26 27 28373 of age (Table 3). The universal bed-net coverage (i.e. one net for every two individuals of a household) 29 30374 in 2015 was 52.7% in the administrative post of Magude-Sede, 59.9% in Motaze, 44.1% in Panjane, 31 32375 52.7% in Mahele and 52.1% in Mapulanguene. Finally, the reported coverage of IRS during the 33 34 35376 previous 12 months was 49% in Magude-Sede, 79% in Motaze, 41.1% in Panjane, 61% in Mahele, 36 37377 and 30.7% in Mapulanguene (Table 2 and Sup.Table 4). http://bmjopen.bmj.com/ 38 39 40 41 42378 Discussion 43 44 45379 Overall, the Demographic and Health Platform established in Magude offered detailed insight on the on October 1, 2021 by guest. Protected copyright. 46 47380 baseline socio-demographic profile of Magude district prior to the Magude project. It revealed that the 48 49381 number of residents in Magude in 2015 (52,804 individuals and 10,965 households) was similar to 50 51 52382 the number reported by the 2007 national census conducted by the National Institute of Statistics 53 54383 (INE) (53,229 individuals and 11,408 households) [9] but did not coincide with the projected 55 56 57384 population for Magude in 2015 estimated by the INE to be 62,000) [10], which is not surprising 58 59 60 18

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 50

1 2 3 4 385 acknowledging the limitations of population projections. The census update conducted in 2016

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 386 identified individuals who had been missed by the baseline census, and recorded the births, deaths, 7 8 387 and in-migrations since baseline. As a result, the population of Magude in 2016 assuming that those 9 10 11388 censed in 2015 who were not found in 2016 were still living in the district, was 61,868 individuals, 12 13389 which is similar to the projection by INE for this year (62,924). 14 15 16390 17 18391 The usefulness of a DemographicFor peer and Health review Surveillance System only in demographic and bio-medical 19 20392 research in African countries has been previously described [18]. Having a precise source of 21 22 23393 population data is crucial for the correct estimation of the risk of disease and coverage of health 24 25394 indicators in a population, as well as for planning purposes of public health interventions [19]. This 26 27 28395 platform has shown that detailed population data collected at individual level at one point in time still 29 30396 misses a number of individuals, who can be captured after a later update. However, all the census 31 32397 rounds faced the challenge of finding individuals at home, or informants of unavailable individuals, to 33 34 35398 identify whether a member lives in the area despite not being found during a specific census round. 36 37399 To identify or collect information about absent individuals, field supervisors attempt to arrange http://bmjopen.bmj.com/ 38 39 40400 interviews with them at their places of work, or interview their family members at their households 41 42401 during weekends. The same phenomenon is expected to take place when implementing malaria 43 44402 interventions in the community (such as rounds of IRS or of mass drug administrations), which 45 on October 1, 2021 by guest. Protected copyright. 46 47403 generally challenges the estimation of intervention coverage. 48 49404 50 51 405 Household-level data indicates that the households in Magude are typical of a rural area with 52 53 54406 suboptimal conditions regarding access to water, electricity, and sanitation. The majority of houses 55 56407 were relatively homogeneous, built with non-durable materials, such as cane and adobe, and did not 57 58 59 60 19

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 50 BMJ Open

1 2 3 4 408 have access to clean water, conditions indicative of a low socio-economic status [20]. The age and

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 409 sex composition of Magude’s population is typical of rural Mozambique and sub-Saharan Africa, 7 8 410 composed primarily by children and young individuals and a decreasing proportion of adults as age 9 10 11411 increases [21]. Another important aspect of the structure of the population of Magude is the lower sex 12 13412 ratio (number of males in relation to that of females) among young and middle-aged adults, which has 14 15 16413 been reported, in the neighbouring district of Manhiça, to be related to male labour migration to 17 18414 Maputo city or South AfricaFor and peerhigher male reviewmortality [21]. only 19 20415 21 22 23416 A large proportion of the heads of households and of the overall population of Magude reported not 24 25417 having received a formal education, a major risk factor for health outcomes. Additionally, most 26 27 28418 individuals reported having occupations related to agriculture or fishing, which are usually carried out 29 30419 early in the morning until around noon. These aspects should be taken into consideration when 31 32420 planning mobilization campaigns and community-based interventions, as visits might have to be 33 34 35421 conducted at certain times of day and/or venues outside the households to be able to find these 36 37422 individuals. http://bmjopen.bmj.com/ 38 39 40423 41 42424 Magude is also subject to significant mobility and migration, as more than half of the population 43 44425 reported travelling frequently outside of the district. The destinations most reported by travelers living 45 on October 1, 2021 by guest. Protected copyright. 46 47426 in Magude were Maputo and South Africa, followed by Gaza and Inhambane provinces. This 48 49427 information is useful to evaluate the risk of malaria importation from other areas due to travel or 50 51 428 migration [22]. Maputo city and South Africa are areas with lower malaria burden than Magude [23]; 52 53 54429 however, the districts surrounding Magude (Manhiça, Moamba), the provinces of Gaza, Inhambane 55 56430 and the rest of Mozambique have higher malaria prevalence estimates than Magude district [2], and 57 58 59 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 50

1 2 3 4 431 could be a source of imported infections [24]. These places are commonly visited by all age groups,

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 432 whereas the lower-endemic areas are mainly visited by <30 year-olds. This age pattern of migration 7 8 433 is typical of the rural areas of Southern Mozambique, where migration rates are higher among 9 10 11434 individuals aged 20 to 45 years and their children, and decrease with increasing age when individuals 12 13435 establish as permanent residents in an area. 14 15 16436 17 18437 The population of MagudeFor has accesspeer to ten review health facilities, oneonly of them with an inpatient ward with 19 20438 57 beds. Thus there are 11.8 beds per 10,000 population, a ratio that is higher than the national level 21 22 23439 (7 beds per 10,000 in Mozambique) [25]. While most respondents indicated seeking health care 24 25440 primarily from formal health facilities, the majority of the population is scattered throughout the district 26 27 28441 as far as 38 kms away from a HF and it is difficult to access health facilities or APEs due to limited 29 30442 roads and lack of transportation. This has implications for the delivery of public health interventions, 31 32443 especially for those that rely on the passive detection of cases to control transmission in the 33 34 35444 community. 36 37445 http://bmjopen.bmj.com/ 38 39 40446 Data between 2010 and 2014 from the outpatient and inpatient department of the Magude Sede 41 42447 health center indicated that malaria was the main cause of hospitalization and later discharge with a 43 44448 clear seasonal pattern. The population of Magude is also burdened by HIV, anemia, pneumonia, and 45 on October 1, 2021 by guest. Protected copyright. 46 47449 diarrhea, similar to its neighboring districts of Manhiça and Chokwé [26–33]. Inpatient information was 48 49450 retrospectively collected, and thus, its quality and accuracy depends on the data at the time at which 50 51 451 it was collected. It may also be biased by disease-specific interventions or changes in diagnostics, 52 53 54452 referral protocols of severe cases to other districts, or reporting practices. Additionally, cases reported 55 56453 by BES are difficult to interpret given their lack of disaggregation in presumed or confirmed, diagnostic 57 58 59 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 50 BMJ Open

1 2 3 4 454 tool used or place of detection of the case. Thus, they are subject to changes in RDT use and stock-

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 455 outs, in care seeking behavior to HFs or APEs, or reporting inaccuracies. In fact, cases reported by 7 8 456 the APEs are usually excluded by those who report in BES, and neither the total number of fevers nor 9 10 11457 the number of individuals tested are reported. Thus, a stronger surveillance system than BES was 12 13458 thought to be required, and consequently established in Magude in 2015,to fully capture the clinical 14 15 16459 malaria profile of a district aiming to eliminate malaria [4,34]. 17 18460 For peer review only 19 20461 The age-specific self-reported history of fever in the past 30 days (7.7%-12.1%) were slightly lower 21 22 23462 than the average percentage of fevers reported in the past 2 weeks in Maputo Province, which was 24 25463 15% according to the most recent malaria indicator survey [2]. Approximately half of the households 26 27 28464 in Magude reported owning one bed net for every two individuals (universal coverage). However, only 29 30465 a fourth of the population reported sleeping under a bed net the previous night. These findings mirror 31 32466 the estimates reported for Maputo Province in 2011 by the Demographic Health Survey, which found 33 34 35467 a reported bed net usage rate of 27% in Maputo Province [35]. This information is indicative of an 36 37468 area where bed net owners do not necessarily use the net as a preventive tool against malaria during http://bmjopen.bmj.com/ 38 39 40469 the rainy season, despite this being a prevalent disease in the district. Other demographic and socio- 41 42470 economic risk factors that have been associated with LLIN use, such as households with more 43 44471 children, an undereducated household head, or households far away from the health facilities, might 45 on October 1, 2021 by guest. Protected copyright. 46 47472 explain the low bed net usage in Magude [36–38]. 48 49473 50 51 474 The reported IRS coverage in the area of Motaze was higher, which corresponds with the focal IRS 52 53 54475 campaign conducted by the NMCP that took place only there in September of 2014. The coverage 55 56476 reported in other areas was probably the consequence of recall bias, as a district-wide IRS campaign 57 58 59 60 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 50

1 2 3 4 477 took place in 2013. Overall, the coverage reported for LLIN and IRS throughout the district was below

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 478 the World Health Organization recommended coverage of >80% [39], leaving a significant proportion 7 8 479 of the population unprotected by any of the standard preventative measures. The identification of 9 10 11480 gaps in such vector control coverages called for a strong community engagement campaign focused 12 13481 on the use of the bed nets and participation in the yearly rounds of IRS. 14 15 16 17 18482 Conclusion For peer review only 19 20 21483 Through the establishment of a Demographic and Health Platform in Magude in 2015, which was 22 23484 updated in 2016, and a baseline assessment of the relevant health indicators, it was possible to fully 24 25485 characterize a district where malaria elimination interventions were to be deployed and evaluated. 26 27 28486 Magude represents a typical rural district of Mozambique characterized by limited social and 29 30487 economic infrastructures, which had to be considered for the design and operationalization of the 31 32 33488 community-based interventions. This study also revealed a low education level among a large 34 35489 proportion of the population and identified agriculture as the main economic activity in the district. 36 http://bmjopen.bmj.com/ 37490 These socio-demographic characteristics suggest that a strong community engagement would be 38 39 40491 necessary for the implementation of a malaria elimination project. Half of the population of Magude 41 42492 reported travelling outside of the district to areas of high and low malaria transmission intensity, 43 44 45493 showing that malaria importation will likely be a source of continuous transmission even if interruption on October 1, 2021 by guest. Protected copyright. 46 47494 of local transmission is achieved. Also, the poor access to health care services, as well as to core 48 49495 malaria vector control interventions such as IRS and bed nets distributed by the NMCP, indicated that 50 51 52496 these aspects needed to be integrated within the malaria elimination program planned for the district, 53 54497 in order to maximize the impact of the interventions aiming to interrupt transmission. Overall, this 55 56 57498 district represents the reality of the majority of malaria endemic areas in Sub-Saharan Africa, where 58 59 60 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 50 BMJ Open

1 2 3 4 499 elimination is most needed, and where it is most pressing to demonstrate the feasibility of elimination

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 500 strategies. 7 8 501 9 10 11 12 13502 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 50

1 2 3 4 503 Abbreviations

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 504 APEs – Agentes Polivalentes Elementais (Community Health Workers) 8 9 505 BES – Boletim Epidemiológico Semanal (Weekly Epidemiological Bulletin) 10 11 506 CISM – Manhiça Health Research Center 12 13 14507 DHP – Demographic and Health Platform 15 16508 GPS – Global Positioning System 17 18 For peer review only 19509 HDSS – Health and Demographic Surveillance System 20 21510 HF – Health Facility 22 23511 HIV – Human Immunodeficiency virus 24 25 26512 INE – Instituto Nacional de Estatística (National Institute of Statistics) 27 28513 IPTp – Intermittent Preventive Treatment in pregnancy 29 30 31514 IRS – indoor residual spraying 32 33515 ISGlobal – Barcelona Institute of Global Health 34 35516 LLINs – Long-Lasting Insecticide treated Nets 36 37 http://bmjopen.bmj.com/ 38517 MDAs – Mass Drug Administration 39 40518 NMCP - National Malaria Control Program of Mozambique 41 42 43519 RDTs – Rapid Diagnostic Tests 44 45520 TV – Television on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 50 BMJ Open

1 2 3 4 521 Declarations

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 522 Acknowledgements 9 10523 We would like to thank the community of Magude for participating in this study, and the team of field 11 12 524 workers who have collected the data presented here. We would also like to acknowledge the 13 14 15525 Administrative and District Health Authorities of Magude for their collaboration and for providing some 16 17526 of the information also included in this article. We thank everyone who supported this study directly 18 For peer review only 19 20527 or indirectly through fieldwork, or analysis support. 21 22 23528 Ethics approval and consent to participate 24 25 26529 The Demographic and Health Platform protocol, consent forms and questionnaires were approved by 27 28530 the CISM internal ethics committee and written consent from the health authorities was also sought 29 30531 prior to its implementation. Meetings were held with community leaders and with general members of 31 32 33532 the community of Magude to inform them about the DHP operations and the malaria elimination 34 35533 project as a whole. A written informed consent was obtained from all household heads to record 36 37 http://bmjopen.bmj.com/ 534 household-level information, as well as from all individuals providing individual information. Informed 38 39 40535 consents for children under the age of 18 years were sought from their parents or primary caretakers. 41 42536 The collection of district health surveillance data was conducted under the protocol that aimed to 43 44 45537 evaluate the impact of the Magude project on malaria transmission, which was approved by CISM’s on October 1, 2021 by guest. Protected copyright. 46 47538 internal ethical committee, the Ethics Committee of the Hospital Clínic of Barcelona, and the Ministry 48 49 539 of Health National Bioethics Committee of Mozambique (IRB00002657). 50 51 52 53540 Competing Interests 54 55541 The authors declare that they have no competing interests 56 57 58 59 60 26

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 50

1 2 3 4 542 Funding

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 543 Funding was provided by the Bill and Melinda Gates Foundation and the Fundación “la Caixa” 7 8 9 544 Partnership for the Elimination of Malaria in Southern Mozambique (OPP1115265). Q. B. is an ICREA 10 11545 (Institut Catal. de la Recerca i Estudis Avan.ats; Catalan Government) Research Professor. ISGlobal 12 13546 is a member of the CERCA Programme, Generalitat de Catalunya. 14 15 16 17547 Author’s Contributions 18 For peer review only 19548 BG: participated in the study design and fieldwork, supported in the data cleaning and data analysis 20 21 22549 process, and wrote the draft of this article 23 24550 AN: participated in the study design, in study analyses, in the interpretation of results and writing of 25 26551 this article 27 28 29552 HMS: cleaned and analyzed the data, and contributed to the writing of this article 30 31553 HM: led the implementation of field activities and data collection process, and participated in the 32 33 34554 interpretation of results 35 36555 EJ: participated in the study design and implementation of field activities and data collection. 37 http://bmjopen.bmj.com/ 38556 CG: participated in the interpretation of results and writing of this article 39 40 41557 LC: participated in the study design, interpretation of results and writing of this article 42 43558 FA: designed the data collection tools, supported the implementation of field activities and data 44 45 on October 1, 2021 by guest. Protected copyright. 46559 cleaning. 47 48560 EM: participated in the study design, and interpretation of results 49 50561 QB: participated in the study design, interpretation of results and writing of this article 51 52 53562 NRR: participated in the interpretation of results and writing of this article 54 55563 PLA: participated in the study design, interpretation of results and writing of this article 56 57 58 59 60 27

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 50 BMJ Open

1 2 3 4 564 PA: participated in the study design, field implementation interpretation of results and writing of this

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 565 article 7 8 566 FS: participated in the study design, supervised field activities, interpretation of results and writing of 9 10 11567 this article 12 13568 CS: conceived the study, participated in interpretation of results and writing of this article 14 15 16569 17 18 For peer review only 19570 Availability of data 20 21571 The datasets collected as part of the Demographic Health Platform and analysed during the current 22 23 24572 study are available from the corresponding author upon written request. 25 26573 The data presented in figures 2a and 2b belongs to the Ministry of Health of Mozambique and is 27 28 29574 considered as third-party data which can be accessed by contacting the following individuals 30 31575 appointed by the Ministry of health: 32 33576  To gain access to Inpatient Department data collected by the Ministry of health please contact 34 35 36577 Dr Baltazar Candrinho, Head of NMCP Mozambique, phone number:+258828665730 or e- 37 http://bmjopen.bmj.com/ 38578 mail: [email protected] 39 40 41579  To gain access to the data collected through the Boletim Epidemiologico Semanal (BES) 42 43580 please contact Dr Lorna Gujral, Department of Epidemiology/MOH/Mozambique, phone 44 on October 1, 2021 by guest. Protected copyright. 45581 number: +25882 3250800 or e-mail: [email protected] 46 47 48582 49 50583 51 52 53 54 55 56 57 58 59 60 28

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 50

1 2 3 4 584 References

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 585 8 9 586 1. World Health Organization. World malaria report 2018. :210. 10 11 12587 2. Ministério da Saúde (MISAU), Instituto Nacional de Estatística (INE), ICF Internacional. Inquérito 13 14588 de Indicadores de Imunização, Malária e HIV/SIDA em Moçambique 2015. Maputo. Moçambique. 15 16 17589 Rockville, Maryland, EUA: INS, INE e ICF International; 2015. 18 For peer review only 19 20590 3. World Health Organization, others. A framework for malaria elimination. 2017 [cited 2017 Mar 28]; 21 22591 Available from: http://apps.who.int/iris/handle/10665/254761 23 24 25592 4. Aide P, Candrinho B, Galatas B, Munguambe K, Guinovart C, Luis F, et al. Setting the scene and 26 27593 generating evidence for malaria elimination in Southern Mozambique. Malar J. 2019;18:190. 28 29 30594 5. Instituto Nacional de Estatistica. Relatório Final do Inquérito ao Orçamento Familiar- IOF 2014/15 31 32 33595 [Internet]. 2015. Available from: http://www.ine.gov.mz/operacoes-estatisticas/inqueritos/inquerito- 34 35596 sobre-orcamento-familiar 36 37 http://bmjopen.bmj.com/ 38597 6. Maharaj R, Moonasar D, Baltazar C, Kunene S, Morris N. Sustaining control: lessons from the 39 40598 Lubombo spatial development initiative in southern Africa. Malar J [Internet]. 2016 [cited 2016 Oct 41 42 43599 17];15. Available from: http://malariajournal.biomedcentral.com/articles/10.1186/s12936-016-1453-9 44 45 on October 1, 2021 by guest. Protected copyright. 46600 7. Ministerio de Administração Estatal. Perfil do distrito de Magude, Provincia de Maputo (2014). 47 48601 Ministerio de Administração Estatal; 2014. 49 50 51602 8. Alonso P, Saúte F, Aponte J, Gómez-Olivé F, Nhacolo A, Thompson R, et al. Manhiça demographic 52 53603 surveillance system, Mozambique. Popul Health Surviv INDEPTH Sites. 2002;1:295–308. 54 55 56604 9. Instituto Nacional de Estatística. Estatísticas do Distrito de Magude. 2010. 57 58 59 60 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 50 BMJ Open

1 2 3 4 605 10. Instituto Nacional de Estatistica. População Projectada por distritos, Maputo Província

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 606 2007_2040.xls — Instituto Nacional de Estatistica [Internet]. [cited 2018 May 29]. Available from: 7 8 607 http://www.ine.gov.mz/estatisticas/estatisticas-demograficas-e-indicadores-sociais/projeccoes-da- 9 10 11608 populacao/populacao-projectada-por-distritos-maputo-provincia-2007_2040.xls/view 12 13 14609 11. Portugaliza HP, Galatas B, Nhantumbo H, Djive H, Murato I, Saúte F, et al. Examining community 15 16610 perceptions of malaria to inform elimination efforts in Southern Mozambique: a qualitative study. Malar 17 18 611 J [Internet]. For 2019 peer [cited review 2019 Aug only 21];18. Available from: 19 20 21612 https://malariajournal.biomedcentral.com/articles/10.1186/s12936-019-2867-y 22 23 24613 12. R, Foundation for Statistical Computing. R: A language and environment for statistical computing 25 26614 [Internet]. Vienna, Austria; Available from: http://www.R-project.org/ 27 28 29615 13. Hijmans R, Cruz M, Rojas E, Guarino L. DIVA-GIS, version 1.4. A geographic information system 30 31 616 for the management and analysis of genetic resources data. 2001; 32 33 34617 14. QGIS Development Team (2017) [Internet]. Open Source Geospatial Foundation Project; 35 36 37618 Available from: http://qgis.osgeo.org http://bmjopen.bmj.com/ 38 39 40619 15. Alkire S, Santos ME. Acute Multidimensional Poverty: A New Index for Developing Countries. 41 42620 SSRN Electron J [Internet]. 2010 [cited 2020 Feb 8]; Available from: 43 44

621 http://www.ssrn.com/abstract=1815243 on October 1, 2021 by guest. Protected copyright. 45 46 47622 16. Coates MM, Kamanda M, Kintu A, Arikpo I, Chauque A, Mengesha MM, et al. A comparison of 48 49 50623 all-cause and cause-specific mortality by household socioeconomic status across seven INDEPTH 51 52624 network health and demographic surveillance systems in sub-Saharan Africa. Glob Health Action. 53 54 625 2019;12:1608013. 55 56 57 58 59 60 30

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 50

1 2 3 4 626 17. Ndima SD, Sidat M, Give C, Ormel H, Kok MC, Taegtmeyer M. Supervision of community health

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 627 workers in Mozambique: a qualitative study of factors influencing motivation and programme 7 8 628 implementation. Hum Resour Health. 2015;13:63. 9 10 11629 18. Bocquier P, Sankoh O, Byass P. Are health and demographic surveillance system estimates 12 13 14630 sufficiently generalisable? Glob Health Action. 2017;10:1356621. 15 16 17631 19. Tatem AJ, Campiz N, Gething PW, Snow RW, Linard C. The effects of spatial population dataset 18 For peer review only 19632 choice on estimates of population at risk of disease. Popul Health Metr. 2011;9:4. 20 21 22633 20. Pons-Duran C, González R, Quintó L, Munguambe K, Tallada J, Naniche D, et al. Association 23 24634 between HIV infection and socio-economic status: evidence from a semirural area of southern 25 26 27635 Mozambique. Trop Med Int Health. 2016;21:1513–21. 28 29 30636 21. Nhacolo AQ, Nhalungo DA, Sacoor CN, Aponte JJ, Thompson R, Alonso P. Levels and trends of 31 32637 demographic indices in southern rural Mozambique: evidence from demographic surveillance in 33 34638 Manhiça district. BMC Public Health [Internet]. 2006 [cited 2017 Jun 6];6. Available from: 35 36 37639 http://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-6-291 http://bmjopen.bmj.com/ 38 39 40640 22. Ruktanonchai NW, Bhavnani D, Sorichetta A, Bengtsson L, Carter KH, Córdoba RC, et al. 41 42641 Census-derived migration data as a tool for informing malaria elimination policy. Malar J. 43 44

642 2016;15:273. on October 1, 2021 by guest. Protected copyright. 45 46 47643 23. Moonasar D, Maharaj R, Kunene S, Candrinho B, Saute F, Ntshalintshali N, et al. Towards malaria 48 49 50644 elimination in the MOSASWA (Mozambique, South Africa and Swaziland) region. Malar J [Internet]. 51 52645 2016 [cited 2016 Aug 25];15. Available from: 53 54 646 http://malariajournal.biomedcentral.com/articles/10.1186/s12936-016-1470-8 55 56 57 58 59 60 31

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 50 BMJ Open

1 2 3 4 647 24. Ruktanonchai NW, Smith DL, De Leenheer P. Parasite sources and sinks in a patched Ross–

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 648 Macdonald malaria model with human and mosquito movement: Implications for control. Math Biosci. 7 8 649 2016;279:90–101. 9 10 11650 25. World Health Organization. GHO | By country | Mozambique - statistics summary (2002 - present) 12 13 14651 [Internet]. WHO. [cited 2018 May 29]. Available from: 15 16652 http://apps.who.int/gho/data/node.country.country-MOZ?lang=en 17 18 For peer review only 19653 26. González R, Augusto OJ, Munguambe K, Pierrat C, Pedro EN, Sacoor C, et al. HIV Incidence 20 21654 and Spatial Clustering in a Rural Area of Southern Mozambique. PloS One. 2015;10:e0132053. 22 23 24655 27. Feldblum PJ, Enosse S, Dubé K, Arnaldo P, Muluana C, Banze R, et al. HIV prevalence and 25 26 27656 incidence in a cohort of women at higher risk for HIV acquisition in Chókwè, southern Mozambique. 28 29657 PloS One. 2014;9:e97547. 30 31 32658 28. Deus N de, João E, Cuamba A, Cassocera M, Luís L, Acácio S, et al. Epidemiology of Rotavirus 33 34659 Infection in Children from a Rural and Urban Area, in Maputo, Southern Mozambique, before Vaccine 35 36 37660 Introduction. J Trop Pediatr. 2018;64:141–5. http://bmjopen.bmj.com/ 38 39 40661 29. Nhampossa T, Mandomando I, Acacio S, Quintó L, Vubil D, Ruiz J, et al. Diarrheal Disease in 41 42662 Rural Mozambique: Burden, Risk Factors and Etiology of Diarrheal Disease among Children Aged 0- 43 44

663 59 Months Seeking Care at Health Facilities. PloS One. 2015;10:e0119824. on October 1, 2021 by guest. Protected copyright. 45 46 47664 30. Aguilar R, Moraleda C, Quintó L, Renom M, Mussacate L, Macete E, et al. Challenges in the 48 49 50665 diagnosis of iron deficiency in children exposed to high prevalence of infections. PloS One. 51 52666 2012;7:e50584. 53 54 55667 31. Garcia-Basteiro AL, Miranda Ribeiro R, Brew J, Sacoor C, Valencia S, Bulo H, et al. Tuberculosis 56 57668 on the rise in southern Mozambique (1997-2012). Eur Respir J. 2017;49. 58 59 60 32

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 50

1 2 3 4 669 32. Lanaspa M, O’Callaghan-Gordo C, Machevo S, Madrid L, Nhampossa T, Acácio S, et al. High

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 670 prevalence of Pneumocystis jirovecii pneumonia among Mozambican children <5 years of age 7 8 671 admitted to hospital with clinical severe pneumonia. Clin Microbiol Infect Off Publ Eur Soc Clin 9 10 11672 Microbiol Infect Dis. 2015;21:1018.e9-1018.e15. 12 13 14673 33. Roca A, Sigauque B, Quinto L, Morais L, Berenguera A, Corachan M, et al. Estimating the 15 16674 vaccine-preventable burden of hospitalized pneumonia among young Mozambican children. Vaccine. 17 18 675 2010;28:4851–7. For peer review only 19 20 21676 34. Barclay VC, Smith RA, Findeis JL. Surveillance considerations for malaria elimination. Malar J. 22 23 24677 2012;11:304. 25 26 27678 35. Ministerio da Saude (MISAU), Instituto Nacional de Estatística (INE) e ICF International (ICFI). 28 29679 Moçambique Inquérito Demográfico e de Saúde 2011. Calverton Md USA MISAU INE E ICFI. 2011; 30 31 32680 36. Pinchoff J, Hamapumbu H, Kobayashi T, Simubali L, Stevenson JC, Norris DE, et al. Factors 33 34681 Associated with Sustained Use of Long-Lasting Insecticide-Treated Nets Following a Reduction in 35 36 37682 Malaria Transmission in Southern Zambia. Am J Trop Med Hyg. 2015;93:954–60. http://bmjopen.bmj.com/ 38 39 40683 37. García-Basteiro AL, Schwabe C, Aragon C, Baltazar G, Rehman AM, Matias A, et al. 41 42684 Determinants of bed net use in children under five and household bed net ownership on Bioko Island, 43 44

685 Equatorial Guinea. Malar J. 2011;10:179. on October 1, 2021 by guest. Protected copyright. 45 46 47686 38. Atieli HE, Zhou G, Afrane Y, Lee M-C, Mwanzo I, Githeko AK, et al. Insecticide-treated net (ITN) 48 49 50687 ownership, usage, and malaria transmission in the highlands of western Kenya. Parasit Vectors. 51 52688 2011;4:113. 53 54 55 56 57 58 59 60 33

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 50 BMJ Open

1 2 3 4 689 39. World Health Organization, World Health Organization, Global Malaria Programme. Global

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 690 technical strategy for malaria, 2016-2030 [Internet]. 2015 [cited 2018 Mar 5]. Available from: 7 8 691 http://apps.who.int/iris/bitstream/10665/176712/1/9789241564991_eng.pdf?ua=1 9 10 11692 12 13 14 15 16693 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 34

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 50

1 2 3 4 694 Tables

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 695 8 9 696 Table 1: Summary of Magude Population in 2015 and 2016. 10 11 Magude Mapulan- 12 District Level Motaze Panjane Mahele 13 Sede guene 14 15 N % N % N % N % N % N % 16 Households 13. 17 10965 8011 73.1 1471 627 5.7 377 3.4 479 4.4 18 2015 For peer review4 only 19 20 Households 12. 11960 8520 71.3 1441 645 5.4 803 6.7 547 4.6 21 2016 1 22 23 Population 2015 14. 24 52802 38534 73.0 7421 3122 5.9 1784 3.4 1934 3.7 25 1 26 Residents 13. 27 48448 91.8* 35346 73.0 6605 2930 6.0 1695 3.5 1868 3.9 28 6 29 30 Non- 19. 4133 7.8* 2970 71.9 816 191 4.6 89 2.2 66 1.6 31 residents 7 32 33 Unclassified 221 0.4* 218 99.5 0 - 1 0.5 0 - 0 - 34 35 Population 2016 13. 61868 44203 71.4 8149 3576 5.8 3499 5.7 2404 3.8 36 2 37 http://bmjopen.bmj.com/ 38 Residents 56943 12. 39 92.1* 40623 71.4 7275 3361 5.9 3329 5.8 2322 4.1 40 ± 8 41 Non- 18. 42 4535 ø 7.3* 3266 72.0 842 206 4.5 143 3.2 76 1.7 43 residents 6 44 45 Unclassified 390 § 0.6* 314 80.9 32 8.2 9 2.3 27 7.0 6 1.5 on October 1, 2021 by guest. Protected copyright. 46 Demographic 12. 47 3078 2204 71.6 390 186 6.0 168 5.5 129 4.2 48 Events 2015-16 7 49 50 Births 13. 51 1687 54.8 1203 71.4 220 92 5.5 91 5.4 80 4.7 0 52 53 Immigration 54 721 23.4 521 72.3 60 8.3 51 7.1 60 8.3 29 4.0 55 s 56 Deaths 16. 57 670 21.8 480 71.6 110 43 6.4 17 2.5 20 3.0 58 4 59 60 35

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 50 BMJ Open

1 2 3 4 697 * Column % 5 698 ± 56,943 = 48,448 (censed in 2015) – 616 (deaths) + 1,670 (births) + 1,235 (immigrations) + 6,206 (censed in 2016) BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 699 ø 4,535 = 4,133– 42 (deaths) + 12 (births) + 64 (immigrations) + 368 (censed in 2016) 8 700 § 390 = 221 – 8 (deaths) + 5 (births) + 21 (immigrations) + 151 (censed in 2016) 9 10701 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 36

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 50

1 2 3 4 702 Table 2: Socio-demographic characteristics of heads of households, and household characteristics

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 703 (2015). 7 8 Characteristics of heads of households* N % 9 10 Sex Males 5051 54.6 11 Females 4206 45.4 12 13 < 18 40 0.4 14 18 – 24 520 5.6 15 Age group (years) 16 25 – 64 6919 74.7 17 > = 65 1777 19.2 18 For peer review only No formal education 5460 59.0 19 20 5-7th grade 2611 28.2 21 Education level 8 -12th grade 807 8.7 22 23 University 37 0.4 24 Missing Information 342 3.7 25 26 Single 4007 43.3 27 Married or de-facto union 3275 35.4 28 Marital status 29 Divorced 16 0.2 30 Separated 393 4.3 31 32 Widow 1554 16.8 33 Residents of Magude 9257 84.4 34 Households characteristics 35 36 Average household size** 5 [3-7] 37 http://bmjopen.bmj.com/ Lone-resident households 1171 10.7 38 39 Cane 3506 32.5 40 Cement 2807 26.0 41 42 Wall material of the main Mud bricks 2338 21.6 43 building Adobe 1703 15.8 44 45 Zinc plates 319 3.0 on October 1, 2021 by guest. Protected copyright. 46 Wood 128 1.2 47 48 Traditional latrine 5774 53.5 49 Improved latrine 1144 10.6 50 Type of toilet 51 WC connected to septic tank 233 2.2 52 No latrine 3650 33.8 53 Paraffin 5362 49.6 54 55 Electricity 3321 30.8 Main source of lighting energy 56 Candles 1074 9.9 57 58 Solar panels 502 4.6 59 60 37

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 50 BMJ Open

1 2 3 Other 539 5.0 4

5 Wood logs 6401 59.3 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Coal 740 6.9 7 8 Gas 46 0.4 Kitchen Fuel 9 Electricity 65 0.6 10 11 Paraffin 8 0.1 12 Missing Information 3541 32,8 13 14 Water pumps 3700 34.3 15 Directly from the river 2691 24.9 16 Primary source of drinking Piped water 2218 20.5 17 water 18 For Openpeer well near review the river only1138 10.5 19 Other 1054 9.8 20 21 Vector control tools 22 % of households with >1 ITN 8906 81.2 23 24 Universal ITN coverage (one 5690 53.2 25 net per two household 26 27 members) 28 Yes 5722 52.2 29 30 IRS in past 12 months No 4719 43.0 31 Unknown 524 4.8 32 33704 * or subhead (if head does not live in the household) 34 35705 **Median [Interquartile Range] 36 http://bmjopen.bmj.com/ 37706 38 39 40707 Table 3: Individual-level health and malaria prevention indicators in Magude district 2015). 41 42 Age group in Years < 5 5-14 >15 43 44 N % N % N % 45 on October 1, 2021 by guest. Protected copyright. Had fever in preceding 30 days 46 47 Yes 1043 12.1 1128 7.7 2931 11.6 48 7492 87.2 13325 91.2 21977 87.1 49 No 50 Unknown 55 0.6 154 1.1 319 1.3 51 52 Primary source of health care 53 8552 99.6 14548 99.6 25075 99.4 54 Health Facility 55 Traditional Healer 13 0.2 27 0.2 61 0.2 56 57 Community Health Worker 22 0.3 32 0.2 66 0.3 58 59 60 38

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 50

1 2 3 Slept under a bed net the preceding 4 5 night BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 2330 27.1 3179 21.8 6799 27.0 7 Yes 8 No 6260 72.9 11428 78.2 18428 73.0 9 10 11 12 13708 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 39

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 50 BMJ Open

1 2 3 4 709 Figure Titles and Legends

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 710 8 9 711 Fig 1: Map of Magude district (2015). 10 11 712 A) Administrative and permanent river shape files were obtained from DIVA-GIS ((http://www.diva- 12 13 14713 gis.org/Data), and confirmed with Magude’s key informants. GPS positions of households, health 15 16714 facilities and community health workers obtained directly from the field and mapped using QGIS. B) 17 18 For peer review only 19715 Population pyramid of Magude district (2015) showing the proportion of five—year age and sex 20 21716 groups out of the total population. 22 23717 24 25 26718 Fig 2: Health profile of Magude district prior to the baseline census (2010-2014). 27 28719 A) Most common diseases leading to all-age hospitalization and later discharge in Magude reported 29 30 31720 by the District Health Authorities between 2010-2014. B) Weekly number of outpatient malaria 32 33721 cases observed in Magude between 2010 and 2014 reported through the Weekly Epidemiological 34 35722 Bulletin (BES) and monthly rainfall data obtained from the Climate Hazards Group InfraRed 36 37 http://bmjopen.bmj.com/ 38723 Precipitation with Station data (CHIRPS). 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 40

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 50

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 Fig1 20 21 119x45mm (300 x 300 DPI) 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 43 of 50 BMJ Open

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 Fig2 19 169x56mm (300 x 300 DPI) 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 44 of 50

1 2 3 4 Supplementary Tables

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 S.Table 1: Selected demographic and socioeconomic characteristics of Magude’s population 8 9 (2015). 10 11 Variable / by sex Males Females 12 13 N Col % N Col % 14 Age groups (years) 21775 26673 15 16 < 1 826 3.8 898 3.4 17 1-4 3391 15.6 3475 13.0 18 For peer review only 19 5-14 7345 33.8 7262 27.2 20 15-64 9334 42.9 13205 49.5 21 22 >=65 864 4.0 1824 6.8 23 Education Level (>= 6 years old) 16712 21461 24 Illiterate 7925 47.4 11136 51.9 25 26 5-9 grade 6221 37.2 7616 35.5 27 10-12 grade 1595 9.5 1825 8.5 28 29 University 43 0.3 20 0.1 30 Missing Information 928 5.6 864 4.0 31 32 Occupation (> 18 years old) 5243 5532 33 Farmer or Fisherman 1371 26.1 3909 70.7 34 Salesperson 466 8.9 609 11.0 35 36 Construction work occupations 1159 22.1 39 0.7 http://bmjopen.bmj.com/ 37 Coal maker / Lumberjack 603 11.5 173 3.1 38 39 Guards/Police/Military 539 10.3 45 0.8 40 Health professional 85 1.6 133 2.4 41 42 Teacher 263 5.0 180 3.3 43 Miner 106 2.0 9 0.2 44 45 Other 651 12.4 435 7.9 on October 1, 2021 by guest. Protected copyright. 46 Marital Status (> 14 years old) 10157 14986 47 Single (never married) 6168 60.7 7772 51.9 48 49 Married or de-facto union 3690 36.3 4702 31.4 50 In a polygamic relationship 202 2.0 - - 51 52 Divorced 7 0.1 13 0.1 53 Separated 138 1.4 541 3.6 54 55 Widow 154 1.5 1958 13.1 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 45 of 50 BMJ Open

1 2 3 4 S. Table 2: Modified household poverty index (PI) based on the following deprivation indicators: 1)

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 lack of electricity, 2) lack or sharing of an improved sanitation facility; 3) lack of access to improved 7 drinking water source, or source only available more than 30-minute walk, round trip; 4) dirt, sand or 8 9 dung household floors; 5) dung, wood or charcoal used for cooking fuel; and 6) households that do not 10 own a car or truck AND do not own more than one of the following: radio, TV, telephone, bike, 11 12 motorbike, or refrigerator. 13 14 15 Deprivations Magude Motaze Panjane Mahele Mapulanguene District 16 17 0-2 32.1 9.8 1 0.9 13 25.2 18 3-4 For49.4 peer58.3 review55.1 61 only70.8 52.3 19 5-6 18.5 31.9 43.9 38.1 16.2 22.4 20 21 22 Median 3 4 4 4 3 4 23 Mean 3.1 3.9 4.2 4.2 3.5 3.4 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 46 of 50

1 2 3 S.Table 3: Number of individuals by age group who reported spending one night outside of 4

5 Magude the day before the census visit was performed (2015). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 Men Women All 8 Age Previous night spent 9 N (%) N (%) N (%)* 10 group outside of Magude 11 592 5.1 580 5.0 1172 43,0 12 Yes 13 * 14 0 - 14 No 10874 94.0 10970 94.3 15 NA± 96 0.8 85 0.7 16 17 324 6.3 362 5.6 686 25.1 18 Yes For peer review only * 19 15 – 29 20 No 4774 92.7 6022 93.3 21 NA± 52 1.0 70 1.1 22 23 298 11.3 197 5.3 495 18.1 Yes 24 * 25 30 – 44 26 No 2323 87.8 3497 93.9 27 NA± 24 0.9 30 0.8 28 29 133 8.6 149 4.9 282 10.3 Yes 30 * 31 45 – 64 32 No 1393 90.5 2839 93.8 33 NA± 13 0.8 39 1.3 34 35 Yes 36 4.2 57 3.1 93 3.4* 36 > 65 No 818 94.7 1744 95.6 37 http://bmjopen.bmj.com/ 38 NA± 10 1.2 23 1.3 39 Yes 1383 6.4 1345 5.0 2728 5.6 40 41 20182 92.7 25072 94,0 4525 93.5 All ages No 42 4 43 44 NA± 195 0.9 247 1.0 442 0.9 45 * Percentage of those who reported travelling (2,728) on October 1, 2021 by guest. Protected copyright. 46 47 ± Missing Information 48 49 50 51 52 53 S.Table 4: Household-level health and malaria prevention indicators in Magude district (2015). 54 55 Administrative Magude- 56 57 Motaze Panjane Mahele Mapulanguene 58 Post Sede 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 47 of 50 BMJ Open

1 2 3 Household size 4 [3-6] 4 [3-7] 4 [2-6] 4 [2-6] 3 [2-5.5] 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Members per 7 2.0 [1.4- 2.0 [1.6- 8 net 2.0 [1.3-3.0] 2.0 [1.5-4.0] 2.0 [1.3-4.0] 9 4.0] 5.5] 10 Median [IQR] 11 12 Universal ITN 13 52.7% 59.9% 44.1% 52.7% 52.1% 14 coverage * 15 16 IRS in past 12 N % N % N % N % N % 17 18 For peer review only 19 months 20 21 Yes 3925 49.0 1162 79.0 258 41.1 230 61.0 147 30.7 22 23 No 3627 45.3 293 19.9 352 56.1 133 35.3 314 65.6 24 25 Unknown 459 5.7 16 1.1 17 2.7 14 3.7 18 3.8 26 27 ** One net for every 2 members of the households 28 29 30 31 32 33 34 35 36 37 Supplementary Figures http://bmjopen.bmj.com/ 38 39 40 41 S.Figure 1: Household Assets in Magude district in 2015. 42 43 44 Percentage of households in Magude district with at least one item per asset, and median and 45 on October 1, 2021 by guest. Protected copyright. 46 interquartile range of the assets for which the number of items available was higher than 1. 47 48 49 50 S.Figure 2: Mobility patterns of Magude’s population in 2015. 51 52 Main destinations per age group among those who reported travelling outside of Magude in 53 54 55 2015. 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 48 of 50

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 49 of 50 BMJ Open

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 50 of 50

1 2 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 3 4 Item Check

5 No Recommendation BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title Yes 7 or the abstract Page 1 8 (b) Provide in the abstract an informative and balanced summary of Yes 9 10 what was done and what was found Page 3 11 Introduction 12 Background/rationale 2 Explain the scientific background and rationale for the investigation Yes 13 14 being reported Page 6 15 Objectives 3 State specific objectives, including any prespecified hypotheses Yes 16 Page 6 17 18 Methods For peer review only 19 Study design 4 Present key elements of study design early in the paper Yes 20 Page 7-8 21 Setting 5 Describe the setting, locations, and relevant dates, including periods of Yes 22 23 recruitment, exposure, follow-up, and data collection Page 7 24 Participants 6 (a) Give the eligibility criteria, and the sources and methods of Yes 25 selection of participants Page 8 26 27 Variables 7 Clearly define all outcomes, exposures, predictors, potential NA 28 confounders, and effect modifiers. Give diagnostic criteria, if Pages 8-9 29 applicable 30 Data sources/ 8* For each variable of interest, give sources of data and details of Yes 31 32 measurement methods of assessment (measurement). Describe comparability of Pages 8-9 33 assessment methods if there is more than one group 34 Bias 9 Describe any efforts to address potential sources of bias Yes 35 Page 4 36 37 Study size 10 Explain how the study size was arrived at NA http://bmjopen.bmj.com/ 38 Quantitative 11 Explain how quantitative variables were handled in the analyses. If Yes 39 variables applicable, describe which groupings were chosen and why Page 8 40 41 Statistical methods 12 (a) Describe all statistical methods, including those used to control for Yes 42 confounding Page 8 43 (b) Describe any methods used to examine subgroups and interactions NA 44 (c) Explain how missing data were addressed Page 8 45 on October 1, 2021 by guest. Protected copyright. 46 (d) If applicable, describe analytical methods taking account of Yes 47 sampling strategy Page 8 48 (e) Describe any sensitivity analyses NA 49 50 Results 51 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers Yes 52 potentially eligible, examined for eligibility, confirmed eligible, Page 9 53 included in the study, completing follow-up, and analysed 54 55 (b) Give reasons for non-participation at each stage Yes 56 Page 9 57 (c) Consider use of a flow diagram No 58 59 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, Yes 60 social) and information on exposures and potential confounders Page 9-10 (b) Indicate number of participants with missing data for each variable Yes

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 51 of 50 BMJ Open

1 2 of interest Table 2, 3 S.Table1, 4

5 S.Table3 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Outcome data 15* Report numbers of outcome events or summary measures Yes 7 Pages 9-12 8 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted NA 9 10 estimates and their precision (eg, 95% confidence interval). Make 11 clear which confounders were adjusted for and why they were 12 included 13 (b) Report category boundaries when continuous variables were Yes 14 15 categorized Table2, 16 S.Table1-3 17 (c) If relevant, consider translating estimates of relative risk into NA 18 For peer review only absolute risk for a meaningful time period 19 20 Other analyses 17 Report other analyses done—eg analyses of subgroups and NA 21 interactions, and sensitivity analyses 22 23 Discussion 24 Key results 18 Summarise key results with reference to study objectives Yes 25 Pages 12-15 26 Limitations 19 Discuss limitations of the study, taking into account sources of Yes 27 28 potential bias or imprecision. Discuss both direction and magnitude of Pages 12-15 29 any potential bias 30 Interpretation 20 Give a cautious overall interpretation of results considering objectives, Yes 31 limitations, multiplicity of analyses, results from similar studies, and Pages 12-15 32 33 other relevant evidence 34 Generalisability 21 Discuss the generalisability (external validity) of the study results Yes 35 Pages 12-15 36 37 Other information http://bmjopen.bmj.com/ 38 Funding 22 Give the source of funding and the role of the funders for the present Yes 39 study and, if applicable, for the original study on which the present Page 17 40 article is based 41 42 43 *Give information separately for exposed and unexposed groups. 44 45 on October 1, 2021 by guest. Protected copyright. 46 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 47 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 48 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 49 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 50 51 available at www.strobe-statement.org. 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from

Demographic and health community-based surveys to inform a malaria elimination project in Magude district, Southern Mozambique ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-033985.R2

Article Type: Original research

Date Submitted by the 12-Mar-2020 Author:

Complete List of Authors: Galatas, Beatriz; Instituto de Salud Global Barcelona, Malaria Elimination Initiative; Centro de Investigacao em Saude de Manhica, ISGLobal Nhacolo, Ariel; Centro de Investigacao em Saude de Manhica Marti, Helena; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Munguambe, Humberto; Centro de Investigacao em Saude de Manhica Jamise, Edgar; Centro de Investigaçao em Saúde de Manhiça Guinovart, Caterina; Barcelona Institute for Global Health Cirera, Laia; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Amone, Felimone; Centro de Investigacao em Saude de Manhica Macete, Eusebio; Centro de Investigação em Saúde de Manhiça Bassat, Quique; BARCELONA INSTITUTE FOR GLOBAL HEALTH http://bmjopen.bmj.com/ Rabinovich, Regina; Instituto de Salud Global Barcelona, Malaria Elimination Initiative Alonso, Pedro; WHO, GMP Aide, Pedro; Manhiça Health Research Centre Saute, Francisco; Centro de Investigação em Saúde de Manhiça Sacoor, Charfudin; Centro de Investigacao em Saude de Manhica, Department of Demography

Primary Subject on October 1, 2021 by guest. Protected copyright. Global health Heading:

Secondary Subject Heading: Epidemiology, Global health, Public health

Public health < INFECTIOUS DISEASES, Demography < TROPICAL Keywords: MEDICINE, Magude project, Malaria elimination, Population and health

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 39 BMJ Open

1 2 3 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 39

1 2 3 1 Demographic and health community-based surveys to inform a malaria elimination project in 4 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 5 2 Magude district, Southern Mozambique 6 7 3 8 9 4 Authors 10 11 5 Beatriz Galatas1, 2*¶, Ariel Nhacolo¶1, Helena Martí-Soler1,2, Humberto Munguambe1, Edgar Jamise1, 12 2 1,2 1 1,3 1,2,5,6 13 6 Caterina Guinovart , Laia Cirera , Felimone Amone , Eusébio Macete , Quique Bassat , N. Regina

14 2,7 1,2. 1,4 1 1 15 7 Rabinovich , Pedro L. Alonso , Pedro Aide , Francisco Saute , Charfudin Sacoor 16 8 17 18 9 Affiliations For peer review only 19 20 10 1. Centro de Investigação em Saúde da Manhiça (CISM), Manhiça, Mozambique; 21 22 11 2. Barcelona Institute for Global Health (ISGlobal), Hospital Clínic-Universitat de Barcelona, Barcelona, 23 24 12 Spain. 25 13 3. National Directorate of Health, Ministry of Health, Mozambique 26 27 14 4. National Institute of Health, Ministry of Health, Maputo, Mozambique 28 29 15 5. ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Spain. 30 31 16 6. Pediatric Infectious Diseases Unit, Pediatrics Department, Hospital Sant Joan de Déu (University of 32 33 17 Barcelona), Barcelona, Spain 34 18 7. Harvard T.H. Chan School of Public Health, Boston, MA, USA. 35 36 37 19 *Corresponding Author http://bmjopen.bmj.com/ 38 39 20 E-mail: [email protected] 40 41 21 ¶ These authors contributed equally to this work 42 43 22 44 45 23 Emails of Authors on October 1, 2021 by guest. Protected copyright. 46 24 [email protected] 47 48 25 [email protected] 49 50 26 [email protected] 51 52 27 [email protected] 53 54 28 [email protected] 55 29 [email protected] 56 57 30 [email protected] 58 59 31 [email protected] 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 39 BMJ Open

1 2 3 32 [email protected] 4 5 33 [email protected] BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 34 [email protected] 8 9 35 [email protected] 10 11 36 [email protected] 12 13 14 15 37 Key words 16 17 38 18 For peer review only 19 39 Magude district, census, demographic surveillance site, population, health, malaria control tools, malaria 20 40 elimination 21 22 41 23 24 25 26 42 Word count 27 28 43 5467 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 39

1 2 3 4 44 Abstract

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 45 (Word count: 302) 7 46 Objectives: A Demographic and Health Platform (DHP) was established in Magude in 2015, prior to the 8 9 47 deployment of a project aiming to evaluate the feasibility of malaria elimination in southern Mozambique, 10 11 48 named the Magude project. This platform aimed to inform the design, implementation and evaluation of 12 13 49 the Magude project, through the identification of households, and population; and the collection of 14 15 50 demographic, health and malaria information. 16 17 51 Setting: Magude is a rural district of Southern Mozambique which borders South Africa. It has nine 18 52 peripheral health facilitiesFor and one peer referral health review center with an onlyinpatient ward. 19 20 53 Intervention: A baseline census enumerated and geolocated all the households, and their resident and non- 21 22 54 resident members, collecting demographic and socio-economic information, and data on the coverage and 23 24 55 usage of malaria control tools. Inpatient and outpatient data during the five years (2010-2014) before the 25 26 56 survey were obtained from the district health authorities. The demographic platform was updated in 2016. 27 57 Results: The baseline census conducted in 2015 reported 48,448 (92.1%) residents and 4,133 (7.9%) non- 28 29 58 residents, and 10,965 households. Magude’s population is predominantly young, half of the population has 30 31 59 no formal education and the main economic activities are agriculture and fishing. Houses are mainly built 32 33 60 with traditional non-durable materials and have poor sanitation facilities. Between 2010 and 2014, malaria 34 35 61 was the most common cause of all-age inpatient discharges (representing 20-40% of all discharges), 36

62 followed by HIV (12-22%), and anemia (12-15%). In early 2015, all-age bed-net usage was between 21.8% http://bmjopen.bmj.com/ 37 38 63 and27.1% and the reported coverage of indoor residual spraying varied across the district between 30.7% 39 40 64 and 79%. 41 42 65 Conclusion: This study revealed that Magude has limited socio-economic conditions, poor access to health 43 44 66 care services and low coverage of malaria vector control interventions. Thus, Magude represented an area 45 on October 1, 2021 by guest. Protected copyright. 67 where it is most pressing to demonstrate the feasibility of malaria elimination. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 39 BMJ Open

1 2 3 4 68 Strengths and limitations of this study

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 69  Strengths: 7 70 o A Demographic and Health Platform (DHP) was established in Magude in 2015 to enumerate 8 9 71 and geolocate all the households, and their resident and non-resident members in order to 10 11 72 update the demographic information of the district which was last collected in 2007 during a 12 13 73 national census. 14 15 74 o A second census round was conducted one year later acknowledging that detailed population 16 17 75 data collected at individual level at one point in time may miss individuals or households that 18 76 can be capturedFor after apeer later update. review only 19 20 77 o The census rounds were planned in close collaboration with the community leaders and district 21 22 78 authorities, and counted with intensified training of fieldworkers, and a strong component of 23 24 79 field and data supervision by experienced demographers. 25  26 80 Limitations: 27 81 o Data collected through the DHP could have been affected by inaccuracies during data collection 28 29 82 or entry, or by recall or desirability bias of census participants. 30 31 83 o Inpatient and malaria outpatient information was limited by the quality and accuracy of the 32 33 84 data at the time it was collected, by disease-specific interventions or changes in diagnostics, 34 35 85 referral or reporting practices. 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 39

1 2 3 4 86 Background

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 87 Mozambique is one of the countries with the highest malaria burden in the world [1]. Malaria prevalence 7 88 among children of 6 to 59 months of age is heterogeneous within the country ranging from high 8 9 89 transmission intensity in the North (>50%) to less than 3% in the South, as reported in the latest malaria 10 11 90 survey conducted in 2015 [2]. Lower prevalence in the Southern region may be related to the great progress 12 13 91 against malaria in the past decades, partly as a result of the regional initiatives aiming for malaria 14 15 92 elimination in the area. In line with the vision of a malaria free world established by the World Health 16 17 93 Organization in its Global Technical Strategy for 2016-2030 [3], the National Malaria Control Program of 18 94 Mozambique (NMCP) decidedFor to redefinepeer its strategic review objectives inonly order to include the implementation of 19 20 95 malaria elimination activities in the South. In this context, a malaria elimination project named the Magude 21 22 96 Project was designed and evaluated in Magude district, Maputo province, by the Manhiça Health Research 23 24 97 Center (CISM) and the Barcelona Institute of Global Health (ISGlobal), to assist the NMCP in adopting a 25 26 98 malaria elimination strategy based on local evidence [4]. 27 99 28 29100 Prior to the initiation of the Magude project in 2015, there was limited and outdated information with 30 31101 regards to the number of individuals living in Magude, as well as to their demographic and socio-economic 32 33102 characteristics [5]. Thus, detailed information from the whole district was deemed crucial to inform the 34 35103 elimination strategies that had been planned for the following years in the district. The process of filling in 36

104 this knowledge gap also aimed to identify and contact key leaders at provincial, district, and at community http://bmjopen.bmj.com/ 37 38105 level, to inform and engage them in the activities prior to their deployment. In this context, a Demographic 39 40106 and Health Platform adapted from the Health and Demographic Surveillance System (HDSS) method was 41 42107 established in the district of Magude in February of 2015 with the objective of providing reliable and 43 44108 updated demographic data to inform the project. This platform allowed to plan the activities and to provide 45 on October 1, 2021 by guest. Protected copyright. 109 a sampling frame to measure indicators in the community such as malaria prevalence, and to estimate the 46 47110 coverage of indoor residual spraying (IRS), coverage and usage of long-lasting insecticide treated nets 48 49111 (LLIN) and mass drug administration (MDA) campaigns. The DHP’s permanent identification numbers 50 51112 were used to track individual’s participation in each intervention of the Magude project longitudinally, 52 53113 thus allowing to identify and quantify potential challenges to the project, such as reasons for non- 54 55114 participation, or probable sources of imported infections. Data were also used to accurately measure 56115 prevalence at the community stratified by age groups and place of residence. Overall, these findings were 57 58 59 60 5

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 39 BMJ Open

1 2 3 116 crucial to the design of the Magude project, and offered robust evidence to guide malaria elimination 4 5 117 strategies in Southern Mozambique. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 118 8 9 119 This article presents the demographic, socio-economic and health characteristics of the population of 10 11120 Magude, as well as the coverage of malaria control interventions estimated through the baseline census 12121 conducted between February and June of 2015. It also presents the burden of disease in Magude during the 13 14122 five years (2010-2014) before the DHP, using the inpatient and outpatient data obtained from the district 15 16123 health authorities. A summary of the demographic profile of Magude after updating the census between 17 18124 August and September ofFor 2016 is alsopeer provided review only 19 20125 21 22 23 24126 Methods 25 26 27127 Study area 28 29128 The district of Magude is located in the North-Western part of Maputo Province and borders with the 30129 districts of Massingir, Chókwe and Bilene, from Gaza Province on the North and North-East; with the 31 32130 districts of Manhiça and Moamba, of Maputo Province in the East and South; and with the South African 33 34131 National Kruger Park, in the West (Figure 1A). 35 36132 37 http://bmjopen.bmj.com/ 38133 Magude was selected as an appropriate demonstration area for the malaria elimination project as it was 39 134 expected to pose the types of challenges that the NMCP would face when implementing a malaria 40 41135 elimination campaign in the south of the country. This is, there were more than 13,000 malaria cases 42 43136 reported in the district in 2014, with the majority of cases observed between January and May, suggesting 44 45137 that the epidemiology of malaria in the district was representative of most endemic areas in the country on October 1, 2021 by guest. Protected copyright. 46 47138 with the typical seasonal pattern coinciding with the rainy season [6,7]. Additionally, the socio-economic 48 139 and infrastructural limitations reported for Magude made it sufficiently representative of a rural district of 49 50140 Mozambique, while still at reach of CISM’s facilities (located in Manhiça district), which facilitated the 51 52141 logistics, supervision and quality control processes. Finally, the population of Magude had had very limited 53 54142 exposure to research projects or targeted innovative malaria interventions prior to 2015, having only 55 56143 received the programmatic IRS, LLINs and child immunization campaigns conducted by the government. 57 58144 This allowed facing the challenges of working in an unexposed population that was not biased by previous 59145 activities. 60 6

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 39

1 2 3 146 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 147 Study design 7 8 148 A Demographic and Health Platform (DHP) was established in Magude in February 2015 by CISM, which 9 10149 was adapted from the health and demographic surveillance system (HDSS) previously established in 11 12150 Manhiça district also by CISM [8]. A baseline census was conducted at that time (February to June) to 13151 identify and enumerate all neighborhoods, households and resident and non-resident individuals living in 14 15152 these households. Geographic positioning system (GPS) coordinates were also captured for every 16 17153 household.All individuals were assigned a permanent and unique identification number, linked to the 18 For peer review only 19154 household number where they were first enumerated. A second census round was conducted between 20 21155 August and September of 2016, to review and update all information collected during the baseline census 22 156 and record live births, deaths, and migrations that occurred between the two censuses. New households 23 24157 and new members were enumerated, and their information was collected to update the baseline databases. 25 26158 27 28159 This DHP defined a household as a structure or set of constructions where an individual or group of 29 30160 individuals live and share domestic activities and costs (such as eating and sleeping) and recognize one of 31 32161 them as their superior or chief, regardless of their kinship ties. Individuals were defined as residents if they 33162 had lived and slept in a household within the study area for a period of three months or more or intended 34 35163 to do so. Non-resident members of a household were defined as those who left the district or who had never 36 37164 lived in the study area for a period of three or more months, but who kept their roles in the households as http://bmjopen.bmj.com/ 38 39165 head of household, husband, or other important breadwinners, and paid regular visits to their households 40 41166 in Magude. These individuals leave their households for specific reasons such work, studies, or 42167 imprisonment; and would otherwise reside in the referred household. This category excluded offspring or 43 44168 other members who had left the household to live in their own households outside the study area, 45 on October 1, 2021 by guest. Protected copyright. 46169 irrespective of whether they kept visiting the reference household or not. 47 48 49170 Data collection procedures 50 51171 Baseline and updated data from the census rounds were collected through standardized questionnaires 52 172 using Open Data Kit (https://opendatakit.org/) installed in Android tablets. Data were sent to a secure server 53 54173 at CISM using Wi-Fi. Information collected on households’ socio-economic characteristics included 55 56174 building materials used for the main house, source of water and electricity, and household assets and 57 58175 livestock. At the individual level, the information collected included the name, sex, date of birth, relation 59 60176 to the head of household, education, and occupation of all residents and non-residents of Magude. 7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 39 BMJ Open

1 2 3 177 Information was also collected on malaria prevention tools, including the possession and use of LLINs, and 4 5 178 whether the house had received IRS in the 12 months preceding the census. Individuals were also asked BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 179 about history of fever during the preceding 30 days, and whether they had sought care at a health facility, 8 9 180 community health worker, traditional healer, or none. The specific health facility where the individual 10 11181 sought care was also specified to delineate health facility catchment areas (HFCA) according to the 12182 community. Finally, information was collected on the migration and mobility patterns of all participants, 13 14183 to better characterize the mobility profile of the district, and offer potential information on the sources of 15 16184 new infections if transmission was eventually significantly reduced. 17 18185 For peer review only 19 20186 Administrative and geographical information was obtained directly from the district authorities and from 21 187 the most recent district profile of Magude published in 2014 by the central government authority [7] and 22 23188 district statistics performed by the National Institute of Statistics [9,10]. We retrospectively collected 24 25189 monthly data on inpatient discharges from the referral health center of Magude Sede, as well as weekly 26 27190 outpatient malaria cases reported through the weekly epidemiological bulletin (Boletim Epidemiológico 28 29191 Semanal or “BES” in Portuguese) for the period of 2010-2014. The BES did not distinguish between 30 31192 presumed and confirmed cases, the diagnostic used (rapid diagnostic test or microscopy) or the location of 32193 detection (health facility or in the community). 33 34 35194 Patient and Public Involvement 36 37195 This study counted with the support and involvement of the community of Magude in all of its stages. http://bmjopen.bmj.com/ 38 39196 During its design and planning, meetings were held with the district authorities and community leaders 40 41197 in order to inform them of the purpose of the DHP under the scope of the Magude project, as well as to 42198 landscape the number of villages in Magude that had to be covered by the DHP. The study team worked 43 44199 closely with village chiefs when performing the household visits, which ensured that all district 45 on October 1, 2021 by guest. Protected copyright. 46200 households were covered by the DHP. Reports were submitted to the administrative authorities to 47 48201 support vector control activities, among others. Preliminary findings were also communicated in 49 50202 meetings held with the community prior to the deployment of MDAs, where the findings obtained from 51 203 the study – particularly focusing on malaria – where discussed in detail [11]. 52 53 54204 Data analysis 55 56205 Data management and descriptive analyses were performed using R Software [12]. The percentage 57 58206 distribution of categories within any given variable were calculated considering the observations with 59 60207 missing values in the denominator. No inferential statistics (p-values or 95% confidence intervals) were 8

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 39

1 2 3 208 calculated as the DHP covers the entire district, and therefore offers direct population parameters. The 4 5 209 relationship between outpatient malaria cases and rainfall was evaluated using linear regressions for BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 210 different rainfall lags. Spatial visualization of the data collected on the field and obtained from web-based 8 9 211 open source administrative data bases from DIVA-GIS [13] were performed using QGIS Software [14]. 10 11212 12213 Household’s socio-economic status was measured using a modified version of Oxford’s Poverty and 13 14214 Human Development Initiative Multidimensional Poverty Index (MPI) [15], following the same 15 16215 methodology used by the INDEPTH group to estimate the poverty index of various HDSS sites 17 18216 throughout Africa [16]. InFor summary, peer this method review categorizes households only as “deprived” or “not deprived” 19 20217 based on six deprivation indicators: 1) lack of electricity, 2) lack or sharing of an improved sanitation 21 218 facility; 3) lack of access to improved drinking water source, or source only available more than 30- 22 23219 minute walk, round trip; 4) sand floors in main houses; 5) dung, wood or charcoal used for cooking fuel; 24 25220 and 6) households that do not own a car nor truck and do not own more than one of the following: radio, 26 27221 TV, telephone, bike, motorbike, or refrigerator. A deprivation indicator is generated as lowest 28 29222 depravation (0-2), moderate depravation (3-4), or highest depravation (5-6) [16]. 30 31 32 33223 Results 34 35 36224 The geography of Magude district 37 http://bmjopen.bmj.com/ 2 38225 The district of Magude has an area of 6,961 km and is divided in five administrative posts, namely: Magude- 39 226 sede, Motaze, Panjane, Mahele and Mapulanguene (Figure 1A). The vegetation of Magude is dominated by 40 41227 open forests and savannahs hosting animals such as impalas, warthogs, lions, buffalos and elephants. There 42 43228 is one permanent river (Incomati), which flows through the south-western region of the district and 44 45229 constitutes the main source of water in the area, and three intermittent rivers dependent on rainfall, called on October 1, 2021 by guest. Protected copyright. 46 47230 Massitonto, Uanétse and Mazimuchopes [7]. 48 49231 Demographic and socio-economic characteristics 50 51232 The baseline census of 2015 registered 10,965 households and 52,802 individuals, of whom 48,448 (91.8%) 52 53233 were residents and 4,133 (7.8%) were non-residents. The district had a population density of 6.9 inhabitants 54 55234 per km2, which varies by administrative post. The majority of the population (73%) lives in the Magude- 56 57235 Sede administrative post (the capital of the district); 14.1% in Motaze, 5.9% in Panjane, 3.7% in 58 59 60 9

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 39 BMJ Open

1 2 3 236 Mapulanguene, and 3.4% in Mahele (Table 1). The age and sex structure of the population is dominated by 4 5 237 children and young people, and a sharp reduction of adults, particularly among males (Figure 1B). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 238 8 9 239 The census update conducted between August and October of 2016 registered a population of 61,868 10 11240 individuals. Of the population censed in 2015, 42,792 (81%) were found during the update round, 6,099 12241 (11.5%) were reported to still be in Magude by a family member although field workers did not find them 13 14242 in other households and therefore were not able to confirm and complete the registration of migration. 15 16243 There were an additional 3,244 (6.1%) individuals censed in 2015 who were not found in 2016 and for 17 18244 whom there were no informants,For peerbut were still review considered to be onlyin Magude for this analysis. This update 19 20245 also recorded the death of 670 individuals censed in 2015, 1,687 live births and 721 immigrations since the 21 246 baseline census. An additional 7,325 individuals who were missed in 2015 were censed during this update 22 23247 (Table 1). 24 25248 26 27249 The baseline census indicated that 62% of males and 63.1% of females above the age of 14 reported being 28 29250 married or in de-facto union; and 3% of married men practice polygamy. The prevalence of lack of formal 30 31251 education among those aged 6 and older in Magude is 51.9% among females and 47.4% among males. 32252 Approximately 37.2% of individuals reached 5th to 9th grade, and 9.5% have completed between 10th and 33 34253 12th grade (Sup.Table 1). In 2015 Magude had 31 primary schools offering 1st to 5th grades; 33 offering 1st- 35 36254 7th grades; one secondary school offering 8th-12th grades; and one private higher-education training center. 37 http://bmjopen.bmj.com/ 38255 According to the national census performed in 2007, Xichangana is the main local language of the district 39 40256 and is the mother tongue of 92% of the population of Magude. However, 51% of the population also reports 41257 being able to speak Portuguese (the official language in Mozambique), which is the mother tongue of only 42 43258 3.2% in the district [7]. 44 45259 on October 1, 2021 by guest. Protected copyright. 46 47260 Occupations are unevenly distributed between males and females older than 18 years of age. The majority 48 49261 of the population (26.1% of males and 70.7% of females) relies on subsistence agriculture, fishing or work 50 262 as cane cutters in the sugar plantations within Magude, or in Xinavane, in the nearby district of Manhiça. 51 52263 Other occupations include being a salesperson (8.9% of men and 11% of women), doing construction work 53 54264 (22.1% of men, 0.7% of women), or making coal (11.5% of men and 3.1% of women). A small proportion 55 56265 of the population works as security guards, or work in the public sector, particularly as health professionals, 57 58266 teachers or in the army (Sup.Table 1). 59 267 60 10

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 39

1 2 3 268 Men represent 54.6% of heads of households in Magude. Only 0.4% of household heads are younger than 4 5 269 18 years of age and 5.6% are between the ages of 18-24, while 74.7% are between 25-64 years old, and BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 270 19.2% are older than 65 years of age (Table 2). Forty-three percent of heads of household are single, 35.4% 8 9 271 are married or in de-facto union, 16.8% are widows and 4.3% separated. Fifty-nine percent of heads of 10 th 11272 households reported having no formal education, while 28.2% reported completing up to 5-9 grade, 8.7% 12273 up to 10th to 12th grade, and 0.4% reported having a university degree (Table 2). 13 14 15274 Living conditions in Magude’s households 16 17275 Settlements in Magude are made of individual household compounds that are built in proximity to each 18 For peer review only 19276 other in the central areas of each administrative post, and more spread out in the rest of the district. The 20 21277 median household size is 5, although 1,171 households (10.7%) have only one resident, 40% of whom are 22 278 older than 65 years old. 23 24279 25 26280 The majority of households are traditional round-shaped or rectangular-shaped huts constructed using cane 27 28281 (32.5%), cement (26%), mud bricks (21.6%), or reeds covered by adobe (15.6%) (Table 2). More than half 29 30282 of the households have traditional latrines (53.5%), 10.6% have improved latrines, while 33.8% of 31 32283 households do not have any form of sanitation facility. Flush toilets can only be found in 2.2% of the 33284 households. The primary lighting sources used in the households are paraffin (49.6%), electricity (30.8%), 34 35285 candles (9.9%) and solar panels (4.6%). The majority of households cook using wood logs (59.3%). Water 36 37286 is mainly obtained from pumps (34.3%) or piped water (20.5%), although a quarter of the population http://bmjopen.bmj.com/ 38 39287 collects water directly from the river (24.9%) or from open wells near the river (10.5%) (Table 2). 40 41288 42289 With regards to household assets, 68.5% of households own at least one mobile phone, 36.6% own a radio, 43 44290 32.7% have a television (TV), while 16.8% and 12.2% of households have a fridge and a freezer, 45 on October 1, 2021 by guest. Protected copyright. 46291 respectively. Almost a quarter of households reported having bicycles (23.7%), while 10.8% reported 47 48292 owning cars, 5.3% motorbikes, and 0.7% trucks (Sup.Figure 1). 49 50293 51 294 Households in Magude had a median number of 4 depravations. Twenty-five percent of the households 52 53295 were in the lowest depravation category, with 0-2 deprivation indicators; while 55.3% fell in the 3-4 54 55296 depravation group and 22.4% in the 5-6 depravation group. This distribution was unevenly observed among 56 57297 the different administrative posts. Magude Sede had the highest proportion of households in the lowest 58 59298 depravation group (32%), while Mapulanguene had the highest proportion of households in the moderate 60 11

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 39 BMJ Open

1 2 3 299 depravation category (70.8%). Panjane and Mahele were the administrative posts with the highest 4 5 300 proportion of highly deprived households (Sup. Table 2). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 301 8 9 10302 Mobility patterns 11 12303 The baseline census showed that 5% of residents reported having spent the night before the census outside 13304 of Magude (6.4% of males and 5% of females). Of these, 43% were younger than 14 years old, 25.1% were 14 15305 15-29, 18.1% were 30-44, 10.3% were 45-64 and 3.4% were older than 65 years old (Sup.Table 3). 16 17306 18 For peer review only 19307 Almost half of the population (43% of males and 46.9% of females) did not report frequently travelling 20 308 outside of Magude during the baseline census. Among the 54.4% who did, the places most frequently visited 21 22309 included Maputo City (55.3% of males, 58.3% of females), South Africa (23.1% of males and 16.8% of 23 24310 females), Gaza Province (12.2% of males and 15.9% of females), Inhambane Province (4.3% males and 2.6% 25 26311 females), and other Northern provinces within Mozambique (5.2% of males and 6.3% of females). Younger 27 28312 individuals (less than 15 years old) reported travelling the most to all provinces within the country, 29 313 followed by those between the ages of 15-30; while 15 to 45-year-olds are more likely to travel to South 30 31314 Africa. Individuals older than 45 reported travelling more to provinces northern of Maputo province, as 32 33315 well as abroad (Sup.Figure 2). 34 35 36316 Health and malaria information 37 http://bmjopen.bmj.com/ 38317 The district has 10 functional health facilities– one with an inpatient and maternity ward (Type I) and nine 39 40318 HFs only with maternity wards (Type II). Eight of the ten HFs were active in 2015, and 2 more Type II HFs 41319 were added in 2016 and 2017. The referral HF, located in the Magude-Sede administrative post, has 57 42 43320 beds and three active medical doctors (0.06 doctors per 1,000 inhabitants). The other 9 Type II health 44 45321 facilities have 37 beds in total. The whole district has only three medical doctors, 15 general nurses and 19 on October 1, 2021 by guest. Protected copyright. 46 47322 maternal and child health nurses (Figure 1A). Only 3 HFs have access to piped water, 4 have access to public 48 49323 network electricity, and 5 rely only on solar panels. All health facilities are located on a local main road to 50324 facilitate access, and the median Euclidian distance from households to the nearest health facility is 2.7 km 51 52325 (interquartile range [IQR] 1.4–7.9 km), although households are as close as 15 m or as far as 38.8 km. There 53 54326 is only one ambulance in Magude, which is used for transferring patients within the district, but also to the 55 56327 hospitals in Xinavane, Manhiça, and Maputo. 57 58328 59 60 12

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 39

1 2 3 329 The health system in Mozambique includes Community Health Workers (or APEs, from its acronym in 4 5 330 Portuguese) who are trained by the Ministry of Health to offer primary health services in areas with poor BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 331 access to HFs. They provide diagnosis and treatment for malaria, diarrhea, and pneumonia, and to refer 8 9 332 patients with signs of sickness requiring higher medical attention [17]. In Magude there are 27 APEs 10 11333 distributed throughout the district (Figure 1A). The average distance between households and the nearest 12334 APE is approximately 6.3 km (median of 4.4 km, IQR 2.8–5.4 km). 13 14335 15 16336 All HFs and APEs in Magude are equipped to diagnose malaria through Rapid Diagnostic Tests (RDTs) and 17 18337 light microscopy is onlyFor available peerin the Magude review Sede Type I HF. only They also offer treatment to all positive 19 20338 cases with Artemether-Lumefantrine, the first line treatment in Mozambique. Intermittent Preventive 21 339 Treatment in pregnancy (IPTp) using Sulfadoxine-pyrimethamine is also offered in all HFs of the district. 22 23340 In May of 2014 the NMCP conducted an LLIN universal distribution campaign, with 35,432 bed nets 24 25341 distributed in Magude district. This was followed by a focal IRS campaign between October and December 26 27342 of 2014 using the insecticides deltamethrine and DDT, which was only deployed in the Motaze 28 29343 administrative post, where the malaria burden was highest. 30 31344 32345 Malaria has traditionally been the first cause of disease in the district, responsible for approximately 53% 33 34346 of all consultations reported in 2003 [7]. According to the inpatient discharge data available from January 35 36347 of 2010 to December of 2014, between 20-40% of all-age discharges in the inpatient department were due 37 http://bmjopen.bmj.com/ 38348 to malaria, followed by HIV (12-22%), anemia (12-15%), pneumonia (6-12%), and diarrhea (3-8%), varying 39 40349 by month and year (Fig 2A). The weekly number of malaria cases reported through the BES in Magude 41350 between 2010 and 2014 follows a seasonal pattern with a peak between December and May and a reduction 42 43351 of cases during the dry season (Fig 2B). According to BES data, there were 293 cases per 1000 population 44 45352 reported during the transmission year of July 2011 to June 2012, 247 per 1000 between July 2012 and June on October 1, 2021 by guest. Protected copyright. 46 47353 2013, 252 cases per 1000 between July 2013 and June 2014, and 110 cases per 1,000 between July 2014 and 48 49354 June 2015. [7]. There is a fairly linear association between the rainfall in the preceding 2 months, and the 50 355 number of malaria cases in Magude (linear regression coefficient=13.6, Rho=0.64, p-value<0.001). 51 52356 53 54357 The prevalence of history of fever during the 30 days prior to the baseline census was 12.1% in children 55 56358 under the age of 5, 7.7% among 5-14 year-olds, and 11.6% in individuals older than 15 years. Almost all 57 58359 individuals reported going to a health facility as their primary source of health care (>99%), while a small 59 360 proportion mentioned seeking care first from traditional healers or community health workers. The 60 13

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 39 BMJ Open

1 2 3 361 proportion of individuals who reported sleeping under a bed net the night before was 27.1% among 4 5 362 children under the age of 5, 21.8% in 5-14 year-olds, and 27% among those > 15 years of age (Table 3). The BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 363 universal bed-net coverage (i.e. one net for every two individuals of a household) in 2015 was 52.7% in the 8 9 364 administrative post of Magude-Sede, 59.9% in Motaze, 44.1% in Panjane, 52.7% in Mahele and 52.1% in 10 11365 Mapulanguene. Finally, the reported coverage of IRS during the previous 12 months was 49% in Magude- 12366 Sede, 79% in Motaze, 41.1% in Panjane, 61% in Mahele, and 30.7% in Mapulanguene (Table 2 and 13 14367 Sup.Table 4). 15 16 17 18368 Discussion For peer review only 19 20369 Overall, the Demographic and Health Platform established in Magude offered detailed insight on the 21 22370 baseline socio-demographic profile of Magude district prior to the Magude project. It revealed that the 23 24371 number of residents in Magude in 2015 (52,804 individuals and 10,965 households) was similar to the 25 26372 number reported by the 2007 national census conducted by the National Institute of Statistics (INE) (53,229 27 28373 individuals and 11,408 households) [9] but did not coincide with the projected population for Magude in 29374 2015 estimated by the INE to be 62,000 [10], which is not surprising acknowledging the limitations of 30 31375 population projections. The census update conducted in 2016 identified individuals who had been missed 32 33376 by the baseline census, and recorded the births, deaths, and in-migrations since baseline. As a result, the 34 35377 population of Magude in 2016 assuming that those censed in 2015 who were not found in 2016 were still 36 37378 living in the district, was 61,868 individuals, which is similar to the projection by INE for this year (62,924). http://bmjopen.bmj.com/ 38 379 39 40380 The usefulness of a demographic and health surveillance system in demographic and bio-medical research 41 42381 in African countries has been previously described [18]. Having a precise source of population data is crucial 43 44382 for the correct estimation of the risk of disease and coverage of health indicators in a population, as well as 45 on October 1, 2021 by guest. Protected copyright. 46383 for planning purposes of public health interventions [19]. This platform has shown that detailed population 47 384 data collected at individual level at one point in time still misses a number of individuals, who can be 48 49385 captured after a later update. However, all the census rounds faced the challenge of finding individuals at 50 51386 home, or informants of unavailable individuals, to identify whether a member lives in the area despite not 52 53387 being found during a specific census round. To identify or collect information about absent individuals, 54 55388 field supervisors attempt to arrange interviews with them at their places of work, or interview their family 56 57389 members at their households during weekends. The same phenomenon is expected to take place when 58 59 60 14

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 39

1 2 3 390 implementing malaria interventions in the community (such as rounds of IRS or of mass drug 4 5 391 administrations), which generally challenges the estimation of intervention coverage. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 392 8 9 393 Household-level data indicates that the households in Magude are typical of a rural area with suboptimal 10 11394 conditions regarding access to water, electricity, and sanitation. The majority of houses were relatively 12395 homogeneous, built with non-durable materials, such as cane and adobe, and did not have access to clean 13 14396 water, conditions indicative of a low socio-economic status [20]. The age and sex composition of Magude’s 15 16397 population is typical of rural Mozambique and sub-Saharan Africa, composed primarily by children and 17 18398 young individuals and a Fordecreasing peer proportion reviewof adults as age increases only [21]. Another important aspect of 19 20399 the structure of the population of Magude is the lower sex ratio (number of males in relation to that of 21 400 females) among young and middle-aged adults, which has been reported, in the neighbouring district of 22 23401 Manhiça, to be related to male labour migration to Maputo city or South Africa and higher male mortality 24 25402 [21]. 26 27403 28 29404 A large proportion of the heads of households and of the overall population of Magude reported not having 30 31405 received a formal education, a major risk factor for health outcomes. Additionally, most individuals 32406 reported having occupations related to agriculture or fishing, which are usually carried out early in the 33 34407 morning until around noon. These aspects should be taken into consideration when planning mobilization 35 36408 campaigns and community-based interventions, as visits might have to be conducted at certain times of day 37 http://bmjopen.bmj.com/ 38409 and/or venues outside the households to be able to find these individuals. 39 40410 41411 Magude is also subject to significant mobility and migration, as more than half of the population reported 42 43412 travelling frequently outside of the district. The destinations most reported by travelers living in Magude 44 45413 were Maputo and South Africa, followed by Gaza and Inhambane provinces. This information is useful to on October 1, 2021 by guest. Protected copyright. 46 47414 evaluate the risk of malaria importation from other areas due to travel or migration [22]. Maputo city and 48 49415 South Africa are areas with lower malaria burden than Magude [23]; however, the district of Manhiça 50 416 surrounding Magude, the provinces of Gaza, Inhambane and the rest of Mozambique have higher malaria 51 52417 prevalence estimates than Magude district [2], and could be a source of imported infections [24]. These 53 54418 places are commonly visited by all age groups, whereas the lower-endemic areas are mainly visited by <30 55 56419 year-olds. This age pattern of migration is typical of the rural areas of Southern Mozambique, where 57 58420 migration rates are higher among individuals aged 20 to 45 years and their children, and decrease with 59 421 increasing age when individuals establish as permanent residents in an area. 60 15

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 39 BMJ Open

1 2 3 422 4 5 423 The population of Magude has access to ten health facilities, one of them with an inpatient ward with 57 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 424 beds. Thus there are 11.8 beds per 10,000 population, a ratio that is higher than the national level (7 beds 8 9 425 per 10,000 in Mozambique) [25]. While most respondents indicated seeking health care primarily from 10 11426 formal health facilities, the majority of the population is scattered throughout the district as far as 38 kms 12427 away from a HF and it is difficult to access health facilities or APEs due to limited roads and lack of 13 14428 transportation. This has implications for the delivery of public health interventions, especially for those 15 16429 that rely on the passive detection of cases to control transmission in the community. 17 18430 For peer review only 19 20431 Data between 2010 and 2014 from the outpatient and inpatient department of the Magude Sede health 21 432 center indicated that malaria was the main cause of hospitalization and later discharge with a clear seasonal 22 23433 pattern. The population of Magude is also burdened by HIV, anemia, pneumonia, and diarrhea, similar to 24 25434 its neighboring districts of Manhiça and Chokwé [26–33]. Inpatient information was retrospectively 26 27435 collected, and thus, its quality and accuracy depends on the data at the time at which it was collected. It 28 29436 may also be biased by disease-specific interventions or changes in diagnostics, referral protocols of severe 30 31437 cases to other districts, or reporting practices. Additionally, cases reported by BES are difficult to interpret 32438 given their lack of disaggregation in presumed or confirmed, diagnostic tool used or place of detection of 33 34439 the case. Thus, they are subject to changes in RDT use and stock-outs, in care seeking behavior to HFs or 35 36440 APEs, or reporting inaccuracies. In fact, cases reported by the APEs are usually excluded by those who 37 http://bmjopen.bmj.com/ 38441 report in BES, and neither the total number of fevers nor the number of individuals tested are reported. 39 40442 Thus, a stronger surveillance system than BES was thought to be required, and consequently established in 41443 Magude in 2015, to fully capture the clinical malaria profile of a district aiming to eliminate malaria [4,34]. 42 43444 44 45445 The age-specific self-reported history of fever in the past 30 days (7.7%-12.1%) were slightly lower than on October 1, 2021 by guest. Protected copyright. 46 47446 the average percentage of fevers reported in the previous 2 weeks in Maputo Province, which was 15% 48 49447 according to the most recent malaria indicator survey [2]. Approximately half of the households in Magude 50 448 reported owning one bed net for every two individuals (universal coverage). However, only a fourth of the 51 52449 population reported sleeping under a bed net the previous night. These findings mirror the estimates 53 54450 reported for Maputo Province in 2011 by the Demographic Health Survey, which found a reported bed net 55 56451 usage rate of 27% in Maputo Province [35]. This information is indicative of an area where bed net owners 57 58452 do not necessarily use the net as a preventive tool against malaria during the rainy season, despite this being 59 453 a prevalent disease in the district. Other demographic and socio-economic risk factors that have been 60 16

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 39

1 2 3 454 associated with LLIN use, such as households with more children, household heads with no formal 4 5 455 education, or households far away from the health facilities, might explain the low bed net usage in Magude BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 456 [36–38]. 8 9 457 10 11458 The reported IRS coverage in the area of Motaze was higher, which corresponds with the focal IRS 12459 campaign conducted by the NMCP that took place only there in September of 2014. The coverage reported 13 14460 in other areas was probably the consequence of recall bias, as a district-wide IRS campaign took place in 15 16461 2013. Overall, the coverage reported for LLINs and IRS throughout the district was below the World Health 17 18462 Organization recommendedFor coverage peer of >80% review [39], leaving a significantonly proportion of the population 19 20463 unprotected by any of the standard preventative measures. The identification of gaps in such vector control 21 464 coverages called for a strong community engagement campaign focused on the use of the bed nets and 22 23465 participation in the yearly rounds of IRS. 24 25 26 27466 Conclusion 28 29467 Through the establishment of a Demographic and Health Platform in Magude in 2015, which was updated 30 31468 in 2016, and a baseline assessment of the relevant health indicators, it was possible to fully characterize a 32 33469 district where malaria elimination interventions were to be deployed and evaluated. Magude represents a 34 35470 typical rural district of Mozambique characterized by limited social and economic infrastructures, which 36 37471 had to be considered for the design and operationalization of the community-based interventions. This http://bmjopen.bmj.com/ 38 472 study also revealed a low education level among a large proportion of the population and identified 39 40473 agriculture as the main economic activity in the district. These socio-demographic characteristics suggest 41 42474 that a strong community engagement would be necessary for the implementation of a malaria elimination 43 44475 project. Half of the population of Magude reported travelling outside of the district to areas of high and low 45 on October 1, 2021 by guest. Protected copyright. 46476 malaria transmission intensity, showing that malaria importation will likely be a source of continuous 47 477 transmission even if interruption of local transmission is achieved. Also, the poor access to health care 48 49478 services, as well as to core malaria vector control interventions such as IRS and bed nets distributed by the 50 51479 NMCP, indicated that these aspects needed to be integrated within the malaria elimination program 52 53480 planned for the district, in order to maximize the impact of the interventions aiming to interrupt 54 55481 transmission. Overall, this district represents the reality of the majority of malaria endemic areas in Sub- 56 57482 Saharan Africa, where elimination is most needed, and where it is most pressing to demonstrate the 58483 feasibility of elimination strategies. 59 60 17

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 39 BMJ Open

1 2 3 484 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 485 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 18

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 39

1 2 3 4 486 Abbreviations

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 487 APEs – Agentes Polivalentes Elementais (Community Health Workers) 7 488 BES – Boletim Epidemiológico Semanal (Weekly Epidemiological Bulletin) 8 9 489 CISM – Manhiça Health Research Center 10 11490 DHP – Demographic and Health Platform 12 13491 GPS – Global Positioning System 14 15492 HDSS – Health and Demographic Surveillance System 16 17493 HF – Health Facility 18494 HIV – Human ImmunodeficiencyFor peer virus review only 19 20495 INE – Instituto Nacional de Estatística (National Institute of Statistics) 21 22496 IPTp – Intermittent Preventive Treatment in pregnancy 23 24497 IRS – indoor residual spraying 25 26498 ISGlobal – Barcelona Institute of Global Health 27499 LLINs – Long-Lasting Insecticide treated Nets 28 29500 MDAs – Mass Drug Administration 30 31501 NMCP - National Malaria Control Program of Mozambique 32 33502 RDTs – Rapid Diagnostic Tests 34 35503 TV – Television 36 37 http://bmjopen.bmj.com/ 38 39504 Declarations 40 41 42505 Acknowledgements 43 44506 We would like to thank the community of Magude for participating in this study, and the team of field 45 on October 1, 2021 by guest. Protected copyright. 46507 workers who have collected the data presented here. We would also like to acknowledge the 47 48508 Administrative and District Health Authorities of Magude for their collaboration and for providing some 49 50509 of the information also included in this article. We thank everyone who supported this study directly or 51510 indirectly through fieldwork, or analysis support. 52 53 54511 Ethics approval and consent to participate 55 56512 The Demographic and Health Platform protocol, consent forms and questionnaires were approved by the 57 58513 CISM internal ethics committee and written consent from the health authorities was also sought prior to 59 60514 its implementation. Meetings were held with community leaders and with general members of the 19

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 39 BMJ Open

1 2 3 515 community of Magude to inform them about the DHP operations and the malaria elimination project as a 4 5 516 whole. A written informed consent was obtained from all household heads to record household-level BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 517 information, as well as from all individuals providing individual information. Informed consents for 8 9 518 children under the age of 18 years were sought from their parents or primary caretakers. The collection of 10 11519 district health surveillance data was conducted under the protocol that aimed to evaluate the impact of the 12520 Magude project on malaria transmission, which was approved by CISM’s internal ethical committee, the 13 14521 Ethics Committee of the Hospital Clínic of Barcelona, and the Ministry of Health National Bioethics 15 16522 Committee of Mozambique (IRB00002657). 17 18 For peer review only 19523 Competing Interests 20 21524 The authors declare that they have no competing interests 22 23525 Funding 24 25526 Funding was provided by the Bill and Melinda Gates Foundation and the Fundación “la Caixa” Partnership 26 27527 for the Elimination of Malaria in Southern Mozambique (OPP1115265). Q. B. is an ICREA (Institut Catal. 28 29528 de la Recerca i Estudis Avan.ats; Catalan Government) Research Professor. ISGlobal is a member of the 30 31529 CERCA Programme, Generalitat de Catalunya. 32 33 530 Author’s Contributions 34 35531 BG: participated in the study design and fieldwork, supported in the data cleaning and data analysis process, 36 37532 and wrote the draft of this article http://bmjopen.bmj.com/ 38 39533 AN: participated in the study design, in study analyses, in the interpretation of results and writing of this 40 41534 article 42 43535 HMS: cleaned and analyzed the data, and contributed to the writing of this article 44536 HM: led the implementation of field activities and data collection process, and participated in the 45 on October 1, 2021 by guest. Protected copyright. 46537 interpretation of results 47 48538 EJ: participated in the study design and implementation of field activities and data collection. 49 50539 CG: participated in the interpretation of results and writing of this article 51 52540 LC: participated in the study design, interpretation of results and writing of this article 53 541 FA: designed the data collection tools, supported the implementation of field activities and data cleaning. 54 55542 EM: participated in the study design, and interpretation of results 56 57543 QB: participated in the study design, interpretation of results and writing of this article 58 59544 NRR: participated in the interpretation of results and writing of this article 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 39

1 2 3 545 PLA: participated in the study design, interpretation of results and writing of this article 4 5 546 PA: participated in the study design, field implementation interpretation of results and writing of this BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 547 article 8 9 548 FS: participated in the study design, supervised field activities, interpretation of results and writing of this 10 11549 article 12550 CS: conceived the study, participated in interpretation of results and writing of this article 13 14551 15 16 17552 Availability of data 18 For peer review only 19553 The datasets collected as part of the Demographic Health Platform and analysed during the current study 20 21554 are available from the corresponding author upon written request. 22 555 The data presented in figures 2a and 2b belongs to the Ministry of Health of Mozambique and is considered 23 24556 as third-party data which can be accessed by contacting the following individuals appointed by the Ministry 25 26557 of health: 27 28558  To gain access to Inpatient Department data collected by the Ministry of health please contact Dr 29 30559 Baltazar Candrinho, Head of NMCP Mozambique, phone number:+258828665730 or e- 31 32560 mail: [email protected] 33561  To gain access to the data collected through the Boletim Epidemiologico Semanal (BES) please 34 35562 contact Dr Lorna Gujral, Department of Epidemiology/MOH/Mozambique, phone 36 37563 number: +25882 3250800 or e-mail: [email protected] http://bmjopen.bmj.com/ 38 39564 40 41565 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 39 BMJ Open

1 2 3 4 566 References

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 567 7 568 1. World Health Organization. World malaria report 2018. :210. 8 9 10569 2. Ministério da Saúde (MISAU), Instituto Nacional de Estatística (INE), ICF Internacional. Inquérito de 11 12570 Indicadores de Imunização, Malária e HIV/SIDA em Moçambique 2015. Maputo. Moçambique. Rockville, 13571 Maryland, EUA: INS, INE e ICF International; 2015. 14 15 16572 3. World Health Organization, others. A framework for malaria elimination. 2017 [cited 2017 Mar 28]; 17 573 Available from: http://apps.who.int/iris/handle/10665/254761 18 For peer review only 19 20574 4. Aide P, Candrinho B, Galatas B, Munguambe K, Guinovart C, Luis F, et al. Setting the scene and 21 22575 generating evidence for malaria elimination in Southern Mozambique. Malar J. 2019;18:190. 23 24576 5. Instituto Nacional de Estatistica. Relatório Final do Inquérito ao Orçamento Familiar- IOF 2014/15 25 26577 [Internet]. 2015. Available from: http://www.ine.gov.mz/operacoes-estatisticas/inqueritos/inquerito-sobre- 27 28578 orcamento-familiar 29 30579 6. Maharaj R, Moonasar D, Baltazar C, Kunene S, Morris N. Sustaining control: lessons from the Lubombo 31 32580 spatial development initiative in southern Africa. Malar J [Internet]. 2016 [cited 2016 Oct 17];15. Available 33 34581 from: http://malariajournal.biomedcentral.com/articles/10.1186/s12936-016-1453-9 35 36582 7. Ministerio de Administração Estatal. Perfil do distrito de Magude, Provincia de Maputo (2014). 37 http://bmjopen.bmj.com/ 38583 Ministerio de Administração Estatal; 2014. 39 40584 8. Alonso P, Saúte F, Aponte J, Gómez-Olivé F, Nhacolo A, Thompson R, et al. Manhiça demographic 41 42585 surveillance system, Mozambique. Popul Health Surviv INDEPTH Sites. 2002;1:295–308. 43 44586 9. Instituto Nacional de Estatística. Estatísticas do Distrito de Magude. 2010. 45 on October 1, 2021 by guest. Protected copyright. 46587 10. Instituto Nacional de Estatistica. População Projectada por distritos, Maputo Província 2007_2040.xls 47 48588 — Instituto Nacional de Estatistica [Internet]. [cited 2018 May 29]. Available from: 49 50589 http://www.ine.gov.mz/estatisticas/estatisticas-demograficas-e-indicadores-sociais/projeccoes-da- 51 52590 populacao/populacao-projectada-por-distritos-maputo-provincia-2007_2040.xls/view 53 54591 11. Portugaliza HP, Galatas B, Nhantumbo H, Djive H, Murato I, Saúte F, et al. Examining community 55 56592 perceptions of malaria to inform elimination efforts in Southern Mozambique: a qualitative study. Malar J 57 58593 [Internet]. 2019 [cited 2019 Aug 21];18. Available from: 59 60594 https://malariajournal.biomedcentral.com/articles/10.1186/s12936-019-2867-y 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 39

1 2 3 595 12. R, Foundation for Statistical Computing. R: A language and environment for statistical computing 4 5 596 [Internet]. Vienna, Austria; Available from: http://www.R-project.org/ BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 597 13. Hijmans R, Cruz M, Rojas E, Guarino L. DIVA-GIS, version 1.4. A geographic information system for 8 9 598 the management and analysis of genetic resources data. 2001; 10 11 12599 14. QGIS Development Team (2017) [Internet]. Open Source Geospatial Foundation Project; Available 13600 from: http://qgis.osgeo.org 14 15 16601 15. Alkire S, Santos ME. Acute Multidimensional Poverty: A New Index for Developing Countries. SSRN 17 602 Electron J [Internet]. 2010 [cited 2020 Feb 8]; Available from: http://www.ssrn.com/abstract=1815243 18 For peer review only 19 20603 16. Coates MM, Kamanda M, Kintu A, Arikpo I, Chauque A, Mengesha MM, et al. A comparison of all- 21 22604 cause and cause-specific mortality by household socioeconomic status across seven INDEPTH network 23 605 health and demographic surveillance systems in sub-Saharan Africa. Glob Health Action. 2019;12:1608013. 24 25 26606 17. Ndima SD, Sidat M, Give C, Ormel H, Kok MC, Taegtmeyer M. Supervision of community health 27 28607 workers in Mozambique: a qualitative study of factors influencing motivation and programme 29608 implementation. Hum Resour Health. 2015;13:63. 30 31 32609 18. Bocquier P, Sankoh O, Byass P. Are health and demographic surveillance system estimates sufficiently 33 34610 generalisable? Glob Health Action. 2017;10:1356621. 35 36611 19. Tatem AJ, Campiz N, Gething PW, Snow RW, Linard C. The effects of spatial population dataset choice 37 http://bmjopen.bmj.com/ 38612 on estimates of population at risk of disease. Popul Health Metr. 2011;9:4. 39 40613 20. Pons-Duran C, González R, Quintó L, Munguambe K, Tallada J, Naniche D, et al. Association between 41 42614 HIV infection and socio-economic status: evidence from a semirural area of southern Mozambique. Trop 43 44615 Med Int Health. 2016;21:1513–21. 45 on October 1, 2021 by guest. Protected copyright. 46616 21. Nhacolo AQ, Nhalungo DA, Sacoor CN, Aponte JJ, Thompson R, Alonso P. Levels and trends of 47 48617 demographic indices in southern rural Mozambique: evidence from demographic surveillance in Manhiça 49 50618 district. BMC Public Health [Internet]. 2006 [cited 2017 Jun 6];6. Available from: 51619 http://bmcpublichealth.biomedcentral.com/articles/10.1186/1471-2458-6-291 52 53 54620 22. Ruktanonchai NW, Bhavnani D, Sorichetta A, Bengtsson L, Carter KH, Córdoba RC, et al. Census- 55 621 derived migration data as a tool for informing malaria elimination policy. Malar J. 2016;15:273. 56 57 58622 23. Moonasar D, Maharaj R, Kunene S, Candrinho B, Saute F, Ntshalintshali N, et al. Towards malaria 59 60623 elimination in the MOSASWA (Mozambique, South Africa and Swaziland) region. Malar J [Internet]. 2016 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 39 BMJ Open

1 2 3 624 [cited 2016 Aug 25];15. Available from: http://malariajournal.biomedcentral.com/articles/10.1186/s12936- 4 5 625 016-1470-8 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 626 24. Ruktanonchai NW, Smith DL, De Leenheer P. Parasite sources and sinks in a patched Ross–Macdonald 8 9 627 malaria model with human and mosquito movement: Implications for control. Math Biosci. 2016;279:90– 10 11628 101. 12 13629 25. World Health Organization. GHO | By country | Mozambique - statistics summary (2002 - present) 14 15630 [Internet]. WHO. [cited 2018 May 29]. Available from: 16 17631 http://apps.who.int/gho/data/node.country.country-MOZ?lang=en 18 For peer review only 19632 26. González R, Augusto OJ, Munguambe K, Pierrat C, Pedro EN, Sacoor C, et al. HIV Incidence and Spatial 20 21633 Clustering in a Rural Area of Southern Mozambique. PloS One. 2015;10:e0132053. 22 23 634 27. Feldblum PJ, Enosse S, Dubé K, Arnaldo P, Muluana C, Banze R, et al. HIV prevalence and incidence 24 25635 in a cohort of women at higher risk for HIV acquisition in Chókwè, southern Mozambique. PloS One. 26 27636 2014;9:e97547. 28 29637 28. Deus N de, João E, Cuamba A, Cassocera M, Luís L, Acácio S, et al. Epidemiology of Rotavirus Infection 30 31638 in Children from a Rural and Urban Area, in Maputo, Southern Mozambique, before Vaccine Introduction. 32 33639 J Trop Pediatr. 2018;64:141–5. 34 35640 29. Nhampossa T, Mandomando I, Acacio S, Quintó L, Vubil D, Ruiz J, et al. Diarrheal Disease in Rural 36 37641 Mozambique: Burden, Risk Factors and Etiology of Diarrheal Disease among Children Aged 0-59 Months http://bmjopen.bmj.com/ 38 39642 Seeking Care at Health Facilities. PloS One. 2015;10:e0119824. 40 41643 30. Aguilar R, Moraleda C, Quintó L, Renom M, Mussacate L, Macete E, et al. Challenges in the diagnosis 42 43644 of iron deficiency in children exposed to high prevalence of infections. PloS One. 2012;7:e50584. 44 on October 1, 2021 by guest. Protected copyright. 45645 31. Garcia-Basteiro AL, Miranda Ribeiro R, Brew J, Sacoor C, Valencia S, Bulo H, et al. Tuberculosis on the 46 47646 rise in southern Mozambique (1997-2012). Eur Respir J. 2017;49. 48 49 50647 32. Lanaspa M, O’Callaghan-Gordo C, Machevo S, Madrid L, Nhampossa T, Acácio S, et al. High prevalence 51648 of Pneumocystis jirovecii pneumonia among Mozambican children <5 years of age admitted to hospital 52 53649 with clinical severe pneumonia. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis. 54 55650 2015;21:1018.e9-1018.e15. 56 57 58 59 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 39

1 2 3 651 33. Roca A, Sigauque B, Quinto L, Morais L, Berenguera A, Corachan M, et al. Estimating the vaccine- 4 5 652 preventable burden of hospitalized pneumonia among young Mozambican children. Vaccine. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 653 2010;28:4851–7. 8 9 654 34. Barclay VC, Smith RA, Findeis JL. Surveillance considerations for malaria elimination. Malar J. 10 11655 2012;11:304. 12 13656 35. Ministerio da Saude (MISAU), Instituto Nacional de Estatística (INE) e ICF International (ICFI). 14 15657 Moçambique Inquérito Demográfico e de Saúde 2011. Calverton Md USA MISAU INE E ICFI. 2011; 16 17 658 36. Pinchoff J, Hamapumbu H, Kobayashi T, Simubali L, Stevenson JC, Norris DE, et al. Factors Associated 18 For peer review only 19659 with Sustained Use of Long-Lasting Insecticide-Treated Nets Following a Reduction in Malaria 20 21660 Transmission in Southern Zambia. Am J Trop Med Hyg. 2015;93:954–60. 22 23 661 37. García-Basteiro AL, Schwabe C, Aragon C, Baltazar G, Rehman AM, Matias A, et al. Determinants of 24 25662 bed net use in children under five and household bed net ownership on Bioko Island, Equatorial Guinea. 26 27663 Malar J. 2011;10:179. 28 29664 38. Atieli HE, Zhou G, Afrane Y, Lee M-C, Mwanzo I, Githeko AK, et al. Insecticide-treated net (ITN) 30 31665 ownership, usage, and malaria transmission in the highlands of western Kenya. Parasit Vectors. 2011;4:113. 32 33 34666 39. World Health Organization, World Health Organization, Global Malaria Programme. Global technical 35667 strategy for malaria, 2016-2030 [Internet]. 2015 [cited 2018 Mar 5]. Available from: 36 37668 http://apps.who.int/iris/bitstream/10665/176712/1/9789241564991_eng.pdf?ua=1 http://bmjopen.bmj.com/ 38 39 669 40 41 42 43 44670 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 39 BMJ Open

1 2 3 4 671 Tables

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 672 7 673 Table 1: Summary of Magude Population in 2015 and 2016. 8 9 Magude Mapulan- 10 District Level Motaze Panjane Mahele Sede guene 11 12 N % N % N % N % N % N % 13 Households 2015 10965 8011 73.1 1471 13.4 627 5.7 377 3.4 479 4.4 14 Households 2016 11960 8520 71.3 1441 12.1 645 5.4 803 6.7 547 4.6 15 Population 2015 52802 38534 73.0 7421 14.1 3122 5.9 1784 3.4 1934 3.7 16 17 Residents 48448 91.8* 35346 73.0 6605 13.6 2930 6.0 1695 3.5 1868 3.9 18 Non- For peer review only 4133 7.8* 2970 71.9 816 19.7 191 4.6 89 2.2 66 1.6 19 residents 20 21 Unclassified 221 0.4* 218 99.5 0 - 1 0.5 0 - 0 - 22 Population 2016 61868 44203 71.4 8149 13.2 3576 5.8 3499 5.7 2404 3.8 23 Residents 56943± 92.1* 40623 71.4 7275 12.8 3361 5.9 3329 5.8 2322 4.1 24 Non- 25 4535 ø 7.3* 3266 72.0 842 18.6 206 4.5 143 3.2 76 1.7 26 residents 27 Unclassified 390 § 0.6* 314 80.9 32 8.2 9 2.3 27 7.0 6 1.5 28 Demographic 29 3078 2204 71.6 390 12.7 186 6.0 168 5.5 129 4.2 Events 2015-16 30 31 Births 1687 54.8 1203 71.4 220 13.0 92 5.5 91 5.4 80 4.7 32 Immigrations 721 23.4 521 72.3 60 8.3 51 7.1 60 8.3 29 4.0 33 Deaths 670 21.8 480 71.6 110 16.4 43 6.4 17 2.5 20 3.0 34 35674 * Column % 36675 ± 56,943 = 48,448 (censed in 2015) – 616 (deaths) + 1,670 (births) + 1,235 (immigrations) + 6,206 (censed in 2016) 37676 ø 4,535 = 4,133– 42 (deaths) + 12 (births) + 64 (immigrations) + 368 (censed in 2016) http://bmjopen.bmj.com/ 38677 § 390 = 221 – 8 (deaths) + 5 (births) + 21 (immigrations) + 151 (censed in 2016) 39678 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 26

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 39

1 2 3 679 Table 2: Socio-demographic characteristics of heads of households, and household characteristics (2015). 4 5 Characteristics of heads of households* N % BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Sex Males 5051 54.6 7 Females 4206 45.4 8 < 18 40 0.4 9 18 – 24 520 5.6 10 Age group (years) 11 25 – 64 6919 74.7 12 > = 65 1777 19.2 13 No formal education 5460 59.0 14 5-7th grade 2611 28.2 Education level 15 8 -12th grade 807 8.7 16 University 37 0.4 17 Missing Information 342 3.7 18 For Singlepeer review only4007 43.3 19 Married or de-facto union 3275 35.4 20 Marital status 21 Divorced 16 0.2 22 Separated 393 4.3 23 Widow 1554 16.8 24 Residents of Magude 9257 84.4 25 Households characteristics 26 Average household size** 5 [3-7] 27 Lone-resident households 1171 10.7 28 Cane 3506 32.5 29 Cement 2807 26.0 30 Wall material of the main Mud bricks 2338 21.6 31 32 building Adobe 1703 15.8 33 Zinc plates 319 3.0 34 Wood 128 1.2 35 Traditional latrine 5774 53.5 36 Improved latrine 1144 10.6

Type of toilet http://bmjopen.bmj.com/ 37 WC connected to septic tank 233 2.2 38 No latrine 3650 33.8 39 Paraffin 5362 49.6 40 Electricity 3321 30.8 41 Main source of lighting energy Candles 1074 9.9 42 43 Solar panels 502 4.6 44 Other 539 5.0 45 Wood logs 6401 59.3 on October 1, 2021 by guest. Protected copyright. 46 Coal 740 6.9 47 Gas 46 0.4 Kitchen Fuel 48 Electricity 65 0.6 49 Paraffin 8 0.1 50 Missing Information 3541 32,8 51 Water pumps 3700 34.3 52 Directly from the river 2691 24.9 53 54 Primary source of drinking water Piped water 2218 20.5 55 Open well near the river 1138 10.5 56 Other 1054 9.8 57 Vector control tools 58 % of households with >1 ITN 8906 81.2 59 60 27

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 39 BMJ Open

1 2 3 Universal ITN coverage (one net 5690 53.2 4 per two household members) 5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from Yes 5722 52.2 6 7 IRS in past 12 months No 4719 43.0 8 Unknown 524 4.8 9 680 * or subhead (if head does not live in the household) 10 11681 **Median [Interquartile Range] 12682 13 14683 Table 3: Individual-level health and malaria prevention indicators in Magude district 2015). 15 16 Age group in Years < 5 5-14 >15 17 N % N % N % 18 For peer review only Had fever in preceding 30 days 19 20 Yes 1043 12.1 1128 7.7 2931 11.6 21 No 7492 87.2 13325 91.2 21977 87.1 22 Unknown 55 0.6 154 1.1 319 1.3 23 24 Primary source of health care 25 Health Facility 8552 99.6 14548 99.6 25075 99.4 26 Traditional Healer 13 0.2 27 0.2 61 0.2 27 Community Health Worker 22 0.3 32 0.2 66 0.3 28 29 Slept under a bed net the preceding 30 night 31 Yes 2330 27.1 3179 21.8 6799 27.0 32 No 6260 72.9 11428 78.2 18428 73.0 33 34 35 36684 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 28

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 39

1 2 3 4 685 Figure Titles and Legends

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 686 7 687 Fig 1: Map of Magude district (2015). 8 9 688 A) Administrative and permanent river shape files were obtained from DIVA-GIS ((http://www.diva- 10 11689 gis.org/Data), and confirmed with Magude’s key informants. GPS positions of households, health facilities 12 13690 and community health workers obtained directly from the field and mapped using QGIS. B) Population 14 15691 pyramid of Magude district (2015) showing the proportion of five—year age and sex groups out of the 16 17692 total population. 18693 For peer review only 19 20694 Fig 2: Health profile of Magude district prior to the baseline census (2010-2014). 21 22695 A) Most common diseases leading to all-age hospitalization and later discharge in Magude reported by the 23 24696 District Health Authorities between 2010-2014. B) Weekly number of outpatient malaria cases observed 25 26697 in Magude between 2010 and 2014 reported through the Weekly Epidemiological Bulletin (BES) and 27698 monthly rainfall data obtained from the Climate Hazards Group InfraRed Precipitation with Station data 28 29699 (CHIRPS). 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 29

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 39 BMJ Open

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 Fig1 20 21 119x45mm (300 x 300 DPI) 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 39

1 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 Fig2 19 169x56mm (300 x 300 DPI) 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on October 1, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 39 BMJ Open

1 2 3 Supplementary Tables 4 5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 S.Table 1: Selected demographic and socioeconomic characteristics of Magude’s population 7 (2015). 8 9 Variable / by sex Males Females 10 N Col % N Col % 11 12 Age groups (years) 21775 26673 13 < 1 826 3.8 898 3.4 14 1-4 3391 15.6 3475 13.0 15 16 5-14 7345 33.8 7262 27.2 17 15-64 9334 42.9 13205 49.5 18 For peer review only 19 >=65 864 4.0 1824 6.8 20 Education Level (>= 6 years old) 16712 21461 21 22 Illiterate 7925 47.4 11136 51.9 23 5-9 grade 6221 37.2 7616 35.5 24 10-12 grade 1595 9.5 1825 8.5 25 26 University 43 0.3 20 0.1 27 Missing Information 928 5.6 864 4.0 28 29 Occupation (> 18 years old) 5243 5532 30 Farmer or Fisherman 1371 26.1 3909 70.7 31 32 Salesperson 466 8.9 609 11.0 33 Construction work occupations 1159 22.1 39 0.7 34 35 Coal maker / Lumberjack 603 11.5 173 3.1 36 Guards/Police/Military 539 10.3 45 0.8 37 http://bmjopen.bmj.com/ Health professional 85 1.6 133 2.4 38 39 Teacher 263 5.0 180 3.3 40 Miner 106 2.0 9 0.2 41 42 Other 651 12.4 435 7.9 43 Marital Status (> 14 years old) 10157 14986 44 45 Single (never married) 6168 60.7 7772 51.9 on October 1, 2021 by guest. Protected copyright. 46 Married or de-facto union 3690 36.3 4702 31.4 47 In a polygamic relationship 202 2.0 - - 48 49 Divorced 7 0.1 13 0.1 50 Separated 138 1.4 541 3.6 51 52 Widow 154 1.5 1958 13.1 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 39

1 2 3 S. Table 2: Modified household poverty index (PI) based on the following deprivation indicators: 1) 4

5 lack of electricity, 2) lack or sharing of an improved sanitation facility; 3) lack of access to improved BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 drinking water source, or source only available more than 30-minute walk, round trip; 4) dirt, sand or 7 8 dung household floors; 5) dung, wood or charcoal used for cooking fuel; and 6) households that do not 9 10 own a car or truck AND do not own more than one of the following: radio, TV, telephone, bike, 11 motorbike, or refrigerator. 12 13 14 Deprivations Magude Motaze Panjane Mahele Mapulanguene District 15 16 0-2 32.1 9.8 1 0.9 13 25.2 17 3-4 49.4 58.3 55.1 61 70.8 52.3 18 5-6 For18.5 peer31.9 review43.9 38.1 only16.2 22.4 19 20 21 Median 3 4 4 4 3 4 22 23 Mean 3.1 3.9 4.2 4.2 3.5 3.4 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 39 BMJ Open

1 2 3 S.Table 3: Number of individuals by age group who reported spending one night outside of 4

5 Magude the day before the census visit was performed (2015). BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Men Women All 7 Previous night spent Age group N (%) N (%) N (%)* 8 outside of Magude 9 592 5.1 580 5.0 1172 43,0 Yes 10 * 11 0 - 14 No 10874 94.0 10970 94.3 12 NA± 96 0.8 85 0.7 13 324 6.3 362 5.6 686 25.1 Yes 14 * 15 – 29 15 No 4774 92.7 6022 93.3 16 NA± 52 1.0 70 1.1 17 298 11.3 197 5.3 495 18.1 Yes 18 * 30 – 44 For peer review only 19 No 2323 87.8 3497 93.9 20 NA± 24 0.9 30 0.8 133 8.6 149 4.9 282 10.3 21 Yes * 22 45 – 64 23 No 1393 90.5 2839 93.8 24 NA± 13 0.8 39 1.3 25 Yes 36 4.2 57 3.1 93 3.4* 26 > 65 No 818 94.7 1744 95.6 27 NA± 10 1.2 23 1.3 Yes 1383 6.4 1345 5.0 2728 5.6 28 All ages No 20182 92.7 25072 94,0 45254 93.5 29 NA± 195 0.9 247 1.0 442 0.9 30 * Percentage of those who reported travelling (2,728) 31 32 ± Missing Information 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 39

1 2 3 S.Table 4: Household-level health and malaria prevention indicators in Magude district (2015). 4

5 Administrative Magude- BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from Motaze Panjane Mahele Mapulanguene 6 Post Sede 7 8 Household size 4 [3-6] 4 [3-7] 4 [2-6] 4 [2-6] 3 [2-5.5] 9 Members per net 10 2.0 [1.4-4.0] 2.0 [1.3-3.0] 2.0 [1.6-5.5] 2.0 [1.5-4.0] 2.0 [1.3-4.0] Median [IQR] 11 12 Universal ITN 52.7% 59.9% 44.1% 52.7% 52.1% 13 coverage * 14 15 IRS in past 12 N % N % N % N % N % 16 months 17 18 Yes 3925For 49.0 peer 1162 79.0 review 258 41.1 only230 61.0 147 30.7 19 No 3627 45.3 293 19.9 352 56.1 133 35.3 314 65.6 20 Unknown 459 5.7 16 1.1 17 2.7 14 3.7 18 3.8 21 22 ** One net for every 2 members of the households 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 39 BMJ Open

1 2 3 Supplementary Figures 4

5 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 S.Figure 1: Household Assets in Magude district in 2015. 7 8 Percentage of households in Magude district with at least one item per asset, and median and 9 10 interquartile range of the assets for which the number of items available was higher than 1. 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 39

1 2 3 S.Figure 2: Mobility patterns of Magude’s population in 2015. 4

5 Main destinations per age group among those who reported travelling outside of Magude in 2015. BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on October 1, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 39 BMJ Open

1 2 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 3 4 Item Check

5 No Recommendation BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title Yes 7 or the abstract Page 1 8 (b) Provide in the abstract an informative and balanced summary of Yes 9 10 what was done and what was found Page 3 11 Introduction 12 Background/rationale 2 Explain the scientific background and rationale for the investigation Yes 13 14 being reported Page 6 15 Objectives 3 State specific objectives, including any prespecified hypotheses Yes 16 Page 6 17 18 Methods For peer review only 19 Study design 4 Present key elements of study design early in the paper Yes 20 Page 7-8 21 Setting 5 Describe the setting, locations, and relevant dates, including periods of Yes 22 23 recruitment, exposure, follow-up, and data collection Page 7 24 Participants 6 (a) Give the eligibility criteria, and the sources and methods of Yes 25 selection of participants Page 8 26 27 Variables 7 Clearly define all outcomes, exposures, predictors, potential NA 28 confounders, and effect modifiers. Give diagnostic criteria, if Pages 8-9 29 applicable 30 Data sources/ 8* For each variable of interest, give sources of data and details of Yes 31 32 measurement methods of assessment (measurement). Describe comparability of Pages 8-9 33 assessment methods if there is more than one group 34 Bias 9 Describe any efforts to address potential sources of bias Yes 35 Page 4 36 37 Study size 10 Explain how the study size was arrived at NA http://bmjopen.bmj.com/ 38 Quantitative 11 Explain how quantitative variables were handled in the analyses. If Yes 39 variables applicable, describe which groupings were chosen and why Page 8 40 41 Statistical methods 12 (a) Describe all statistical methods, including those used to control for Yes 42 confounding Page 8 43 (b) Describe any methods used to examine subgroups and interactions NA 44 (c) Explain how missing data were addressed Page 8 45 on October 1, 2021 by guest. Protected copyright. 46 (d) If applicable, describe analytical methods taking account of Yes 47 sampling strategy Page 8 48 (e) Describe any sensitivity analyses NA 49 50 Results 51 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers Yes 52 potentially eligible, examined for eligibility, confirmed eligible, Page 9 53 included in the study, completing follow-up, and analysed 54 55 (b) Give reasons for non-participation at each stage Yes 56 Page 9 57 (c) Consider use of a flow diagram No 58 59 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, Yes 60 social) and information on exposures and potential confounders Page 9-10 (b) Indicate number of participants with missing data for each variable Yes

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 39

1 2 of interest Table 2, 3 S.Table1, 4

5 S.Table3 BMJ Open: first published as 10.1136/bmjopen-2019-033985 on 5 May 2020. Downloaded from 6 Outcome data 15* Report numbers of outcome events or summary measures Yes 7 Pages 9-12 8 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted NA 9 10 estimates and their precision (eg, 95% confidence interval). Make 11 clear which confounders were adjusted for and why they were 12 included 13 (b) Report category boundaries when continuous variables were Yes 14 15 categorized Table2, 16 S.Table1-3 17 (c) If relevant, consider translating estimates of relative risk into NA 18 For peer review only absolute risk for a meaningful time period 19 20 Other analyses 17 Report other analyses done—eg analyses of subgroups and NA 21 interactions, and sensitivity analyses 22 23 Discussion 24 Key results 18 Summarise key results with reference to study objectives Yes 25 Pages 12-15 26 Limitations 19 Discuss limitations of the study, taking into account sources of Yes 27 28 potential bias or imprecision. Discuss both direction and magnitude of Pages 12-15 29 any potential bias 30 Interpretation 20 Give a cautious overall interpretation of results considering objectives, Yes 31 limitations, multiplicity of analyses, results from similar studies, and Pages 12-15 32 33 other relevant evidence 34 Generalisability 21 Discuss the generalisability (external validity) of the study results Yes 35 Pages 12-15 36 37 Other information http://bmjopen.bmj.com/ 38 Funding 22 Give the source of funding and the role of the funders for the present Yes 39 study and, if applicable, for the original study on which the present Page 17 40 article is based 41 42 43 *Give information separately for exposed and unexposed groups. 44 45 on October 1, 2021 by guest. Protected copyright. 46 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 47 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 48 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 49 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 50 51 available at www.strobe-statement.org. 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml