Research ArticleMaternal Attitudes about Objectively Monitored Bednet Use inRural Uganda
Paul J Krezanoski123 Data Santorino45 Nuriat Nambogo5
Jeffrey I Campbell3 and David R Bangsberg6
1Department of Medicine Massachusetts General Hospital Boston MA USA2Department of Pediatrics Massachusetts General Hospital Boston MA USA3Harvard Medical School Boston MA USA4Mbarara University of Science and Technology Mbarara Uganda5Consortium for Affordable Medical Technologies (CAMTech) Mbarara Uganda6Oregon Health Sciences University and Portland State University School of Public Health Portland OR USA
Correspondence should be addressed to Paul J Krezanoski pkrezanoskimghharvardedu
Received 11 May 2016 Revised 17 September 2016 Accepted 25 September 2016
Academic Editor Robert Novak
Copyright copy 2016 Paul J Krezanoski et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
Insecticide-treated bednets (ITNs) are a mainstay of malaria prevention yet poor adherence poses a major barrier to effectiveprevention Self-reports of bednet use suffer from recall and social desirability biases We have designed a device that electronicallyrecords ITN usage longitudinally SmartNet consists of circuits made from a conductive fabric interwoven into the sides and top ofa rectangular ITN Digital sampling of the state of these circuits allows for determining whether the SmartNet is deployed for use orfolded upWe conducted a study among pregnantwomen andwomenwith childrenlt5 years inUganda to determine attitudes aboutobjective bednetmonitoring and SmartNet Fifty womenwere interviewedwith an average age of 27 years and 23 children Twenty-two percent were pregnant Ninety-five percent had used a bednet and 90 reported having a bednet at home After displaying aSmartNet 92 thought it would be easy to use and 100 expressed interest in using SmartNet Concerns about SmartNet includedwashing the net worries about being monitored while asleep and worries about users removing the device components Objectivemonitoring of ITN use appears to be acceptable among women in rural Uganda setting the stage for further SmartNet fieldtesting
1 Introduction
Approximately 34 billion people live at risk of malariainfection throughout the world and most of the more than600000 annual malaria-related deaths occur among Sub-Saharan African children under five years of age [1] TheUnited Nations Development Goals target a 75 reduction inmalaria by 2015 [2] Insecticide-treated bednets (ITNs) are acost-effective means of malaria prevention [1 3] Ownershipof an ITN reducesmortality by 18ndash23 [4] saving 55 lives peryear for every 1000 ITNs distributed [5] In 2013 the WorldHealthOrganization (WHO) recommended that free bednetsbe made available universally for anyone at risk of malaria[6]
Despite the acceptance of bednets as effective tools formalaria prevention there are still significant questions aboutthe factors that influence household decisions to use bednets[7] Most studies assessing the determinants of bednet userely on self-reports from households and individuals [8ndash12] However self-reported use may be prone to both socialdesirability and recall biases [13] which compromise bothour understanding of barriers to bednet use and estimates ofthe true cost-effectiveness of bednets as a malaria preventiontool The effects of these biases due to self-reporting onestimates of bednet use are not well defined in the literatureas there are very few studies which explicitly discuss the dis-crepancy between self-reports and more objective measures[14ndash16]
Hindawi Publishing CorporationMalaria Research and TreatmentVolume 2016 Article ID 8727131 8 pageshttpdxdoiorg10115520168727131
2 Malaria Research and Treatment
Figure 1 Side view of SmartNet
Compared to asking individuals or household represen-tatives about their bednet use more objective means of mea-suring bednet use have been attempted but are problematicSpot checks during sleeping hours [17ndash19] while the mostaccurate are invasive and unlikely to be acceptable to mostpopulations Visually confirming that bednets are mountedabove sleeping areas in households [20ndash24] is also invasiverequiring entry into households and does not ensure that abednet is actually being unfurled at night
Real-time logging of bednet use for later analysis offersthe possibility for more precise understanding of bednetuse behaviors in terms of both accuracy and temporalresolution with less invasion than visual observation of use[25] Furthermore objective bednet use monitoring will be akey to understanding malaria epidemiology and preventionas mosquito biting behavior is known to change with bednetusage [26] With these advantages of objective monitoring inmind we have developed SmartNet an electronic monitor ofbednet use SmartNet uses a microprocessor and conductivefabric sewn into a WHO-approved bednet in order to detectwhether a bednet is unfurled or not The unfurling of thebednet in a household is taken as a proxy of bednet use andconversely the folding up of the bednet is assumed to be aproxy of nonuse The conductive material can be integratedinto any size bednet For this study SmartNet componentswere integrated into white rectangular (51015840 times 71015840 times 71015840) bednetsbought from a local pharmacy (sim$12) The conductive fabricis sewn to form three distinct electric circuits that makecontact with each other when the bednet is folded or tiedup for storage A microprocessor sends an electrical currentevery 15 minutes to check circuit connectivity allowing theSmartNet to distinguish between a net in use or not Time-stamped data is stored on a removable memory card for lateranalysis (Figures 1 2 and 3)
SmartNet is being actively developed through a col-laboration between engineers and health professionals atMassachusetts General Hospital in Boston MA USA andthe Mbarara University of Science and Technology (MUST)at the Consortium for Affordable Medical Technologies(CAMTech) innovation center in Uganda Development ofSmartNet exemplifies a cocreation model for development ofhealth technologies in developing countries [27] Cocreationengages the end-users of technologies in developing countries
Figure 2 Top view of SmartNet
Figure 3 SmartNet microprocessor
to become innovators themselves through top-to-bottominvolvement across the product development lifecycle
It is envisioned that SmartNet could be utilized as aresearch tool for obtaining reliable measures of longitudinalbednet use in target populations or as a sampling tool to vali-date other forms of assessing bednet use that is self-reportsAs a first step to understand the feasibility and acceptability ofelectronic bednet use monitoring we examined the attitudesabout SmartNet among mothers of young children andamong pregnant women in malaria-endemic rural UgandaAfter introducing the women to a model SmartNet andexplaining its functionality we conducted quantitative andsemistructured qualitative interviews focused on impressionsof SmartNet and attitudes about objective bednetmonitoring
2 Methods
This study was conducted in southwestern Uganda at theKinoni Health Center IV in Rugando Subcounty MbararaDistrictThehealth center located 5 kilometers fromMbararaTown serves about 110 patients per day most of whom arerural subsistence farmers Malaria is endemic to the regionand makes up an estimated 20 of outpatient visits (16ndash25visitsday) to the Kinoni Health Center
We recruited 50 participants from July to August 2014The number of participants was based on an attempt tocapture the range of experiences with bednets and attitudes
Malaria Research and Treatment 3
about SmartNet and electronic bednet monitoring Partici-pants were identified among patients waiting to see healthproviders at the clinic They were invited to participate inthe study after their clinic visit was complete A room wasset aside for confidential study visits Informed consent wasgathered from all participants before proceeding to the studyinterviews Since our study focused on those at most risk ofmalaria we recruitedwomen to participate in the study if theywere mothers of at least one child aged five years or youngeror were currently pregnant These recruitment criteria alsoallowed for selection of a study sample which is familiarwith bednets as our previous work suggests that householdswith young children are more likely to own and use bednets[28] Participants were informed that their participation inthis study would not affect future inclusion in SmartNet-related studies No incentive was provided for participationAll consents and surveys were translated into Runyankolethe local language All participants signed a written informedconsent to participate Participants were given the optionto complete the survey in either English or Runyankole Inorder to maintain consistency of interpretation of qualitativeresponses a single research assistant (Nuriat Nambogo) whois fluent in Runyankole and English conducted all interviews
Participants received a structured interview to collectdata on age marital status number of children pregnancystatus education and literacy employment and averagemonthly income Then they answered a questionnaire relat-ing to their prior experience with antimalarial bednetscurrent use of bednets frequency of household bednet useand their opinions about the ease of use of bednets Ourprior work suggests that knowledge about malaria is asignificant indicator of bednet use within a household [29]All participants were shown a SmartNet which was hangingin the interview room and its components and functionalitywere described before participants were asked SmartNet-specific qualitative questions
Participants answered open-ended questions about theirimpressions of SmartNet their ldquolikesrdquo and ldquodislikesrdquo aboutthe device individual concerns and anticipated concernsabout the device from the viewpoint of others within theirhousehold This final line of questioning from the viewpointof others in the household provided a level of abstraction inorder for participants to provide honest comments withoutviolating social norms The responses were written on thesurvey sheet for later translation and coding Themes werederived from a sample of qualitative responses to eachquestion and then each survey was coded independentlyby two members of the research team (Jeffrey I CampbellandNuriat Nambogo) Researchers discussed inconsistenciesin coding and when inconsistencies existed a final codewas established by consensus between the two coders or theprimary author if consensus could not be reached
All data were collected on paper forms and entered man-ually into the Research Electronic Data Capture (REDCap)tool hosted at Massachusetts General Hospital Boston MAUSA [30] The data were cleaned following the completionof data entry Quantitative data were analyzed in Stata 10(Statacorp College Rd TX) and qualitative responses werereviewed and coded in Microsoft Excel
Table 1 Demographic characteristics
Total number of respondents 119899 = 50
Age (mean plusmn SD)27 plusmn 56
Range 17ndash40Pregnant (119899 []) 11 (22)Gestational months of pregnancy (119899 = 11)(mean plusmn SD)
57 plusmn 21Range 3ndash9
Number of children under 5 years of age(mean plusmn SD)
28 plusmn 16Range 0ndash8
Number of children under five years (119899[])
0 5 (10)1 22 (44)2 19 (38)3 4 (8)
Highest education attained (119899 [])Attended up to
Primary school 27 (54)Secondary school 17 (34)Tertiary schooling 2 (4)Never attended school 4 (8)
Literacy (119899 [])Read sample text completely 31 (62)Read sample text partially 7 (14)Unable to read sample text 12 (24)
Able to write name legibly (119899 [])Yes 44 (88)No 6 (12)
Average monthly income (mean plusmn SD) 162250UGXRange 5000ndash1500000
3 Results
31 Demographic Characteristics A total of 50 women wereapproached and all consented to participate Participants hadamean age of 27 years and amean of 23 children Ten percentof the women had no children (550) and 8 (450) hadmore than two children under five years Overall twenty-two percent (1150) of women reported being currentlypregnant
Fifty-four percent (2750) of women had attendedonly primary school and 8 (450) had never attendedschool Seventy-two percent (3650) of women self-identifiedas farmers The average self-reported monthly incomewas 162250 Ugandan Shillings (UGX) (sim$6228) (range5000UGX to 1500000UGX) (Table 1)
4 Malaria Research and Treatment
32 Bednet Ownership and Use Ninety percent (4550) ofwomen reported that their household currently owned atleast one bednet and 80 (4050) reported that at leastone bednet in their household was in use the night beforethe survey Seventy-six percent (3850) of women reportedthat a bednet was in use every day in the previous weekby at least someone in their household Eighty-four per-cent (4250) reported that their householdrsquos ability to usebednets as directed was good or very good and 8 (450)reported their householdrsquos ability as poor or very poor Thehouseholds owned an average of 242 bednets (range 1 to 6)(Table 2)
Self-reported bednet use was high with 96 (4850) ofrespondents reporting that they had used a bednet at somepoint in the past and 66 (3350) reporting using a bednetevery night in the previous week Twenty percent (1050) didnot use a net at all in the last week and 14 (750) had notused a net in the last month Among women who had everused bednets greater than 80 reported that bednets wereeasy or very easy to use and easy or very easy to remember touse However 146 (748) of ever-users reported that theywere difficult to use and 104 (548) reported that they weredifficult to remember to use Reasons for not using bednetsincluded excessive heat while under the net traveling awayfrom home or not owning enough nets
33 Impressions of the SmartNet Device All participants(5050) said theywould be interested in using the SmartNet intheir homes In open-ended responses reasons that womengave included liking the appearance of the SmartNet think-ing the net would be useful and wanting to have a new netNinety-two percent (4650) reported that they thought thedevice would be easy to use (Table 3) Notably many of theresponses to the question ldquoare there particular things you likeabout the SmartNetrdquo pertained to nets in general and wereindependent of monitoring For instance many participantsreported that they would like new nets because their currentnets were old or had holes in them
When queried about how their behavior would change inresponse to using SmartNet 32 (1650) of women reportedthat the SmartNet would not change their net use Noparticipants stated that their net use would decrease Of the68 (3450) who reported an anticipated increase in net usereasons included the fact of beingmonitored being remindedby the device to use the net and the attractive appearance ofthe net
While the majority of responses to the SmartNet demon-stration were positive some participants offered suggestionsor negative impressions of the net Themes included fear ofthe device perceived dislike of being studied when use of bednet is not as directed and dislike of home-visits to check onthe net
34 Perceptions of Objective Monitoring of Bednet Use Themajority of participants did not express personal concernswith objective monitoring When participants were askedldquohow would you feel knowing that your bednet use isbeing monitored by the SmartNetrdquo ninety percent (4550)
Table 2 Bednet ownership and use
Household bednet characteristics 119899 = 50
Household bednet ownership (119899 []) 45 (90)Average number of bednets perhousehold (mean plusmn SD)
242 plusmn 112Range 1ndash6
Household bednet use the night before (119899[]) 40 (80)
Household net use by someone in theweek before (days) (119899 [])
7 38 (76)lt7 6 (12)0 5 (10)No response 1 (2)
Householdrsquos ability to use bednetscorrectly (119899 [])
Very good 19 (38)Good 23 (46)Fair 3 (6)Poor 2 (4)Very poor 2 (4)No response 1 (2)
Individual bednet characteristicsEver used bednet (119899 []) 48 (96)Personally used every night last week (119899[]) 33 (66)
Did not use in last week (119899 []) 10 (20)Did not use in last month (119899 []) 14 (28)Bednets are (119899 [])
Very easy to use 27 (54)Easy to use 13 (26)Neither easy nor hard to use 2 (4)Difficult to use 6 (12)Very difficult to use 1 (2)No response 1 (2)
Remembering to use a bednet is (119899 [])Very easy 30 (60)Easy 13 (26)Neither easy nor hard 1 (2)Difficult 5 (10)Very difficult 0 (0)No response 1 (2)
Agree that bednets are useful (119899 []) 48 (96)Reasons for not using a net (119899 = 25) (119899[])Too hot 8 (16)Traveled away from home 6 (12)Not enough nets 6 (12)Allergicdislike treatment 4 (8)Net damaged or being washed 2 (4)Spouse does not want to use it 1 (2)
of participants reported that they did not believe theirfamily members would be worried about being monitored bySmartNet (Table 4)
When they did mention concerns participants focusedon monitoring of behaviors including sleeping behaviorswhile in bed at night
Malaria Research and Treatment 5
Table 3 Impressions of SmartNet device
Interested in using (119899 []) 50 (100)Perceived to be easy to use (119899 []) 46 (92)
Positive impressionsTheme ExampleLooksfeels nice or attractive ldquoI love the yellow color and the designs on itrdquo
Looks easy to use ldquoI can hang it well on my walls because it looks easy to hang It is also light enoughto move withrdquo
Would increase usage ldquoI would be more accurate in putting [up] my netrdquo
Would help to learn how to use nets ldquoI want the researchers to monitor me and know if I use my net well or not whichcan help me improve in case I donrsquot use it wellrdquo
Help with malaria prevention in the community ldquoIf these nets are given out it would reduce on the malaria cases especially invillagesrdquo
Help learn about net use ldquoThey [researchers] will be able to find out how best people can use their netsrdquoNegative impressions
Theme Example
Appearance ldquoSome people may not like this color of the net since they have children who candirty itrdquo
Treatment ldquoIf it has a lot of chemicals it will disturb merdquoNet would be difficult to hang ldquoThe squared net might be difficult to some people in hangingrdquo
Device could be damaged ldquoIn case the device is not covered and it stays open people may spoil it especiallychildrenrdquo
People who misuse nets may not like the SmartNet ldquoFor those who donrsquot use the nets they may not like it when you come to check ontheir bednets as you monitor themrdquo
The device could negatively affect health ldquoThey may fear that it [SmartNet] might negatively affect their healthrdquo
Dislike of home visits ldquoSome people who are not hygienically good in their bedrooms may not like youentering their bedroomsrdquo
Not able to wash the net ldquoThey may fear washing the net if it is to be washedrdquoFear that the device may be stolen ldquoFearing that thieves may steal the device from themrdquo
Table 4 Impressions of objective monitoring of bednet use
Personally not worried about monitoring (119899[]) 45 (90)
Family would not be worried aboutmonitoring (119899 []) 43 (86)
Positive impressionsTheme ExampleMonitoring would improve use ldquoBecause I would want to get good results about the userdquo
Provides feedback ldquoIf I can be given feedback on how I use my net it can help me improve in case Idonrsquot use it wellrdquo
Benefits of participating in a research study ldquoIt would mean that the researchers always will remember to check on me and visitme at homerdquo
Negative impressionsTheme Example
Monitoring while sleeping ldquoPeople might think that they device will be monitoring them as well when they aresleeping in their beds under the netrdquo
Monitoring of misuse ldquoSome people may not like to be monitored especially those that at times do notuse their netsrdquo
Change current behaviors ldquoFor people who get discomfort with heat at night and they remove their nets theymay fear to do this because of being monitoredrdquo
Concern about continuous monitoring ldquoI would be thinking every time they are looking at me and measuring me while inmy bedrdquo
Concern about recording private behavior ldquoI would think that the device is recording me while in bed with my husbandrdquo
6 Malaria Research and Treatment
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
Compared to asking individuals or household represen-tatives about their bednet use more objective means of mea-suring bednet use have been attempted but are problematicSpot checks during sleeping hours [17ndash19] while the mostaccurate are invasive and unlikely to be acceptable to mostpopulations Visually confirming that bednets are mountedabove sleeping areas in households [20ndash24] is also invasiverequiring entry into households and does not ensure that abednet is actually being unfurled at night
Real-time logging of bednet use for later analysis offersthe possibility for more precise understanding of bednetuse behaviors in terms of both accuracy and temporalresolution with less invasion than visual observation of use[25] Furthermore objective bednet use monitoring will be akey to understanding malaria epidemiology and preventionas mosquito biting behavior is known to change with bednetusage [26] With these advantages of objective monitoring inmind we have developed SmartNet an electronic monitor ofbednet use SmartNet uses a microprocessor and conductivefabric sewn into a WHO-approved bednet in order to detectwhether a bednet is unfurled or not The unfurling of thebednet in a household is taken as a proxy of bednet use andconversely the folding up of the bednet is assumed to be aproxy of nonuse The conductive material can be integratedinto any size bednet For this study SmartNet componentswere integrated into white rectangular (51015840 times 71015840 times 71015840) bednetsbought from a local pharmacy (sim$12) The conductive fabricis sewn to form three distinct electric circuits that makecontact with each other when the bednet is folded or tiedup for storage A microprocessor sends an electrical currentevery 15 minutes to check circuit connectivity allowing theSmartNet to distinguish between a net in use or not Time-stamped data is stored on a removable memory card for lateranalysis (Figures 1 2 and 3)
SmartNet is being actively developed through a col-laboration between engineers and health professionals atMassachusetts General Hospital in Boston MA USA andthe Mbarara University of Science and Technology (MUST)at the Consortium for Affordable Medical Technologies(CAMTech) innovation center in Uganda Development ofSmartNet exemplifies a cocreation model for development ofhealth technologies in developing countries [27] Cocreationengages the end-users of technologies in developing countries
Figure 2 Top view of SmartNet
Figure 3 SmartNet microprocessor
to become innovators themselves through top-to-bottominvolvement across the product development lifecycle
It is envisioned that SmartNet could be utilized as aresearch tool for obtaining reliable measures of longitudinalbednet use in target populations or as a sampling tool to vali-date other forms of assessing bednet use that is self-reportsAs a first step to understand the feasibility and acceptability ofelectronic bednet use monitoring we examined the attitudesabout SmartNet among mothers of young children andamong pregnant women in malaria-endemic rural UgandaAfter introducing the women to a model SmartNet andexplaining its functionality we conducted quantitative andsemistructured qualitative interviews focused on impressionsof SmartNet and attitudes about objective bednetmonitoring
2 Methods
This study was conducted in southwestern Uganda at theKinoni Health Center IV in Rugando Subcounty MbararaDistrictThehealth center located 5 kilometers fromMbararaTown serves about 110 patients per day most of whom arerural subsistence farmers Malaria is endemic to the regionand makes up an estimated 20 of outpatient visits (16ndash25visitsday) to the Kinoni Health Center
We recruited 50 participants from July to August 2014The number of participants was based on an attempt tocapture the range of experiences with bednets and attitudes
Malaria Research and Treatment 3
about SmartNet and electronic bednet monitoring Partici-pants were identified among patients waiting to see healthproviders at the clinic They were invited to participate inthe study after their clinic visit was complete A room wasset aside for confidential study visits Informed consent wasgathered from all participants before proceeding to the studyinterviews Since our study focused on those at most risk ofmalaria we recruitedwomen to participate in the study if theywere mothers of at least one child aged five years or youngeror were currently pregnant These recruitment criteria alsoallowed for selection of a study sample which is familiarwith bednets as our previous work suggests that householdswith young children are more likely to own and use bednets[28] Participants were informed that their participation inthis study would not affect future inclusion in SmartNet-related studies No incentive was provided for participationAll consents and surveys were translated into Runyankolethe local language All participants signed a written informedconsent to participate Participants were given the optionto complete the survey in either English or Runyankole Inorder to maintain consistency of interpretation of qualitativeresponses a single research assistant (Nuriat Nambogo) whois fluent in Runyankole and English conducted all interviews
Participants received a structured interview to collectdata on age marital status number of children pregnancystatus education and literacy employment and averagemonthly income Then they answered a questionnaire relat-ing to their prior experience with antimalarial bednetscurrent use of bednets frequency of household bednet useand their opinions about the ease of use of bednets Ourprior work suggests that knowledge about malaria is asignificant indicator of bednet use within a household [29]All participants were shown a SmartNet which was hangingin the interview room and its components and functionalitywere described before participants were asked SmartNet-specific qualitative questions
Participants answered open-ended questions about theirimpressions of SmartNet their ldquolikesrdquo and ldquodislikesrdquo aboutthe device individual concerns and anticipated concernsabout the device from the viewpoint of others within theirhousehold This final line of questioning from the viewpointof others in the household provided a level of abstraction inorder for participants to provide honest comments withoutviolating social norms The responses were written on thesurvey sheet for later translation and coding Themes werederived from a sample of qualitative responses to eachquestion and then each survey was coded independentlyby two members of the research team (Jeffrey I CampbellandNuriat Nambogo) Researchers discussed inconsistenciesin coding and when inconsistencies existed a final codewas established by consensus between the two coders or theprimary author if consensus could not be reached
All data were collected on paper forms and entered man-ually into the Research Electronic Data Capture (REDCap)tool hosted at Massachusetts General Hospital Boston MAUSA [30] The data were cleaned following the completionof data entry Quantitative data were analyzed in Stata 10(Statacorp College Rd TX) and qualitative responses werereviewed and coded in Microsoft Excel
Table 1 Demographic characteristics
Total number of respondents 119899 = 50
Age (mean plusmn SD)27 plusmn 56
Range 17ndash40Pregnant (119899 []) 11 (22)Gestational months of pregnancy (119899 = 11)(mean plusmn SD)
57 plusmn 21Range 3ndash9
Number of children under 5 years of age(mean plusmn SD)
28 plusmn 16Range 0ndash8
Number of children under five years (119899[])
0 5 (10)1 22 (44)2 19 (38)3 4 (8)
Highest education attained (119899 [])Attended up to
Primary school 27 (54)Secondary school 17 (34)Tertiary schooling 2 (4)Never attended school 4 (8)
Literacy (119899 [])Read sample text completely 31 (62)Read sample text partially 7 (14)Unable to read sample text 12 (24)
Able to write name legibly (119899 [])Yes 44 (88)No 6 (12)
Average monthly income (mean plusmn SD) 162250UGXRange 5000ndash1500000
3 Results
31 Demographic Characteristics A total of 50 women wereapproached and all consented to participate Participants hadamean age of 27 years and amean of 23 children Ten percentof the women had no children (550) and 8 (450) hadmore than two children under five years Overall twenty-two percent (1150) of women reported being currentlypregnant
Fifty-four percent (2750) of women had attendedonly primary school and 8 (450) had never attendedschool Seventy-two percent (3650) of women self-identifiedas farmers The average self-reported monthly incomewas 162250 Ugandan Shillings (UGX) (sim$6228) (range5000UGX to 1500000UGX) (Table 1)
4 Malaria Research and Treatment
32 Bednet Ownership and Use Ninety percent (4550) ofwomen reported that their household currently owned atleast one bednet and 80 (4050) reported that at leastone bednet in their household was in use the night beforethe survey Seventy-six percent (3850) of women reportedthat a bednet was in use every day in the previous weekby at least someone in their household Eighty-four per-cent (4250) reported that their householdrsquos ability to usebednets as directed was good or very good and 8 (450)reported their householdrsquos ability as poor or very poor Thehouseholds owned an average of 242 bednets (range 1 to 6)(Table 2)
Self-reported bednet use was high with 96 (4850) ofrespondents reporting that they had used a bednet at somepoint in the past and 66 (3350) reporting using a bednetevery night in the previous week Twenty percent (1050) didnot use a net at all in the last week and 14 (750) had notused a net in the last month Among women who had everused bednets greater than 80 reported that bednets wereeasy or very easy to use and easy or very easy to remember touse However 146 (748) of ever-users reported that theywere difficult to use and 104 (548) reported that they weredifficult to remember to use Reasons for not using bednetsincluded excessive heat while under the net traveling awayfrom home or not owning enough nets
33 Impressions of the SmartNet Device All participants(5050) said theywould be interested in using the SmartNet intheir homes In open-ended responses reasons that womengave included liking the appearance of the SmartNet think-ing the net would be useful and wanting to have a new netNinety-two percent (4650) reported that they thought thedevice would be easy to use (Table 3) Notably many of theresponses to the question ldquoare there particular things you likeabout the SmartNetrdquo pertained to nets in general and wereindependent of monitoring For instance many participantsreported that they would like new nets because their currentnets were old or had holes in them
When queried about how their behavior would change inresponse to using SmartNet 32 (1650) of women reportedthat the SmartNet would not change their net use Noparticipants stated that their net use would decrease Of the68 (3450) who reported an anticipated increase in net usereasons included the fact of beingmonitored being remindedby the device to use the net and the attractive appearance ofthe net
While the majority of responses to the SmartNet demon-stration were positive some participants offered suggestionsor negative impressions of the net Themes included fear ofthe device perceived dislike of being studied when use of bednet is not as directed and dislike of home-visits to check onthe net
34 Perceptions of Objective Monitoring of Bednet Use Themajority of participants did not express personal concernswith objective monitoring When participants were askedldquohow would you feel knowing that your bednet use isbeing monitored by the SmartNetrdquo ninety percent (4550)
Table 2 Bednet ownership and use
Household bednet characteristics 119899 = 50
Household bednet ownership (119899 []) 45 (90)Average number of bednets perhousehold (mean plusmn SD)
242 plusmn 112Range 1ndash6
Household bednet use the night before (119899[]) 40 (80)
Household net use by someone in theweek before (days) (119899 [])
7 38 (76)lt7 6 (12)0 5 (10)No response 1 (2)
Householdrsquos ability to use bednetscorrectly (119899 [])
Very good 19 (38)Good 23 (46)Fair 3 (6)Poor 2 (4)Very poor 2 (4)No response 1 (2)
Individual bednet characteristicsEver used bednet (119899 []) 48 (96)Personally used every night last week (119899[]) 33 (66)
Did not use in last week (119899 []) 10 (20)Did not use in last month (119899 []) 14 (28)Bednets are (119899 [])
Very easy to use 27 (54)Easy to use 13 (26)Neither easy nor hard to use 2 (4)Difficult to use 6 (12)Very difficult to use 1 (2)No response 1 (2)
Remembering to use a bednet is (119899 [])Very easy 30 (60)Easy 13 (26)Neither easy nor hard 1 (2)Difficult 5 (10)Very difficult 0 (0)No response 1 (2)
Agree that bednets are useful (119899 []) 48 (96)Reasons for not using a net (119899 = 25) (119899[])Too hot 8 (16)Traveled away from home 6 (12)Not enough nets 6 (12)Allergicdislike treatment 4 (8)Net damaged or being washed 2 (4)Spouse does not want to use it 1 (2)
of participants reported that they did not believe theirfamily members would be worried about being monitored bySmartNet (Table 4)
When they did mention concerns participants focusedon monitoring of behaviors including sleeping behaviorswhile in bed at night
Malaria Research and Treatment 5
Table 3 Impressions of SmartNet device
Interested in using (119899 []) 50 (100)Perceived to be easy to use (119899 []) 46 (92)
Positive impressionsTheme ExampleLooksfeels nice or attractive ldquoI love the yellow color and the designs on itrdquo
Looks easy to use ldquoI can hang it well on my walls because it looks easy to hang It is also light enoughto move withrdquo
Would increase usage ldquoI would be more accurate in putting [up] my netrdquo
Would help to learn how to use nets ldquoI want the researchers to monitor me and know if I use my net well or not whichcan help me improve in case I donrsquot use it wellrdquo
Help with malaria prevention in the community ldquoIf these nets are given out it would reduce on the malaria cases especially invillagesrdquo
Help learn about net use ldquoThey [researchers] will be able to find out how best people can use their netsrdquoNegative impressions
Theme Example
Appearance ldquoSome people may not like this color of the net since they have children who candirty itrdquo
Treatment ldquoIf it has a lot of chemicals it will disturb merdquoNet would be difficult to hang ldquoThe squared net might be difficult to some people in hangingrdquo
Device could be damaged ldquoIn case the device is not covered and it stays open people may spoil it especiallychildrenrdquo
People who misuse nets may not like the SmartNet ldquoFor those who donrsquot use the nets they may not like it when you come to check ontheir bednets as you monitor themrdquo
The device could negatively affect health ldquoThey may fear that it [SmartNet] might negatively affect their healthrdquo
Dislike of home visits ldquoSome people who are not hygienically good in their bedrooms may not like youentering their bedroomsrdquo
Not able to wash the net ldquoThey may fear washing the net if it is to be washedrdquoFear that the device may be stolen ldquoFearing that thieves may steal the device from themrdquo
Table 4 Impressions of objective monitoring of bednet use
Personally not worried about monitoring (119899[]) 45 (90)
Family would not be worried aboutmonitoring (119899 []) 43 (86)
Positive impressionsTheme ExampleMonitoring would improve use ldquoBecause I would want to get good results about the userdquo
Provides feedback ldquoIf I can be given feedback on how I use my net it can help me improve in case Idonrsquot use it wellrdquo
Benefits of participating in a research study ldquoIt would mean that the researchers always will remember to check on me and visitme at homerdquo
Negative impressionsTheme Example
Monitoring while sleeping ldquoPeople might think that they device will be monitoring them as well when they aresleeping in their beds under the netrdquo
Monitoring of misuse ldquoSome people may not like to be monitored especially those that at times do notuse their netsrdquo
Change current behaviors ldquoFor people who get discomfort with heat at night and they remove their nets theymay fear to do this because of being monitoredrdquo
Concern about continuous monitoring ldquoI would be thinking every time they are looking at me and measuring me while inmy bedrdquo
Concern about recording private behavior ldquoI would think that the device is recording me while in bed with my husbandrdquo
6 Malaria Research and Treatment
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
about SmartNet and electronic bednet monitoring Partici-pants were identified among patients waiting to see healthproviders at the clinic They were invited to participate inthe study after their clinic visit was complete A room wasset aside for confidential study visits Informed consent wasgathered from all participants before proceeding to the studyinterviews Since our study focused on those at most risk ofmalaria we recruitedwomen to participate in the study if theywere mothers of at least one child aged five years or youngeror were currently pregnant These recruitment criteria alsoallowed for selection of a study sample which is familiarwith bednets as our previous work suggests that householdswith young children are more likely to own and use bednets[28] Participants were informed that their participation inthis study would not affect future inclusion in SmartNet-related studies No incentive was provided for participationAll consents and surveys were translated into Runyankolethe local language All participants signed a written informedconsent to participate Participants were given the optionto complete the survey in either English or Runyankole Inorder to maintain consistency of interpretation of qualitativeresponses a single research assistant (Nuriat Nambogo) whois fluent in Runyankole and English conducted all interviews
Participants received a structured interview to collectdata on age marital status number of children pregnancystatus education and literacy employment and averagemonthly income Then they answered a questionnaire relat-ing to their prior experience with antimalarial bednetscurrent use of bednets frequency of household bednet useand their opinions about the ease of use of bednets Ourprior work suggests that knowledge about malaria is asignificant indicator of bednet use within a household [29]All participants were shown a SmartNet which was hangingin the interview room and its components and functionalitywere described before participants were asked SmartNet-specific qualitative questions
Participants answered open-ended questions about theirimpressions of SmartNet their ldquolikesrdquo and ldquodislikesrdquo aboutthe device individual concerns and anticipated concernsabout the device from the viewpoint of others within theirhousehold This final line of questioning from the viewpointof others in the household provided a level of abstraction inorder for participants to provide honest comments withoutviolating social norms The responses were written on thesurvey sheet for later translation and coding Themes werederived from a sample of qualitative responses to eachquestion and then each survey was coded independentlyby two members of the research team (Jeffrey I CampbellandNuriat Nambogo) Researchers discussed inconsistenciesin coding and when inconsistencies existed a final codewas established by consensus between the two coders or theprimary author if consensus could not be reached
All data were collected on paper forms and entered man-ually into the Research Electronic Data Capture (REDCap)tool hosted at Massachusetts General Hospital Boston MAUSA [30] The data were cleaned following the completionof data entry Quantitative data were analyzed in Stata 10(Statacorp College Rd TX) and qualitative responses werereviewed and coded in Microsoft Excel
Table 1 Demographic characteristics
Total number of respondents 119899 = 50
Age (mean plusmn SD)27 plusmn 56
Range 17ndash40Pregnant (119899 []) 11 (22)Gestational months of pregnancy (119899 = 11)(mean plusmn SD)
57 plusmn 21Range 3ndash9
Number of children under 5 years of age(mean plusmn SD)
28 plusmn 16Range 0ndash8
Number of children under five years (119899[])
0 5 (10)1 22 (44)2 19 (38)3 4 (8)
Highest education attained (119899 [])Attended up to
Primary school 27 (54)Secondary school 17 (34)Tertiary schooling 2 (4)Never attended school 4 (8)
Literacy (119899 [])Read sample text completely 31 (62)Read sample text partially 7 (14)Unable to read sample text 12 (24)
Able to write name legibly (119899 [])Yes 44 (88)No 6 (12)
Average monthly income (mean plusmn SD) 162250UGXRange 5000ndash1500000
3 Results
31 Demographic Characteristics A total of 50 women wereapproached and all consented to participate Participants hadamean age of 27 years and amean of 23 children Ten percentof the women had no children (550) and 8 (450) hadmore than two children under five years Overall twenty-two percent (1150) of women reported being currentlypregnant
Fifty-four percent (2750) of women had attendedonly primary school and 8 (450) had never attendedschool Seventy-two percent (3650) of women self-identifiedas farmers The average self-reported monthly incomewas 162250 Ugandan Shillings (UGX) (sim$6228) (range5000UGX to 1500000UGX) (Table 1)
4 Malaria Research and Treatment
32 Bednet Ownership and Use Ninety percent (4550) ofwomen reported that their household currently owned atleast one bednet and 80 (4050) reported that at leastone bednet in their household was in use the night beforethe survey Seventy-six percent (3850) of women reportedthat a bednet was in use every day in the previous weekby at least someone in their household Eighty-four per-cent (4250) reported that their householdrsquos ability to usebednets as directed was good or very good and 8 (450)reported their householdrsquos ability as poor or very poor Thehouseholds owned an average of 242 bednets (range 1 to 6)(Table 2)
Self-reported bednet use was high with 96 (4850) ofrespondents reporting that they had used a bednet at somepoint in the past and 66 (3350) reporting using a bednetevery night in the previous week Twenty percent (1050) didnot use a net at all in the last week and 14 (750) had notused a net in the last month Among women who had everused bednets greater than 80 reported that bednets wereeasy or very easy to use and easy or very easy to remember touse However 146 (748) of ever-users reported that theywere difficult to use and 104 (548) reported that they weredifficult to remember to use Reasons for not using bednetsincluded excessive heat while under the net traveling awayfrom home or not owning enough nets
33 Impressions of the SmartNet Device All participants(5050) said theywould be interested in using the SmartNet intheir homes In open-ended responses reasons that womengave included liking the appearance of the SmartNet think-ing the net would be useful and wanting to have a new netNinety-two percent (4650) reported that they thought thedevice would be easy to use (Table 3) Notably many of theresponses to the question ldquoare there particular things you likeabout the SmartNetrdquo pertained to nets in general and wereindependent of monitoring For instance many participantsreported that they would like new nets because their currentnets were old or had holes in them
When queried about how their behavior would change inresponse to using SmartNet 32 (1650) of women reportedthat the SmartNet would not change their net use Noparticipants stated that their net use would decrease Of the68 (3450) who reported an anticipated increase in net usereasons included the fact of beingmonitored being remindedby the device to use the net and the attractive appearance ofthe net
While the majority of responses to the SmartNet demon-stration were positive some participants offered suggestionsor negative impressions of the net Themes included fear ofthe device perceived dislike of being studied when use of bednet is not as directed and dislike of home-visits to check onthe net
34 Perceptions of Objective Monitoring of Bednet Use Themajority of participants did not express personal concernswith objective monitoring When participants were askedldquohow would you feel knowing that your bednet use isbeing monitored by the SmartNetrdquo ninety percent (4550)
Table 2 Bednet ownership and use
Household bednet characteristics 119899 = 50
Household bednet ownership (119899 []) 45 (90)Average number of bednets perhousehold (mean plusmn SD)
242 plusmn 112Range 1ndash6
Household bednet use the night before (119899[]) 40 (80)
Household net use by someone in theweek before (days) (119899 [])
7 38 (76)lt7 6 (12)0 5 (10)No response 1 (2)
Householdrsquos ability to use bednetscorrectly (119899 [])
Very good 19 (38)Good 23 (46)Fair 3 (6)Poor 2 (4)Very poor 2 (4)No response 1 (2)
Individual bednet characteristicsEver used bednet (119899 []) 48 (96)Personally used every night last week (119899[]) 33 (66)
Did not use in last week (119899 []) 10 (20)Did not use in last month (119899 []) 14 (28)Bednets are (119899 [])
Very easy to use 27 (54)Easy to use 13 (26)Neither easy nor hard to use 2 (4)Difficult to use 6 (12)Very difficult to use 1 (2)No response 1 (2)
Remembering to use a bednet is (119899 [])Very easy 30 (60)Easy 13 (26)Neither easy nor hard 1 (2)Difficult 5 (10)Very difficult 0 (0)No response 1 (2)
Agree that bednets are useful (119899 []) 48 (96)Reasons for not using a net (119899 = 25) (119899[])Too hot 8 (16)Traveled away from home 6 (12)Not enough nets 6 (12)Allergicdislike treatment 4 (8)Net damaged or being washed 2 (4)Spouse does not want to use it 1 (2)
of participants reported that they did not believe theirfamily members would be worried about being monitored bySmartNet (Table 4)
When they did mention concerns participants focusedon monitoring of behaviors including sleeping behaviorswhile in bed at night
Malaria Research and Treatment 5
Table 3 Impressions of SmartNet device
Interested in using (119899 []) 50 (100)Perceived to be easy to use (119899 []) 46 (92)
Positive impressionsTheme ExampleLooksfeels nice or attractive ldquoI love the yellow color and the designs on itrdquo
Looks easy to use ldquoI can hang it well on my walls because it looks easy to hang It is also light enoughto move withrdquo
Would increase usage ldquoI would be more accurate in putting [up] my netrdquo
Would help to learn how to use nets ldquoI want the researchers to monitor me and know if I use my net well or not whichcan help me improve in case I donrsquot use it wellrdquo
Help with malaria prevention in the community ldquoIf these nets are given out it would reduce on the malaria cases especially invillagesrdquo
Help learn about net use ldquoThey [researchers] will be able to find out how best people can use their netsrdquoNegative impressions
Theme Example
Appearance ldquoSome people may not like this color of the net since they have children who candirty itrdquo
Treatment ldquoIf it has a lot of chemicals it will disturb merdquoNet would be difficult to hang ldquoThe squared net might be difficult to some people in hangingrdquo
Device could be damaged ldquoIn case the device is not covered and it stays open people may spoil it especiallychildrenrdquo
People who misuse nets may not like the SmartNet ldquoFor those who donrsquot use the nets they may not like it when you come to check ontheir bednets as you monitor themrdquo
The device could negatively affect health ldquoThey may fear that it [SmartNet] might negatively affect their healthrdquo
Dislike of home visits ldquoSome people who are not hygienically good in their bedrooms may not like youentering their bedroomsrdquo
Not able to wash the net ldquoThey may fear washing the net if it is to be washedrdquoFear that the device may be stolen ldquoFearing that thieves may steal the device from themrdquo
Table 4 Impressions of objective monitoring of bednet use
Personally not worried about monitoring (119899[]) 45 (90)
Family would not be worried aboutmonitoring (119899 []) 43 (86)
Positive impressionsTheme ExampleMonitoring would improve use ldquoBecause I would want to get good results about the userdquo
Provides feedback ldquoIf I can be given feedback on how I use my net it can help me improve in case Idonrsquot use it wellrdquo
Benefits of participating in a research study ldquoIt would mean that the researchers always will remember to check on me and visitme at homerdquo
Negative impressionsTheme Example
Monitoring while sleeping ldquoPeople might think that they device will be monitoring them as well when they aresleeping in their beds under the netrdquo
Monitoring of misuse ldquoSome people may not like to be monitored especially those that at times do notuse their netsrdquo
Change current behaviors ldquoFor people who get discomfort with heat at night and they remove their nets theymay fear to do this because of being monitoredrdquo
Concern about continuous monitoring ldquoI would be thinking every time they are looking at me and measuring me while inmy bedrdquo
Concern about recording private behavior ldquoI would think that the device is recording me while in bed with my husbandrdquo
6 Malaria Research and Treatment
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
32 Bednet Ownership and Use Ninety percent (4550) ofwomen reported that their household currently owned atleast one bednet and 80 (4050) reported that at leastone bednet in their household was in use the night beforethe survey Seventy-six percent (3850) of women reportedthat a bednet was in use every day in the previous weekby at least someone in their household Eighty-four per-cent (4250) reported that their householdrsquos ability to usebednets as directed was good or very good and 8 (450)reported their householdrsquos ability as poor or very poor Thehouseholds owned an average of 242 bednets (range 1 to 6)(Table 2)
Self-reported bednet use was high with 96 (4850) ofrespondents reporting that they had used a bednet at somepoint in the past and 66 (3350) reporting using a bednetevery night in the previous week Twenty percent (1050) didnot use a net at all in the last week and 14 (750) had notused a net in the last month Among women who had everused bednets greater than 80 reported that bednets wereeasy or very easy to use and easy or very easy to remember touse However 146 (748) of ever-users reported that theywere difficult to use and 104 (548) reported that they weredifficult to remember to use Reasons for not using bednetsincluded excessive heat while under the net traveling awayfrom home or not owning enough nets
33 Impressions of the SmartNet Device All participants(5050) said theywould be interested in using the SmartNet intheir homes In open-ended responses reasons that womengave included liking the appearance of the SmartNet think-ing the net would be useful and wanting to have a new netNinety-two percent (4650) reported that they thought thedevice would be easy to use (Table 3) Notably many of theresponses to the question ldquoare there particular things you likeabout the SmartNetrdquo pertained to nets in general and wereindependent of monitoring For instance many participantsreported that they would like new nets because their currentnets were old or had holes in them
When queried about how their behavior would change inresponse to using SmartNet 32 (1650) of women reportedthat the SmartNet would not change their net use Noparticipants stated that their net use would decrease Of the68 (3450) who reported an anticipated increase in net usereasons included the fact of beingmonitored being remindedby the device to use the net and the attractive appearance ofthe net
While the majority of responses to the SmartNet demon-stration were positive some participants offered suggestionsor negative impressions of the net Themes included fear ofthe device perceived dislike of being studied when use of bednet is not as directed and dislike of home-visits to check onthe net
34 Perceptions of Objective Monitoring of Bednet Use Themajority of participants did not express personal concernswith objective monitoring When participants were askedldquohow would you feel knowing that your bednet use isbeing monitored by the SmartNetrdquo ninety percent (4550)
Table 2 Bednet ownership and use
Household bednet characteristics 119899 = 50
Household bednet ownership (119899 []) 45 (90)Average number of bednets perhousehold (mean plusmn SD)
242 plusmn 112Range 1ndash6
Household bednet use the night before (119899[]) 40 (80)
Household net use by someone in theweek before (days) (119899 [])
7 38 (76)lt7 6 (12)0 5 (10)No response 1 (2)
Householdrsquos ability to use bednetscorrectly (119899 [])
Very good 19 (38)Good 23 (46)Fair 3 (6)Poor 2 (4)Very poor 2 (4)No response 1 (2)
Individual bednet characteristicsEver used bednet (119899 []) 48 (96)Personally used every night last week (119899[]) 33 (66)
Did not use in last week (119899 []) 10 (20)Did not use in last month (119899 []) 14 (28)Bednets are (119899 [])
Very easy to use 27 (54)Easy to use 13 (26)Neither easy nor hard to use 2 (4)Difficult to use 6 (12)Very difficult to use 1 (2)No response 1 (2)
Remembering to use a bednet is (119899 [])Very easy 30 (60)Easy 13 (26)Neither easy nor hard 1 (2)Difficult 5 (10)Very difficult 0 (0)No response 1 (2)
Agree that bednets are useful (119899 []) 48 (96)Reasons for not using a net (119899 = 25) (119899[])Too hot 8 (16)Traveled away from home 6 (12)Not enough nets 6 (12)Allergicdislike treatment 4 (8)Net damaged or being washed 2 (4)Spouse does not want to use it 1 (2)
of participants reported that they did not believe theirfamily members would be worried about being monitored bySmartNet (Table 4)
When they did mention concerns participants focusedon monitoring of behaviors including sleeping behaviorswhile in bed at night
Malaria Research and Treatment 5
Table 3 Impressions of SmartNet device
Interested in using (119899 []) 50 (100)Perceived to be easy to use (119899 []) 46 (92)
Positive impressionsTheme ExampleLooksfeels nice or attractive ldquoI love the yellow color and the designs on itrdquo
Looks easy to use ldquoI can hang it well on my walls because it looks easy to hang It is also light enoughto move withrdquo
Would increase usage ldquoI would be more accurate in putting [up] my netrdquo
Would help to learn how to use nets ldquoI want the researchers to monitor me and know if I use my net well or not whichcan help me improve in case I donrsquot use it wellrdquo
Help with malaria prevention in the community ldquoIf these nets are given out it would reduce on the malaria cases especially invillagesrdquo
Help learn about net use ldquoThey [researchers] will be able to find out how best people can use their netsrdquoNegative impressions
Theme Example
Appearance ldquoSome people may not like this color of the net since they have children who candirty itrdquo
Treatment ldquoIf it has a lot of chemicals it will disturb merdquoNet would be difficult to hang ldquoThe squared net might be difficult to some people in hangingrdquo
Device could be damaged ldquoIn case the device is not covered and it stays open people may spoil it especiallychildrenrdquo
People who misuse nets may not like the SmartNet ldquoFor those who donrsquot use the nets they may not like it when you come to check ontheir bednets as you monitor themrdquo
The device could negatively affect health ldquoThey may fear that it [SmartNet] might negatively affect their healthrdquo
Dislike of home visits ldquoSome people who are not hygienically good in their bedrooms may not like youentering their bedroomsrdquo
Not able to wash the net ldquoThey may fear washing the net if it is to be washedrdquoFear that the device may be stolen ldquoFearing that thieves may steal the device from themrdquo
Table 4 Impressions of objective monitoring of bednet use
Personally not worried about monitoring (119899[]) 45 (90)
Family would not be worried aboutmonitoring (119899 []) 43 (86)
Positive impressionsTheme ExampleMonitoring would improve use ldquoBecause I would want to get good results about the userdquo
Provides feedback ldquoIf I can be given feedback on how I use my net it can help me improve in case Idonrsquot use it wellrdquo
Benefits of participating in a research study ldquoIt would mean that the researchers always will remember to check on me and visitme at homerdquo
Negative impressionsTheme Example
Monitoring while sleeping ldquoPeople might think that they device will be monitoring them as well when they aresleeping in their beds under the netrdquo
Monitoring of misuse ldquoSome people may not like to be monitored especially those that at times do notuse their netsrdquo
Change current behaviors ldquoFor people who get discomfort with heat at night and they remove their nets theymay fear to do this because of being monitoredrdquo
Concern about continuous monitoring ldquoI would be thinking every time they are looking at me and measuring me while inmy bedrdquo
Concern about recording private behavior ldquoI would think that the device is recording me while in bed with my husbandrdquo
6 Malaria Research and Treatment
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
Interested in using (119899 []) 50 (100)Perceived to be easy to use (119899 []) 46 (92)
Positive impressionsTheme ExampleLooksfeels nice or attractive ldquoI love the yellow color and the designs on itrdquo
Looks easy to use ldquoI can hang it well on my walls because it looks easy to hang It is also light enoughto move withrdquo
Would increase usage ldquoI would be more accurate in putting [up] my netrdquo
Would help to learn how to use nets ldquoI want the researchers to monitor me and know if I use my net well or not whichcan help me improve in case I donrsquot use it wellrdquo
Help with malaria prevention in the community ldquoIf these nets are given out it would reduce on the malaria cases especially invillagesrdquo
Help learn about net use ldquoThey [researchers] will be able to find out how best people can use their netsrdquoNegative impressions
Theme Example
Appearance ldquoSome people may not like this color of the net since they have children who candirty itrdquo
Treatment ldquoIf it has a lot of chemicals it will disturb merdquoNet would be difficult to hang ldquoThe squared net might be difficult to some people in hangingrdquo
Device could be damaged ldquoIn case the device is not covered and it stays open people may spoil it especiallychildrenrdquo
People who misuse nets may not like the SmartNet ldquoFor those who donrsquot use the nets they may not like it when you come to check ontheir bednets as you monitor themrdquo
The device could negatively affect health ldquoThey may fear that it [SmartNet] might negatively affect their healthrdquo
Dislike of home visits ldquoSome people who are not hygienically good in their bedrooms may not like youentering their bedroomsrdquo
Not able to wash the net ldquoThey may fear washing the net if it is to be washedrdquoFear that the device may be stolen ldquoFearing that thieves may steal the device from themrdquo
Table 4 Impressions of objective monitoring of bednet use
Personally not worried about monitoring (119899[]) 45 (90)
Family would not be worried aboutmonitoring (119899 []) 43 (86)
Positive impressionsTheme ExampleMonitoring would improve use ldquoBecause I would want to get good results about the userdquo
Provides feedback ldquoIf I can be given feedback on how I use my net it can help me improve in case Idonrsquot use it wellrdquo
Benefits of participating in a research study ldquoIt would mean that the researchers always will remember to check on me and visitme at homerdquo
Negative impressionsTheme Example
Monitoring while sleeping ldquoPeople might think that they device will be monitoring them as well when they aresleeping in their beds under the netrdquo
Monitoring of misuse ldquoSome people may not like to be monitored especially those that at times do notuse their netsrdquo
Change current behaviors ldquoFor people who get discomfort with heat at night and they remove their nets theymay fear to do this because of being monitoredrdquo
Concern about continuous monitoring ldquoI would be thinking every time they are looking at me and measuring me while inmy bedrdquo
Concern about recording private behavior ldquoI would think that the device is recording me while in bed with my husbandrdquo
6 Malaria Research and Treatment
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
ldquoI would think that the device is recording mewhile in bed with my husbandrdquo
ldquoI would be thinking every time they are lookingat me and measuring me while in my bedrdquo
More participants described how other people might beconcerned with monitoring compared to those who reportedbeing concerned themselves Some other concerns focusedaround fear of detecting net misuse
ldquoSome people may not like to be monitoredespecially those that at times do not use theirnetsrdquo
ldquoFor people who get discomfort with heat atnight and they remove their nets they may fearto do this because of being monitoredrdquo
Several participants reported that monitoring would helpthem to improve their net usage and said they wouldappreciate feedback that the net could provide For instanceone participant noted that the desire to ldquoget good resultsrdquowould help her to improve her net use A few participantsalso mentioned the perceived benefits of participating ina research study involving SmartNet including frequentinteraction with study staff
4 Discussion
In this mixed-method survey mothers and pregnant womenin rural Uganda who were well acquainted with bednetsexpressed favorable views regarding using SmartNet forobjective monitoring of their householdsrsquo bednet use Whilea minority of participants expressed concern about beingmonitored at night while they slept all participants werenevertheless willing to use a SmartNet in their home
SmartNet was developed out of a concern that self-reported use of bednets is inaccurate and temporally impre-cise Most studies of bednet use rely on self-reported useWhile some studies have relied on more objective mea-surements to reduce bias the majority of these have beenone-time assessments with visual observation of hangingbednets or individuals under bednets Since effective malariaprevention requires consistent correct bednet use to protectagainst mosquitoes every night one-time surveys of use areunable to capture a temporally complete adherence recordThe ideal adherence monitor for bednet use must capturenightly adherence data We are aware of only one other studythat has used an objective nightly adherence tool [21] Thisstudy found that reports of use the night before correlatedwell with objective use data fromamotion detector Howeverthere were significant discrepancies in reports about netuse a week prior and 2ndash4 weeks prior confirming thepresence of recall bias This study did not report on theviews of participants about the monitoring of their bednetuse
While multiple studies have used qualitative methods toexamine perceptions of malaria and bednet use throughoutSub-Saharan Africa [31 32] we are unaware of any study
which has made as its primary aim the exploration ofstudy subjectsrsquo attitudes towards objective monitoring ofbednet use Perceptions of the monitoring itself are par-ticularly important when developing tools for electronicadherence monitoring in order to understand how a toolwill be used and to ensure that monitoring is culturallyacceptable
An important question is whether the monitoring itselfwill change behaviors This is the well-known ldquoHawthorneeffectrdquo whereby individuals modify their behavior merelyas a result of being monitored For instance women in thestudy reported that they thought the use of SmartNet wouldimprove their regular use of bednets in a desire to ldquogetgood resultsrdquo Determining how objective bednet monitoringinfluences bednet behavior is an important area of futureresearch
Data gathered in our study highlighted several areas fordesign modifications for SmartNet and for similar objec-tive bednet monitors In particular participants reporteda fear that the uncovered logging device for capturinguse data could be damaged Future improvements couldinclude better protection and concealment of the loggingdevice In addition some participants reported a dislikeof household visits to gather net use data Improvementsin battery life or potentially wireless transmission of datacould mitigate this concern Predeployment studies suchas this one are crucial components of a cocreation ethic[22] whereby real-time data about user impressions oftechnologies are fed back into design improvements Furtherrefinements of the SmartNet device are already underway inUganda at the CAMTech Innovation laboratory (httpwwwmassgeneralcenterforglobalhealthorgcamtech) offering anenhanced responsiveness to local views of the SmartNettechnology
There are several limitations of this study The partic-ipants did not actually use the SmartNet so these resultsare perceptions of participants based on theoretical useFurther design improvement and perceptions data will comefrom acceptability components of longer field studies amonghouseholds using SmartNet devices Because of the focus onbednet use to prevent maternal and child malaria men werenot interviewed and may hold different views of bednetsobjective monitoring of bednet use and SmartNet in par-ticular Future studies will explore this in more detail Theresponses to this predeployment attitudes survey could beprone to social desirability bias providing an overly positiveimpression of the device Women were told before joiningthis study that their answers would make no difference inultimately being chosen for use of a SmartNet in futurestudies Nevertheless some of the perceptions of SmartNetsuggest that women indicated interest in SmartNet as ameansof obtaining a free bednet perhaps inflating the apparentacceptability of the technology The study participants weredrawn from a convenience sample of women visiting onehealth center in the community of interest and therefore thestudy results may suffer from selection bias related to thischoice Finally SmartNet measures whether or not a bednetis unfurled but does not measure who is under the bednet orwhether the bednet is in perfect use
Malaria Research and Treatment 7
5 Conclusion
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
Objective monitoring of ITN use appears to be acceptableamong women in rural Uganda Concerns about monitoringfocused on the appearance of the device and the idea of beingmonitored while sleepingThese results will inform SmartNetdevice improvements and set the stage for future field-basedfeasibility testing
Ethical Approval
The study was approved by the Mbarara University of Scienceand Technology and the Partners Human Research Commit-tee Institutional Review Boards
Consent
All participants signed a written informed consent to partic-ipate
Competing Interests
The authors declare no competing interests
Authorsrsquo Contributions
Paul J Krezanoski Data Santorino Jeffrey I Campbell andDavid R Bangsberg conceived the study Paul J KrezanoskiData Santorino Nuriat Nambogo and Jeffrey I Campbelldesigned the study protocol Paul J Krezanoski Data San-torino Jeffrey I Campbell and David R Bangsberg carriedout the analysis and interpretation of data Paul J Krezanoskidrafted the manuscript Data Santorino Nuriat NambogoJeffrey I Campbell and David R Bangsberg critically revisedthe manuscript for intellectual content All authors read andapproved the final manuscript Paul J Krezanoski is theguarantor of the paper
Acknowledgments
The SmartNet team wishes to thank the study participantsDr Robert Ssekimpi (the one in-charge of the KinoniHealth Center) and the Mbarara District Health Office forsupporting this study Additional thanks go to the studentvolunteers atMUSTCAMTech the local tailors instrumentalin preparing the SmartNet devices and Michael Yang forassistance in preparing the manuscript for publication Thiswork was supported through generous support from theThrasher Research Foundation Early Career Award (PaulJ Krezanoski) and the American Academy of PediatricsResident Research Grant (Paul J Krezanoski)
References
[1] World Health Organization World Malaria ReportWHO Geneva Switzerland 2013 httpwwwwhointirisbitstream106659700819789241564694 engpdf
[2] United Nations Millenium Development Goals CombatHIVAIDS Malaria and Other Diseases United Nations New
York NY USA 2014 httpwwwunorgmillenniumgoalsaidsshtml
[3] A D Flaxman N Fullman M W Otten Jr et al ldquoRapidscaling up of insecticide-treated bed net coverage in Africaand its relationship with development assistance for healtha systematic synthesis of supply distribution and householdsurvey datardquo PLoS Medicine vol 7 no 8 article e1000328 2010
[4] S S Lim N Fullman A Stokes et al ldquoNet benefits amulticountry analysis of observational data examining asso-ciations between insecticide-treated mosquito nets and healthoutcomesrdquo PLoS Medicine vol 8 no 9 Article ID e10010912011
[5] C Lengeler ldquoInsecticide-treated bed nets and curtains forpreventing malariardquo Cochrane Database Of Systematic Reviewsno 2 Article ID CD000363 2004
[6] World Health Organization WHO Recommendations forAchieving Universal Coverage with Long-Lasting InsecticidalNets for Malaria Control WHO Geneva Switzerland 2013httpwwwwhointmalariapublicationsatozwho recom-mendations universal coverage llinspdf
[7] J Pulford M W Hetzel M Bryant P M Siba and I MuellerldquoReported reasons for not using a mosquito net when one isavailable a review of the published literaturerdquoMalaria Journalvol 10 article 83 2011
[8] M Belay and W Deressa ldquoUse of insecticide treated nets bypregnant women and associated factors in a pre-dominantlyrural population in northern Ethiopiardquo Tropical Medicine ampInternational Health vol 13 no 10 pp 1303ndash1313 2008
[9] S D Fernando R R Abeyasinghe G N L Galappaththy NGunawardenaACRRanasinghe andLCRajapaksa ldquoSleep-ing arrangements under long-lasting impregnated mosquitonets differences during low and high malaria transmissionseasonsrdquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 103 no 12 pp 1204ndash1210 2009
[10] J Skarbinski J J Massaga A K Rowe and S P Kachur ldquoDis-tribution of free untreated bednets bundled with insecticide viaan integrated child health campaign in Lindi Region Tanzanialessons for future campaignsrdquo American Journal of TropicalMedicine and Hygiene vol 76 no 6 pp 1100ndash1106 2007
[11] M A Kulkarni J Vanden Eng R E Desrochers et alldquoContribution of integrated campaign distribution of long-lasting insecticidal nets to coverage of target groups andtotal populations in malaria-endemic areas in MadagascarrdquoAmerican Journal of Tropical Medicine and Hygiene vol 82 no3 pp 420ndash425 2010
[12] S I Becker-Dreps A K Biddle A Pettifor et al ldquoCost-effectiveness of adding bed net distribution for malaria pre-vention to antenatal services in Kinshasa Democratic Republicof the Congordquo The American Journal of Tropical Medicine andHygiene vol 81 no 3 pp 496ndash502 2009
[13] K Yohannes J M Dulhunty C Kourleoutov et al ldquoMalariacontrol in central Malaita Solomon Islands 1 The use ofinsecticide-impregnated bed netsrdquo Acta Tropica vol 75 no 2pp 173ndash183 2000
[14] J Skarbinski C A Winston J J Massaga S P Kachur and AK Rowe ldquoAssessing the validity of health facility-based data oninsecticide-treated bednet possession and use comparison ofdata collected via health facility and household surveysmdashLindiregion and Rufiji district Tanzania 2005rdquo Tropical Medicineand International Health vol 13 no 3 pp 396ndash405 2008
[15] E L Korenromp J Miller R E Cibulskis M K Cham DAlnwick and C Dye ldquoMonitoring mosquito net coverage for
8 Malaria Research and Treatment
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
malaria control in Africa possession vs use by children under5 yearsrdquo Tropical Medicine and International Health vol 8 no8 pp 693ndash703 2003
[16] J Wong M P Shah D Mwandama J E Gimnig K ALindblade and D P Mathanga ldquoHome visits to assess the reli-ability of caregiver-reported use of insecticide-treated bednetsby children in Machinga District Malawirdquo American Journal ofTropical Medicine and Hygiene vol 92 no 4 pp 825ndash827 2015
[17] H Iwashita G Dida K Futami et al ldquoSleeping arrangementand house structure affect bed net use in villages along LakeVictoriardquoMalaria Journal vol 9 no 1 article 176 2010
[18] J A Alaii W A Hawley M S Kolczak et al ldquoFactors affectinguse of permethrin-treated bed nets during a randomizedcontrolled trial in Western Kenyardquo The American Journal ofTropical Medicine and Hygiene vol 68 no 4 supplement pp137ndash141 2003
[19] C Frey C Traore M De Allegri B Kouyate and O MullerldquoCompliance of young children with ITN protection in ruralBurkina FasordquoMalaria Journal vol 5 article 70 2006
[20] P J Krezanoski A B Comfort and D H Hamer ldquoEffect ofincentives on insecticide-treated bed net use in sub-SaharanAfrica a cluster randomized trial in Madagascarrdquo MalariaJournal vol 9 no 1 article 186 2010
[21] G Fink and F Masiye ldquoAssessing the impact of scaling-up bednet coverage through agricultural loan programmesevidence from a cluster randomised controlled trial in KateteZambiardquo Transactions of the Royal Society of Tropical Medicineand Hygiene vol 106 no 11 pp 660ndash667 2012
[22] S Spencer A D Grant P Piola et al ldquoMalaria in campsfor internally-displaced persons in Uganda evaluation of aninsecticide-treated bednet distribution programmerdquo Transac-tions of the Royal Society of Tropical Medicine and Hygiene vol98 no 12 pp 719ndash727 2004
[23] S Gerstl S Dunkley A Mukhtar et al ldquoLong-lastinginsecticide-treated net usage in eastern Sierra Leonemdashthesuccess of free distributionrdquo Tropical Medicine amp InternationalHealth vol 15 no 4 pp 480ndash488 2010
[24] A Wolkon J L Vanden Eng K Morgah et al ldquoRapid scale-upof long-lasting insecticide-treated bed nets through integrationinto the national immunization program during child healthweek in Togo 2004rdquoThe American Journal of Tropical Medicineand Hygiene vol 83 no 5 pp 1014ndash1019 2010
[25] B G Koudou D Malone and J Hemingway ldquoThe use ofmotion detectors to estimate net usage by householders in rela-tion to mosquito density in central Cote drsquoIvoire preliminaryresultsrdquo Parasites and Vectors vol 7 no 1 article 96 2014
[26] T L Russell N J Govella S Azizi C J Drakeley S PKachur and G F Killeen ldquoIncreased proportions of outdoorfeeding among residual malaria vector populations followingincreased use of insecticide-treated nets in rural TanzaniardquoMalaria Journal vol 10 article 80 pp 1475ndash2875 2011
[27] A Caldwell A Young J Gomez-Marquez and K R OlsonldquoGlobal health technology 20rdquo IEEE Pulse vol 2 no 4 pp 63ndash67 2011
[28] P J Krezanoski A B Comfort A C Tsai and D R BangsbergldquoHouseholds with young children and use of freely distributedbednets in rural Madagascarrdquo International Health vol 6 no 1Article ID iht033 pp 29ndash34 2014
[29] P J Krezanoski A C Tsai D H Hamer A B Comfortand D R Bangsberg ldquoHousehold malaria knowledge and its
association with bednet ownership in settings without large-scale distribution programs evidence from rural MadagascarrdquoJournal of Global Health vol 4 no 1 Article ID 010401 2014
[30] P A Harris R Taylor R Thielke J Payne N Gonzalez andJ G Conde ldquoResearch electronic data capture (REDCap)-a metadata-driven methodology and workflow process forproviding translational research informatics supportrdquo Journalof Biomedical Informatics vol 42 no 2 pp 377ndash381 2009
[31] S P Kachur P A Phillips-Howard A M Odhacha T K Rue-bush A J Oloo and B L Nahlen ldquoMaintenance and sustaineduse of insecticide-treated bednets and curtains three yearsafter a controlled trial in western Kenyardquo Tropical Medicine ampInternational Health vol 4 no 11 pp 728ndash735 1999
[32] L P Toe O Skovmand K R Dabire et al ldquoDecreasedmotivation in the use of insecticide-treated nets in a malariaendemic area in Burkina Fasordquo Malaria Journal vol 8 no 1article 175 2009
