UBC Faculty Research and Publications

The effect of an interactive weekly mobile phone messaging on retention in prevention of mother to child… Awiti, Patricia O; Grotta, Alessandra; van der Kop, Mia; Dusabe, John; Thorson, Anna; Mwangi, Jonathan; Belloco, Rino; Lester, Richard; Ternent, Laura; Were, Edwin; Ekström, Anna M Jul 11, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12911_2016_Article_321.pdf [ 678.84kB ]
JSON: 52383-1.0366841.json
JSON-LD: 52383-1.0366841-ld.json
RDF/XML (Pretty): 52383-1.0366841-rdf.xml
RDF/JSON: 52383-1.0366841-rdf.json
Turtle: 52383-1.0366841-turtle.txt
N-Triples: 52383-1.0366841-rdf-ntriples.txt
Original Record: 52383-1.0366841-source.json
Full Text

Full Text

STUDY PROTOCOL Open AccessThe effect of an interactive weeklymobile phone messaging on retentionin prevention of mother to childtransmission (PMTCT) of HIV program:study protocol for a randomizedcontrolled trial (WELTEL PMTCT)Patricia Opondo Awiti1*, Alessandra Grotta2, Mia van der Kop3,4,5, John Dusabe1, Anna Thorson1,Jonathan Mwangi1, Rino Belloco2, Richard Lester6, Laura Ternent7, Edwin Were8 and Anna Mia Ekström1AbstractBackground: Improving retention in prevention of mother to child transmission (PMTCT) of HIV programs iscritical to optimize maternal and infant health outcomes, especially now that lifelong treatment is immediateregardless of CD4 cell count). The WelTel strategy of using weekly short message service (SMS) to engage patients incare in Kenya, where mobile coverage even in poor areas is widespread has been shown to improve adherence toantiretroviral therapy (ART) and viral load suppression among those on ART. The aim of this study is to determine theeffect of the WelTel SMS intervention compared to standard care on retention in PMTCT program in Kenya.Methods: WelTel PMTCT is a four to seven-centers, two-arm open randomized controlled trial (RCT) that will beconducted in urban and rural Kenya. Over 36 months, we plan to recruit 600 pregnant women at their first antenatalcare visit and follow the mother-infant pair until they are discharged from the PMTCT program (when infant is aged24 months). Participants will be randomly allocated to the intervention or control arm (standard care) at a 1:1 ratio.Intervention arm participants will receive an interactive weekly SMS ‘How are you?’ to which they are supposed torespond within 24 h. Depending on the response (ok, problem or no answer), a PMTCT nurse will follow-up and triageany problems that are identified.The primary outcome will be retention in care defined as the proportion of mother-infant pairs coming for infant HIVtesting at 24 months from delivery. Secondary outcomes include a) adherence to WelTel; (b) adherence to antiretroviralmedicine; (c) acceptance of WelTel and (d) cost-effectiveness of the WelTel intervention.Discussion: This trial will provide evidence on the effectiveness of mHealth for PMTCT retention. Trial results and thecost-effectiveness evaluation will be used to inform policy and potential scale-up of mHealth among mothers livingwith HIV.Trial registration: ISRCTN98818734; registered on 9th December 2014Keywords: Mobile health (m-health), Retention, HIV/PMTCT, Antiretroviral therapy (ART), Kenya* Correspondence: Patricia.Awiti@ki.se1Department of Public Health Sciences, Karolinska Institutet, 171 77Stockholm, SwedenFull list of author information is available at the end of the article© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Awiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 DOI 10.1186/s12911-016-0321-4BackgroundHIV and maternal and newborn health in KenyaIdentification and rollout of the most cost-effective in-terventions to eliminate HIV infections and AIDS deathamong women and children is a key priority for Kenya.According to the AIDS Response Progress Report 2014,the number of people living with HIV (PLHIV) in Kenyahas increased by 15 % over the last 4 years, reaching 1.6million cases in 2013, with an estimated incidence of100000 new HIV infections in 2014 [1]. Women representabout 57 % of all PLHIV in Kenya and close to 1 %(101,000) of all Kenyan children are living with HIV, cor-responding to 10 % of all PLHIV [1]. As many as 11 000Kenyan children were estimated to be newly infected in2013, mainly through mother to child transmission(MTCT). Thus, Kenya’s commitment to eliminateMTCT of HIV by 2015 will not be achieved, but im-provements have been made. Over the last decade, thenumber of HIV-infected pregnant women in need forPMTCT in Kenya has also declined, albeit too slowly,from 98,000 in 2004 to 79,000 in 2013 [1]. Since 2013,Kenya has adopted PMTCT Option B+, the WorldHealth Organization guideline recommending all preg-nant women living with HIV receive immediate HIVtreatment for life regardless of immune defense (CD4count), a strategy called PMTCT Option B+ [2].Antiretrovirals (ARVs) during pregnancy, delivery andbreastfeeding, and for the infant 6 weeks post-deliverycan reduce the risk of transmission from 35 % to <2 %in low-income countries [3]. Key strategies to finallyeliminate MTCT include increased knowledge ofPMTCT, increased involvement from the male partner,universal attendance of antenatal care (ANC) by preg-nant women, universal testing of pregnant women forHIV and provision of ARVs from early pregnancythroughout the breastfeeding period, and facility delivery[4]. Kenya’s progress on these goals is uneven. While theproportion of pregnant women tested for HIV has in-creased from 68 to 92 % in the last 5 years [1], only 5 %of male partners accompanied their pregnant partner toANC [1]. PMTCT coverage (the number of pregnantwomen living with HIV started on ARVs before delivery)declined in Kenya from 86 % in 2010 to 73 % in 2013,i.e. 58,000 out of 79,000 pregnant women living withHIV were offered PMTCT services [1]. This was partlydue to the multiple challenges of implementing OptionB+, which is much more resource demanding.Only 50 % in need of PMTCT were given ARVswithin 6 weeks of their HIV diagnosis [1], a servicedelay that may cause avoidable MTCT. Only 45 % ofall HIV-exposed infants were tested for HIV, i.e. themajority of children were lost to follow up leading topreventable child deaths [1]. Because of these short-comings, the proportion of HIV-exposed children whobecame HIV-infected has halted at 14 % in the last3 years. If Kenya is to achieve the global target ofeliminating MTCT of HIV, initially aimed for 2015,this trend needs to be reversed through new andmore effective interventions.Mobile phone use for information and health servicestrengthening in KenyaThe rapid expansion in mobile phone technology inAfrica has created new opportunities for informationsharing and service delivery where other infrastruc-ture, such as cable connectivity and constant electri-city supply is inadequate. In 2013, 82 % of Kenyanhouseholds owned a mobile phone [5], reaching100 % of 20–29 year-olds i.e. the age-period whenmost women give birth [6]. In fact, Kenya has theworld’s highest proportion of cell phone owners(80 %) who use mobile banking, and 60 % of Kenyansliving on less than $2,50 i.e. under the poverty line,have mobile phones [6].Mobile technology for health (mHealth) is increas-ingly being used to overcome shortcomings in infor-mation systems, laboratory equipment and humanresource capacity in low-income countries. Withinmaternal and child health care (MCH), mobile tech-nology has been used to link ANC with pregnantwomen and new mothers [7], to remind communityhealth workers in rural areas [8] and to submit sur-veillance reports on disease outbreaks and delivery ofservices [9]. Other uses of mHealth include registra-tion of records [10] and monitoring of drug procure-ments [11, 12].Several trials have highlighted the potential of mHealthto improve HIV services e.g. in adherence to ART, reten-tion in ART care, and even for receipt of laboratory HIVtest results [13–15]. A recent Cochrane review, whichincluded the WelTel RCT in Kenya found compellingevidence that weekly text messages to non-pregnantHIV infected patients are effective [16] in improvingART adherence.Whether a weekly interactive SMS intervention (theWelTel model) improves adherence and retention inPMTCT care and encourages life-long ARVs is un-known. The situation for pregnant women living withHIV in a country like Kenya is often highly complicated.Many women do not feel comfortable disclosing theirHIV status to a partner or family members in fear of be-ing stigmatized and socially isolated [17], which greatlyinfluences adherence to ART and PMTCT. Furthermore,most women are newly diagnosed with HIV during preg-nancy, often asymptomatic and have little time to adjustto the idea of living with HIV before they must start onARVs. Our previous research shows that competing bur-dens, including breadwinning responsibilities, stigma,Awiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 2 of 8feelings of guilt and fear of transmitting HIV to thebaby, fear of abandonment and violence from the malepartner also affects women’s capacity to adhere toPMTCT [18].The WelTel trial remains the only SMS adherenceintervention study with the lowest risk of bias ranking.The WelTel service has other advantages in that it islow cost since it uses SMS to check-in on patients, canwork with minimal literacy since problems can befollowed up with voice calls and extends into the popu-lation that do not own their own phones since sharedaccess is sufficient, and by being open-ended check-insallows almost any problem to be triaged. We will assessthe effectiveness, including evaluation of costs of weeklyinteractive SMS reminders to improve the retention ofpregnant women and mother-infant pairs in PMTCTcare in Kenya. Ultimately this will result in betterPMTCT coverage and reduced infant HIV infections.Research hypothesisThe weekly interactive mobile phone SMS (WelTel) isan effective as well as cost-effective method to improvethe retention of women living with HIV and their new-borns’ in PMTCT care (women living with HIV and theirHIV exposed infants who successfully complete the pro-gram when infant is aged 24 months).Study objectivesPrimary objectiveTo determine effectiveness of the WelTel SMS inter-vention on retention of women living with HIV andtheir newborns’ in PMTCT care in urban and ruralKenya.Secondary objectives1. To assess adherence to the WelTel SMSintervention among pregnant women and newlydelivered mothers living with HIV.2. To determine adherence to single componentsof PMTCT among pregnant women and newlydelivered mothers living with HIV (ARVs,facility-based delivery, early infant HIV testing andexclusive breastfeeding).3. To explore facilitators for and barriers to usingWelTel SMS in order to inform any improvementson the model for PMTCT among pregnant womenand newly delivered mothers living with HIV as wellas PMTCT staff.4. To evaluate costs from a payer’s perspective, of theWelTel SMS for retaining women living with HIVand HIV-exposed infants in clinical follow-up until24 months post-delivery (discharge from PMTCT).Methods/designTrial designThe WelTel PMCT study is a 4–7 center two-arm openrandomized controlled trial in which the intervention isallocated in a 1:1 ratio (Fig. 1 trial design).Study settingThe study is in western Kenya and involves 4–7 facilitiesthat are among over 192 facilities providing PMTCT ser-vices located in the catchment of Academic Model Provid-ing Access to Health Care (AMPATH) – a large HIVComprehensive Care Program run under the auspices ofMoi University School of Medicine, located in Eldoret.These facilities implement Option B+ regimen of PMTCT.The approximate coverage of mobile phones in westernKenya is around 78 % [19]. The research setting has beencarefully selected to represent urban and rural mixes thathave high antenatal HIV percentage prevalence (10–15 %),about twice that of the national prevalence of 6 %.Study populationThe study population will consist of: (i) pregnantwomen living with HIV aged 18 and over presentingat ANC for a first visit in the current pregnancy atFig. 1 WelTel PMTCT trial designAwiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 3 of 8the selected clinics; and (ii) newborns delivered tothese women living with HIV. Women who will bepregnant and will be diagnosed with HIV infectionwill be referred to a male or female research assistantto complete a checklist for eligibility. Determinationof HIV infection will be based on two repeated Deter-mine or Colloidal Gold tests for women newly diag-nosed during the current pregnancy, or, based onreferral from the comprehensive care clinic for thosewith known HIV infection and on antiretroviral ther-apy (ART) or pre-ART). Individuals must fulfill allthe inclusion criteria, provide consent to participateand complete an interviewer-administered question-naire before they are randomized to either the controland intervention groups.Inclusion criteria1. Women aged 18 years or above2. Evidence of pregnancy3. Evidence of HIV infection4. Resident of the PMTCT clinic catchment area andplans to remain residents from recruitment until24 months after delivery5. Willing to be followed-up from recruitment until24 months after delivery6. Owning a mobile phone or having access to amobile phone7. Able to text message in Kiswahili or have someonein close contact that they trust to read and respondto a text message8. Willing to receive text messages from the PMTCTclinic staff9. Able and willing to provide informed consentInterventionParticipants in the intervention group will register theirphone numbers in the WelTel system (online or viaSMS) and then receive a weekly short text messagequestion in Kiswahili “Mambo?” (Kiswahili for “How areyou?”) asking about their general wellbeing (Fig. 2). Themessage will be sent on a fixed day of the week and willallow the patient to respond within 24 h either that theyare well for example “ok” or “sawa” or that they have aproblem (for example “problem” or “shida”). A femalestudy coordinator will be in charge of centrally monitor-ing the WelTel SMS platform, which automaticallysends the messages and registers responses from the par-ticipants and categorizes them. All participants who re-spond “problem” or who do not respond will be directlylinked to a regular PMTCT nurse at the woman’s clinicto assist with identified problems. Problems that cannotbe immediately resolved by the nurse follow routine pro-cedures at the clinic and are normally referred to thePMTCT clinical officer at the respective facility who willthen decide if the patient needs to visit the facility orshould receive a follow up phone call. The study coord-inator will follow up with the respective PMTCT nursesto record action taken, which is entered directly into theWelTel platform logs as notes. Patients who will not re-spond to the SMS within 24 h will be traced (first bytelephone then at households) within the defaulterFig. 2 WelTel SMS interventionAwiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 4 of 8tracing outreach program in routine PMTCT care. Atenrollment, the participants will be informed that theweekly SMS support service does not replace routineclinic services, and that all appointments made byPMTCT staff should be honored and all emergenciesshould be handled by usual means. A WelTel SMS plat-form technician will handle all technical problems thatmay arise.OutcomesPrimary outcomeRetention in PMTCT care is defined as the proportionof women living with HIV and their HIV-exposed infantsthat remain in care until infants are aged 24 monthsmeasured from when the pair is enrolled in the programfrom the woman’s first visit at ANC until 24 monthsafter birth.Secondary outcomes1. Adherence to the WelTel SMS interventionmeasured as the proportion of women living withHIV who do not respond to the SMS within 7 days.2. Adherence to (i) ARVs measured as the proportionof women living with HIV who do not respond tothe SMS within 7 days when suspected to be out ofARVs (i.e. failure to pick up ARVs that can cover herabsence); (ii) facility-based delivery measured as theproportion of women who deliver in hospital;(iii) early infant HIV testing defined as the proportionof HIV-exposed infants who are tested for HIV within8 weeks of birth measured as HIV –exposed infantswith known HIV status at age 10weeks and (iii)exclusive breastfeeding defines as feeding theHIV-exposed infant only milk from the mother’sbreast.3. Participant perceived facilitators and barriers of theWelTel SMS intervention (assessed as perceivedreasons and challenges of use as well as suggestionsto possible solutions).4. Cost compared with (i) effectiveness (levels ofretention) from a payers’ perspective and (ii) qualityof life (differences in pregnant women’s QALYsbetween the two trial arms.Sample sizeThe primary outcome of the study is defined as the pro-portion of mother-newborn retained in PMTCT care at24 months; assuming i) a power of 80 %, ii) a two-sidedtest (alpha = 0.05), iii) and based on prior knowledge, aproportion retained in the control group of about 30 %,a sample size of 300 participants in each arm for a totalof 600 subjects, with 5 % dropout rate (i.e. women whodecide to withdraw from the study) in both the controland intervention arms was estimated to detect a 11 %difference (the smallest detectable difference) in theprimary outcome between the intervention and con-trol arms that is the difference that it would be im-portant to detect is 11 %, computed as theproportion retained in the intervention arm minusthe proportion retained in the control arm i.e. theproportion retained in the intervention group minusthe proportion retained in the control group. Calcula-tions were performed using Stata 14.1.RecruitmentA PMTCT nurse at the selected clinics will inform allconsecutive pregnant women identified as living withHIV at their first ANC visit about the study. ThePMTCT nurse will then refer these clients to the re-search assistant who will assess their eligibility andprovide detailed information about the study. Individ-uals who are eligible will be invited to participate inthe study and the research assistant will seek theirconsent.Participants in the intervention arm who own or haveaccess to mobile phones will be registered directly ontothe WelTel platform with their phone number.Randomization and allocationEligible and consenting patients will be randomized tothe intervention and control arms using a 1:1 allocationratio. To ensure balance between the arms throughoutthe trial, we adopted a permuted-block randomizationscheme. The block size will be concealed until the trialis over. Randomization will be performed separately ateach clinic. We will use opaque sealed envelopes to as-sign participants to the intervention and control arms.The randomization list was be generated at the Karo-linska Institutet (Stockholm, Sweden) by an independentstatistician.BaselineAn interviewer- facilitated baseline questionnaire will beadministered after recruitment and allocation. Questionsrecord information on participants’ social and demo-graphic characteristics; time of HIV diagnosis, time onARV, disclosure of HIV status, HIV care and social sup-port, mobile phone use as well as costs for accessingcare.Follow upFollow up visits will occur at 6 and 24 months postpar-tum, at which time research assistant will administer thefollow-up questionnaire. The follow up questionnairewill capture information on participants’ missed appoint-ments, engagement with health workers and satisfactionwith care, mobile phone access and using the WelTelAwiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 5 of 8intervention, and health related status/quality of lifeusing the EuroQol 5-dimensional (EQ-5D) utility scoresand twelve item short form survey (SF-12) standardizedtools.To investigate quality of life, we will first develop a pa-tient generated index (PGI) to identify areas in thewomen’s lives that are affected by their HIV infectionduring pregnancy and nursing period; periods when therisk of transmitting HIV is high. A PGI is an individual-ized patient reported instrument that allows the re-spondent to state, weight and rate areas of importanceto the patients’ lives that are affected by their illness.The PGI will also enable the investigators to assess thevalidity of the EQ5D and SF12 in this population andcontext.Data collection and managementAll outgoing and incoming text messages will be auto-matically recorded on the WelTel platform. The plat-form also captures all of the “problems” noted byparticipants, instances of non-response, and actionstaken in relation to participants’ ‘problem’ responses andnon-responses. Platform data will be backed up every7 days.All questionnaires will be paper-based and the datamanager will then enter data into a database at the cen-tral office on an ongoing basis. A data manager willcheck the forms for completeness and quality will beverified by re-checking a random sample of 10 % of thedata. Any problems that arise will be resolved promptly.Participant files will be stored in a locked office at thetrial sites.Data on attendance, HIV care clinical indicators likeviral load and CD4 cell count as well as treatment regi-men will be collected medical records. We will also col-lect information on demographic characteristics (age,education level, marital status and parity) of all screenedpatients/potential trial participants.Qualitative research using in –depth interviews willalso be performed to discover ‘how’ and ‘why’ the inter-vention works to improve retention in PMTCT program.Purposive sampling will be used to identify participantsfor qualitative interviews. Face-to-face interactions willbe used to build trust during the interview process andto enhance free interaction between the researcher andthe participants [20]. All conversations will be recordedwith permission from the respondents and the inter-views will be performed at a place and in a language pre-ferred by the respondent.All qualitative research tools will be developed inEnglish, then translated to Kiswahili, and then backtranslated to English. The interview guide will be avail-able in both languages.AnalysisStatistical methodsBaseline participants’ characteristics will be reported sep-arately by treatment arm. Baseline characteristics in-clude: age, parity (nulliparous versus previous birth),marital status (single, married or cohabiting, divorced/widowed), ARV exposure (experienced vs. naive), dur-ation of known HIV diagnosis (newly vs. previously diag-nosed), age (18–29, 30–39, 40–49, ≥50 years of age),phone ownership (owned vs. shared), level of education(none, primary, secondary, post-secondary), distancefrom clinic (≤1 h vs. >1 h), number of children bornafter HIV diagnosis, on ARV at enrolment (yes, no), timeon ARV at enrolment (≤6 months, 7–12 months,≥13 months) and HIV status disclosure (yes, no). Wewill report the mean (standard deviation [SD]) or me-dian (first quartile, third quartile) for continuous vari-ables, and count and percentages for categoricalvariables. All analyses are by intention-to-treat i.e.according to the study group to which women wereoriginally allocated regardless of subsequent interventionreceived and per protocol. Other more statisticalmethods will be used to take into account switching.For the primary outcome, we will compare theproportion of mothers living with HIV and their HIV-exposed infants in the program at 24 months post-birth in the intervention vs. control arm using bothparametric (Chi-Square) and exact (Fisher) statisticaltests. Secondary outcomes will also be comparedbetween arms, with t-tests for normally distributedvariables, Mann Whitney-U tests for non-normallydistributed variables, and Chi-Square and Fisher exacttests for categorical variables. The relative risk (RR)for PMTCT retention with 95 % confidence intervalswill be computed and the number needed to treat toprevent one non-retained mother-infant pair will alsobe estimated. Concerning the secondary outcomes,average treatment effects (ATE) will be computed forcontinuous outcomes and RRs for categorical out-comes. For both primary and secondary outcomes,log-linear or linear regression models will be used toprovide effect estimates adjusted for potential imbal-ances in baseline participants’ characteristics, if re-quired. We will repeat the analysis of primary andsecondary outcomes within such subgroups (in rela-tion to socio-demographics) to assess the homogen-eity of the intervention effect across pre-determinedsubgroups of patients. Stratified RRs and ATEs willbe computed. Regression models, which include theintervention allocation and subgroup-defining vari-ables and their interaction, will be applied to assesseffect modification across groups; all statistics testswill be run based on two side p-values and values<0.05 will be considered statistically significant. AllAwiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 6 of 8statistical analyses will be again performed with Stataversion 14.1 (Stata Corporation, College Station, TX,USA).Qualitative process evaluationQualitative data will be analyzed using content analysis,guided by Graneheim and Lundman [21]. First, the tran-scribed material is read a number of times to get a gen-eral sense of the material by a group of researchers.Using the open code software for qualitative research,meaning units, which are key phrases in the text, areidentified, condensed and outlined. Codes will then beascribed to each meaning unit. The codes will be com-pared and grouped into sub-categories. This comparisonwill be performed consistently to identify emerging cat-egories that will be further compared, re-organized andmerged into sub-themes and one overarching theme.The coding and analysis process will be deductive in na-ture and involve key members of the research team.Economic evaluationWe will perform an economic evaluation to assess Wel-Tel SMS from a healthcare payer perspective. The pri-mary outcome of the cost effective analysis (CEA) willbe the incremental cost per additional mother-infantpairs that remain in PMTCT until infant is aged24 months. A secondary outcome is averted infant HIV-infections at cessation of breastfeeding. For the cost-utility analysis the outcome will be quality adjusted lifeyears (QALYs), based on responses to the SF-12 andEQ5D. For both analyses (CEA and CUA) we will reportthe incremental cost-effectiveness ratios (ICERs) thatwill be computed as the ratio of the incremental costs toprovide WelTel SMS over usual care and incrementaleffects e.g. cost per additional mother-infant pairs thatremain in PMTCT until infant is aged 24 months, costaverted infant HIV-infections at cessation of breastfeed-ing and cost per QALY.A secondary analysis will also consider the incrementalcost per averted deaths by bringing infants lost to followup back and enabling treatment. Thus, we will deter-mine ICER for cost per averted deaths. Deterministicsensitivity analysis will be to address uncertainty.DiscussionThis trial provides an opportunity to test whether theWelTel SMS intervention is effective in improving re-tention in PMTCT care. Further, we will be able to de-termine whether the interactive WelTel text-messagingintervention, by engaging patients with the PMTCTclinic on a weekly basis, is a cost-effective way to im-prove retention in this critical stage of care, potentiallyhelping to eliminate pediatric HIV infections.AbbreviationsAIDS, acquired immune deficiency syndrome; ANC, antenatal care; ART,antiretroviral therapy; ARVs, antiretroviral drugs; CEA, cost effectivenessanalysis; CUA, cost utility analysis; HIV, human immunodeficiency virus;ICERs, incremental cost effectiveness ratios; IDIs, in-depth interviews; IREC,Institutional review and ethics committee; MCH, maternal and child health;MTCT, mother to child transmission; PIP, partners in prevention – MoiUniversity; PLHIV, people living with HIV; PMTCT, prevention of mother tochild transmission; QALYs, quality-adjusted life years; RCT, randomised clinicaltrial; RR, relative risk; SMS, short message service; UNAIDS, Joint UnitedNations programme on AIDS; WelTel SMS, Weekly mobile phone shortmessage serviceAcknowledgementsWe thank the patients who participated in pre-testing the data collectiontools, the staff at Partners in Prevention (PIP) Moi University for organizingfield logistics for the trial. Special thanks go to WelTel Kenya for providingreduced costs, software and technical support.FundingThe Swedish Research Council (Vetenskaprådet) supported the trial protocolreported in this publication. The funder had no role in the design of thestudy and will not have any role during its execution, analyses, interpretationof data, or decision to submit results. The content of this publication is solelythe responsibility of the authors and does not necessarily represent theofficial views of the Swedish Research Council.Availability of data and materialThere is no data shared in this protocol. However, any data collected will beavailable for others interested in the area of research.Authors’ contributionsPA is a co-investigator and leads the trial implementation. She is part of theteam that conceived the study; contributed to study design and also leadscoordinates fieldwork for the trial implementation. AT is part of the team thatconceived the study and also contributed in study design. AME is the leadinvestigator, conceived the study and also contributed to the study design.JD contributed to the study design and field implementation. M van der Kcontributed to study design and drafting of the manuscript. EW is aco-investigator and contributed to study design. RL contributed to the studydesign of the trial. LT contributed to the economic evaluation and studydesign. AG contributed to the statistical aspects and study design of thetrial. RB contributed to statistical aspects of the trial. JM contributed tostudy design of the trial. All authors contributed to the development studyprotocol and approved the final manuscript.Authors’ informationAnna Thorson, Anna Mia Ekström, Edwin Were, Richard Lester andJonathan Mwangi are clinicians working with the management of HIVinfections as well as public health issues of HIV. Patricia Awiti is asociologist and public health researcher working with issues of HIV andwomen and child health including reproductive health. Mia van derKop is an epidemiologist working with HIV issues among others.John Dusabe is an environmental scientists and public health researcher.Richard Lester, Mia van der Kop and John Dusabe are experienced indesigning and implementing mHealth interventions. Laura Ternent is aneconomist with experiences of performing costing of evaluations withinhealth sector. Alessandra Grotta and Rino Bellocco are statisticians andexperienced in performing clinical trials. Most authors have experiencesof working in low-income countries in Africa.Competing interestsRichard T Lester is the founder and paid consultant for the WelTelInternational mHealth Society, a non-profit non-governmental mHealthorganization with the goal of scaling up evidence-based mHealth solutionsand is funded in part by Grand Challenges Canada. For the remainingauthors no conflicts of interest were declared.Consent for publicationNot applicable.Awiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 7 of 8Ethics approval and consent to participateThe original study protocol, consent forms and data collection tools wereapproved by the Moi University Institutional Research and Ethics Committee(IREC 1292) in Kenya. Any modifications to the original trial protocol,including the data collection tools, will be submitted as amendmentsto the institutional review boards, and requisite approvals obtained.Ethical approval will be renewed annually.Ethical considerations and consent to participateStrict criteria for obtaining informed consent have been established forhealth behavior research. Informed consent will be sought in writing and allcontent pertaining to voluntary participation and description of the trial willbe presented in English and Kiswahili. Research team members are sensitiveto the fact that of potential participants may be newly diagnosed with HIVat the time of enrolment. They will give all eligible participants that arenewly diagnosed until their next clinic appointment to decide whetherto enroll in the study. The risk of breach of confidentiality resulting fromthe text messaging intervention will be minimized since the content ofthe text message will not include language related to HIV. In addition,the research phone number will not be displayed on the participant’scell phone when a text is received unless the participant actively installsa “true caller” software that reveals the number from which an SMS orcall originates.Each participant will be assigned a unique study ID (the outpatient numberwith additional serialization that denote participation codes), which will beused in all subsequent data collection and analyses. This ID number will notbe released to individuals outside of the core research team.HarmsIn this trial all expected adverse events include those directly involved withthe intervention such as unintended HIV infection disclosure and to a lesserextent disagreement relating to privacy in a relationship. Potential harms willbe discussed during the consent process and also outlined are channels fordealing with such events. A study coordinator will be in charge ofdocumenting all adverse events as per the requirements of the ethics reviewboard.DisseminationAll findings will be published in peer-reviewed journals. Additional disseminationwill occur in conferences, workshops and specified feedback sessions planned atmid and end of the trial targeting involved staff, participants and local andnational stakeholders including providers and policymakers.Author details1Department of Public Health Sciences, Karolinska Institutet, 171 77Stockholm, Sweden. 2Department of Medical Epidemiology and Statistics,Karolinska Institutet, 171 77 Stockholm, Sweden. 3Division of Infectiousdiseases, University of British Columbia, Vancouver, Canada. 4Division ofGlobal Health, University of British Columbia, Vancouver, Canada.5Department of Public Health Sciences, 171 77 Stockholm, Sweden.6Department of Medicine, University of British Columbia, Vancouver, Canada.7Newcastle University, Tyne and Wear NE1 7RU, UK. 8Department ofReproductive Health, Moi University, P.O.BOX 3900–30100, Eldoret, Kenya.Received: 13 January 2016 Accepted: 8 June 2016References1. Kenya National AIDS Control Council. Kenya AIDS response progressreport - progress towards zero. Nairobi, Kenya: Kenyan Ministry of Health;2014.2. World Health Organisation. Use of antiretroviral drugs for treating pregnantwomen and preventing HIV infection in infants. Geneva: WHO/HIV;2012. 2012.6.3. Joint United Nations Programme on HIV/AIDS UNICEF World HealthOrganization. Towards Universal Access: Scaling Up Priority HIV/AIDSInterventions in the Health Sector: Progress Report. Geneva, Switzerland:World Health Organization; 2007.4. World Health Organization. Consolidated guidelines on the use ofantiretroviral drugs for treating and preventing HIV infection. Geneva:World Health Organization; 2013.5. Pew Research Center. Emerging nations embrace internet, mobiletechnology. Accessible at: http://www.pewglobal.org/files/2014/02/Pew-Research-Center-Global-Attitudes-Project-Technology-Report-FINAL-February-13-20147.pdf [last Accessed 29 Sept 2015], 2014.6. Mas, I. and D. Radcliffe. Mobile payments go viral: M-PESA in Kenya.Washington DC: World Bank; 2010.7. Tamrat T, Kachnowski S. Special delivery: an analysis of mHealth in maternaland newborn health programs and their outcomes around the world.Matern Child Health J. 2012;16(5):1092–101.8. Shieshia M, Noel M, Andersson S, Felling B, Alva S, Agarwal S, et al.Strengthening community health supply chain performance through anintegrated approach: Using mHealth technology and multilevel teams inMalawi. J Glob Health. 2014;4(2):020406.9. Nsubuga P, White ME, Thacker SB, Anderson MA, Blount SB, Broome CV, etal. Public health surveillance: a tool for targeting and monitoringinterventions. Dis Control Priorities Dev Ctries. 2006;2:997–1018.10. Doctor HV, Olatunji A. Bridging the communication gap: Successes andchallenges of mobile phone technology in a health and demographicsurveillance system in northern Nigeria. Online journal of public healthinformatics. 2012; 4(3) doi: 10.5210/ojphi.v4i3.4288.11. Barrington J, Wereko-Brobby O, Ward P, Mwafongo W, Kungulwe S.SMS for Life: a pilot project to improve anti-malarial drug supplymanagement in rural Tanzania using standard technology. Malar J.2010;9(298):1–9.12. Githinji S, Kigen S, Memusi D, Nyandigisi A, Mbithi AM, Wamari A, et al.Reducing stock-outs of life saving malaria commodities using mobile phonetext-messaging: SMS for life study in Kenya. PLoS One. 2013;8(1):e54066.13. Lester RT, Ritvo P, Mills EJ, Kariri A, Karanja S, Chung MH, et al. Effects of amobile phone short message service on antiretroviral treatmentadherence in Kenya (WelTel Kenya1): a randomised trial. Lancet. 2010;376(9755):1838–45.14. Pop-Eleches C, Thirumurthy H, Habyarimana JP, Zivin JG, Goldstein MP,De Walque D, et al. Mobile phone technologies improve adherence toantiretroviral treatment in a resource-limited setting: a randomizedcontrolled trial of text message reminders. AIDS. 2011;25(6):825. London,England.15. Mbuagbaw L, van der Kop ML, Lester RT, Thirumurthy H, Pop-Eleches C,Ye C, et al. Mobile phone text messages for improving adherence toantiretroviral therapy (ART): an individual patient data meta-analysis ofrandomised trials. BMJ Open. 2013;3(12):e003950.16. Horvath, T, Azman H, Kennedy GE, Rutherford GW. Mobile phone textmessaging for promoting adherence to antiretroviral therapy in patientswith HIV infection. The Cochrane Library; 201217. Turan JM, Miller S, Bukusi E, Sande J, Cohen C. HIV/AIDS and maternity carein Kenya: how fears of stigma and discrimination affect uptake andprovision of labor and delivery services. AIDS Care. 2008;20(8):938–45.18. Ngarina M, Tarimo EA, Naburi H, Kilewo C, Mwanyika-Sando M, Chalamilla G,et al. Women’s Preferences Regarding Infant or Maternal AntiretroviralProphylaxis for Prevention of Mother-To-Child Transmission of HIV duringBreastfeeding and Their Views on Option B+ in Dar es Salaam, Tanzania.PLoS One. 2014;9(1):e85310.19. Kenya Ministry of Information Science and Technology. Nairobi, Kenya:Kenya ICT report; 2015.20. Richie J, Lewis J. Qualitative research practice. A guide for social sciencestudents and researchers. London: SAGE Publications; 2003.21. Graneheim UH, Lundman B. Qualitative content analysis in nursing research:concepts, procedures and measures to achieve trustworthiness. Nurse EducToday. 2004;24(2):105–12.Awiti et al. BMC Medical Informatics and Decision Making  (2016) 16:86 Page 8 of 8


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items