Final Report Summary - INTHEC (Health Education and Community Integration: Evidence based Strategies to increase equity, integration and effectiveness of reproductive health services for poor communities in Sub Saharan Africa.)
Executive Summary:
EXECUTIVE SUMMARY
IntHEC is a 4 year programme which aims to design and evaluate an adolescent reproductive health (ARH) programme in Niger and Tanzania. It is implemented by a consortium of 8 research, NGO and government ministry partners from 4 countries.
The specific activities were as follows:
1. Situation analysis: Conduct a situation analysis of current community and implementer experiences of existing ARH programmes in Tanzania and Niger, identifying priority areas of weakness in RH service provision and opportunities for strengthened service uptake and integration.
2. Intervention Development: Develop and implement an innovative package of pilot interventions that are feasible, equitable and appropriately designed to address identified weaknesses and opportunities.
3. Process evaluation: Document and evaluate the processes of these new interventions with special reference to gender and social equity.
4. Impact evaluation: Document and evaluate the impact of these new interventions and their effects on ARH provision, uptake, and effectiveness with special reference to gender and social equity.
5. Policy research and Advocacy: Support the development and implementation of feasible, effective, and equitable RH interventions in sub-Saharan Africa through effective collaboration between key research institutions, international authorities, government sectors and leading non-governmental organisations (NGOs).
In both Tanzania and Niger, discussion based intervention were developed and implemented. In Mwanza, interventions were supported by an SMS referral intervention.
Interventions were evaluated using a range of activities including baseline and follow up cluster randomized household surveys.
Overall evaluation found that the interventions were very popular at the community level and felt to be of value.
However, mystery client studies found overall that in health units, infrastructural limitations which were beyond the scope of the intervention continued to be a major bar to adolescent friendly service provision and somewhat overshadowed improvements arising from intervention activities.
Conversely, qualitative and quantitative evaluation found the SMS intervention was highly acceptable to both drug stores and their clients and did promote health facility attendance
A notebook assisted household survey was successfully conducted. Preliminary data suggest that the intervention did indeed improve uptake of VCT and STI services in Niger. Preliminary data suggest overall rates of STI in Tanzania appeared to be higher in comparison communities but further adjustment of the data is necessary to interpret the significance of this.
Project Context and Objectives:
PROJECT CONTEXT AND OBECTIVES
The project aims to improve the reproductive health of poor people in sub-Saharan Africa. The overall objective of the project is to improve the delivery of reproductive health (RH) services in Tanzania and Niger by successfully engaging policy-makers and programmers in the generation of new evidence about effective ways to strengthen the provision, uptake, equity and effectiveness of adolescent reproductive health (ARH) programmes.
The research was implemented in two regions in Niger (Say and Aguie) and Tanzania (Mwanza and Iringa). The national governments in both countries are implementing ARH activities that are being rolled out throughout each country. These interventions were implemented using strategies, and within contexts that are typical of ARH service provision elsewhere in sub-Saharan Africa. The results of our research are therefore likely to be applicable to several other countries.
The research integrated a community psychological approach into the Medical Research Council (MRC) best practice framework for developing complex interventions to develop the intervention package in three stages: (i) situation analysis; (ii) design of pilot interventions to address issued identified in situation analysis; (iii) small-scale pre-testing and formative evaluation of interventions. This aimed to ensure feasibility, acceptability and likelihood of effectiveness. These interventions were implemented in 18 intervention wards (9 in Iringa and 9 in Mwanza) for 12 months. Full scale intervention evaluation comprised process evaluation through a series of in-depth qualitative studies and impact evaluation through a population based community randomised trial. A combination of quantitative, qualitative and participatory methods was used to complete each stage of the research.
SPECIFIC OBJECTIVES
1. Conduct a situation analysis of current community and implementer experiences of existing ARH programmes in Tanzania and Niger, identifying priority areas of weakness in RH service provision and opportunities for strengthened service uptake and integration.
2. Develop and implement an innovative package of pilot interventions that are feasible, equitable and appropriately designed to address identified weaknesses and opportunities.
3. Document and evaluate the processes of these new interventions and their effects on ARH provision, uptake, and effectiveness with special reference to gender and social equity.
4. Support the development and implementation of feasible, effective, and equitable RH interventions in sub-Saharan Africa through effective collaboration between key research institutions, international authorities, government sectors and leading non-governmental organisations (NGOs).
Project Results:
DESCRIPTION OF THE MAIN SCIENTIFIC AND TECHNICAL RESULTS
The study was implemented in 18 wards in each of 2 Regions in each country. Within each Region 9 wards were randomised to receive the intervention and 9 wards randomised to comparison groups (a ward is an administrative unit that is roughly equivalent to 6-8 villages).
The overall study design is presented in Figure 1:
Figure 1- Overall study Design:
The study took place in 3 broad phases:
1. Phase 1- Situation analysis which included a literature review; baseline studies in schools, communities and health facilities and a household survey
2. Phase 2- Intervention development based on qualitative findings from the situation analysis and subsequent Intervention implementation
3. Phase 3- Evaluation comprising process evaluation of intervention implementation and evaluation of the intervention impact through the baseline and follow up household surveys.
Development and testing of the intervention and evaluation tools and strategies were tested in the Validity Villages, Formative wards and Pretest wards. The interventions were then implemented in randomly assigned intervention communities with pre and post impact evaluation through baseline and follow up household surveys. Detailed process evaluation was conducted in a subset of intervention and evaluation wards.
A BASELINE SITUATION ANALYSIS was conducted and included:
1. A scientific literature review on ARH was conducted in both Niger and Tanzania from March to April 2010 to identify relevant impact and process indicators. A review of international and national ARH laws and policies provided an overview of the range of policies relevant to ARH in the two countries. Linking with the ARH indicators review identified key performance standards for key actors within ARH service delivery which can be used as benchmarks for ARH services evaluation. The review also laid the groundwork for the “Policy engagement, advocacy and dissemination” work package (WP5).
2. Development of an Evaluation Framework. A theoretical evaluation framework was developed. Key concepts of research and intervention (for example activity systems, mediating moments) were agreed amongst all partners (during the first partners workshop in Liverpool and subsequent participatory research development workshops). The initial Evaluation Framework was further refined and the relationship between studies and analysis model clarified. See Fig 2.
Figure 2- Complementary baseline data collection studies:
The evaluation framework comprised five baseline studies including cluster randomised household surveys, close to community service provider consultations, policy reviews, mystery client studies and adolescent consultations.
In these studies we used qualitative, quantitative and action research techniques in health facilities, school and community settings. The objective of the qualitative baseline studies was to determine the range, quality and integration of Adolescent Reproductive Health (ARH) promotion and treatment interventions currently implemented within the health and education sectors and within the general community.
Key Baseline Qualitative Findings
A number of factors limiting ARH access to services were identified in both Tanzania and Niger as follows:
Customary norms and values have a strong and significant adverse influence on young people’s reproductive health seeking behaviour as well as the provision of health services. These community views are strongly shared by teachers and health providers and limit adolescents’ access to reproductive health services such that
• Adolescents are unable to express their ARH needs:
• Young people do not visit health centres because of disapproval from health providers
• There is an absence of dialogue between caregivers and adolescents, which inhibits STI/HIV/AIDS prevention.
• There is an absence of appropriate monitoring of adolescents’ RH needs and behaviours
In School Settings
• SRH related topics are taught in most schools, but do not adequately address pregnancy prevention, negotiation of safer sex, and how to deal with peer pressure or pressure for sex
• Teachers were willing, within the context of the research, to talk about their own reproductive health needs and how they think SRH can be enforced within the teaching community, as opposed to among the pupils.
Integration
• Teachers had a view that there must be a continuum – from family to community in teaching or discussing with adolescents on SRH needs and services.
• The majority of parents are against SRH education for pupils, because of fears that it may promote sexual activity. However, a minority did argue for more widespread condom availability.
• Sexual violence and abuse towards adolescent girls perpetrated by adult men, including teachers was, acknowledged.
Infrastructural limitations
• Serious understaffing of health staff and long queues of patients in the health unit.
• Irregular or erratic supply of SRH materials and products in public and government run health facilities.
• Restrictive working schedules limit service accessibility to young people
• Inadequate supply of drugs and contraceptives at health care facilities.
Key Baseline Quantitative Findings
As predicted the majority religion in Niger was Muslim, and majority ethnic group was Zarma in Tillabery and Haussa in Maradi. In Tanzania, the majority ethnic group in Mwanza was Sukuma, but the ethnic distribution in Iringa was much more varied.
The large sample size means that within each country, there were statistically significant differences between Regions in almost all the basic demographic variables, but examination of the percentages suggests this may exaggerate the social significance of many of these differences. However, religious observance did seem to be substantially higher in Tillabery than Maradi for both men and women.
Distribution of basic demographics and for the primary outcome indicators, is similar between intervention and comparison arms, with the exception of an apparently greater number of home births among intervention groups among 15-19 year old girls in Mwanza.
However, uptake of the primary outcome was universally low with only 10 subjects or fewer with STI symptoms reporting uptake of STI services among males or females either sex in either the 15-29 or 20-34 age of group.
INTERVENTION DESIGN AND IMPLEMENTATION:
An a priori intervention design framework was developed (See Figure 3) which aimed to capture the three intervention settings (health services, community and schools) and the relatinships between them.
Figure 3: Intervention Framework
In each country, data from the situation analysis was used to develop multifaceted interventions that relied on reflexive discussion at grass roots level.
Phase 2- INTERVENTION DEVELOPMENT AND IMPLEMENTATION IN TANZANIA:
The intervention in Tanzania comprised:
(i) Case scenario based facilitated discussions led by Facility-Own Resource Persons (FORPS)
(ii) Text messaging intervention to facilitate referral to government health facilities
(I) FORPS INTERVENTION: In Tanzania a cadre of individuals known as Facility-Own Resource Persons (FORPS) were elected according to specific criteria and trained to facilitate scenario-based discussions within their three respective settings. Prototypes of interventions were developed and refined by the NIMR team in collaboration with the Ministry of Health and Social Welfare (MoHSW), and two of the key Non Governmental Organisations (NGOs) working in Tanzania at the time: Chama Cha Uzazi na Malezi Bora Tanzania – (Kiswahili for “Family Planning and Reproductive Health Association”, UMATI FHI360 .
Researchers and the Intervention team in NIMR initiated cascade consultations meetings with stakeholders from the district level to sub-village level and local partners from local NGO’s. The consultation meetings with the stakeholders involved discussions on intervention setup (framework), types of intervention, key actors, resource requirements of each intervention and the justifications for various proposed intervention.
The FORPS intervention was supported by Youth Familiarisation Days in both Mwanza and Iringa and a text messaging intervention to enhance integration between community (drug stores) and formal health facility provision of reproductive health services in Mwanza only.
FORPs were community members or representatives from health or education sectors, and community facilities who are selected by their peers to facilitate reflexive discussion meetings with teachers, parents, influential people, health staff in their settings and also facilitating youth familiarization day in the health setting.
In brief FORPS were selected and trained to facilitate a series of discussions in schools, health units and the community as follows:
• Participation of the health workers in adolescents “discussion groups”, in order to better understand the reality of the lives of young people;
• Regular same gender discussions between health workers, so they can discuss issues related to gender arising in the line of their duty: for example, persons reticent to consult an agent from the opposite sex.
• Parents’ day to be used as a network channel between parents and teachers to discuss issues of wellbeing with their children and any SRH issues.
A participatory and interactive training of the FORPS used lectures, group discussions, role plays, brainstorming, case study covered the following topics:
• Brief presentation of the IntHEC project.
• ASRH situation in Tanzania.
• The importance of reflection/reflective practice.
• Communication skills with young people.
• Responding to difficulty questions asked by adolescents
• Delivery of family planning services,
• SRH rights and community responsibility in delivery of SRH.
• National policies/guidelines on SRH service delivery in school, and Health setting.
• Familiarization with different health, school and community intervention manuals.
• Development of action Plan.
Case scenarios and key factors identified in the situation analysis formed the basis for facilitated discussions. These discussions sought to help teachers and health workers in each setting to identify the mediating moments at which adverse health/service uptake outcomes occur and facilitate remedial action in terms of individual or facility level practice.
(ii) TEXT MESSAGING INTERVENTION
The text messaging intervention aimed to facilitate uptake of government services for SRH symptoms by pairing drug stores with government health facilities. This intervention worked in tandem with a training of drug stores initiative that was being implemented by the Tanzania Food and Drug Agency (TDFA).
After a process of pretesting and evaluation the intervention was rolled out in Mwanza.
All dispensaries in the intervention wards were selected to participate in the intervention. These health units are partnered with 3 drug stores of their choice each according to the following criteria: (i) within their catchment areas; (ii) trusted, and (iii) have ability to transfer to TFDA training programme when it is rolled out in Mwanza.
The district medical officers mobilised clinical officers participating on the intervention within their respective districts. In turn, the clinical officers mobilised drug store attendants from their selected drug stores. Training took place at the district headquarters. Its main content was on use of the system. Major topics included the benefits of using the system, how phone numbers are registered onto the system, how the system works including step-by-step of how to deal with patients at all levels, how to ask and record names and age, symptoms, appointments, use of the toll-free number (15543), how to receive and pass on the password to the patient, how to retrieve it at the dispensary/health centre when the patient comes and how to read the cue card. To facilitate record keeping and avoid loss of information in case text messages are accidentally erased from users’ phones, dispensaries and health centres were also taught how to keep record of and updated Text Message Report form, which they refer too every time a text message comes and every time a referred patient reports to the facility. Cue cards were distributed and certificates awarded to all trained participants.
Phase 2- INTERVENTION DEVELOPMENT AND IMPLEMENTATION NIGER
Using the results from the situation analysis, Discussions were facilitated using vignettes that were developed using representative scenarios derived from the situation analysis. These vignettes were filmed and used to produce a DVD which was used as a discussion starters.
Three primary interventions were developed. These interventions followed a discussion starter model based on a series of vignettes developed from the Study 1 analysis. The vignettes were refined into small scenarios that were depicted in a DVD. These video Vignettes were then to be used as discussion starters for the three main interventions as follows:
1. Training of service providers (health agents and teachers) ;
2. Discussion groups with the health workers and community members;
3. Open talk groups with adolescents and young.
A series of discussions conducted from May to August 2012 were used to finalise the interventions.
Seventy two teachers were trained within IntHEC. Subsequently one thousand four hundred teachers have been trained in SRH as part of the Ministry of Health and Ministry of Education joint policy after the teachers’ lack of SRH knowledge was highlighted within the IntHEC project.
Topics from the IntHEC training were introduced in the national teacher-training manual and the manual/curriculum will be shared with LSTM.
In addition to the training 24 discussion groups and 48 open talk groups were held in each Region.
PHASE 3- INTERVENTION EVALUATION: The major component of intervention was evaluated through qualitative process evaluation studies, mystery client studies and a follow up cluster-randomised household survey. The follow up Mystery Client Studies aimed to assess and compare the qualities of adolescent health services in the intervention and comparison communities to identify practices that limit service quality and assess the degree to which the intervention had ameliorated adverse practices and environments identified at baseline.
1. Process Evaluation Studies:
• A Process evaluation sub study was conducted in 3 intervention and 1 comparison wards in each study Region. Three focus group discussions and 121 individual interviews were conducted. In two phases in April 2013.
Thematic analysis revealed the following:
A. Training: The interventions were implemented slightly differently in each district. Those in SAY tended to stick very rigidly to the training protocols, where those in Ague were less consistent.
The students were very interested in the course and the teachers more relaxed in their style. Students reported a better understanding of Sexual and Reproductive health after the training
B. Discussion groups: All the planned group sessions were conducted.
It was found that these discussion groups promoted involvement of local leaders, and intensified debate on issues such as early marriage. The discussion groups also raised awareness of issues such as the risk to young girls of petty trade (talla). Involvement of the local leaders was felt to be a particularly positive feature
C. Open talk groups
The Youth associations for young men (fadas) were the basic unit for Intervention. Overall, the male groups were unreliable and scheduling proved difficult. Girls groups operated much better. However, participants reported a decrease in risk behaviours and promotion of protective behaviours (exaltation of abstinence, condoms demand at the youth friendly centers, their use by the boys, and their requirement by the girls for any sexual intercourse)
Differences between Say and Aguie: Overall, implementation went as planned in Say, but was more improvised in Aguie- e.g. Aguie- Innovated- included a range of people beyond HW and Teachers such as NGO members, Radio “melting pot”.
2. Stakeholder analysis
An external researcher was commissioned to conduct a stakeholder analysis among LASDEL, MoPH and UNFPA staff and community implementers and participants. This comprised a series of 35 interviews and discussions conducted in October 2013.
Interviews were recorded and transcribed in French. These 35 qualitative interviews were content analysed by UCL.
A synthesis of results was presented to Nigerien partners who then fed back their views in a recorded group discussion at the final workshop in Liverpool.
The study examined implementers and community recipient perspectives of the research-action.
Overall, the study recipients at community level were welcoming and supportive of the intervention; however implementers’ views were more varied: on the one hand the research experience was seen to be a good one, conversely IntHEC was perceived to be very work intensive, putting a strain on any other commitments they may have.
3. Follow up Mystery Client Studies
The follow up Mystery Client Studies aimed to assess and compare the qualities of adolescent health services in the intervention and comparison communities to identify practices that limit service quality and assess the degree to which the intervention had ameliorated adverse practices and environments identified at baseline (see also Work Package 1 Period 1 and 2 reports).
In Niger the study was conducted in February 2014 in 8 Health Units (CSI) (2 intervention and 2 comparison per Region). The protocol was revised in collaboration with LSTM and UCL.
16 young people were given a general training that addressed
• Introduction of the IntHEC programme (objectives, intervention strategies, etc.);
• Introduction of adolescent reproductive health;
• Objectives of the mystery client study and the expected results;
• Explanation of the different scenarios;
• Use of digital recorders.
The mystery clients were then allocated the following scenario which they
• Scenario 1: A young boy who wants to get some condoms goes to the CSI male /female nurse. He is received by the male /female nurse to whom he explained his request.
• Scenario 2: A young boy having had occasional sexual intercourse weeks ago, has discussion on Aids with friends. He is concerned about his serological status and decided to meet the CSI male /female nurse to have reliable information on the HIV transmission as well as the protection means. He is received by the male /female nurse to whom he explains his problem.
• Scenario 3: A young school girl has a menstruation delay and pregnancy signs goes to the mid-wife. She asks the assistance of the mid-wife and does not want her parents to be informed about her situation.
• Scenario 4: A young girl needs information on the different contraceptive methods. She goes to the CSI health agent to have the information she needs.
In Tanzania key activities for the follow up Mystery Client Study were completed as follows:
• Revision of mystery client study protocol to incorporate lessons learnt from implementation at base line
• Consultation with regional and district health authorities for approval
• Mobilization of Health workers to about the study (2 months prior to the study)
• Fieldwork started in April 2014 in Mwanza and ended in May 2014 in Iringa regions.
• Mystery Clients (MC) conducted interviews in 24 wards (12 in Mwanza and 12 in Iringa).
MC simulated 3 Reproductive health scenarios as follows:
• Condom request
• Seeking information regarding STIs
• General Information on Family planning
The condom request scenario was only simulated by male MC and the general family planning scenario was simulated by female MC’s only. Information seeking on STIs simulated by male and female MCs
A total of 35 visits were conducted by mystery clients from 25 health facilities in Mwanza and Iringa region. Three of the health facilities were the Health Centres and the remaining 22 were dispensaries. All mystery clients were interviewed by the interviewers immediately after their visit (this aided in their recall of events). A total of 19 exit interviews were conducted in Mwanza and 16 interviews were conducted in Iringa region.
Out of 35 recorded consultations, 3 were excluded due to unclear recordings. Thus, a total of 32 recorded consultations and de-brief sessions were transcribed and translated. For the purpose of this report, the research team reviewed and analysed 32 transcripts. Transcripts from the recorded consultations (32) and from de-brief interviews (32) with young people were analysed in parallel.
4. Follow up Household Survey
The follow up Household Survey (HHS) aimed to assess the impact of the intervention at Household level.
In both countries the follow up HHS used an adapted version of the questionnaire that had been implemented at baseline. The protocols were modified after consultation with the Project Advisory Committee. In particular, it was agreed that different designs be used in the two countries as befits their circumstances. In Niger the follow up HHS combined a cohort study with an additional cross sectional component – while in Tanzania the HHS adopted a cross-sectional design. The decision to implement the different components was based on:
- Scarcity of data on ARH in Niger and increased possibilities of statistical analyses if longitudinal design was pursued
- Increased feasibility of follow up of Phase 1 participants in Niger (because of special measures taken by Lasdel during Phase 1 including photography of participants.
- Efficiency of analysis using cross-sectional design in Tanzania.
In both countries the baseline protocol and tools (survey field guide, fieldwork report, lost-to-follow up calculation scheme) were revised in collaboration with UCL, LSTM and LASDEL. Likewise, a short “lost to follow up” questionnaire was added to the main HHS questionnaire in Niger.
The follow-up HHS questionnaire was revised over several months taking into account lesson learned during phase 1, baseline statistical analysis and study design modification. The number of questions was decreased and the details of the questions modified to address follow up objectives, including intervention exposure and corrected with back translation from French to English.
The statistical analysis plan for the cross sectional survey developed by NIMR/LSTM for Tanzania was modified by UCL and formed the basis of a completed a statistical analysis of Niger data (see below)
INTERVENTION EVALUATION RESULTS:
MYSTERY CLIENT STUDIES suggested that overall, the infrastructural limitations such as availability of space, unfriendly working hours, lack of supplies adversely affected the provision of services to adolescents. These factors were beyond the scope of the intervention and somewhat overshadowed its effectiveness.
Confidentiality
The health facilities lacked enough space to provide privacy services to the adolescents who went to seek the SRH services. Some health workers did not see the adolescents in privacy. Most often they kept the door or window open, which allowed other patient who were seated outside to hear the conversation between the providers and their clients.
Attitudes
Although many health workers would encourage boys to use condom they (health workers) still expressed paternalistic sentiments or religious morals. Most of them would provide them mixed and at time contradictory messages. The health workers would provide condom but alongside they advise the adolescent to abstain from sex.
THE HOUSEHOLD SURVEY: Suggested that there were significant differences at the end of the intervention implementation period as follows:
In Niger, (i) a higher proportion of adolescent girls accessed health services for STI services in intervention wards than in control wards, but this difference disappeared after adjustment by age and risk stratum, but overall government health service use did appear to be significantly greater amongst control groups in Aguie after adjustment by age and risk stratum. (ii) A higher proportion of women in intervention wards used VCT or prenuptial counselling. This difference was observed in both age groups in Say and only in 15-19 year olds in Aguié. (iii) A higher proportion of adolescent women use modern contraception in intervention wards compared to the control wards of Say region. When adjusted for age and risk stratum, this difference remained significant.
In Tanzania, there was no evidence of increased uptake of services in either Iringa or Mwanza when stratified by subgroups, however, the proportion of participants overall who reported STI symptoms was greater in comparison communities in both Iringa and Mwanza.
Potential Impact:
SOCIETAL IMPLICATIONS The data are preliminary and should be interpreted with caution, but do suggest that such complex, community based interventions are feasible and acceptable. The uptake of the intervention within the Ministries is encouraging and 1,400 teachers have been trained in SRH as part of the Ministry of Health and Ministry of Education joint policy in Niger after the teachers’ lack of SRH knowledge was highlighted within the IntHEC project. Topics from the IntHEC training were introduced in the national teacher-training manual and the manual/curriculum will be shared with LSTM.
LESSONS LEARNT AND RECOMMENDATIONS
This discussion based intervention did appear to have an impact on key outcomes relating to contraceptive uptake in some settings. However this is context specific and more research is needed
(i) In relation to impact of different religious and cultural influences on both intervention implementation and uptake
(ii) The specific situation of very young married adolescent girls in conservative settings.
(iii) The study had a focus on Ethical issues and the following issues in particular warrant further research:
• The high prevalence of child marriage meant that a number of married adolescents were below the internationally recognized age of majority, but were emancipated by their marriage. The ethical implications of their social and biological developmental status need to be explored.
• The use of mystery client studies is critical to capturing the quality of services, especially amongst adolescents. There is an urgent need to further explore both health worker and mystery client perceptions of the acceptability and validity of these studies.
• Field researcher perceptions of what are “ethical” issues is key to capturing the realities of research ethics in the field. Further research in this area is necessary if we are to be better able to frame the key ethics research questions in this area.
• More research is needed to explore the mechanisms through which ethics is or can be made performative in this and other forms of research
• Methodological commentary on the relationship between ethics / ethical research practice and reflexivity
EXECUTIVE SUMMARY
IntHEC is a 4 year programme which aims to design and evaluate an adolescent reproductive health (ARH) programme in Niger and Tanzania. It is implemented by a consortium of 8 research, NGO and government ministry partners from 4 countries.
The specific activities were as follows:
1. Situation analysis: Conduct a situation analysis of current community and implementer experiences of existing ARH programmes in Tanzania and Niger, identifying priority areas of weakness in RH service provision and opportunities for strengthened service uptake and integration.
2. Intervention Development: Develop and implement an innovative package of pilot interventions that are feasible, equitable and appropriately designed to address identified weaknesses and opportunities.
3. Process evaluation: Document and evaluate the processes of these new interventions with special reference to gender and social equity.
4. Impact evaluation: Document and evaluate the impact of these new interventions and their effects on ARH provision, uptake, and effectiveness with special reference to gender and social equity.
5. Policy research and Advocacy: Support the development and implementation of feasible, effective, and equitable RH interventions in sub-Saharan Africa through effective collaboration between key research institutions, international authorities, government sectors and leading non-governmental organisations (NGOs).
In both Tanzania and Niger, discussion based intervention were developed and implemented. In Mwanza, interventions were supported by an SMS referral intervention.
Interventions were evaluated using a range of activities including baseline and follow up cluster randomized household surveys.
Overall evaluation found that the interventions were very popular at the community level and felt to be of value.
However, mystery client studies found overall that in health units, infrastructural limitations which were beyond the scope of the intervention continued to be a major bar to adolescent friendly service provision and somewhat overshadowed improvements arising from intervention activities.
Conversely, qualitative and quantitative evaluation found the SMS intervention was highly acceptable to both drug stores and their clients and did promote health facility attendance
A notebook assisted household survey was successfully conducted. Preliminary data suggest that the intervention did indeed improve uptake of VCT and STI services in Niger. Preliminary data suggest overall rates of STI in Tanzania appeared to be higher in comparison communities but further adjustment of the data is necessary to interpret the significance of this.
Project Context and Objectives:
PROJECT CONTEXT AND OBECTIVES
The project aims to improve the reproductive health of poor people in sub-Saharan Africa. The overall objective of the project is to improve the delivery of reproductive health (RH) services in Tanzania and Niger by successfully engaging policy-makers and programmers in the generation of new evidence about effective ways to strengthen the provision, uptake, equity and effectiveness of adolescent reproductive health (ARH) programmes.
The research was implemented in two regions in Niger (Say and Aguie) and Tanzania (Mwanza and Iringa). The national governments in both countries are implementing ARH activities that are being rolled out throughout each country. These interventions were implemented using strategies, and within contexts that are typical of ARH service provision elsewhere in sub-Saharan Africa. The results of our research are therefore likely to be applicable to several other countries.
The research integrated a community psychological approach into the Medical Research Council (MRC) best practice framework for developing complex interventions to develop the intervention package in three stages: (i) situation analysis; (ii) design of pilot interventions to address issued identified in situation analysis; (iii) small-scale pre-testing and formative evaluation of interventions. This aimed to ensure feasibility, acceptability and likelihood of effectiveness. These interventions were implemented in 18 intervention wards (9 in Iringa and 9 in Mwanza) for 12 months. Full scale intervention evaluation comprised process evaluation through a series of in-depth qualitative studies and impact evaluation through a population based community randomised trial. A combination of quantitative, qualitative and participatory methods was used to complete each stage of the research.
SPECIFIC OBJECTIVES
1. Conduct a situation analysis of current community and implementer experiences of existing ARH programmes in Tanzania and Niger, identifying priority areas of weakness in RH service provision and opportunities for strengthened service uptake and integration.
2. Develop and implement an innovative package of pilot interventions that are feasible, equitable and appropriately designed to address identified weaknesses and opportunities.
3. Document and evaluate the processes of these new interventions and their effects on ARH provision, uptake, and effectiveness with special reference to gender and social equity.
4. Support the development and implementation of feasible, effective, and equitable RH interventions in sub-Saharan Africa through effective collaboration between key research institutions, international authorities, government sectors and leading non-governmental organisations (NGOs).
Project Results:
DESCRIPTION OF THE MAIN SCIENTIFIC AND TECHNICAL RESULTS
The study was implemented in 18 wards in each of 2 Regions in each country. Within each Region 9 wards were randomised to receive the intervention and 9 wards randomised to comparison groups (a ward is an administrative unit that is roughly equivalent to 6-8 villages).
The overall study design is presented in Figure 1:
Figure 1- Overall study Design:
The study took place in 3 broad phases:
1. Phase 1- Situation analysis which included a literature review; baseline studies in schools, communities and health facilities and a household survey
2. Phase 2- Intervention development based on qualitative findings from the situation analysis and subsequent Intervention implementation
3. Phase 3- Evaluation comprising process evaluation of intervention implementation and evaluation of the intervention impact through the baseline and follow up household surveys.
Development and testing of the intervention and evaluation tools and strategies were tested in the Validity Villages, Formative wards and Pretest wards. The interventions were then implemented in randomly assigned intervention communities with pre and post impact evaluation through baseline and follow up household surveys. Detailed process evaluation was conducted in a subset of intervention and evaluation wards.
A BASELINE SITUATION ANALYSIS was conducted and included:
1. A scientific literature review on ARH was conducted in both Niger and Tanzania from March to April 2010 to identify relevant impact and process indicators. A review of international and national ARH laws and policies provided an overview of the range of policies relevant to ARH in the two countries. Linking with the ARH indicators review identified key performance standards for key actors within ARH service delivery which can be used as benchmarks for ARH services evaluation. The review also laid the groundwork for the “Policy engagement, advocacy and dissemination” work package (WP5).
2. Development of an Evaluation Framework. A theoretical evaluation framework was developed. Key concepts of research and intervention (for example activity systems, mediating moments) were agreed amongst all partners (during the first partners workshop in Liverpool and subsequent participatory research development workshops). The initial Evaluation Framework was further refined and the relationship between studies and analysis model clarified. See Fig 2.
Figure 2- Complementary baseline data collection studies:
The evaluation framework comprised five baseline studies including cluster randomised household surveys, close to community service provider consultations, policy reviews, mystery client studies and adolescent consultations.
In these studies we used qualitative, quantitative and action research techniques in health facilities, school and community settings. The objective of the qualitative baseline studies was to determine the range, quality and integration of Adolescent Reproductive Health (ARH) promotion and treatment interventions currently implemented within the health and education sectors and within the general community.
Key Baseline Qualitative Findings
A number of factors limiting ARH access to services were identified in both Tanzania and Niger as follows:
Customary norms and values have a strong and significant adverse influence on young people’s reproductive health seeking behaviour as well as the provision of health services. These community views are strongly shared by teachers and health providers and limit adolescents’ access to reproductive health services such that
• Adolescents are unable to express their ARH needs:
• Young people do not visit health centres because of disapproval from health providers
• There is an absence of dialogue between caregivers and adolescents, which inhibits STI/HIV/AIDS prevention.
• There is an absence of appropriate monitoring of adolescents’ RH needs and behaviours
In School Settings
• SRH related topics are taught in most schools, but do not adequately address pregnancy prevention, negotiation of safer sex, and how to deal with peer pressure or pressure for sex
• Teachers were willing, within the context of the research, to talk about their own reproductive health needs and how they think SRH can be enforced within the teaching community, as opposed to among the pupils.
Integration
• Teachers had a view that there must be a continuum – from family to community in teaching or discussing with adolescents on SRH needs and services.
• The majority of parents are against SRH education for pupils, because of fears that it may promote sexual activity. However, a minority did argue for more widespread condom availability.
• Sexual violence and abuse towards adolescent girls perpetrated by adult men, including teachers was, acknowledged.
Infrastructural limitations
• Serious understaffing of health staff and long queues of patients in the health unit.
• Irregular or erratic supply of SRH materials and products in public and government run health facilities.
• Restrictive working schedules limit service accessibility to young people
• Inadequate supply of drugs and contraceptives at health care facilities.
Key Baseline Quantitative Findings
As predicted the majority religion in Niger was Muslim, and majority ethnic group was Zarma in Tillabery and Haussa in Maradi. In Tanzania, the majority ethnic group in Mwanza was Sukuma, but the ethnic distribution in Iringa was much more varied.
The large sample size means that within each country, there were statistically significant differences between Regions in almost all the basic demographic variables, but examination of the percentages suggests this may exaggerate the social significance of many of these differences. However, religious observance did seem to be substantially higher in Tillabery than Maradi for both men and women.
Distribution of basic demographics and for the primary outcome indicators, is similar between intervention and comparison arms, with the exception of an apparently greater number of home births among intervention groups among 15-19 year old girls in Mwanza.
However, uptake of the primary outcome was universally low with only 10 subjects or fewer with STI symptoms reporting uptake of STI services among males or females either sex in either the 15-29 or 20-34 age of group.
INTERVENTION DESIGN AND IMPLEMENTATION:
An a priori intervention design framework was developed (See Figure 3) which aimed to capture the three intervention settings (health services, community and schools) and the relatinships between them.
Figure 3: Intervention Framework
In each country, data from the situation analysis was used to develop multifaceted interventions that relied on reflexive discussion at grass roots level.
Phase 2- INTERVENTION DEVELOPMENT AND IMPLEMENTATION IN TANZANIA:
The intervention in Tanzania comprised:
(i) Case scenario based facilitated discussions led by Facility-Own Resource Persons (FORPS)
(ii) Text messaging intervention to facilitate referral to government health facilities
(I) FORPS INTERVENTION: In Tanzania a cadre of individuals known as Facility-Own Resource Persons (FORPS) were elected according to specific criteria and trained to facilitate scenario-based discussions within their three respective settings. Prototypes of interventions were developed and refined by the NIMR team in collaboration with the Ministry of Health and Social Welfare (MoHSW), and two of the key Non Governmental Organisations (NGOs) working in Tanzania at the time: Chama Cha Uzazi na Malezi Bora Tanzania – (Kiswahili for “Family Planning and Reproductive Health Association”, UMATI FHI360 .
Researchers and the Intervention team in NIMR initiated cascade consultations meetings with stakeholders from the district level to sub-village level and local partners from local NGO’s. The consultation meetings with the stakeholders involved discussions on intervention setup (framework), types of intervention, key actors, resource requirements of each intervention and the justifications for various proposed intervention.
The FORPS intervention was supported by Youth Familiarisation Days in both Mwanza and Iringa and a text messaging intervention to enhance integration between community (drug stores) and formal health facility provision of reproductive health services in Mwanza only.
FORPs were community members or representatives from health or education sectors, and community facilities who are selected by their peers to facilitate reflexive discussion meetings with teachers, parents, influential people, health staff in their settings and also facilitating youth familiarization day in the health setting.
In brief FORPS were selected and trained to facilitate a series of discussions in schools, health units and the community as follows:
• Participation of the health workers in adolescents “discussion groups”, in order to better understand the reality of the lives of young people;
• Regular same gender discussions between health workers, so they can discuss issues related to gender arising in the line of their duty: for example, persons reticent to consult an agent from the opposite sex.
• Parents’ day to be used as a network channel between parents and teachers to discuss issues of wellbeing with their children and any SRH issues.
A participatory and interactive training of the FORPS used lectures, group discussions, role plays, brainstorming, case study covered the following topics:
• Brief presentation of the IntHEC project.
• ASRH situation in Tanzania.
• The importance of reflection/reflective practice.
• Communication skills with young people.
• Responding to difficulty questions asked by adolescents
• Delivery of family planning services,
• SRH rights and community responsibility in delivery of SRH.
• National policies/guidelines on SRH service delivery in school, and Health setting.
• Familiarization with different health, school and community intervention manuals.
• Development of action Plan.
Case scenarios and key factors identified in the situation analysis formed the basis for facilitated discussions. These discussions sought to help teachers and health workers in each setting to identify the mediating moments at which adverse health/service uptake outcomes occur and facilitate remedial action in terms of individual or facility level practice.
(ii) TEXT MESSAGING INTERVENTION
The text messaging intervention aimed to facilitate uptake of government services for SRH symptoms by pairing drug stores with government health facilities. This intervention worked in tandem with a training of drug stores initiative that was being implemented by the Tanzania Food and Drug Agency (TDFA).
After a process of pretesting and evaluation the intervention was rolled out in Mwanza.
All dispensaries in the intervention wards were selected to participate in the intervention. These health units are partnered with 3 drug stores of their choice each according to the following criteria: (i) within their catchment areas; (ii) trusted, and (iii) have ability to transfer to TFDA training programme when it is rolled out in Mwanza.
The district medical officers mobilised clinical officers participating on the intervention within their respective districts. In turn, the clinical officers mobilised drug store attendants from their selected drug stores. Training took place at the district headquarters. Its main content was on use of the system. Major topics included the benefits of using the system, how phone numbers are registered onto the system, how the system works including step-by-step of how to deal with patients at all levels, how to ask and record names and age, symptoms, appointments, use of the toll-free number (15543), how to receive and pass on the password to the patient, how to retrieve it at the dispensary/health centre when the patient comes and how to read the cue card. To facilitate record keeping and avoid loss of information in case text messages are accidentally erased from users’ phones, dispensaries and health centres were also taught how to keep record of and updated Text Message Report form, which they refer too every time a text message comes and every time a referred patient reports to the facility. Cue cards were distributed and certificates awarded to all trained participants.
Phase 2- INTERVENTION DEVELOPMENT AND IMPLEMENTATION NIGER
Using the results from the situation analysis, Discussions were facilitated using vignettes that were developed using representative scenarios derived from the situation analysis. These vignettes were filmed and used to produce a DVD which was used as a discussion starters.
Three primary interventions were developed. These interventions followed a discussion starter model based on a series of vignettes developed from the Study 1 analysis. The vignettes were refined into small scenarios that were depicted in a DVD. These video Vignettes were then to be used as discussion starters for the three main interventions as follows:
1. Training of service providers (health agents and teachers) ;
2. Discussion groups with the health workers and community members;
3. Open talk groups with adolescents and young.
A series of discussions conducted from May to August 2012 were used to finalise the interventions.
Seventy two teachers were trained within IntHEC. Subsequently one thousand four hundred teachers have been trained in SRH as part of the Ministry of Health and Ministry of Education joint policy after the teachers’ lack of SRH knowledge was highlighted within the IntHEC project.
Topics from the IntHEC training were introduced in the national teacher-training manual and the manual/curriculum will be shared with LSTM.
In addition to the training 24 discussion groups and 48 open talk groups were held in each Region.
PHASE 3- INTERVENTION EVALUATION: The major component of intervention was evaluated through qualitative process evaluation studies, mystery client studies and a follow up cluster-randomised household survey. The follow up Mystery Client Studies aimed to assess and compare the qualities of adolescent health services in the intervention and comparison communities to identify practices that limit service quality and assess the degree to which the intervention had ameliorated adverse practices and environments identified at baseline.
1. Process Evaluation Studies:
• A Process evaluation sub study was conducted in 3 intervention and 1 comparison wards in each study Region. Three focus group discussions and 121 individual interviews were conducted. In two phases in April 2013.
Thematic analysis revealed the following:
A. Training: The interventions were implemented slightly differently in each district. Those in SAY tended to stick very rigidly to the training protocols, where those in Ague were less consistent.
The students were very interested in the course and the teachers more relaxed in their style. Students reported a better understanding of Sexual and Reproductive health after the training
B. Discussion groups: All the planned group sessions were conducted.
It was found that these discussion groups promoted involvement of local leaders, and intensified debate on issues such as early marriage. The discussion groups also raised awareness of issues such as the risk to young girls of petty trade (talla). Involvement of the local leaders was felt to be a particularly positive feature
C. Open talk groups
The Youth associations for young men (fadas) were the basic unit for Intervention. Overall, the male groups were unreliable and scheduling proved difficult. Girls groups operated much better. However, participants reported a decrease in risk behaviours and promotion of protective behaviours (exaltation of abstinence, condoms demand at the youth friendly centers, their use by the boys, and their requirement by the girls for any sexual intercourse)
Differences between Say and Aguie: Overall, implementation went as planned in Say, but was more improvised in Aguie- e.g. Aguie- Innovated- included a range of people beyond HW and Teachers such as NGO members, Radio “melting pot”.
2. Stakeholder analysis
An external researcher was commissioned to conduct a stakeholder analysis among LASDEL, MoPH and UNFPA staff and community implementers and participants. This comprised a series of 35 interviews and discussions conducted in October 2013.
Interviews were recorded and transcribed in French. These 35 qualitative interviews were content analysed by UCL.
A synthesis of results was presented to Nigerien partners who then fed back their views in a recorded group discussion at the final workshop in Liverpool.
The study examined implementers and community recipient perspectives of the research-action.
Overall, the study recipients at community level were welcoming and supportive of the intervention; however implementers’ views were more varied: on the one hand the research experience was seen to be a good one, conversely IntHEC was perceived to be very work intensive, putting a strain on any other commitments they may have.
3. Follow up Mystery Client Studies
The follow up Mystery Client Studies aimed to assess and compare the qualities of adolescent health services in the intervention and comparison communities to identify practices that limit service quality and assess the degree to which the intervention had ameliorated adverse practices and environments identified at baseline (see also Work Package 1 Period 1 and 2 reports).
In Niger the study was conducted in February 2014 in 8 Health Units (CSI) (2 intervention and 2 comparison per Region). The protocol was revised in collaboration with LSTM and UCL.
16 young people were given a general training that addressed
• Introduction of the IntHEC programme (objectives, intervention strategies, etc.);
• Introduction of adolescent reproductive health;
• Objectives of the mystery client study and the expected results;
• Explanation of the different scenarios;
• Use of digital recorders.
The mystery clients were then allocated the following scenario which they
• Scenario 1: A young boy who wants to get some condoms goes to the CSI male /female nurse. He is received by the male /female nurse to whom he explained his request.
• Scenario 2: A young boy having had occasional sexual intercourse weeks ago, has discussion on Aids with friends. He is concerned about his serological status and decided to meet the CSI male /female nurse to have reliable information on the HIV transmission as well as the protection means. He is received by the male /female nurse to whom he explains his problem.
• Scenario 3: A young school girl has a menstruation delay and pregnancy signs goes to the mid-wife. She asks the assistance of the mid-wife and does not want her parents to be informed about her situation.
• Scenario 4: A young girl needs information on the different contraceptive methods. She goes to the CSI health agent to have the information she needs.
In Tanzania key activities for the follow up Mystery Client Study were completed as follows:
• Revision of mystery client study protocol to incorporate lessons learnt from implementation at base line
• Consultation with regional and district health authorities for approval
• Mobilization of Health workers to about the study (2 months prior to the study)
• Fieldwork started in April 2014 in Mwanza and ended in May 2014 in Iringa regions.
• Mystery Clients (MC) conducted interviews in 24 wards (12 in Mwanza and 12 in Iringa).
MC simulated 3 Reproductive health scenarios as follows:
• Condom request
• Seeking information regarding STIs
• General Information on Family planning
The condom request scenario was only simulated by male MC and the general family planning scenario was simulated by female MC’s only. Information seeking on STIs simulated by male and female MCs
A total of 35 visits were conducted by mystery clients from 25 health facilities in Mwanza and Iringa region. Three of the health facilities were the Health Centres and the remaining 22 were dispensaries. All mystery clients were interviewed by the interviewers immediately after their visit (this aided in their recall of events). A total of 19 exit interviews were conducted in Mwanza and 16 interviews were conducted in Iringa region.
Out of 35 recorded consultations, 3 were excluded due to unclear recordings. Thus, a total of 32 recorded consultations and de-brief sessions were transcribed and translated. For the purpose of this report, the research team reviewed and analysed 32 transcripts. Transcripts from the recorded consultations (32) and from de-brief interviews (32) with young people were analysed in parallel.
4. Follow up Household Survey
The follow up Household Survey (HHS) aimed to assess the impact of the intervention at Household level.
In both countries the follow up HHS used an adapted version of the questionnaire that had been implemented at baseline. The protocols were modified after consultation with the Project Advisory Committee. In particular, it was agreed that different designs be used in the two countries as befits their circumstances. In Niger the follow up HHS combined a cohort study with an additional cross sectional component – while in Tanzania the HHS adopted a cross-sectional design. The decision to implement the different components was based on:
- Scarcity of data on ARH in Niger and increased possibilities of statistical analyses if longitudinal design was pursued
- Increased feasibility of follow up of Phase 1 participants in Niger (because of special measures taken by Lasdel during Phase 1 including photography of participants.
- Efficiency of analysis using cross-sectional design in Tanzania.
In both countries the baseline protocol and tools (survey field guide, fieldwork report, lost-to-follow up calculation scheme) were revised in collaboration with UCL, LSTM and LASDEL. Likewise, a short “lost to follow up” questionnaire was added to the main HHS questionnaire in Niger.
The follow-up HHS questionnaire was revised over several months taking into account lesson learned during phase 1, baseline statistical analysis and study design modification. The number of questions was decreased and the details of the questions modified to address follow up objectives, including intervention exposure and corrected with back translation from French to English.
The statistical analysis plan for the cross sectional survey developed by NIMR/LSTM for Tanzania was modified by UCL and formed the basis of a completed a statistical analysis of Niger data (see below)
INTERVENTION EVALUATION RESULTS:
MYSTERY CLIENT STUDIES suggested that overall, the infrastructural limitations such as availability of space, unfriendly working hours, lack of supplies adversely affected the provision of services to adolescents. These factors were beyond the scope of the intervention and somewhat overshadowed its effectiveness.
Confidentiality
The health facilities lacked enough space to provide privacy services to the adolescents who went to seek the SRH services. Some health workers did not see the adolescents in privacy. Most often they kept the door or window open, which allowed other patient who were seated outside to hear the conversation between the providers and their clients.
Attitudes
Although many health workers would encourage boys to use condom they (health workers) still expressed paternalistic sentiments or religious morals. Most of them would provide them mixed and at time contradictory messages. The health workers would provide condom but alongside they advise the adolescent to abstain from sex.
THE HOUSEHOLD SURVEY: Suggested that there were significant differences at the end of the intervention implementation period as follows:
In Niger, (i) a higher proportion of adolescent girls accessed health services for STI services in intervention wards than in control wards, but this difference disappeared after adjustment by age and risk stratum, but overall government health service use did appear to be significantly greater amongst control groups in Aguie after adjustment by age and risk stratum. (ii) A higher proportion of women in intervention wards used VCT or prenuptial counselling. This difference was observed in both age groups in Say and only in 15-19 year olds in Aguié. (iii) A higher proportion of adolescent women use modern contraception in intervention wards compared to the control wards of Say region. When adjusted for age and risk stratum, this difference remained significant.
In Tanzania, there was no evidence of increased uptake of services in either Iringa or Mwanza when stratified by subgroups, however, the proportion of participants overall who reported STI symptoms was greater in comparison communities in both Iringa and Mwanza.
Potential Impact:
SOCIETAL IMPLICATIONS The data are preliminary and should be interpreted with caution, but do suggest that such complex, community based interventions are feasible and acceptable. The uptake of the intervention within the Ministries is encouraging and 1,400 teachers have been trained in SRH as part of the Ministry of Health and Ministry of Education joint policy in Niger after the teachers’ lack of SRH knowledge was highlighted within the IntHEC project. Topics from the IntHEC training were introduced in the national teacher-training manual and the manual/curriculum will be shared with LSTM.
LESSONS LEARNT AND RECOMMENDATIONS
This discussion based intervention did appear to have an impact on key outcomes relating to contraceptive uptake in some settings. However this is context specific and more research is needed
(i) In relation to impact of different religious and cultural influences on both intervention implementation and uptake
(ii) The specific situation of very young married adolescent girls in conservative settings.
(iii) The study had a focus on Ethical issues and the following issues in particular warrant further research:
• The high prevalence of child marriage meant that a number of married adolescents were below the internationally recognized age of majority, but were emancipated by their marriage. The ethical implications of their social and biological developmental status need to be explored.
• The use of mystery client studies is critical to capturing the quality of services, especially amongst adolescents. There is an urgent need to further explore both health worker and mystery client perceptions of the acceptability and validity of these studies.
• Field researcher perceptions of what are “ethical” issues is key to capturing the realities of research ethics in the field. Further research in this area is necessary if we are to be better able to frame the key ethics research questions in this area.
• More research is needed to explore the mechanisms through which ethics is or can be made performative in this and other forms of research
• Methodological commentary on the relationship between ethics / ethical research practice and reflexivity
