Introduction As the world is making progress towards elimination of mother-to-child transmission of HIV, poor coverage of PMTCT services in Nigeria remains a major challenge. In order to address this, scale-up was planned with activities organized into 3 phases. This paper describes the process undertaken in eight high burden Nigerian states to rapidly close PMTCT coverage gaps at facility and population levels between February 2013 and March 2014. Methods Activities were grouped into three phases-pre-assessment phase (engagement of a wide range of stakeholders), assessment (rapid health facility assessment, a cross sectional survey using mixed methods conducted in the various states between Feb and May 2013 and impact modelling), and post-assessment (drawing up costed state operational plans to achieve eMTCT by 2015, data-driven smart scale-up). Results Over a period of 10 months starting June 2013, 2044 facilities were supported to begin provision of PMTCT services. This increased facility coverage from 8% to 50%. A 246% increase was also recorded in the number of pregnant women and their families who have access to HIV testing and counselling in the context of PMTCT. Similarly, access to antiretrovirals for PMTCT has witnessed a 152% increase in these eight states between October 2013 and October 2014 Conclusion A data-driven and participatory approach can be used to rapidly scale-up PMTCT services at community and facility levels in this region. These results present us with hope for real progress in Nigeria. We are confident that the efforts described here will contribute significantly to eliminating new pediatric HIV infection in Nigeria.
The intervention was grouped into three phases–pre-assessment phase, assessment, and post-assessment. We describe the design and methods adopted for each phase below. The key features are summarized in Box 1. The FHI 360 Office of International Research Ethics (OIRE) determined that this project does not meet the regulatory definition of research and/or research involving human subjects as defined under the Department of Health and Human Services Code of Federal Regulations [45 CFR part 46.102(d)(f)]. FHI 360 OIRE Project #:899555. Recognizing the important role of having a government-led process for sustainability, the governments of these eight states were engaged at the highest political/administrative levels on the need to rapidly scale-up PMTCT services. Advocacy visits were paid to the state governors or in their absence, the deputies or ‘first ladies’ (wife of the Governor who wields significant influence especially on maternal health issues at state level). We prepared advocacy packs which included past achievements of the state government in the area of PMTCT/HIV services as well as data showing the remaining scale/burden of HIV and low coverage of services. The need for a government directive to ensure full scale government-led scale-up was one of the advocacy goals. Once the governors were engaged, they issued directives to the respective health ministries to ensure PMTCT scale-up activities. The scale-up process was designed to be state-led and evidence-driven supported by multiple stakeholders. To achieve a state-led process, we worked with state health ministry officials to understand the gaps and importance of closing them. Scale-up planning meetings were subsequently held in each state. Attendees included directors general and project managers of state AIDS control agencies, state AIDS program coordinators from the ministries of health, directors of primary health, directors of regulatory bodies for the private health sector, primary health care control boards/agencies, donor coordination departments, departments of economic planning and budgeting, among others. At these meetings, attendees reviewed the objectives of the assessments, health facility lists, geographical maps, distribution of logistics hubs, assessment tool, composition of assessment teams and general logistics details. Detailed implementation plans were drawn up following these meetings. The number of teams, tools and hubs were planned such that assessments were conducted over one week in each state. The assessment tools were designed to collect both quantitative and qualitative data and were accompanied by an assessment guide. While a scale-up plan was available at national level, there were no such comprehensive plans at state and local government area (LGA) levels. Not only were there no guiding documents, there was a dearth of state-level information regarding the status of the health system and understanding of the bottlenecks to expand PMTCT services. In response, the eight states conducted rapid health facility assessments (R-HFA) with PEPFAR technical and financial support. The aim of these assessments was to build evidence at the local level and provide a baseline to plan the rapid scale up of services. The R-HFA was a cross sectional survey using mixed methods conducted in the various states between Feb and May 2013. Data was collected in broad areas categorized as human resource (HR), infrastructure, service availability, utilization of services, enabling environment and community support systems. The components of these broad areas are described in more details below. There were 182 LGAs in the eight 12+1 states. The strengthening integrated delivery of HIV/AIDS services (SIDHAS) project supported HIV services in 150 of these 182 LGAs. The assessment took place in all 150 SIDHAS-supported LGAs in the eight of the 12+1 states. The sampling frame was a total listing of all health facilities in each state as available from different government agencies that kept a form of register of health facilities. The inclusion criterion was all facilities with antenatal care (ANC) services, as in principle these health facilities could provide PMTCT services if equipped with the proper technical and human expertise. Excluded from the assessment were health facilities that were already providing PMTCT services–meaning ARVs for PMTCT—or had concluded plans to initiate PMTCT services in 2013. After applying this criteria, 5935 health facilities were included. Of these, 1759 could not be assessed due to unavailability of comprehensive facility lists in some states, ocean tidal fluctuations, terrain challenges and communal unrest (Fig 1). Quantitative information across seven domains was collected: facility health linkages (distance and travelling time to facilities offering comprehensive HIV services), health human resource complement (number of staff allocated and available for different cadres), client flow (data showing utilization of services at the health facility including out-patient department, ante-natal clinic and labour ward), scope of services provided (outpatient, inpatient, laboratory, drug dispensing, referrals, etc), community support systems (attached ward development committees, community development associations, etc), current infrastructure (electricity, water, toilets, incinerator, computer, filing cabinets, furniture, medical equipment, drug storage facilities, etc), and future prospects for expansion (availability of space to set up service delivery points or provide additional services). Geospatial location of the facilities was determined using GPS devices. Key informant interviews (KIIs) with health workers were used to explore community birth site options, perceived reasons for preferred choice of birth site, factors influencing facility patronage, and the extent of community participation in service delivery. Data was collected using Google Forms, exported to MS Excel and analysed using SPSS. The data was then validated with a wide group of state level stakeholders. The human resource complement was measured against the stipulated national standard for a PMTCT site—at least one doctor, at least one nurse midwife, two or more Community health workers (CHEWS/JCHEWS), Pharmacist or Pharmacy technician, Lab scientist and Record officer [12]. From June to September 2013, stakeholders from national, state and local government levels as well as development partners, civil society, traditional institutions and the public converged at state-level workshops to review the R-HFA data and agree on key steps to ensure that 90% of HIV positive pregnant and lactating women have access to ARVs and other PMTCT services. The results of the R-HFA informed areas of focus for each state. In one of the states, Anambra, UNICEF also conducted a bottleneck analysis which focused on identifying the bottlenecks in effective implementation of PMTCT at the local government level; further enriching the gap analysis and planning for specific activities in that state. In most states there was a discussion around–“with two years to the 2015 deadline, should the operational plans focus on what states think is achievable or should the plans cover what needs to be done to achieve elimination of new pediatric infection”. In seven of eight states, consensus was built to plan for “what needs to be done” and to commit resources to achieving set goals. Consequently, key deliberations were held to modify the national HR and service delivery requirements, with consideration for resources at the state level while not compromising service quality. The minimum state-specific HR complement required per health facility was iteratively refined to determine the number of health workers required to achieve the desired population coverage. The iterative process meant that when a minimum complement was set, it was examined against the possible coverage that will be achieved. If the agreed complement did not achieve a desired coverage, it was refined until a desired coverage level was attained. State-level technical experts then generated a set of key activities required to meet the respective state “elimination” goals as contained in the National PMTCT guidelines [13] and which derived from global eMTCT goals [5]. The activities were relevant to the state context, comprehensive [13] and had strong demand creation components–mobilization, sensitization, media messaging, community level activities, etc. Activities were subsequently refined and costed while an estimation of the impact of implementation to scale was modelled. The operational plans covered the period 2013 to 2015, and had resource commitments from partners at the table including all levels of government—local, state and federal as well as donor agencies. The full list and cost of all the activities are contained in the state level operational plans for eMTCT which are published online [14]. We constructed a model to estimate the impact of implementation of the operational plans. The objective was to understand (in terms of lives saved, infections averted) if targets were met and plan was implemented to scale. The base case scenario was if current levels of coverage were maintained across three main targets: 1) reduce HIV incidence by 50% among women of reproductive age by 2015; 2) reduce unmet need by 90% of HIV positive women (by increasing voluntary FP use) by 2015; and 3) increase access to antiretroviral treatment (ART) to 90% of HIV positive pregnant women by 2015. The alternate scenario is when these targets are met. The estimated impact is the difference between the two scenarios. The full definitions of variables and assumptions that went into the model are described elsewhere. [6] With the completion of the costed eMTCT operational plans, implementation began in earnest in October 2013. A key first step in smart expansion is the use of available evidence to prioritize areas with the highest MTCT potential (maternal HIV) and widest PMTCT coverage gaps. The LGA was chosen as the intervention unit. To prioritize the facilities, we created a rank order matrix applied at two levels–to prioritize LGAs and within LGAs, to prioritize facilities. The rank order was designed to select LGAs with the highest PMTCT coverage gaps, high HIV prevalence and large population size. The HIV prevalence, estimated population of women of reproductive age group and fertility rate was used to derive an estimated number of HIV positive pregnant women for each LGA. This represented the burden of potential MTCT in each LGA. A rank (rank 1) was assigned to each LGA for this burden–the higher the burden, the higher the rank. A second rank (rank 2) was assigned for PMTCT service coverage gap. Using data from the assessment, we calculated the proportion of sites with ANC services that did not provide PMTCT services. The higher this proportion, the higher the rank 2 value. In essence, rank 2 was assigned to give higher rank to LGAs with wider service coverage gaps. We then summed up both ranks 1 & 2 to give a final rank sum that was used to determine which LGAs were prioritized for the earlier phases of scale-up. An example of the prioritization in Abia State is shown in Table 1. Within LGAs, R-HFA data was used to rank the facilities by the number of clinical health workers available, the number of antenatal attendees at the health facility in the 12 months preceding the assessment and the number of deliveries. The higher the numbers reported for these three elements, the more likelihood of being selected for the earlier phase of scale-up. In essence, the rank order prioritization was aimed at selecting facilities that had more HR, were patronized by more pregnant women, had no other facility providing PMTCT in the vicinity and were located in a higher HIV burden LGA. Services were then scaled-up in a phased approach. The equitable spread was mapped using geographic information system data from the R-HFA. Once a clearer picture for service expansion was painted, the next stage was to roll out services. First a list of health workers to be trained was generated for each selected facility. This was done by the different government agencies overseeing the various levels of health workers in the health system. Government agencies regulating the private-for-profit health sector generated the private health facility training lists in conjunction with proprietors. Trainings were followed by “site activation”. Activation in the scale-up context involved a two-day post-training onsite hands-on support to respective facilities. Activation teams spent this time deploying activation kits, setting up PMTCT services and getting health workers acquainted with service flow. The activation kits included HIV rapid test kits, ARVs, patient education and information materials, data collection and reporting tools. Further, mobilization and sensitization activities were conducted in communities around the health facilities with the aim of generating demand for these services. The training and activation phase for each facility was designed to be led by multi-disciplinary teams and concluded within a month. This meant that the team of health workers from each facility went through requisite training courses and their facility was visited for the two-day site activation all within one month. Resources were deployed accordingly to work within this timeline. Thus, concurrent training and site activation batches were set up in a production line fashion. As sites were trained by the multiple training teams at different training venues, they were passed on to activation teams who provided onsite mentorship for commencement of PMTCT services as well as supply of essential commodities. The typical timeline is depicted in Fig 2 below. Fig 2 shows the typical timeline of activities within a single state. Activities could however stretch into longer periods especially for such activities that required securing appointments or scheduling with political office holders. Activities were not being run simultaneously in all 8 states, therefore, activities in ‘week 1’ could in real dates, when aggregated, have spanned a total of eight to 12 weeks across the eight different states. Once training and activation were completed, the new PMTCT sites were handed over to site mentors. A quality mentorship and accountability program (Q-MAP) was instituted to provide routine mentorship of health care workers in these newly activated sites. These site mentors comprised of experienced multidisciplinary PMTCT technical experts (clinicians, pharmacists, laboratory scientists, monitoring & evaluation officers) from the supporting NGO and Government staff. They paid routine visits–at least fortnightly in the first two months and monthly thereafter–to review guidelines, standard operating procedures, client folders, and service delivery challenges. A unique feature of the roll-out phase was the inclusion of the private and non-formal health sector. The expansion incorporated specific strategies to engage privately owned health facilities while the non-formal health sector included traditional birth attendants and other community leaders. Further emphasis was made on strong demand creation given the low rates of health facility utilization for maternal and child health services. This paper reports data from health system processes. The subjects are not directly human but the health system processes that led to increased access for the population. Informed consent was not obtained as no clinical records were retrieved for this paper. Only routine health system data–service statistics–are presented. The routine data has no identifiers and is not linked to individual persons. The FHI 360 Office of International Research Ethics (OIRE) has determined this project does not meet the regulatory definition of research and/or research involving human subjects as defined under the Department of Health and Human Services Code of Federal Regulations [45 CFR part 46.102(d)(f)]. Given this determination, further review and approval of this project is not required. FHI 360 OIRE Project #:899555.