Project design and approach
Connect operates as a randomized cluster trial (RCT) allocating 50 villages to intervention (villages receiving WAJA) and 51 to comparison (villages not receiving WAJA). All 101 villages were located in Kilombero, Ulanga and Rufiji districts of Tanzania where the IHI Health and Demographic Surveillance Systems (HDSS) monitored population dynamics since 1996 to allow for an evaluation of the project’s impact on mortality [31, 32]. Further, the Connect economic evaluation aims to assist the MoHSW to determine and model the cost of recruitment, training and deployment of WAJAs as envisioned in the MMAM [27] and corresponding ongoing policy development.
By August 2012, 113 WAJA were trained and deployed in the 50 villages of intervention. Training and deployment were pursued in two phases, with WAJA deployed to villages in August 2011 (Batch one = 57 WAJA) and August 2012 (Batch two = 56 WAJA). Analysis included two phases of WAJAs deployment. The project provides funds to districts for human resources, and supplies are directly provided by the project. WAJAs link communities to the health system by providing basic curative and preventive health services, health education, promotion and referral to the health care facilities [32].
Recruitment of WAJA
Recruitment of WAJAs involved community sensitization that informed the community about the project and necessary qualifications for selected individuals for WAJA training. Selection criteria included residence in the community, at least four years of secondary education, and minimum pass in at least two subjects. Other recruitment activities included posting advertisements to notice boards and other important places in communities. Applicants sent their applications to the village executive officer. With the help of the district education office, the village committee screened applicants’ certificates. At a subsequent meeting, qualified candidates were asked to brief community members about their suitability for the position, after which community members voted to select the WAJA(s) for their area. One to four WAJAs were selected from each community, based on village population size.
Training design and process
The WAJAs training was held at the TTCIH and all students from Rufiji, Ulanga and Kilombero settled at the training center in Ifakara town. The training spanned nine months and was divided into two sections. The first involved theoretical training that included basic clinical skills, community health skills such as communication and advocacy, counseling, disease prevention, health education, nutrition, primary health care and human biology. A field practice component was also incorporated in the first section of the training, conducted in the villages surrounding the training institution with support from WAJA trainers. The second training section was a field attachment, done in the WAJAs’ home villages. This involved practicing their skills under the close supervision of trainers and a clinical officer from the nearest health facility. WAJAs were then invited to the TTCIH for a final writing and oral examination to test competencies gained during the training. The TTCIH incurred a number of preliminary costs when preparing to accommodate the training such as the renovation of dormitories, hiring a new building for dormitories outside the school, developing curriculum and capacity building for trainers.
WAJA’s deployment
Deployment proceeded immediately after WAJA graduated and passed their final examination. Community members organized a village meeting where the project presented a fully trained WAJA with the necessary equipment (bicycle, mobile phone, weighing machines, flip charts, and malaria rapid diagnostic test kits) and medicines (Amoxycillin, Artemeter Lumefantrine (ALu) as an Artemesinin-based Combination Therapy (ACT) known as Coartem, oral rehydration salts (ORS), Zinc oxide plaster etc.). Communities provided a cupboard for storing WAJA medicines and supplies. The district health administration was responsible for providing WAJA with an employment appointment letter and administering all human resource matters, including WAJA salaries (channeled to the district by the project, but paid by district authorities) and social security benefits. WAJAs were supervised by intervention coordinators from the project and by the district health administration. A selected HW from a nearby health facility provided technical advice and a village executive officer oversaw WAJA activities at the community level. These supervisors met with WAJA at least once every three months. WAJA phones were also connected to close user group service (CUG), allowing free calls between them and their supervisors for immediate support.
Data sources
Data on training costs were obtained from the actual expenditures related to schooling costs at the TTCIH. These were extracted from expenditure accounts and improved by in-depth interviews with TTCIH management to provide necessary clarifications. All information was confirmed by records review and interviews of relevant TTCIH and Connect project staff. Data on the cost of deployment and annual operation costs of WAJAs were extracted from the Connect project records. Information was gathered from expenditure records of Connect interventions and the actual cost incurred for specific activities like training, salary, supervision, medicines and supplies. The number of WAJAs per village and village population size were obtained from project records. Resources expended for on year of WAJA operations were tracked from the expenditure accounts and records of quantities of items supplied directly to the WAJAs. The national population and village data were drawn from the Tanzania 2012 Census.
The relative training costs of comparable MoHSW operations were gathered from the Clinical Officer Training Centre of Kilosa in Morogoro, a public institution that specializes in two-year training certificates for Assistant Clinical Officers, and three-year diplomas for the lowest level paramedics, Clinical Officers, who provide preventive and treatment services at primary health care facilities. Public Clinical Officer Training Centres were selected due to their future responsibility in scaling up the WAJA training program should it be adopted as policy. In-depth interviews with management and financial officers were conducted at the Clinical Officer Training Centre. These, in combination with records review, helped determine the costs of all pre-service training, logistics and operations associated with certificate and/or diploma procedures for one academic year. Information from the Kilosa Training Centre in Morogoro Region, the public institution that trains the lowest level of paramedics, was similarly gathered. The training costs of Assistant Clinical Officers were ultimately used as a benchmark, as this cadre is most similarly comparable to the education level required for the WAJAs.
Data analysis
Analysis involved valuing each resources used in every activity related to WAJA training, deployment and annual running costs. Interview responses were coded and grouped into specific groups to clarify mentioned categories. Actual costs from the records were then combined based on this information.
The unit training cost of WAJA was calculated as the total cost of establishing WAJAs divided by the number of WAJAs trained by the project. The unit training cost of Assistant Clinical Officers was calculated as the annual cost of training divided by the number of trainees.
Modeling scale-up training cost using project expenditure and clinical training centre information followed, and was developed according to seven assumptions :i) The training of WAJA will be undertaken by existing government training institutions and will be within their existing competence; ii) Tuition allowance, training facilities, and utilities are to be included into an overall tuition fee; iii) Curriculum and identity card costs are included in tuition fees; iv) Meals are provided for students undergoing training; v) Transport costs for WAJA to travel to training and return to their villages are not included in the estimates; vi) Students will have two months of field attachment; and vii) The district health system is the unit of planning. After deploying WAJAs to the health system at the village level, resources to run the program for a full year were tracked and related costs recorded to determine annual running costs for the program.
Our estimation was based on the number of WAJAs required to serve: 1) A standard-sized district in Tanzania; and 2) The entire country. This estimate was based on the population of the villages as assessed in Connect project. Villages with populations less than 1,000 people received 1 WAJA; villages with a population of 1000 to 4000 received 2 WAJAs; villages with a population of 4000 to 7000 received 3 WAJAs; and those with more than 7000 received 4 WAJAs. To determine the standard size of a district, villages in project districts were summed and averaged to have a mean number of villages in a “standard-sized district”.
The trial is registered with the International Standard Randomized Controlled Trial Register (number ISRCTN96819844).