Attachment J.1.2. ME.3.A.6 Tanzania routine distribution costing report_USAID APPROVED.docx

Cost Analysis Series Routine Facility-Based Distribution in Mainland Tanzania Prepared by: Janna Wisniewski and Joshua Yukich October 2017

This report is made possible by the generous support of the American people through the United States Agency for International Development (USAID) and the President’s Malaria Initiative (PMI) under the terms of Cooperative Agreement AID-OAA-A-14-00057. The contents do not necessarily reflect the views of USAID, PMI, or the United States government.

Contents

Abbreviations

Abstract The cost-effectiveness of long-lasting insecticidal nets (ITN) in reducing morbidity and mortality is now well-documented. To increase the utilization of ITNs in mainland Tanzania, the government has initiated distribution of free ITNs to pregnant women and children receiving routine immunizations in public health centers and dispensaries. This study was conducted to estimate both the economic and financial costs of net distribution from the provider perspective during the first twelve months (May 2016 through April 2017). Cost data were collected retrospectively from financial and operational records kept by VectorWorks, the NMCP, and interviews with program managers, implementers, and health facility staff involved in routine facility-based distribution activities in mainland Tanzania. The financial cost per net distributed was estimated to be USD 7.50, and the economic cost per net distributed was estimated to be USD 7.95. Financial cost to international donors to fully fund the procurement and distribution was estimated to be USD 3.44 per ITN and the Purchase/Discounted value per ITN was USD 2.64. The remainder of these costs ($1.42) were contributed by the Government of Tanzania.

Background and Purpose The cost-effectiveness of long-lasting insecticidal nets (ITN) in reducing morbidity and mortality is now well documented. To increase utilization of ITNs in mainland Tanzania, the government has initiated distribution of free ITNs to pregnant women and children receiving routine immunizations in public health centers and dispensaries. PMI VectorWorks, a global USAID-funded project implemented by the Johns Hopkins Center for Communication Programs, supports this facility-based distribution. During the study period, the program was phased into seven regions throughout Tanzania.

Methods The cost analysis followed existing guidelines for the cost evaluation of ITN programs, as well as methodologies similar to previous studies in order to maximize comparability.

Intervention Description A description of the intervention was developed based on direct observation, document reviews, and key informant interviews with stakeholders at the NMCP and partner organizations.

Time Frame and Perspective Data collected on costs and outputs covered the time-period between May 2016 and April 2017. The study uses the provider perspective; household-level costs were not included. In cases for which the cost data provided did not align with the study timeframe, costs were standardized to a one year time period.

Types of Costs Included All direct costs of the Tanzanian facility-based distribution channels to the provider were included, including the cost of commodities, health promotion, transport, and payment of salaries and per diems.

Data Collection Cost data were collected retrospectively from financial and operational records kept by PMI VectorWorks, the NMCP, and interviews with program managers, implementers, and health facility staff involved in routine facility-based distribution activities in mainland Tanzania. Data were collected by examining agencies’ financial records, including budgets, expenditure records, reports, receipts, and invoices, as well as operational records such as reports and published documents. Data was collected in June 2017.

When costs or resource inputs were not reflected in financial or other records, information was collected through stakeholder interviews. The ingredients approach was used whenever possible, meaning that activities or inputs were first determined, and then a unit cost was assigned, allowing for an estimated total cost. If the information was deemed too sensitive or was not available in adequate detail, aggregated expenditure was generally used instead. In cases in which salary information could not be shared, the WHO-CHOICE database3 was used for valuation purposes after conversion to non-purchasing power parity-adjusted U.S. dollars.

Resource use was valued at four levels: (1) nationally or internationally, (2) the region level, (3) the district level, and (4) the health facility level. Resources were valued based on the reported expenditures or budgets, and in the case of shared personnel, on salary plus any fringe benefits.

Interviews were conducted with the NMCP, Mtwara regional office, two districts within Mtwara (Mtwara District Council and Mkindani), and two health facilities within each of the two districts. Sites were selected by PMI VectorWorks program staff in consultation with regional and district staff. Average costs for each activity were calculated from the selected areas and applied to all other regions, districts and facilities included in the program as if these were mean costs per unit (i.e. the observed units were treated as if they were typical units).

Cost Classification and Adjustments Costs have been divided into capital and recurrent[footnoteRef:1] costs, based on the lifetime of the good or service being purchased. Capital costs have been discounted in the economic analysis using lifetimes and discount rates determined through stakeholder interviews, expert information, and past literature. Varying discount rates and lifetimes were examined in the sensitivity analysis. Both financial and economic analyses were conducted to show (1) financial costs[footnoteRef:2]—what the actual expenses of running a program would be, and (2) economic costs[footnoteRef:3]—the value of all resource use over the study period. In the financial analysis, capital costs have not been discounted and are instead incurred in full at the time of the purchase. Costs were collected in U.S. dollars or in Tanzanian Shillings (TSH). Costs collected in TSH were converted to U.S. dollars based on official yearly average exchange rates for the period during which the costs were incurred. They are reported here in 2016 USD. A distinct financial cost called donor budget cost was also calculated. This cost was determined by excluding resources contributed by the local government or community and calculating a financial cost of the remaining resources. It is intended to reflect an estimate of budget requirements for international or other donor agencies to fund similar projects. [1: Capital costs are costs incurred to purchase goods or services with a life span greater than one year. Recurrent costs are costs incurred for goods or services lasting less than one year.] [2: Financial costs are a measure of expenditures as they were incurred by the provider of the intervention.] [3: Economic costs are a measure of resource use by the provider of the intervention rather than financial expenditure. As such they include a valuation of donated goods as well as annuitization of capital goods.]

Due to the fact that the seven regions began distribution at different times throughout the first year of the project, district, region, and facility costs are adjusted to reflect shorter periods of operation as necessary.

Outputs The primary output measure used in the analysis was the number of nets distributed to beneficiaries. This data was collected from PMI VectorWorks records disaggregated by district.

The number of nets distributed was used to calculate a second combined output measure—person-years of protection (PYP), which assumes that one net can cover two people for a period of three years in the base case scenario.

The primary output was also used to calculate a third combined output measure—cost per treated-net year (TNY), which assumes a lifetime of three years per net in base case scenario calculations.

Base Case Scenario In this analysis, the base case scenario relies on the following set of assumptions: a discount rate of 3% has been applied to economic capital costs, and each ITN provides three years of protection for two people. The cost of nets was based on the Free on Board (FOB) price of the nets.

Results Intervention Description Between May 2016 and April 2017, USAID procured 872,652 ITNs and distributed 793,320 of them through routine facility-based distribution channels. USAID currently imports the health facility ITNs to Tanzania through the Global Health Supply Chain (GHSC) project[footnoteRef:4], and the PMI VectorWorks project contracts Simba Logistics to transport them from the port to one of four private warehouses. After clearing customs (done by GHSC), PMI VectorWorks manages the nets from their arrival at the central warehouses through to their transport to health facilities (i.e. health centers and dispensaries). A private-sector logistics partner is contracted to physically transport the ITNs to health facilities (in contrast, other countries often transport ITN to the district level, which are then responsible for transporting them to health facilities, and for finding resources to do so). One major advantage of this system is that ITNs are not left at a half way point with little certainty of when they will make it to the intended facilities, reducing the chances of stock-outs and opportunities for diversion. The project began with a “Smart Push” strategy in which each health facility received an initial 6-month supply of ITNs. The numbers that each facility received were planned based on patient volume data. Following the “Smart-Push,” facilities report quarterly on the number of nets they have distributed and are re-stocked accordingly. The nets are given free-of-charge to pregnant women at their first antenatal care visit, and to children under five when they receive their nine-month measles vaccination. There is no means-test for recipients to receive an ITN. [4: Prior to October 2016, USAID DELIVER was the project responsible for procurement.]

Outputs The total numbers of ITNs distributed in each district during the study period are shown in Table 1.

Table 1. Number of ITNs distributed through routine facility-based channels, by district (May 2016-April 2017)

TOTAL

793,320

Overall financial and economic costs, and cost per treated net-year Table 2 shows the financial and economic costs overall and per TNY. The financial cost per net distributed was estimated to be USD 7.50, and the economic cost per net distributed was estimated to be USD 7.95. The costs per treated net-year were substantially lower than the costs per net distributed. Given the assumption of a three-year net life, this is to be expected.

Table 2. Financial and economic costs

Distribution cost per ITN

Abbreviation: TNY, treated-net year Cost per person-year of protection The financial and economic cost per PYP is shown in Table 3. The costs per TNY were significantly lower than overall cost per net distributed. Given the assumption of a three-year net life and two persons protected per net, this is unsurprising.

Table 3. Financial and economic cost per person-year of protection Total number of ITNs issued

Costs The total financial and economic costs of Tanzania’s routine facility-based distribution scheme are shown in Table 4. The total financial cost for the system was nearly USD 6.0 million. The total economic cost for the system estimated to be almost USD 6.3 million.

Table 4. Financial and economic costs by contributor

Figure 1 illustrates the breakdown of financial costs of Tanzania’s routine facility-based distribution scheme by contributor. The purchase of ITNs (which is managed by USAID) comprised the second largest portion of the financial costs, at 35 percent. Transport (managed by PMI VectorWorks [Johns Hopkins Center for Communication Programs and partners]) contributed the most at 46 percent of financial costs, with the various levels of government contributing the remaining 19 percent. The majority of the government’s financial costs were incurred at the facility level.

Figure 1. Breakdown of financial costs by contributor level

Figure 2 shows the economic costs by contributor. The proportions were largely similar to those in the financial analysis, with ITNs constituting 35 percent of economic costs. Donors were estimated to contribute 47 percent of economic costs, and the government contributed 18 percent.

Figure 2. Breakdown of economic costs by contributor level

Table 5 displays the allocations of financial costs by activity and cost item. The cost of purchasing ITNs, which constituted 35 percent of the total financial costs, is not included in this table. Outside of procurement, therefore, the management and operations (of which personnel and indirect cost made up the largest percentages) and ITN distribution (contracted to Simba Logistics) were the two largest cost activities, at about one third of total financial costs each. Training, planning and supervision, and monitoring and evaluation made up 13, 10, and 8 percent of financial costs not related to ITN purchase, respectively. Overall, the largest cost items were personnel (44 percent) and transport (42 percent). This data is shown graphically in Figure 3.

Table 5. Percentage of total financial costs by activity and cost item

Personnel

Transport

Indirect

Other

Supplies

Figure 3. Percentage of total financial costs by activity, broken out by cost element (including ITN)

Sensitivity Analysis One-way sensitivity analysis was used to determine whether any of the main assumptions involved in cost estimation were responsible for substantial changes in the estimated financial and economic cost per net distributed. The main uncertain parameters which were addressed in sensitivity analysis were the discount rate, useful life of an ITN, number of person years protected by an ITN. The costs at various levels of government were also tested, as these costs were determined largely through interviews rather than cost reports. The results from the sensitivity analysis are summarized in Table 6.

Changes to the discount rate do not seem to have large impacts on the outcomes of interest. Changing the total net years impacts person years of protection but not the financial or economic cost of distributing a net. Unsurprisingly, facility costs have the largest impact on all outcomes. This occurs most dramatically when considering the financial and economic cost per net distributed, which, if facility cost were doubled, would increase by USD 1.09 and USD 1.08 respectively, for a total of USD 8.59 and 9.03.

Table 6. One-way sensitivity analysis of selected parameters

Abbreviations: PYP, person-years of protection; TNY, treated-net year (i.e. median net lifespan).

Discussion and Recommendations This study estimates the cost of the first year of a routine facility-based ITN scheme in mainland Tanzania. Findings indicate that the main drivers of financial costs are the purchase and importation of ITNs followed by operational costs which include moving nets to health facilities, training and technical support, supervision, and monitoring and evaluation. Donors contributed an estimated USD 6.08 per net; USD 2.64 of which covered purchase and importation. This leaves a donor financial cost of USD 3.44 per ITN distributed. This amount is more than double the amount estimated for the ITN distribution through the CD system in Zanzibar, which was USD 1.46. However, this estimate is in line with the financial costs of other delivery systems for ITNs[endnoteRef:1]. The White et al review found that the “median financial cost of protecting one person for one year was USD 2.20 (range 0.88-9.54) for ITNs.” This result is similar to the base case analysis of USD 1.25 per person year of protection. [1: White, M et al (2011) Costs and cost-effectiveness of malaria control interventions - a systematic review. Malaria Journal 10:337]

Opportunities for reducing costs are limited because routine facility-based distribution, though revised, is not new to Tanzania and has been generally well-integrated into the existing health system. Some efficiencies might be gained by incorporating what is currently stand-alone supportive supervision for routine facility-based distribution into multi-topic, comprehensive integrated supportive supervision activities. Similarly, health worker orientations could be incorporated into integrated supportive supervision, particularly since best practices for ITN stock management can be applied with other commodities.

Excluding the cost of the purchase of nets, the Tanzanian government appears to be contributing approximately USD 1 for every USD 2.56 worth of financial and economic resources given by the foreign donors. However, as stated above, the government costs rely heavily on interview data rather than cost reports. Future costing studies would benefit from improved information as to government salary schedules.

Conclusions The financial cost per net distributed was estimated to be USD 7.50, and the economic cost per net distributed was estimated to be USD 7.95. Financial cost to international donors to finance the procurement and distribution of ITNs was estimated to be USD 6.08 per ITN. The government contributes a considerable percentage of the resources needed, particularly in the areas of supervision, monitoring and evaluation, and health education with patients and communities, and distribution to patients.

NMCP Region District Facility LLINs JHU and partners 1.0093084669994999E-2 1.4091161329226699E-2 2.05553300591821E-2 0.144675855586746 0.351938139453286 0.45864642890156399

NMCP Region District Facility LLINs JHU and partners 9.5293453215320803E-3 1.3626105685395899E-2 1.9781238758641701E-2 0.13678792361851899 0.35241424647193298 0.467861140143979 image2.png image3.png image1.jpeg