This investment case projected the financial requirements of the malaria program to reach malaria elimination by 2030 and values the economic and financial returns of reducing malaria transmission compared to alternative scenarios. To accomplish this, the investment case leveraged multiple methodologies and data sources. The study design incorporated a variety of quantitative methods: numerical and regression techniques to develop a transmission model to predict the epidemiological impact of various interventions used for malaria control and elimination and economic analysis to estimate the cost and economic impact of the interventions nationally and regionally. The economic analysis was informed by the outputs of a transmission model. All monetary figures are expressed in 2018 constant US dollars (USD) [8].
This investment case projected the financial requirements of the malaria program to reach malaria elimination by 2030 and values the economic and financial returns of reducing malaria transmission compared to alternative scenarios. To accomplish this, the investment case leveraged multiple methodologies and data sources. The study design incorporated a variety of quantitative methods: numerical and regression techniques to develop a transmission model to predict the epidemiological impact of various interventions used for malaria control and elimination and economic analysis to estimate the cost and economic impact of the interventions nationally and regionally. The economic analysis was informed by the outputs of a transmission model. All monetary figures are expressed in 2018 constant US dollars (USD) [8].
Epidemiological model framework
A spatially explicit, compartmental, nonlinear, ordinary differential equation model is an extension of previously published models and have been implemented in R and C++ [9, 10]. The model simulated a range of malaria interventions and estimated their impact on the transmission of Plasmodium falciparum malaria between 2018 and 2030 nationally and in the three ecological zones in Ghana (Coastal, Forest and Savanna). Historical data from 2012-2018 was used to parameterize and fit the model. The key transmission features and drivers of transmission in the model included:
- Infection classes dependent on the level of severity of infection
- Development and loss of immunity against clinical infection
- Superinfection
- Subnational climatic variation (seasonality)
- Importation of infection
More details on the model have been published elsewhere [9, 10, 11].
Interventions modelled included:
- Passive Case detection (routine diagnosis and treatment in health facilities and the community)
- Vector Control:
- Distribution of LLINs
- IRS
- Health System Strengthening (supportive supervision, training for improved malaria testing and treatment and supply chain management support)
- Social and Behavioral Change (SBC) for improved health seeking behavior
- Seasonal Malaria Chemoprevention (SMC)
Data sources used were:
- Data from the NMCP (monthly incidence and deaths by district from the Health Management Information System (HMIS))
- WHO World Malaria Reports and Annexes
- Peer reviewed literature
- Expert opinion (for assumptions where data was unavailable)
Four scenarios and two reverse scenarios were developed in collaboration with the NMCP:
- Baseline scenario: existing set of malaria control activities as implemented in 2018 with coverage levels of 2018
- Fully funded response (FFR): fully funding the activities and coverage levels in the NSP
- Better use of nets: added to the “fully funded response” through the use of SBC (social and behavioral change) to increase the usage of LLINs
- Increase treatment seeking from 73% to 90% (through SBC): added to the “better use of nets” scenario.
- Reverse scenarios: Reduce the amount of funding for the implementation of activities from the 2018 baseline coverage. Where:
- Reverse 1: cutting out IRS, SMC and LLINs by 50%
- Reverse 2: cutting out IRS, SMC (LLINs remain at 2018 levels)
Table 1. describes the scenarios in detail.
Economic analysis
Using a societal perspective and cost of illness approach [11, 12], the economic burden of malaria in 2018 was evaluated. Specifically, (i) direct health system costs, (ii) direct household costs, and (iii) indirect costs were estimated. Table 2 illustrates the framework used.
Direct health system costs
To facilitate the gathering of direct cost data, an interview guide and data entry sheet were developed to collect existing costing data, identify gaps, and locate additional data to fill gaps. These interviews were conducted in a semi-structured format with key malaria program personnel who were familiar with program spending patterns and records. Data on government and external spending were collated. National health system costs outside of vertical malaria program expenditures were included as much as possible to obtain the total actual cost to the health system in Ghana. When expenditures were unavailable, budget figures, National Health Accounts (NHA) and secondary sources such as peer-reviewed or grey literature or deduction were used. Costs of treating outpatient and inpatient malaria cases were obtained from the NHA [13, 14].
Individual costs were extracted and aggregated to obtain estimates of the costs of each intervention. The cost of each scenario was estimated using a cost estimation model fed by outputs of the transmission model. The cost of each scenario was then used to obtain the incremental or additional cost of a fully funded response compared to the baseline. All costs were discounted at 7%. The discount rate used was based on the inflation rate and the expert opinion of economists in-country. Additional table 1.1 contains the cost inputs used in the analysis.
Direct household costs
Malaria exacts a significant financial burden on households. Malaria patients often pay for transportation to access health facilities, diagnostic services, and medicines. In Ghana, although testing and treatment for malaria are free, prepaid or covered by the NHIS, malaria patients still incur out-of-pocket expenditures (OOP) for transport, food and other expenses not covered through the public sector. To estimate direct household costs on malaria, the number of reported OP and IP malaria cases in 2018 was multiplied by the mean OOP spending (separately for OP and IP cases). Data on OOP was obtained from published literature [15, 16].
Indirect Costs
The economic impact of malaria extends beyond the health system. Patients forego income while recovering from malaria, and caregivers looking after ill children and the elderly also lose out on potential earnings. Society also incurs an indirect cost due to premature deaths through losses in lifetime productivity and in the social value people place in living longer, healthier lives.
To evaluate the economic impact of malaria-related morbidity, the foregone income of malaria patients and caregivers was calculated. The gross domestic product (GDP) per capita per day was obtained from 2018 GDP estimates from World Bank Data [5]. The resulting figure was used as a proxy for the average income per capita and multiplied by the duration of OP and IP illness from published literature and the number of reported OP and IP cases. In addition, the effect of reduced productivity from “presenteeism” was calculating by assuming that adults retuning to work would be 50% less productive for an additional six days. This assumption was made based on interviews in Ghana.
A full income accounting approach was used to quantify the economic impact of premature death as postulated by the Lancet Commission on health [17]. Assuming 40 years as the average age of malaria-related adult deaths and 2.5 years as the average death amongst children under 5 years, the average remaining life expectancy of males and females was multiplied by the value of each additional life year (VLY). Life expectancy was retrieved from the Central Statistics Service [18]. One VLY was assumed to be 4.2 times the 2018 GDP per capita of Ghana [17].
Cost savings from reduced public and private expenditures on malaria are likely to spur consumer spending and create new businesses thus injecting more money into the local economy. Throughout the process, overall disposable incomes increase, creating more markets for local businesses. These induced responses result in an economic multiplier or “ripple” effect. A 2011 USAID report [19] estimated that income multipliers in West Africa lay between 1.58 and 2.43. An average multiplier of 2 was therefore used for the purposes of this analysis.
Economic benefits estimation
The mortality and morbidity averted from malaria elimination were obtained by subtracting the estimated cases and deaths in the fully funded scenarios from the corresponding outputs of the “business as usual” scenario. Similarly, the excess cases and deaths in the reverse scenarios were calculated by subtracting from the corresponding outputs of the “business as usual” scenario. These health benefits were calculated using the methodology and inputs previously outlined.
Direct costs averted to the health system includes costs associated with diagnosis and treatment costs of IPs and OPs;
- Direct cost averted to the individual households is out-of-pocket (OOP) expenditures for seeking care; and
- Indirect cost averted to the society due to patients’ lost productivity due to premature death and morbidity and caregivers’ reduced economic output.
The benefits of investing and not investing in malaria control and elimination were estimated as the sum of the direct cost savings to the health system from reduced use of outpatient and inpatient health services and reduction in cost of delivering malaria control activities; the direct cost savings to households; and the indirect cost savings of reduced morbidity and mortality from malaria calculated above.
The Net Present Value (NPV) was calculated to obtain the present value of the future revenue generated from elimination using standard economic techniques. The purpose was to give a true picture of the financial value of an investment made today whereby savings would be accrued in the future [12]. The timeframe used for calculating the NPV was 11 years and a 7% discount rate was applied as before.
Return on investment
To calculate the ROI from malaria investments, the NPV of the benefits of reduced transmission were subtracted from the incremental cost of elimination. The resulting figure was divided by the incremental cost of the fully funded response (compared to baseline). The ROI is interpreted as the economic return from every additional dollar spent on malaria above the business as usual scenario.
Financial gap
Various sources were consulted to estimate past, present, and future financing for malaria. Projected financing was estimated using projected figures from GOG, the Global Fund and PMI. Many of malaria services are covered under the NHIS via the health insurance levy. These estimated resources are included in under “domestic financing” (obtained from the NMCP).
Sensitivity analysis
A stochastic sensitivity analysis on the epidemiological and cost outputs of the malaria transmission model was performed. The minimum, median, and maximum malaria cases and deaths predicted by the model for each scenario were used to calculate the minimum, median, and maximum costs. Three hundred random samples were drawn, which generated a range of costs. From the range of costs generated, the minimum, maximum, median, mean, and other percentiles are presented.
Data collection, tools and analysis
A worksheet was been developed in Microsoft Excel® to facilitate the organization of cost data. Analysis of the cost data was conducted in Microsoft Excel to estimate the current and future costs of the malaria activities in each scenario. All quantitative data records (no identifying information), were stored in Microsoft Excel spreadsheets on encrypted, password-protected computers. Data was collected in August 2019.
Ethical approval
Ethical approval for this study was obtained from the Ethical Review Committee (ERC) of the Ghana Health Service prior to data collection (GHS/RDD/ERC Ref No. 1913445).
Study limitations
A number of known and unknown factors limit the findings of this report. Due to time and resource constraints, the transmission model estimated sub-national malaria transmission based on three climatic zones. Ideally, higher levels of spatial heterogeneity would be modelled to provide to enable subnational estimates of interventions and costs.
The costs of interventions have been estimated based on available data from the NMCP and proxies when data were unavailable. For example, the costs of outpatients, in-patients and health worker salaries were estimated from the National Health Accounts (NHA). Separating out the cost of interventions in integrated systems is challenging and the analysts have relied on country-level partners to arrive at disaggregated costs. This report utilized reported cases from the HMIS and estimated cases and deaths from WHO World Malaria Reports. The wide variation in these two estimates of burden makes it harder to be sure of the resources required to eliminate the disease.
As Ghana moves closer to elimination the impact of active surveillance on both the epidemiology and cost will need to be incorporated. This was not included due to a lack of historical data to enable fitting the model for impact or cost. The savings observed may well be offset by the increased costs of active surveillance required in elimination settings. At the same time, targeting of interventions rather than ubiquitous coverage to the entire country may reduce the costs of elimination and the financing gap. Without subnational estimates of incidence and coverage, targeted interventions are difficult to estimate and cost. Without an informed and complete understanding of the detailed current cartography of malaria risk and prevalence, future projections of the cost of eliminating malaria face an overwhelming uncertainty.
While employee absenteeism was included in the estimates of benefits, the analysis did not include the economic benefits conferred by reductions in school absenteeism and subsequent improvements in cognitive development due to the lack of empirical evidence to enable converting these estimates to wages earned. Other benefits not included include potential benefits on tourism, the impact of economic development and housing improvements on malaria transmission as well as regional or cross border externalities.