Over the past two years we assessed in detail five insecticide-treated net (ITN) distribution programs and two indoor residual spraying (IRS) programs operating at large or national scales in sub-Saharan Africa. The ITN programs were chosen to represent the major distribution systems being used in sub-Saharan Africa at present, as well as to represent a geographically diverse selection of countries. The two IRS programs were chosen largely based on scale and the accessibility of cost data. The ITN programs represent the following models: (1) free ITN delivery through integrated vaccination campaigns (Togo), (2) free ITN delivery through routine services and at the community level (Eritrea), (3) highly subsidized ITN delivery through routine services in the frame of a social marketing program (Malawi), (4) largely commercial sector promotion (Senegal), (5) commercial sector promotion through social marketing combined with vouchers to reach high-risk groups (Tanzania). The two IRS programs represent (1) a national program funded locally (KwaZulu-Natal, South Africa) and (2) an international intervention funded by donors and a public-private partnership (Lubombo Spatial Development Initiative, Southern Mozambique). Methodological details and data sources are given in the full report to be released soon.

In addition to operational descriptions of each program, cost and cost-effectiveness analyses were conducted. Costs were measured locally or derived from existing studies and focused on the provider perspective, supplemented by direct costs to users for net procurement. Effectiveness was measured by combining outputs with standard effectiveness indicators derived from the existing Cochrane review on ITNs.  Effectiveness for IRS was also based on these data because of the lack of sufficient IRS-specific data and the fact that past research had shown the effectiveness of the two interventions to be similar in several African settings.

This is the first time that large-scale comparative data are available on vector control programmes.

The results indicate that all the programs are cost-effective interventions and are attractive within low-income-country settings.  However, the deployment of long-lasting ITNs is clearly more cost-effective than IRS.  Among ITN distribution models, all except one are within a relatively narrow range of cost-effectiveness.  Summary results are shown in tables 1 and 2.

Use of conventional ITNs.  Annualized economic costs per conventional ITN distributed varied by about a factor of three across sites, i.e. from USD 3.23 per net in Togo to USD 8.05 in Senegal (less variation was seen in pure provider costs).  The cost per net-year of protection ranged from USD 1.43 in Eritrea to USD 6.05 in Senegal. Under this scenario, the cost per DALY averted ranged from USD 36 to USD 89, and the cost per death averted from USD 1,174 to USD 2,926 (in Togo and Senegal, respectively).  

Use of long-lasting insecticidal nets (LLIN).  For an LLIN giving 3 years of protection, the costs per net distributed were generally higher than for conventional nets because of the higher purchase cost for these nets.  However, the cost per treated-net year of protection was generally significantly lower, ranging from USD 1.48 in Eritrea to USD 2.64 in Senegal.  As a result, the cost per DALY averted ranged from USD 16 in Eritrea to USD 29 in Senegal and the cost per death averted ranged from USD 539 to USD 960.  In almost all cases the cost per death and DALY averted were even lower when a LLIN with a 5 year life and 7 USD price was assumed.   Clearly, the use of LLIN is a very attractive public health investment.

Costs per person-year protection by IRS were similar if the protection of the whole population was considered: the costs were USD 3.27 in KwaZulu Natal and 3.90 in Mozambique.  However, the cost per under five child protected was substantially higher for IRS than for ITNs, because IRS can not be targeted only to high-risk age groups in which expected health benefits are maximized: USD 23.96 (KZN) and USD 21.63 (Mozambique).  As a result, cost per DALY averted (USD 119-132) and per death averted (USD 3,933-4,357) were also substantially higher for IRS compared to ITNs.  Unfortunately we were not able to include the benefits of protection for adults as the effects of vector control in such groups are poorly quantified.  However, in highly endemic areas under five children account for the vast majority of the malaria disease burden and effects from adult protection are likely to be minimal in comparison.   

Programs differed greatly in their potential for rapid ITN scale up, with Togo clearly demonstrating characteristics of a good “catch-up” strategy while the other four models were predominantly “keep-up” strategies - though all programs produced meaningful changes in ITN coverage and usage.
 
Sensitivity analysis indicated that program parameters with the largest effects on overall cost and cost-effectiveness ratios for ITN programs were those related to either the length of protection offered by ITNs (or re-treatment kits) or with the usage rates of nets by children, additionally the cost of ITNs also had a large effect. These results suggest the clear potential for improving the cost-effectiveness of programs by switching to long-lasting insecticide-treated nets (LLINs) and stronger netting, targeting nets to highly vulnerable groups, and heeding the importance of behavioral factors associated with net programs.  ITNs and insecticide commonly accounted for approximately 60% of the total costs of programs.

For IRS programs, sensitivity analysis revealed that the most important parameters were either those relating to the compliance/acceptability of spraying or with insecticide choice and the number of spraying rounds required per annum. For these programs, insecticide and staff costs accounted for the largest share of overall economic costs.

Clearly, some strategies were more cost-effective than others but each strategy also brings specific advantages, both in the short-term and in the long-term.  It is the consideration of these many mainly local factors, based on the cost-effectiveness results presented in this report, which should guide national planers.

Clearly, both IRS and ITN vector control programs are excellent public health investments and should be expanded rapidly. This is a time of unprecedented opportunities for malaria control, with growing global interest and resources, as well as increased commitment by endemic country governments. It is time to bring down substantially the unacceptable burden of disease due to malaria.

Suggestions for further reading:

Tan-Torres Edejer T, Baltussen R, Adam T, Hutubessy R, Acharya A, Evans D, Murray C, 2003.  Making Choices in Health: WHO Guide to Cost-Effectiveness Analysis. Geneva: World Health Organization.

Lengeler, C. & Sharp, B, 2003. Indoor Residual Spraying and Insecticide-Treated Nets. In Global Health Council (Ed.), Reducing Malaria's Burden: evidence of effectiveness for decision makers (pp. 17-24). Global Health Council.

 

Joshua Yukich

Fabrizio Tediosi

Christian Lengeler