Cost and cost effectiveness of long-lasting insecticide-treated bed nets - a model-based analysis

Abstract

Background: The World Health Organization recommends that national malaria programmes universally distribute long-lasting insecticide-treated bed nets (LLINs). LLINs provide effective insecticide protection for at least three years while conventional nets must be retreated every 6-12 months. LLINs may also promise longer physical durability (lifespan), but at a higher unit price. No prospective data currently available is sufficient to calculate the comparative cost effectiveness of different net types. We thus constructed a model to explore the cost effectiveness of LLINs, asking how a longer lifespan affects the relative cost effectiveness of nets, and if, when and why LLINs might be preferred to conventional insecticide-treated nets. An innovation of our model is that we also considered the replenishment need i.e. loss of nets over time.

Methods: We modelled the choice of net over a 10-year period to facilitate the comparison of nets with different lifespan (and/or price) and replenishment need over time. Our base case represents a large-scale programme which achieves high coverage and usage throughout the population by distributing either LLINs or conventional nets through existing health services, and retreats a large proportion of conventional nets regularly at low cost. We identified the determinants of bed net programme cost effectiveness and parameter values for usage rate, delivery and retreatment cost from the literature. One-way sensitivity analysis was conducted to explicitly compare the differential effect of changing parameters such as price, lifespan, usage and replenishment need.

Results: If conventional and long-lasting bed nets have the same physical lifespan (3 years), LLINs are more cost effective unless they are priced at more than USD 1.5 above the price of conventional nets. Because a longer lifespan brings delivery cost savings, each one year increase in lifespan can be accompanied by a USD 1 or more increase in price without the cheaper net (of the same type) becoming more cost effective. Distributing replenishment nets each year in addition to the replacement of all nets every 3-4 years increases the number of under-5 deaths averted by 5-14% at a cost of USD 17-25 per additional person protected per annum or USD 1080-1610 per additional under-5 death averted.

Conclusions: Our results support the World Health Organization recommendation to distribute only LLINs, while giving guidance on the price thresholds above which this recommendation will no longer hold. Programme planners should be willing to pay a premium for nets which have a longer physical lifespan, and if planners are willing to pay USD 1600 per under-5 death averted, investing in replenishment is cost effective.

Figure 1

Equally cost effective bed nets of different price, lifespan and type. Each bed net choice is as cost effective as the base case, the 3-year LLIN priced at USD 4. Programme cost is USD 1.71 per net delivered (and retreated) per annum, USD 0.85-0.86 per person (USD 3.05-3.06 per under-5) protected per annum, USD 554-556 per under-5 death averted and USD 16.8-16.9 per DALY averted.

Figure 2

Relationship between usage rate and programme cost effectiveness when programme cost is constant. The figure illustrates how cost effectiveness changes depending on how frequently nets are used. The base case is 50% usage among the general population and 70% among under 5s, which gives LLIN cost effectiveness as USD 0.86 per person protected and USD 3.06 per under-5 protected. If conventional nets are used instead (with the same physical lifespan and purchase price and base case assumptions about retreatment), cost effectiveness is USD 1.16 per person protected. A higher usage rate implies a lower cost per person protected.