Goodman and her colleagues (Goodman et al., 2004) studied the functioning of the retail market for antimalarials in rural areas of Tanzania (parts of the Kilombero, Ulanga, and Rufiji districts) in 2001/2002. As in much of sub-Saharan Africa, most antimalarials in the study sites are acquired directly through a retail shop. In this case, 54 percent of people seeking treatment for fever went to a shop, compared with the 38 percent who went to a health facility. There were no pharmacies in these areas, only drug stores and general shops or stalls where drugs are sold. General shops (one for every 273 people) far out numbered drug stores (one for every 4570 people), but the proportion of general stores that stocked antimalarials was low and falling, with the result that 85 percent of retail antimalarial sales were through drug stores.

Drug stores were competing with the public sector on drug availability, stocking effective antimalarials when government facilities had only chloroquine (when it was no longer effective), or had run out of antimalarials. However, price competition among private retailers was weak, with drug store markups for antimalarials ranging from two to more than six times in relation to an international reference price (although there is no firm rule about what price ratio is appropriate, a ratio greater than two is often taken as excessive). The reasons are most likely the obvious ones: there are few outlets, so little reason to keep prices low, and people tend to seek care close to home, further reducing competition. There also may be some implicit collusion to avoid antagonizing fellow traders.

The high prices (and other factors) contributed to poor quality care for people seeking treatment for malaria through the private sector. People often bought only partial courses of antimalarials, or bought only antipyretics, which are cheaper.

Unless preemptive steps are taken, the same problems of high retail markups and poor quality treatment are likely to occur when ACTs subsidized high in the distribution chain flow through the system. In the absence of major system changes, Goodman and her colleagues identified other measures that could be used to improve access, including:

• Encouraging over-the-counter sale of prepackaged ACTs with clear labelling and locally tested instructions

• Widespread distribution of a single branded product, or a limited number of quality-marked brands

• Retail price maintenance by price labelling and consumer education campaigns

• Encouraging sale of products in a wide range of outlets, including drug shops and general stores in remote areas

• A positive role for regulators in improving knowledge and encouraging regular stocking

It will be important, as well, to monitor prices and their effects on access to ACTs, once a subsidy is in place.