This study explored three dimensions of access to anti-malarials through the retail sector in a rural district in Kenya following the implementation of a nation-wide anti-malarial subsidy programme. Accessibility, availability and affordability anti-malarials in the retail sector was compared to the public sector facilities in the same district.
In this study, 'accessibility' is defined as physical access to an outlet where anti-malarials may be obtained. Alba et al reported that physical access was the most important factor in determining whether a patient got an effective anti-malarial. Patients living in a village with a shop or health facility were four times more likely to get an ACT within 48 hours of onset of fever . Based on the results presented here, anti-malarials may be more accessible through the retail sector when considering the abundance of malaria medicine retailers, their proximity to households, and longer opening hours than public facilities. On average, households travel twice the distance to attend a public health facility than to reach a malaria medicine retailer. Most public health facilities do not offer weekend opening hours. Hospitals and health centres do open on Saturdays and Sundays (hospitals only), but these facilities are few and located near the peri-urban centre, away from the rural communities. In contrast, 85% of the malaria medicine retailers offer Saturday opening hours and nearly half open on Sunday, a factor shown to contribute to treatment seeking in the retail sector in other studies .
AL has been the recommended first line treatment for uncomplicated malaria in Kenya since 2006. SP is still widely available and used, but resistance is common; more than 80% of infections are resistant  resulting in 22-40% treatment failure [21, 22]. Forty-four percent of malaria medicine retailers had AL in stock on the day of the survey, and an additional 20% reported stocking it but were out of stock that day. Thirty-nine percent of malaria medicine retailers reported that AL was the most commonly sold drug. In contrast, a study in a neighboring district showed almost no retail sales of AL in 2008, only two years before our study .
Only seven malaria medicine retailers were completely stocked out of anti-malarials on the day of the survey, although a quarter reported stockouts as a regular occurrence. SP and quinine were the most frequently stocked drug, together making up over 60% of all the drugs observed. On a positive note, very few malaria medicine retailers sold uncombined artemisinin and none reported carrying chloroquine.
In public facilities, AL, SP, and quinine injections were available. SP is provided to facilities for Intermittent Preventive Treatment in pregnancy (IPT-p) although it cannot be ruled out that SP is being given for febrile episodes. Although only five out of 13 public facilities reported regular stockouts of anti-malarials, anti-malarial shortages are a frequent problem in rural facilities in Kenya .
Prices of anti-malarials varied widely between types and between retailers. The same brandname sometimes varied by an order of magnitude between retailers. It is possible that medicine retailers procured the same brandname from different wholesalers or other outlets leading to price variability. It is likely that prices in the retail sector are not 'fixed' at the level of the consumer, particularly in the less formal retail shops, but rather the price may be adjusted based on the client's ability to pay. Other studies have shown that the decision about which drug to sell is often dictated by what the patient can afford. In our study, medicine retailers reported separating blister packs of pills and selling only the number of tablets that a client can afford. Overall, less effective anti-malarials like SP and amodiaquine were significantly less expensive than artemisinin combination therapies. Non-AL ACT was more expensive and less common. Uncombined artemisinin was more expensive than AL or other types of ACT. In contrast to our results, a study in Burundi found ACT in private retail shops to be less expensive than quinine and amodiaquine even without subsidies . In Kenya, the minimum wage is 2.5 USD per day. Although the average price of AL is roughly equivalent to one-day's wages, a criteria for affordability proposed by the World Health Organization, unemployment in Kenya is estimated to be 40% and most casual workers found in our study area do not work every day of the month. In the WHDSS area, the high price of ACT is undoubtedly still an obstacle to accessing effective treatment.
Pharmacies in Kenya are required to be registered and only registered pharmacies are permitted to sell prescription medicines. Even registered pharmacies in Kenya are not legally permitted to sell ACT over-the-counter without a prescription and unregistered medicine retailers are not permitted to sell ACT at all, although both practices are common. Under the current policy, the AMFm subsidy would only benefit patients who have a prescription either from a private facility or from a public facility where AL was unavailable. There is some misalignment between current policy in Kenya and the intention of the AMFm subsidy. Tanzania has designated a special cadre of accredited shops that are permitted to sell ACT over the counter. Establishment of Accredited Drug Dispensing Outlets (ADDOs) and provision of subsidized ACT through this channel has greatly improved and expanded access to ACT [8, 19]. However, the strategy hasn't significantly reduced the percent of older children and adults who purchase SP for their fever  and only 30% of drug shops carried AL . The survey used for this report did not specifically identify registered pharmacies, but according to the Kenya Pharmacy and Poisons Board there are only six pharmacies registered in the area, indicating that the large majority of medicine shops in our study were probably unregistered. Formalizing the sale of ACT over-the-counter in Kenya through specialized or trained drug vendors may increase the accessibility and availability of subsidized AL.
Providing subsidized ACT through the retail sector is intended to reduce the cost of effective drugs to below that of ineffective therapies and increase the number of fevers treated with an appropriate anti-malarial. A randomized controlled trial in Kenya demonstrated a dramatic improvement in the percent of children under-5 who received AL for their fever after deployment of subsidized AL in the retail sector . A pilot evaluation of subsidized ACT provided to wholesalers in three districts in Tanzania also showed dramatic improvements in the number of patients who purchased ACT .
This is the first report of the effect of the international subsidy of ACT through the retail sector. Country-wide retail sector subsidies for AL through the AMFm programme began in Kenya in August 2010, five months before our survey. Between August and November 2010, 1.2 million treatment courses of subsidized AL (a single brandname - Artefan) were procured and delivered to Kenya. National media campaigns were used to raise public awareness about the subsidy and the correct pricing. The price of a treatment course of AL under the subsidy is intended to be 0.5 USD (40 Ksh). In this study, subsidized AL was significantly less expensive than other brands, but still almost three times as expensive as SP. However, the subsidy is applied at the level of the wholesaler, and retailers set their own prices. In our survey, only one shop was selling subsidized AL at the recommended price.
Eleven percent of malaria medicine retailers in the study area stocked subsidized AL. Nearly all retailers with subsidized AL did not carry non-subsidized brands (either in-stock or reported out-of-stock) which suggests that these retailers may not have been selling AL prior to the subsidy and the subsidy may have expanded the availability of AL. In addition, retailers with subsidized AL were distributed throughout the study area (Figure 1) and were not concentrated in the town centre.
It is interesting to speculate about how accessibility, availability, and affordability impact treatment-seeking decisions for febrile illnesses. Clearly, accessing anti-malarials through medicine retailers is more convenient than attending a public health facility for most families. On the other hand, treatment is free in public facilities, although families must also weigh transportation costs and time investment to attend the public facility . It is also possible that patients perceive 'free drugs' as less desirable or ineffective [26, 27]. During frequent drug shortages in public facilities, patients receive a prescription and must buy the drug from a retailer. When drugs are not available in facilities, many patients probably bypass the facility and go directly to the retailer.
Our results also demonstrate an impact in the reverse direction - customer demand, preference, and resources influence which drugs are stocked in retail shops and which drug a customer purchases. This is in agreement with an exit survey of shop customers, which showed that the majority of patients who visited a shop specifically asked for an anti-malarial, but only 16% asked for an ACT. Asking for an ACT significantly increased the likelihood of receiving one . Shops are responsive to customer demands and preferences which suggests that leveraging customer awareness and demand could have a significant positive impact on the effectiveness of the AMFm subsidy. This also underscores the importance of continuing public awareness campaigns and health education messages in the implementation of AMFm.