Expanding access to parasite-based malaria diagnosis through retail drug shops in Tanzania: evidence from a randomized trial and implications for treatment

Malaria prevalence among children under 5 years old in Tanzania declined from 18% in 2007 to 9% in 2011, whereas 2-week fever prevalence remained approximately 20% over the same time period, from 19 to 20% [1, 2]. This reduction in the proportion of fevers caused by malaria is likely due at least in part to Tanzania’s recent scale-up of prevention and treatment measures [3‐5]. Presumptive treatment of fevers with artemisinin-based combination therapy (ACT) thus will increasingly result in incorrect malaria diagnoses, prescription and wastage of inappropriate medications, and subsequent delays in obtaining effective treatment for the true cause of illness [6‐9].

The Tanzania National Malaria Control Programme (NMCP) case management guidelines require suspected malaria cases to receive parasite-based diagnosis by microscopy or rapid diagnostic test (RDT) prior to treatment with anti-malarial drugs [10, 11]. In practice, however, up to 54% of those seeking treatment for suspected malaria in Tanzania first visit the private retail sector, where malaria diagnostics are currently prohibited from being sold and administered [12]. As a result, parasite-based testing is not received by many suspected malaria patients.

The Tanzanian private retail sector includes both unregistered outlets and a network of more than 6000 registered shops or Accredited Drug Dispensing Outlets (ADDOs) (duka la dawa muhimu or DLDM in Kiswahili) regulated by the Pharmacy Council. Unlike unregistered outlets, ADDOs are permitted to stock and sell both over-the-counter medicines and certain classes of prescription medications, including ACT [13‐15]. To obtain a Pharmacy Council permit each year, ADDOs owners must meet certain conditions related to the premises, training and certification of a dispenser who may or may not be the owner, and the products stocked. Like all other private retail outlets, RDTs may not be sold or performed at ADDOs, however.

Malaria RDTs have been safely administered by non-medical personnel in several previous settings [16‐18]. The introduction of RDTs into ADDOs has the potential to improve fever case management by increasing availability, access, and use of parasite-based diagnosis. It is unclear, however, whether customers will be willing to pay the extra cost for diagnosis, whether ADDO owners would encourage RDT use, whether RDTs might require subsidization to encourage uptake in the private sector, and whether treatment choices would adhere to test results. To investigate these questions, this pilot evaluated the operational feasibility of selling RDTs through ADDOs and measured changes in suspected malaria patient case management that occurred as a result of making RDTs available at two different prices.

Ninety-five percent of ADDO dispensers passed the 2-day training. In total, 310 dispensers from 262 ADDOs, 164 (147 ADDOs) from Kilombero in the lower-priced or “subsidized” district and 146 (115 ADDOs) from Kilosa the higher priced or “unsubsidized” district, were certified and allowed to sell and perform RDTs. The number of outlet surveys and exit interviews completed are presented in Table 1.

The results from this study suggest that ADDOs offer an important opportunity for improving malaria case management in Tanzania towards the national goal of 80% parasite-based diagnosis of suspected malaria patients. Two-thirds of suspected malaria patients who sought treatment at ADDOs with RDT-certified dispensers purchased RDTs. Improved testing rates through ADDOs in this pilot project resulted in improved targeting of ACTs to patients with malaria, with 90% of malaria-positive and 93% of malaria-negative ADDO-tested patients making appropriate treatment decisions according to their test results.

These results add to a growing body of evidence that malaria case management can be performed well outside of the formal public health sector. A pilot in Ghana that measured vendor-reported adherence to test results after the introduction of RDTs to 28 Licensed Chemical Sellers (private retail sector shops) found 89% of positives and 3% of negatives received ACT [24]. Vendors of 59 licensed drug shops in Uganda also reported high uptake (98%) and favorable adherence to test results (ALu was prescribed to 98% of positives and 1% of negatives) in a small-scale pilot study [25]. The higher uptake observed in these studies compared with this one may be attributable to self-reporting by drug shop vendors compared with customer interviews, a lower $0.20 price point (Uganda) compared to $0.31–0.67 in Tanzania, and/or variance in length of training and frequency of supervision. Not all studies have found such positive outcomes, however: a different study in Uganda with 92 shops selling RDTs that measured adherence through household surveys found only 32% of positives purchased an ACT medicine (7% of negatives) at a median $0.40 per test [16].

This pilot was successful in improving case management practices for several reasons. First, ADDOs are a highly utilized source for anti-malarials in Tanzania: ADDOs were the first place treatment was sought for 75% of interviewed patients, emphasizing the need for access to parasite-based diagnosis at this point of care. Second, nearly all ADDOs in the study area participated in the trainings and RDT program. ADDOs were kept engaged in the study through quarterly supervision and monitoring visits, which also allowed for frequent quality assessments on safety protocol. Third, proximity of the region to Dar es Salaam and relatively good road infrastructure allowed for existing supply chains to be utilized, creating a more realistic market scenario. Finally, negotiations with RDT suppliers and agreements with wholesalers and ADDOs on mark ups, ensured affordable recommended retail prices and enabled comparison of price points.

These results provided data to help guide several open questions surrounding an introduction of RDTs to the private retail market. A similar proportion of patients were willing to pay the higher ‘unsubsidized’ price of a 1100 Tanzanian Shillings or USD $0.67 for a RDT compared to the lower unsubsidized price of 500 Tanzania Shillings or USD $0.32 for the subsidized RDT, indicating that a subsidy may not be necessary to encourage testing uptake. Additionally, similar to the Ghana pilot [24], an increase in the purchase of antibiotics among those receiving a negative test result was not observed in this study.

The study was also subject to several limitations. A significant pre-intervention difference in parasite-based diagnostic rates between intervention (19%) and control (3%) districts suggests inherent differences in health facility testing uptake that may confound the effect of the intervention. A second limitation was the unforeseen mobility of ADDO dispensers, periodically changing jobs work in different ADDOs during the study period. While dispensers were certified in this study, it was unclear if the certification extended to the ADDO when the dispenser was not present. Dispenser movement complicated follow-up and suggests future consideration for specifying that dispensers and ADDOs be certified, and requiring ADDOs that sell RDTs to have a certified dispenser present. Additional follow-up challenges arose when ADDOs closed temporarily or permanently, or were not reachable for stock delivery or enumeration due to flooded roads and bridges. Also, although dispensers were trained to refer severe and negative cases to the nearest health facility, this study was not resourced to follow the outcomes of these patients. In practice, it is unlikely that all negative cases will complete the referral pathway, underscoring the importance of including diagnosis and treatment of non-malaria febrile illness and recognition of danger signs in ADDO dispenser training. Lastly, the post-intervention survey sampled from ADDOs that participated in the training in the intervention areas, and while these were the majority of the ADDOs in these districts, they are not representative of all ADDOs in the study area.

As Tanzania uses the results of this study to inform policy decisions, it will be important to consider how results might vary in different endemic settings and in more urban areas. Expansion of this program will also require evaluation of how best to ensure long-term routine supervision and training, as well as waste management for sharps and cassettes. Finally, a critical challenge to scale-up will be integration of the private retail sector into the routine surveillance system for capturing malaria case data. This information will be imperative for strategic decision-making as malaria prevalence declines across the country.

This pilot study contributes to a growing evidence base that introducing RDTs to the private retail sector in low resource settings can increase parasite-based diagnostic rates for malaria and adherence to test results when dispensers are trained and supervised. The pilot study also showed that RDTs can be introduced under real market conditions; utilizing existing supply chains without subsidizing the cost of the RDT. While the intervention piloted here was limited to a rural setting in Tanzania’s specialized ADDO network, it provides a basis for policy decisions on scaling up RDT access across Tanzania to increase parasite-based diagnosis and rational use of ACT through heavily used existing channels. The results presented here support the goals for Tanzania’s National Strategic Plan and may have applications in similar settings.