The malaria rapid diagnostic tests monitored in this study were exposed to conditions above manufacturer recommendations in three of the four countries studied, though for only brief periods. However, in all countries, conditions within these supply lines were above 30°C for considerable periods- beyond the WHO recommendations for stability of pharmaceuticals that are delivered through similar supply chains , and thus for many other commercially-available malaria RDTs. While the countries appear to have selected RDTs with stated stability, in keeping with general transport and storage conditions in this limited survey, the results raise questions over adequacy of supply line management for other medical commodities. In the four countries, drugs and RDTs are stored under the same conditions; this illustrates the importance of understanding transport and storage conditions in RDT product selection .
Whether the periods of temperature exposure recorded here affect RDT quality depends on the stability of the individual products. Previous studies have shown a high variability in RDT stability, both for HRP2 detecting test lines and particularly for test lines detecting pLDH [1–4]. Stability can vary at times between lots of specific products [1, 3, 15], but real-time stability data on a range of products lot-tested through the WHO-FIND malaria RDT evaluation programme suggests that significant failures in tests stored within manufacturer recommendations are not common . The products in use by public health services in the countries involved in this study had maximum storage recommendations ranging from 40°C to 45°C, well above the former ICH recommendations, but only exceeded these temperatures for relatively short periods of time. However, these temperatures are above the recommendations for many other malaria RDTs; only 13 of the 50 RDTs tested in the Round 3 of the WHO Product Testing Programme had recommended storage temperatures of 40°C or above (FIND, unpublished data).
Other commodities (with the exception of vaccines) are generally stored together in the same facilities at central, regional, and peripheral levels. For example, certain tablet formulations of anti-malarials (e.g., artemether-lumefantrine, artesunate, amodiaquine, sulphadoxine, pyrimethamine), and HIV antiretrovirals (e.g., lopinavir, ritonavir) listed as essential drugs should not be stored above 30°C .
These data raise a number of issues. Firstly, diagnostics (and drugs) clearly need to be selected taking into account the expected exposure to heat and humidity. While humidity is normally addressed by the moisture-proof packaging in which the product is delivered, measuring temperature using electronic monitors is relatively easy and cheap. This knowledge should inform procurement criteria. As openly-available data on real-time stability is limited, it seems advisable for countries to require real-time heat stability data from manufacturers on which product storage recommendations should be based. It has been recommended to store RDTs at a central level for as long as possible on the assumption that peripheral storage is less controlled . High temperatures recorded at central facilities in the four countries suggest that logistics planning should take actual storage data into account, and that managing conditions at central storage facilities should be taken more seriously. While storage standards for in-vitro diagnostic are less well defined, WHO standards for temperature stability of pharmaceuticals were exceeded at central storage over 23% of the time in two countries .
Clearly, published standards for pharmaceuticals and diagnostics in these climatic conditions are inadequate. Without substantial resources devoted to controlled transport and storage, temperatures of 30°C are routinely exceeded. There does appear to be a case here for investment in temperature control of central medical storage facilities. However, RDTs are a relatively high volume commodity, making controlled-temperature transport and storage at remote locations often impractical. Refrigeration is unnecessary for current products, and can also pose some risks, as shown in this study where one RDT was exposed to temperatures below −3°C in Burkina Faso. RDT storage outside the manufacturer-recommended temperature could shorten RDT product shelf life, yet this is difficult to detect. When RDTs are known to have been subjected to extreme conditions for some time, re-testing a batch withdrawn from the field against parasite panels is possible (e.g. lot testing laboratories in the Philippines and in Cambodia, , or comparing with microscopy, but both can be logistically difficult. Rejecting a batch on the basis of transport conditions without evidence of poor performance is costly. The development of positive control wells, based on lyophilized parasite antigens, holds promise to address clinic-level quality control dilemmas for malaria RDTs .
Lower cost solutions are available for remote-area storage, such as underground storage or the use of evaporative cooler boxes . Storage using evaporative cooling has been demonstrated for malaria RDTs in Afghanistan  and Cambodia , reducing temperatures from as much as 37°C to 23°C without use of electricity, well within the common range for storage of non-vaccine medical commodities. Moreover, simple measures during transport, such as loading and discharging vehicles in the shade or at night, reduce the probability of deterioration [21, 24]. In the end, the performance of the product at end-user level is the most important measure of impact of previous environmental damage; this is difficult to sustain for both diagnostics and drugs. Whereas this is currently being addressed for malaria, monitoring the availability of the active ingredient of a drug in a remote location is likely to remain a major challenge.
Due to the complexities of coordinating logistics throughout transport chains, only segments of the transport chain were recorded in this study. A larger study would likely reveal a wider range of conditions and problems. Since only a small sample of transport and storage was collected in each country, the results cannot be considered fully representative of the conditions under which medical commodities are transported and stored in any of these countries. However, they illustrate the lack of control in supply chains and potential for environmental damage of products. Ideally, the condition of RDTs should also have been tested; the countries concerned procure RDTs with relatively high recommended maximum storage temperatures so it is unclear whether the relatively short periods when they are exposed to higher temperatures would have been significant. However, the study results raise concern with respect to RDT stability between production lots (pre-release real-time testing of production lots is obviously not possible [1, 3], as well as the effect of their co-storage with other perishable medical commodities, as discussed above.