This is the first study to estimate the survival time of an ATSB product during seasonal deployment to control malaria in Africa. ATSB stations installed with excellent protection had median survival times > 218 days, which is longer than the seasonal deployment period in this setting. The overall median survival time for ATSB stations was 149 days, meaning 50% of the ATSB stations were still hanging and in good condition after approximately 5 months. For comparison, the seasonal ATSB deployment period in this Zambian setting was approximately seven months (deployment in November and removal in June). There were large differences in median survival times by cluster, and according to the characteristics of the structure and location on the structure where ATSB stations were installed.
ATSB stations may be appropriate for areas with seasonal malaria such as in this Zambia setting, particularly if installed on structures offering a good level of protection such as those with mud walls and thatch roofs. The majority of households in the trial area (over 70%) had thatch roofs10.
The most common reasons for ATSB station replacement were holes/tears, mold growth, and leaking bait. These results are consistent with data from ATSB intervention monitoring implemented during the main trial in which this sub study was conducted 12. According to Arnzen et al., the study site has yearly rainfall of approximately 1000mm with temperatures ranging between 18.9 and 30.3 Celsius10. Such weather patterns likely supported the significant mold growth on the ATSB stations. In addition, the substantial precipitation likely contributed to bait leakage and subsequent depletion of the ATSB stations.
Households in this area usually maintain agricultural plots near their living structures, cultivating maize and cassava10. In rural African settings, households often contend with rodent infestations attracted to food stores kept in and around the household. Such infestations can result in significant economic losses for households23 and may also have contributed to the prevalence of rodent damage (holes and tears) to ATSB stations in this setting.
ATSB stations hung on structures with roofs made of cement, iron sheet or wood planks or which were positioned without roof overhang had shorter median survival times. Thatch roofed structures in this area tend to have roof overhangs extending beyond the walls of the structures on all sides, which may have shielded the ATSB stations against rainwater, sun damage, dust, and wind. In the overall trial area, 72% of the dwelling structures were roofed with thatch10. In this sample, it was evident that wood plank and cement were not common roof materials in this community. This contributed to fewer observations for these roof types and low precision of ATSB survival time for these types of structures.
In general, the protection level at which ATSB stations were installed was largely determined by the architectural style of the structure where they were hung. Stations without protection were typically installed in the center of very high walls, or where the roof had a very small overhang, these tended to be on structures without thatch roofs. The survival period of ATSB stations in these positions were shorter than those with moderate or excellent protection. ATSB stations hung closer to the ground survived better than those hung higher up on the walls. Typically, in this setting, structures with thatch roofing were shorter in height, and had a substantial roof overhang, while structures with high walls generally had limited roof overhangs or the roofline was too high to reach, hence the longer survival of the ATSB stations under shorter structures. This is in contrast to the findings by Diarra et al. (2021)4 where ATSB stations hung at 1.8m above the ground had limited damages, compared to stations hung at lower levels.
Large differences in the survival of ATSB stations were observed among the different clusters. While these differences may be partly attributable to differences in housing style between clusters, it is also likely that other environmental, structural, or socioeconomic differences contributed to variations in longevity of the ATSB stations between clusters. The use of a shared frailty model for the estimation of associations with survival is likely necessary to account for correlation in frailty22 of individual ATSB stations within study clusters that share characteristics (including environmental exposures, foraging animals, weather elements, community cultural practices); however, additional clustering at the structure and household level was not accounted for in our analysis.
Earlier prototypes of the Westham ATSB stations used in Mali (Westham Innovations LTD., Tel Aviv, Israel)5 maintained their attraction and killing effect while lowering biting rates and reducing entomological inoculation rates while deployed in the field for more than 6 months2,5. This suggests that subsequent prototypes such as the Sarabi v1.2 ATSB stations that were used in the Zambia trial could remain bio-efficacious and maintain physical integrity longer than the earlier ATSB models, as vector control products continue to be refined and improved. For comparison, the study area in Mali experienced temperature range of 25.1–32.4 Celsius and annual rainfall average of 1098mm5, against the approximately 1000mm precipitation in western Zambia10.
Limitations of this study include the convenience sample used to purposively select the households and structures. The characteristics of housing for bait stations followed in this study may not have been representative of the wider ATSB trial area or of all of western Zambia. The Zambia Malaria Indicator Survey (MIS) 2021, indicated that 18.8% of the surveyed households across Zambia, and 61% across Western Province, had structures that were constructed using rudimentary wall materials of wood with mud plaster, compared to 35.9% in this study site (unpublished MIS 2021). Within the wider trial area, over 70% of households had thatch roofs, but only 47.7% of households in this study had thatch roofs suggesting that the Sarabi v1.2 ATSB stations may survive longer in the trial area than the average for our whole sample. Since ATSB station survival time is associated with structure architecture and installation site characteristics, the overall time to failure for the Sarabi v1.2 ATSB stations in this study should be interpreted with caution. However, the failure times associated with different levels of protection may be useful in combination with information about local architecture in guiding decisions around deployment to maximize the durability of this new tool. Another limitation of this study was right-censoring for some ATSB stations. The ATSB stations were removed from all households after 218 days, the end of the seasonal transmission period and end of the study period, resulting in right-censoring of a substantial number of ATSB stations especially among those deployed under excellent levels of protection. For these stations the estimate of median survival time cannot be made as the median survival time for ATSB stations with excellent protection exceeded the study duration. All that is known is that the ATSB stations hung under excellent level of protection median survival time exceeded 218 days. In addition, the precision of measurement of time to follow-up in this study was somewhat limited by interval censoring since ATSB stations were visited once per month and it was assumed that the failure or censoring occurred at the time of the last observation.
ATSB stations that were removed and replaced by the community ATSB monitors were classified as “removed by monitor” as their reason for failure and replacement, without reclassifying their specific reason for removal. It is not known whether the ATSB stations that were “absent on visit” could have been removed by household owners. However, destruction by humans may be limited to adults due to the high position at which the ATSB stations were placed, which was out of reach of children. Additionally, factors such as the suitability of the environment for rodent habitation, temperature and humidity fluctuations may influence the rate at which physical integrity of ATSB stations deteriorates (Kyomuhangi, et al.)24.
This study was focused on ATSB stations’ physical integrity as defined by criteria that were used during the first large-scale deployment of an ATSB product in a cRCT for malaria control in sub-Saharan Africa. The extent to which these criteria for replacement are associated with loss of ATSB efficacy, bio-efficacy, attractancy or other potential concerns including community acceptability, is not known. It is also important to note that ATSB station deployment in the study setting was not associated with a statistically significant reduction in malaria incidence or prevalence (Ashton et al., in preparation)25.