In order to establish ZeroVector® DL as a viable long-lasting alternative for IRS, it must demonstrate equivalent or superior levels of bioefficacy, acceptability, durability and logistical feasibility to currently available products. In this study, superior efficacy was achieved by DL compared to IRS throughout the trial. In both Mali and Vietnam, complete mosquito mortality using DL was observed at ten and fifteen months post-installation, respectively. By contrast, loss of IRS activity was reported after only one month in Vietnam and by six months in Equatorial Guinea and Ghana. These results are explained by ZeroVector® DL containing a long-lasting insecticide reservoir, while IRS has finite residual activity. However, this trial was only conducted for one year and further evaluations must be undertaken to determine how efficacious DL remains after three to four years of field use.
The DL was received with high levels of acceptability, from both study participants and external visitors, as evidenced by numerous positive responses to entomological and aesthetic questions. Although the impact of DL on local vector populations was not empirically measured, perceived reductions in mosquito density, mosquito bites and nuisance insects were described from all study sites and across all time points. In addition, when asked if they would like to retain their DL even if it had no effect against insect vectors, the majority of participants agreed with this statement. This result suggests DL may be accepted by users based on its aesthetics alone. The importance of both perceived entomological efficacy and decorative value, as key determinants of household acceptability, has been previously demonstrated during early field evaluations of prototype wall linings
. Most participants in this study also did not express objections to the effect of DL on their indoor environment, reporting no changes in indoor temperature or light levels. An overwhelming majority of respondents wanted to keep their DL at the end of the trial and were satisfied with their lining colour. When offered a choice of vector control product in Equatorial Guinea, all intervention groups preferred DL to LLITCs and IRS.
Regarding durability, after one year, a minority of households reported failed nail fixings and/or damage to the DL, with one severe case described (burning of DL in Equatorial Guinea). The latter was unexpected as ZeroVector® DL has been subjected to flammability tests conducted by an independent agency (The Govmark Organization Inc.) and is considered non-flammable (documentation available from manufacturer on request). No household reported actively repairing failed fixings or lining tears, which may indicate that either the DL was not genuinely valued or passive recipients were not focused on making routine repairs at this stage of the intervention. The method and quality of DL cleaning varied widely between households and study sites; in Ghana all respondents reported cleaning their wall lining, while no such activity was described from Equatorial Guinea or Vietnam. In future, the installation of DL should be complemented with information about care and maintenance.
Currently, ZeroVector® DL is manufactured in large rolls (2.3 × 100 m) that require an external installation team to correctly cut for particular house dimensions and fix with nails to avoid wastage and ensure optimal physical durability. In Mali and Vietnam it took three individuals approximately 60-75 minutes to install DL in one house. This is comparable to the time taken by a single person to apply IRS to one house (~60 minutes) but with the advantage that DL may only need to be repeated every three to four years, not every six to twelve months. Additionally, in Equatorial Guinea, where DL was installed using bamboo rods sourced from local materials, this subset of houses reported fewer failed fixings and lining damage than those which received DL installed with nails. These observations suggest that with such high levels of acceptability and adequate supervision during installation, DL has the potential to be widely implemented at the community-level when supported by local technical and logistical infrastructure. Future improvements could include the production of DL rolls for specific house dimensions that are consistent among individual countries, cultures or communities.
The high DL efficacies reported in this study were supported by a loss of only 23% of deltamethrin content after fifteen months, a similar rate to recent field evaluations of LLINs. After twelve months of field use, a loss of 30% of alphacypermethrin from Interceptor® LLINs was reported
, and 17% and 11% of deltamethrin from PermaNet® 2.0 and PermaNet® 3.0, respectively
. In addition, the DL demonstrated equivalent levels of efficacy in the WHO bioassays to the LLITCs in Equatorial Guinea, suggesting it remains as effective as other long-lasting netting products for at least one year in the field. However, unlike DL, the high LLITC bioefficacies did not result in as great a perceived reduction in mosquitoes. More respondents with DL described declines in mosquito numbers and bites than users of LLITCs, and more LLITC households reported using alternate vector control products which could be interpreted as dissatisfaction with their intervention.
While this present study does not seek to establish the mode of action of DL on vector populations, growing evidence suggests that DL is more comparable to IRS than LLINs. Experimental hut trials in Burkina Faso indicated mosquito mortality was the principal activity of permethrin-impregnated DL, and only slight inhibition of An. gambiae entry and feeding was observed
. Similar reports from Benin where bendiocarb-treated wall coverings also did not impact on blood feeding
 suggest that wall linings afford little to no personal protection when used by only a minority of houses
. Instead, it is anticipated that DL will control malaria transmission when applied widely at the community-level, through its effect on mosquito density and longevity. In support of this, recent studies from India achieved a mass population effect on malaria vectors using high coverage of deltamethrin ITPS in temporary urban labour settlements
 and tribal villages
There are several weaknesses in the reported study design, which need to be considered when interpreting the data. The questionnaires were designed to ask direct questions with standardized delivery to ensure consistency of responses between countries. It could be argued that some questions were too specific, or not sufficiently open, so that interviewees delivered the answers they anticipated the questioner wanted. In all study sites the interviewers were associated with the installation process, making it probable that a proportion of respondents’ answers were not as objective as they might have been had the interviewers been entirely unknown. Entomological indices, including indoor and outdoor resting catches of mosquitoes and blood meal ELISAs, are needed to corroborate the reported entomological effects of DL. This study lacked a control (i.e., a DL with no insecticide) and it is possible that even users of untreated DL would have perceived a decline in mosquito biting if that is what they expected to see. Wall installations may also have coincided with seasonal changes in mosquito biting and as such results are not adjusted for respondents’ preconceptions. Nevertheless, all study sites reported fewer mosquito numbers and bites at all time points across the year, suggesting that this was an observable phenomenon.
While IRS and LLITCs acted as control interventions, to a certain extent this trial would have benefitted from an untreated control had such a material been available. An alternative approach would have been to offer both treated DL and untreated DL as ‘aesthetic interventions’ with no mention of their control capabilities and then probe for entomological observations or perceptions during follow-up interviews. This study design would also allow the fourteen adverse reactions to DL to be attributed to this product and not to a ‘placebo effect’. However, there was reluctance to use an untreated control by authorities in some countries at a stage when the product was completely unknown. Finally, WHO cone bioassays are designed to measure the residual activity of an insecticide-treated substrate and not its efficacy against free flying vectors. Comparisons of indoor and outdoor mosquito densities, human landing catches or light trap collections, between intervention and control clusters in a community randomized trial, would be required to substantiate these basic efficacy results.