The primary purpose of this study was to assess the field performance of the polyester-based LLIN Interceptor®
with respect to the criteria set for WHOPES evaluations. Accordingly, the methodology closely followed the WHOPES guidelines for LLIN testing [6
] with the following modifications: i) sample size of 40 nets instead of the recommended 30 per time point was used; ii) at baseline only three instead of the recommended 18 samples were taken per net to establish within-net variability.
Insecticide concentration at baseline for the LLIN was 6.48 g/kg (192.0 mg/m2
) very close to the value given in the specifications for a 75 denier net of this product of 6.7 g/kg [13
]. Between- and within-net variation expressed as relative standard variation was 11.0% and 3.9% respectively, well below the 25% allowed for between net variation and the 20% allowed for within net variation [13
]. Even when the range of values for the 10 baseline nets was considered (5.3-7.1 g/kg) this was within a ± 25% range of the target dose (5.0-8.4 g/kg). After three years of field use bio-assay results showed that 83.3% of sampled nets had at least a 60-minute knock-down rate of 95% or a 24-hour mortality rate of at least 80% and median alpha-cypermethrin residue was 73.8 mg/m2
or 38% of the initial dose with 94% of net samples having at least 15 mg/m2
. In contrast, conventionally treated nets lost insecticide content rapidly with only 7% of initial dose left after two years of use (median 1.6 mg/m2
) and none of the samples having at least 15 mg/m2
although still 75% showed satisfactory bio-assay results. Monitoring net utilization and conditions over the study period demonstrated a tendency of reducing use with increasing deterioration but after three years still 86% of LLIN were regularly used and up to two years (when all conventional ITN were collected) no difference in use rates between LLIN and conventionally treated nets were found suggesting a high level of acceptability of the product. All these findings are in agreement with the conditions set by WHOPES [6
] and allow the conclusion that the tested LLIN fulfilled the criteria to be qualified as an LLIN. It must be kept in mind, however, that results may vary from site to site depending on harshness of conditions the nets are used in. With an average washing frequency in the study area of less than two washes per year, a climate with temperatures not exceeding 35°C and reasonable housing conditions this site has to be considered as moderately stressful on the nets. Washing frequencies reported in other studies range between 1.2 and 5.6 per year [15
] when measured rather than assessed as "intention to wash" [20
] and climatic and socio-economic conditions such as those in Western Uganda are found in many places where LLIN are applied. The fact that households in the study area were using LLIN for 8-9 years may have increased the use rate and, hence, the stress on the nets, but also may have resulted in a more careful handling based on passed experiences. This suggests that a similar performance of the Interceptor®
LLIN can be expected in a significant proportion of interventions sites were a net culture already exists although it cannot be excluded that the LLIN performs less well in more extreme conditions.
Loss rate of nets (attrition) observed in this study was lower than expected so that sufficient nets were available after 42 months of field use to allow an additional evaluation of field performance. The sample after three and a half years included 21 LLIN so was slightly lower than the recommended 30 [6
] but still sufficient to allow a basic statistical analysis. Chemical content results demonstrate a continuous linear decline of insecticide with a median 56.2 mg/m2
alpha-cypermethrin remaining and 81% of samples still giving > 15 mg/m2
(see also Figure ). Bio-assay results had 71% of LLIN still with optimal biological effectiveness and comparison with results from the preceding year shows a constant rate of decline between year 2-3 and 3-4 of 12% (Table ). This strongly suggests that protective effectiveness of the LLIN does not dramatically deteriorate the year following the three year cut-off chosen by WHOPES for evaluation purposes although some acceleration of the decline was observed.
Testing of LLIN at the Kyenjojo field site has been going on since the year 2000 and findings of this study regarding socio-economic environment, net use and washing habits do not differ from those previously reported [12
]. Also the rate of physical deterioration of the 75 denier polyester Interceptor®
LLIN was similar to that found earlier for another 75 denier polyester LLIN product. As shown in Table , 63% of nets had any holes after 3 years of follow-up with a mean simple hole index (sHI) of 10.3. In the previously reported studies the respective values had been 79% and 10.7 sHI in the first and 69% and 15.6 sHI in the second [12
To date three studies on the field performance of Interceptor®
have been published. Banek et al.
] studied the LLIN in Liberia using a randomized allocation design of LLIN and conventionally treated nets in a returning refugee setting. Nets were followed for 12 months with six assessments of chemical residue. Mean concentration of alpha-cypermethrin at baseline was 180 mg/m2
(95% CI 152, 208), which gradually declined to 126 mg/m2
(113, 139) after 12 months. This is a 20% loss within the first year and very similar to the finding from Western Uganda. Physical deterioration of the 75 denier LLIN within the first year was also very similar with 26.6% of nets showing any hole in Liberia compared to 25.7% in Uganda.
The other two studies were undertaken in India. Sharma et al.
] studied the Interceptor®
LLIN in 19 villages in Odessa State, India and measured net performance with monthly bio-assay tests (WHO cone) using Anopheles culicifacies
and Anopheles fluviatilis
. After seven months of regular use by the villagers knock-down rates against the two vectors were 70-80% with 100% mortality on all tested nets. The authors could also demonstrate a significant reduction of vector densities in villages allocated the LLIN compared to untreated or no nets. Dev and co-workers [9
] assessed acceptability and side effects in communities in Assam, northeast India. They found that in spite of 9% of users reporting initial and transient effects such as eye irritation acceptability and satisfaction was very high with 80% of users reporting a reduction in visible mosquitoes in the houses. This was confirmed by assessments of vector indoor densities of Anopheles minimus
which were reduced to zero in the LLIN villages.
Conventionally treated nets in Western Uganda lost insecticide quickly (93% within two years and after approximately three washes) but still had surprisingly high knock-down and mortality rates with 75% of nets still showing optimal performance even at low levels of insecticide. This is, however, in keeping with results reported in the literature for alpha-cypermethrin. Adams et al
] tested low doses of various insecticides in Malawi and demonstrated that even a dose of only 6.5 mg/m2
alpha-cypermethrin resulted in a 93% mortality rate in Anopheles gambiae s.s
. Similarly high bio-assay results were obtained by Graham et al
in Pakistan [22
] after 21 washes and a target dose of 15 mg/m2
achieving a 49% mortality rate for Anopheles stephensi
. In The Gambia, Miller et al
] observed a 85% reduction in insecticide content after three washes which is very close to the findings in Uganda. In contrast, Jawara and co-workers [24
] found An. gambiae s.l
. mortality of only 8% after two washes and a target dose of 40 mg/m2
alpha-cypermethrin in The Gambia, but these nets where mainly made of cotton which may have altered the performance.
Considering the consistency of results within the study site in Uganda regarding net use, washing and net performance, variables for similar LLIN products as well as the favorable comparisons with other results for the Interceptor® LLIN and nets conventionally treated with alpha-cypermethrin it appears that the results presented are reliable and valid for the assessment of the performance of the Interceptor® in the environment of Western Uganda.
The second interest of this study was to explore the methodological aspects of determining "useful life" or durability of a given LLIN product. Although insecticide treated nets have been used for over 20 years, this aspect has been very much neglected and only has come into focus of discussions with the scale-up of mass-distribution campaigns for LLIN. While some progress has recently been made to better understand the concept of LLIN durability and how it can be measured [25
], one of the key issues yet to be solved, is how best to assess the physical condition of the nets. While counting holes in the net has been frequently used to describe the textile integrity [19
], an agreement on how such counts are best, and in a standardized fashion, summarized into a single measure that allows to distinguish between "good" or "serviceable" nets and those that are unlikely to fulfil their protective function, is still lacking.
In this study, a hole index was applied that attempts to improve the previous methodology [12
] by applying a weight to the count of three different size categories of holes which is approximately proportionate to the average hole surface area of each hole category. Such an approach has the advantage that the resulting index value for a net corresponds to the total torn net surface and has since been adopted by WHOPES in a slightly modified definition [31
]. This value was then used to categorize the nets into good, intermediate and poor physical condition. The cut-offs used are based on the little literature available on this topic and which shows that, if treated with insecticide, nets with holes are still effective [30
] and can provide at least some protection even with a hole surface area of 0.24 m2
], which is equivalent to a pHI of 700, i.e. a value more than twice the cut-off used for poor condition in this study.
Categorization of the physical condition of the nets can then be combined with the retention/attrition rates and the results of the bioassay to provide an estimate of the proportion of nets that are still present, in acceptable physical condition and with functional insecticidal protection, hence an assessment of the "useful life" of the LLIN. Applying this approach to the data from this study and assuming that half of the lost nets were thrown away due to wear and tear, would then suggest that 76.9% of the LLIN were fit for use (good or acceptable condition, pHI < 300) and protective after three years and 61.6% after three and a half years. Assuming further that "useful life" is measured by the median survival time of a product in the field, then the Interceptor® LLIN in Western Uganda demonstrated a "useful life" of at least three and a half years and most likely close to four years. There is certainly more work needed to fine-tune cut-offs and definitions, but having a better measure of physical condition of nets, as suggested here, is an important step towards establishing the durability of LLIN.