Household surveys
A total of 705 households were identified in the baseline survey; 70% had mud walls and 75% palm thatched roofs. Other roofing materials included corrugated iron, some walls were made of brick. The mean ages of household heads in ICON Maxx and Iconet households were 43 and 52 years respectively. Most householders were farmers (33% ICON Maxx, 55% Iconet); the remainder were employed as fishermen, teachers, nurses, students or unemployed. The majority (65% ICON Maxx, 78% Iconet) had received 7 years of primary school education, 7% and 20% had received secondary education and others had not gone to school at all.
Both the ICON Maxx nets and Iconet CTNs were well-accepted by the communities. Reported net use was 100%. Respondents indicated using their nets year-round and every night. Almost 90% of surveyed nets were found hanging above beds and 10% were observed suspended over floor mattresses. 72% and 78% of sampled populations stated their reason for using nets was to protect themselves from mosquito biting while 13% and 28% stated for protection from malaria. The frequency of net washing was found not to differ between ICON Maxx net surveys. Estimated washing frequency was 4 times per year (Table 1).
All respondents in all surveys reported washing their nets in cold water. Nobody reported rubbing nets against rocks or on washing stones. Nets were pre-soaked by 18-40% of respondents; soaking times ranged from 10 min to 4 hours. Nets were reported washed using commercial bar soap (30-85%), detergent powder (14-50%) or both (14-32%). Most nets (92-98%) were rinsed after washing and most (92-98%) were dried outdoors.
Despite householders reporting high frequency of ICON Maxx net washing, it was observed that only 45% and 16% were scored as clean at 12 months and 36 months respectively and 19% and 22% were scored as very dirty at 24 months and 36 months respectively. There was no association between the alpha-cypermethrin content remaining on the nets at 36 months and the reported number of washes (F1,48 = 1.2, P=0.30). Nor was there any association between the reported number of washes over 36 months and the proportion of nets failing the cone bioassay criterion (F1, 48 = 0.3, P = 0.85).
Physical integrity of nets in cross-sectional surveys
The same brand of 100-denier nets was used in ICON Maxx and Iconet CTN arms. In the baseline survey there were no holes or open seams on any of the sampled nets in the ICON Maxx or CTN arms. After 6 months, approximately half of the nets (52% of ICON Maxx nets, 56% of Iconet CTNs) had at least one hole (these were mainly of the smallest size category) (Table 2b); the mean number of holes per net was 4.3 for ICON Maxx nets and 5.5 for CTN (Table 2a). The number of holes increased between 6 and 12 months to a mean of 9 for ICON Maxx arm and 7 for Iconet arm; the majority of holes were always found in the lower part of the panels (Table 2b). Comparison of physical integrity between ICON Maxx and CTN nets confirmed there was no difference between either arm at 12 months when the CTNs were disused (Tables 2a, 2b). After 24 months, while most ICON Maxx nets (84%) had a least one hole, these remained of the smallest hole category (Tables 2a, 2b); the mean number of holes per net had increased to 14.5 (Tables 2a, 2b), the mean hole index (HI) was 589, the median HI was 197 (IQR = 352) and the geometric mean number was 71 (Table 3). After 36 months, most hole indices had increased: 82% of nets were holed (Table 2a, b), the majority were still size 1 (59%), the mean HI was 740, the median HI was 417 (IQR = 615) and the geometric mean number was 59 (Table 3).
The age of nets (number of months of use) was positively associated with net HI (F1,345 = 9.31, P = 0.002). There was no association between the reported number of washes per net and net HI (R2 = 0.015, P<0.4027), suggesting that frequency of reported washing was not associated with net durability. Nor did any differences in type of washing agent used have any association with net durability (F= 0.03, P< 0.969).
While the mean number of holes per net, the hole index and hole area showed an increasing trend between 0 and 36 months, no more than 13% of holes were ever greater than size 2 (Tables 2 and 3).
Net efficacy through bioassay
Baseline cone bioassay tests on ICON Maxx and CTN (Iconet) after treatment but before distribution resulted in 100% knockdown and 100% 24-hour mortality on all pieces tested (Figure 2 & 3). After six months of use the mean percentage mortality (± 95% CI) was significantly greater on ICON Maxx than on CTN (87.2%, CI 82-92 vs 63.9%, CI 56-87, p<0.0001); similarly, mean percentage knockdown on ICON Maxx was significantly greater than that on the CTN (97.7%, CI 96-99 vs 86%, CI 78-94, p<0.004) (Figure 2 & 3). The survey after 12 months of use continued to show differences between ICON Maxx and CTN treatments in mean percentage mortality (93%, CI 86-96, vs 72%, CI 62-81, p<0.0004) and mean percentage knockdown (97%, CI 94-99, vs 85%, CI 78-91, p<0.0005) (Figures 2 & 3). With respect to pass rate, 1 ICON Maxx and 8 CTN failed the cone after 6 months of use (Figure 4). Some nets that failed the cone test subsequently passed the tunnel test criteria (1 ICON Maxx and 2/8 CTN), producing combined test pass rates of 100% for ICON Maxx and 80% (24/30) for CTN at 6 months (Figure 4, Table 4). After 12 months of use, whilst the combined test pass rates remained high for ICON Maxx nets at 96.7% (29/30), it was much lower for the CTN at 66.7% (20/30) (Figure 4, Table 4). Bioassays on the CTN were discontinued forthwith. Subsequent surveys focused on the ICON Maxx nets. After 18 months fewer ICON Maxx nets passed the cone test (70%, 21/30) but combined cone and tunnel testing produced an overall pass rate of 96.7 % (29/30) similar to the pass rate at 12 months (Figure 4, Table 4). After 24 months the combined pass rate remained high at 90% (27/30). After 30 months, although fewer nets passed the cone test criteria (33%, 10/30), the majority of nets that failed the cone tests achieved the tunnel test criterion (80%, 16/20) producing an overall pass rate of 86.7% (26/30). At 36 months, although only 26% (13/50) of ICON Maxx nets passed the cone criterion, 86.5% (32/37) of nets that failed the cone achieved the tunnel test criterion producing an overall pass rate of 90% (45/50) (Figure 4). The incremental decrease in pass rate over the full 36 months was small but significant (χ2trend = 11, P = 0.001). The incremental decrease in cone test pass rate was highly significant over the full 36 months (χ2trend = 34, P = 0.001).
Analysis of chemical content and insecticide retention (Figures 5 and Table 5)
At baseline, the mean lambda-cyhalothrin content of ICON Maxx treated nets was 60.1 mg AI/m2 (Figure 5a); this was very close to the target of 62 mg AI/m2. At baseline the mean lambda-cyhalothrin content among CTN was 12.7 mg AI/m2, well within the acceptable limits of the target dose of 15 mg AI/m2 (Figure 5b). After 12 months of household use the lambda-cyhalothrin content of ICON Maxx had decreased to 28.9 mg AI/m2 corresponding to 52% loss of the baseline AI content (Figure 5a); the content of CTN decreased to 4 mg AI/m2 after 12 months corresponding to 68% of baseline AI content (Figure 5b). After 36 months the mean lambda-cyhalothrin content on the surveyed ICON Maxx nets was 15.8 mg AI/m2 (n = 50, RSD = 93.4%) corresponding to 73.7% loss of the original content (Table 5).
The mean lambda-cyahalothrin concentration on ICON Maxx treated nets that passed the cone bioassay criteria at 36 months was 30.4 (21.4-39.4) mg AI/m2, while content on nets that failed the cone criteria was 10.7 (7.5-13.9) mg AI/m2; the difference in AI content between failing nets and those passing cone test criteria was significant (F1,48 = 26.1, P=0.0001).
The mean concentration on nets that failed the tunnel test criteria was 5.7 mg/m2 and on those that passed was 16.9 mg/m2; the difference in AI content between pass and fail was significant (t = 2.5, P= 0.009).
Adverse effects among staff treating nets and families using nets
Three attendants were responsible for treating nets at the start of the project at a rate of 60 nets per person per day for 2 weeks. All attendants who treated the nets reported sneezing and facial itching, and one reported fever. The adverse effects were more common after treating with Iconet (CTN) than after treating with ICON Maxx, even though the treatment dose was higher for ICON Maxx. One of the attendants regularly reported irritation to facial skin (paraesthesia). The effect took about 3 hours to subside on each occasion. The discomfort was not so severe that the individual took time off work. All proper precautions were taken while treating the nets including wearing of masks and gloves.
Of the 60 households included in the first week and first month post-treatment surveys, only a small proportion reported experiencing any adverse effects and only during the first few days of net use. Similar proportions of Iconet CTN users (10%) and ICON Maxx net users (6%) reported these effects which included bad odour, sneezing, skin itching, nasal discharge and facial itching. The effects were transient and did not deter users from continuing to use the nets. No adverse effects were reported after one month of use. During the 6 months survey the interviewees reported that all symptoms stopped after the net had been washed once. At no stage did any of the adverse events require medical attention.
Net attrition and survivorship rate
Net survivorship due to loss of integrity (accumulated holes) caused by physical deterioration or damage fell from 100% to 78% after 12 months, to 70% after 24 months and to 68% after 36 months (Figure 7 & Table 2b). A very high proportion of nets distributed (72%, 189/264) were lost not due to loss of integrity but to more mundane reasons such as moving house to outside the study area, hut collapse, or nets being given away during the 36 months. For most of the nets that were lost to follow up this occurred between 0 and 24 months.
In the comparison of physical integrity of nets between the cohort-longitudinal surveys and cross-sectional surveys, the majority of holes in the cohort surveys were found in the lower part of the panels (Table 3 & Figure 6). In all three cohort surveys the hole indices were significantly lesser (Z = 2.46; P = 0.014 for 12th month, Z = 2.6; P = 0.009 for 24th month and Z = 6.14; P = 0.001 for 36th month surveys) than in the cross-sectional surveys after the corresponding periods of use (Table 3 & figure 6). This may reflect cohort members’ self-awareness that they were being monitored more closely than other recipients of ICON Maxx treated nets.
Attrition and physical integrity of the ICON Maxx-treated nets were monitored as recommended in the WHO LLIN testing guidelines. As a home-treatment kit, it should be noted that net integrity and hole index were never part of the product claim of ICON Maxx. Nevertheless, it was important to compare net integrity and bio-efficacy as part of the evaluation of ICON Maxx and to correlate the formulation performance with polyester net condition, as well as consider other types of material that ICON Maxx might be called upon to treat.