This survey was designed to evaluate ITN ownership, usage, and equity following the distribution of free LLINs in Manica and Sofala provinces, central Mozambique. Vouchers redeemable for free LLINs were distributed as part of the second phase of an immunization campaign and were to be redeemed for free LLINs approximately two months afterwards. By providing free LLINs to children under five years of age, the campaign was intended to increase HH ownership and usage of ITNs in these two provinces. These results show increased HH ownership and usage by children under five years of age, 50.2% and 60.3%, respectively, when data for both provinces are combined. The campaign also served to improve equity of ITN ownership in this region of Mozambique.
This survey found that 87.8% of children under five years of age went to an immunization post in September 2005, but only 57% of children received vouchers for LLINs because, among other reasons, the campaign limited the number of vouchers to one per HH. Approximately 50% of children under five years of age received an LLIN. Using vouchers may achieve lower coverage rates compared with direct ITN distribution during immunization campaigns, because the extra step of voucher exchange at an ITN redemption post is required [8
]. In addition, limited financial resources often do not allow for voucher redemption sites to be as numerous as immunization posts during the immunization campaign, meaning that caregivers needed to go greater distances to reach a redemption post. Despite these limitations, free distribution of ITNs using vouchers is a commonly used alternative in instances when bed nets are not available at the time of the immunization campaign or it is not feasible to coordinate simultaneously the logistics of both immunization campaigns and bed net distribution [5
HH ownership of bed nets of any kind increased from 20.6% to 55.1% in Manica, with an improved equity ratio. In Sofala, HH ownership of any bed nets increased to 59.6%, also with a higher equity ratio. These proportions are lower than those observed in Niger (>80%), where similar surveys were conducted one and nine months after an immunization campaign and bed net distribution [8
]. The difference between the results of these campaigns can in part be explained by the higher proportion of HHs with a child under five years of age in Niger (75%) compared with Mozambique (65%). In addition, Niger was reported to have a pre-existing bed net usage culture, which might have contributed to increased awareness and interest in having an LLIN [8
]. Nonetheless, this difference highlights one of the inherent limitations of ITN distribution via mass immunization campaigns, which is the targeting of a subset of HHs, i.e., those with children eligible for the vaccination [5
]. There is no doubt that higher levels of ITN ownership and usage can be achieved with universal distributions. As national malaria control programmes move towards universal access to ITNs, innovative strategies will need to be put in place to attain and maintain high levels of coverage among different age groups [14
This campaign in Mozambique was successful in increasing HH ownership of ITNs to 50.2% in both provinces combined. It was not possible to determine the insecticide treatment status of bed nets present in the HH prior to the campaign. However, considering the proportions of ITNs among non-campaign bed nets at the time of the survey (53 ITNs among 173 non-campaign bed nets or 30.6%), the pre-campaign HH ownership of ITNs can be estimated at 8.3% (27.2% times 30.6%), lower than what was observed post-campaign. These findings are consistent with previous studies in the literature, which suggest that integrated campaigns to distribute ITNs are able to rapidly increase HH ownership of ITNs [5
]. However, in order to maintain and further increase these high ownership rates, it is recommended that other strategies, such as ITN distribution at antenatal clinics and during children's regular immunization visits, be undertaken as complementary efforts [16
The results of this survey show that HH ownership was achieved more equitably in Sofala than in Manica. This difference in equity could be related to the fact that pre-campaign HH ownership of bed nets of any kind was more equitable in Sofala (35.6%, equity ratio = 0.71) than in Manica (20.6%, equity ratio = 0.10). In addition, this finding could also have resulted from differences in the distribution between the two provinces, such as the higher rate of post-distribution visits by Mozambican Red Cross volunteers in Sofala to promote usage, which might have contributed to higher rates of LLIN retention.
When ITN usage by children under five years of age (campaign and RBM target group) was considered, the usage rate for both provinces combined was above the 60% initial Abuja target for vulnerable populations. In addition, the adherence rate (children sleeping under an ITN given at least one ITN in the HH) was greater than 95%. Different factors may have contributed to this finding. First, one can expect that people who made a trip to the voucher redemption post were more aware of the benefits of sleeping under an ITN and consequently more prone to adhere to such behaviour. Second, this survey was conducted late in the rainy season, with a likely increase in mosquito population, a factor known to increase usage of bed nets [18
]. A survey in Togo during the rainy season, nine months after LLIN distribution, found an increased usage rate, 69.5% versus 52.8% in the dry season, in HHs with at least one ITN (Wolkon A, unpublished data). Third, this higher adherence rate could in part be attributed to the work done by volunteers involved in social mobilization, voucher and LLIN distributions, and post-campaign follow-up visits.
This survey has several limitations. First, given the cross-sectional design of the survey, it was not possible to understand how ITN ownership and usage may vary over time. Second, our evaluation was focused on the end results of the campaign, i.e., ITN ownership and usage, and provided little information about improving the overall process of the campaign. Analysis of the campaign itself might have yielded important lessons for improving future distribution strategies. Heavy rains made some of the initially selected EAs inaccessible, especially in Sofala. This might have contributed to an overestimation of our indicators, since reachable areas, which could coincide with areas that also had an easier access during the distribution efforts, were more easily accessible. Finally, because children under five years of age were eligible to collect an LLIN at the distribution post even without presenting a voucher, it was not possible to assess the effect a voucher-based bed net distribution might have had on reducing LLIN uptake by the target population.
In conclusion, this voucher-based LLIN distribution in Sofala and Manica appears to have been an effective strategy to rapidly scale up ITN ownership and usage. The figures for HH ownership of ITNs, although encouraging, are below the then-Abuja targets, and efforts to achieve higher rates should be considered in the future. The overall project did achieve the then-Abuja target for ITN usage by children under five years of age when the provinces were combined, but achieved this target only for Sofala when data from each province are analyzed separately. Integration of ITN distribution with immunization campaigns presents an important opportunity for reaching malaria control strategies among young children. Whenever feasible, bed nets should be distributed directly at the time of vaccine administration to help maximize immediate ownership and usage indicators among children under five years of age. As countries move toward universal access to ITNs, other strategies may need to be developed to reach older population groups [16