Gender differences in the use of insecticide-treated nets after a universal free distribution campaign in Kano State, Nigeria: post-campaign survey results

The post-campaign evaluation survey took place at the end of 2009 in Kano State, in the northwest region of Nigeria, and included 24 local government areas (LGA), which were covered by two independent campaign waves. During Wave 1, which took place in May 2009, the campaign covered 21 LGAs, while Wave 2 (July 2009) covered 23 LGAs.

This study was a cross-sectional household survey with a stratified two-stage cluster sampling design. The strata were the two areas covered in the different campaign waves and each stratum was considered a survey domain. The size of the sample was estimated based on confidence intervals (alpha-error) of 95%, statistical power (1-beta-error) of 80%, and a design effect of 1.75. It was assumed an anticipated nonresponse of 5% and an average household size of 5 people. Children under 5 constituted an estimated 20% of the population, while 3.5% of the female population was currently pregnant.

Stage one included the selection of clusters using systematic sampling with probability proportionate to size (PPS). Stage two included the selection of households by field team mapping small villages, randomly selecting 17 households and using the same household definition used in the distribution campaign, which was "a wife with her direct dependents". A compound was divided into several households depending on the number of wives and the husband was assigned to the first wife's household. Larger communities (more than 120 compounds) used the equal size section-approach.

Clusters were defined as urban or rural, based on their categorization in the 2006 census, not their urban/rural stratification. A sample of 30 clusters, with an estimated 17 households per stratum, resulted in a total sample of 1,020 households in the campaign area, including 4,602 individuals. Data collection took place from 19 October to 4 November, 2009 and involved a questionnaire adapted from the Malaria Indicator Survey [13]. The ITN was first identified before asking a standard question (‘Was this net slept under by any person last night?’) to the respondent for determining ITN use in the household. Further details on the survey methodology are presented elsewhere [7].

The total sample included 4,602 participants; however, only individuals (3,056) living in households with at least one ITN and sleeping in the households the night before the survey visit (de facto population) were included in the analysis. It should be noted that in Wave 1, the official definition of household used was from the 2008 NDHS. This definition stated that a household was “a person or a group of persons, related or unrelated, who live together and share common cooking and eating arrangements”. However, during the Wave 2 campaign, each wife, with her direct dependents, was defined as a household in polygamous households. For the purpose of this analysis, the authors used the NDHS’ definition of household and adapted the data from Wave 2.

The determination of net ownership consisted of computing two key malaria indicators: (a) the percentage of individuals living in households with at least one ITN and (b) the percentage of individuals using an ITN in households with at least one ITN, by sex and background characteristics.

Binary response logistic regression was used to assess the effect of sex on ITN use among individuals living in households with at least one ITN. The model controlled for covariates reported by other studies to be associated with net use, including household wealth quintiles, campaign waves (Wave 1– May 2009 and Wave 2 – July 2009), age of individual, place of residence, education of the head of household, polygamous households, and ratio of ITNs to household members. The campaign wave covariate was chosen to determine whether messages from campaign waves varied by gender and affected ITN use. Age was selected since different age groups have dissimilar needs for or beliefs about using ITNs. Place of residence was chosen to see if the universal campaign messages were accessible in both rural and urban areas. Since northern Nigeria is comprised of a predominantly traditional, conservative population where the head of the household is often the decision-maker, the education level of the head of household was considered. The polygamous household variable was also measured since the ITN distribution strategy during the campaign considered every wife as a household unit; however, in the analysis, the official definition of a household was used in order to be comparable to those used as demographic health survey indicators. Finally, previous studies have shown that more nets in a household lead to better net usage, thus analysing the ratio of ITNs to household members was essential. The statistical significance was tested at the 95% level.

The study sample included 4,602 individuals, 51% of which were female and 49% of which were male; the majority of respondents were between 15 and 49 years old. Sixty-six percent of those surveyed lived in rural areas with an even distribution of females and males. The majority of the study sample (65%) reported having no formal education, while 19% completed primary level and 16% completed secondary level.

Of the 4,602 individuals surveyed in the post-campaign survey, 66% lived in households that owned at least one ITN, with no difference in ownership between females (67%) and males (66%). Wave 2 (July 2009) revealed better results than Wave 1 (May 2009), overall. There was no difference by sex in the proportion of individuals living in households that owned at least one ITN in either rural or urban areas. Individuals belonging to households in the second and fourth wealth quintiles were the most likely to own at least one ITN (71% and 70%, respectively), followed by the middle (68%) and lowest (64%) quintiles. The highest wealth quintile had the lowest (61%) ITN ownership (Table 1).

Overall, ITN use among individuals living in households with at least one ITN was 53%; however, there was a significant difference in use between females and males (57% vs 53%, p<0000.1). Consistently, females reported significantly higher ITN use than males in Wave 1 (52% vs 46%, p=0.017) and Wave 2 (62% vs 51%, p<0.0001). Similar patterns showing high female ITN use were observed by place of residence with higher ITN use among females compared to males in both urban (56% vs 44% p<0.0001) and rural (58% vs 52%, p=0.004) environments. Furthermore, females had significantly higher use of ITNs compared to males when analysed by wealth quintiles, particularly in the lowest (57% vs 46%, p=0.012) and fourth (58% vs 46%, p=0.003) quintiles. When the ratio of one ITN per two persons in a household was met, net use was equally high among women and men. In contrast, when the ratio was not met, females used nets more than the males (53% vs 48%, p<0.001). Age was also associated with ITN use. Among individuals below 15 years, the use of ITN was similar among females and males; however, at the age of 15 years and above, use of ITN was higher among females compared to males (Table 2).

The results of the logistic regression show that females are more likely to use ITNs compared to males after controlling for a number of potential confounders, including: household wealth quintiles, campaign wave, age, place of residence, education of the head of household, polygamous status, and ratio of ITNs to household members (OR: 1.5, 95% CI: 1.3-1.7).

Among the covariates included in the model, the following were significantly associated with ITN use: campaign wave, age, education and ratio of ITNs to household members. After Wave 2 (July 2009), individuals were more likely to use ITNs compared to Wave 1 – May 2009 (OR: 1.4, 95% CI: 1.2-1.6). Age shows a significant effect on ITN use. Children under five years had higher odds of using an ITN compared to individuals aged 25 years or older (OR 1.4, 95% CI 1.2-1.6). Individuals aged 15–25 years were significantly less likely to use an ITN compared to individuals aged 25 years or older (OR: 0.6, 95% CI: 0.5-0.7). Education level of the head of household was associated with ITN use: in households with a head that had no formal education, ITN use was lower. The number of ITNs in the household was a strong predictor of ITN use; individuals living in households with at least one ITN for every two members were 2.5 times more likely to use an ITN compared to individuals living in households with less than one ITN for every two members. Such background characteristics as household wealth, place of residence, and polygamous households were not associated with ITN use (Table 3).