Reproductive health indicators in Nigeria remain among the poorest in the world. Modern contraceptive prevalence rate has remained practically at the same level since 2003. Whereas contraceptive use is higher in urban compared with rural areas, the majority of urban women are not using a modern contraceptive method. Low contraceptive use in the country results from unfavorable ideational variables, including knowledge, values, beliefs, and norms. Ideational segmentation allows program planners to identify groups that are homogenous with respect to ideational variables and which have similar informational and skills strengthening needs. In this study, we conducted latent class analyses with longitudinal data collected from a sample of women in 2010/2011 and 2014 to identify distinct ideational segments among women of reproductive age in six Nigerian cities at the two points in time. The analysis revealed three ideational segments at baseline: (a) Inefficacious, misinformed and unsupported, labeled "Disempowered," 58.8%; (b) Efficacious, misinformed, and unsupported, labeled "Unsupported," 31.8%; and (c) efficacious, supported but misinformed, labeled "Skeptics, 9.4%. These groups differ in terms of sociodemographic characteristics and contraceptive use and intention to use. At endline, four ideational segments were identified: (a) Inefficacious, misinformed and unsupported, labeled "Disempowered," 28.9%; (b) Efficacious, supported but misinformed, labeled "Skeptics," 28.5%; (c) Efficacious, informed but unsupported, labeled "Unsupported," 27.0%; and (d) efficacious, informed and supported, labeled "Empowered," 15.6%. Exposure to Nigerian Urban Reproductive Health Initiative communication interventions was one of the most significant predictors of improvements in ideational characteristics over time. Messages tailored to the informational needs of the various ideational segments and that address their worldview are needed to increase contraceptive use.
Keywords: Nigeria; contraceptive use; ideation; segmentation; urban.