Table 3.
Multilevel parameter estimates and average odds ratios for pregnancy termination
| PARAMETER | ESTIMATE | S. E | AOR | 95% CI of AOR | ||
|---|---|---|---|---|---|---|
| Const | −4.57 | 0.186 | ||||
| Year (2003) | ||||||
| 2008 | −0.13 | 0.105 | 0.88 | [0.71, | 1.08] | |
| 2013 | −0.41 | 0.188 | 0.66 | [0.46, | 0.96] | * |
| 2018 | 0.16 | 0.103 | 1.17 | [0.96, | 1.44] | |
| Age group (15–19) | ||||||
| 20–24 | 0.85 | 0.076 | 2.34 | [2.02, | 2.72] | * |
| Education (none) | ||||||
| Primary | 0.35 | 0.089 | 1.42 | [1.19, | 1.69] | * |
| Secondary | 0.12 | 0.094 | 1.13 | [0.94, | 1.36] | |
| Marital status (never married) | ||||||
| Currently married | 1.10 | 0.102 | 3.00 | [2.46, | 3.67] | * |
| Living with a partner | 1.14 | 0.162 | 3.13 | [2.28, | 4.30] | * |
| Previously married | 0.64 | 0.206 | 1.90 | [1.27, | 2.84] | * |
| Watch TV (Never) | ||||||
| less than once a week | 0.11 | 0.095 | 1.12 | [0.93, | 1.34] | |
| at least once a week | 0.31 | 0.081 | 1.36 | [1.16, | 1.60] | * |
| Ethnicity (Hausa) | ||||||
| Igbo | −0.40 | 0.165 | 0.67 | [0.49, | 0.93] | * |
| Yoruba | −0.66 | 0.172 | 0.52 | [0.37, | 0.72] | * |
| Others | −0.08 | 0.089 | 0.92 | [0.78, | 1.10] | |
| Number of children (none) | ||||||
| One | −0.25 | 0.077 | 0.78 | [0.67, | 0.91] | * |
| Two | −0.50 | 0.094 | 0.61 | [0.50, | 0.73] | * |
| 3 and more | −0.84 | 0.121 | 0.43 | [0.34, | 0.55] | * |
| Age at first sex (18+) | ||||||
| Never had sex | −5.55 | 1.060 | 0.00 | [0.00, | 0.03] | * |
| Less than 15 | 0.84 | 0.094 | 2.32 | [1.93, | 2.78] | * |
| 15–17 | 0.50 | 0.790 | 1.65 | [0.35, | 7.76] | |
| Contextual factors | ||||||
| contracep prev_cluster | 1.43 | 0.231 | 4.18 | [2.66, | 6.57] | * |
| Random effects | ||||||
| Level 3 - State | 0.10* | 0.033 | (ICC = 0.03) | |||
| Level 2 - Cluster | 0.22* | 0.056 | (ICC = 0.09) | |||
* - significant at 5% level (p < 0.05)