Table 2

Multivariate correlates and predictors of switching to more and less effective methods during COVID-19 among women at risk of unintended pregnancy, by country

FactorBurkina FasoKenya
More effective method*Less effective method†More effective method*Less effective method†
(n=300)(n=508)(n=977)(n=1765)
AOR95% CIAOR95% CIAOR95% CIAOR95% CI
Sociodemographic factors
Residence
 Urbanrefrefrefref
 Rural1.220.344.320.890.243.341.190.731.941.130.642.01
Age (years)
 25–34refrefrefref
 15–191.420.395.091.360.355.290.860.431.720.900.441.84
 35–490.240.070.812.340.806.830.900.581.410.980.631.52
Education
 Lowerrefrefrefref
 Higher2.370.826.831.760.535.931.160.721.870.750.501.13
Wealth
 Lowrefrefrefref
 Middle13.182.0285.8513.41.8398.250.650.401.050.940.571.54
 High9.651.8350.785.250.8034.490.740.421.310.850.471.53
Reproductive health
Parity (births)
 0–1refrefrefref
 2–33.490.6917.655.510.7043.40.690.341.430.580.331.05
 4+8.741.4353.241.890.2613.50.780.341.800.490.250.96
Emotional response to potential pregnancy
 Very happyrefrefrefref
 Happy2.630.4914.143.100.4322.30.890.41.960.840.371.91
 Mixed feelings0.390.082.001.480.258.761.300.712.371.090.552.19
 A little unhappy0.410.062.722.410.5710.190.860.391.870.920.431.95
 Very unhappy1.470.278.012.160.489.671.210.602.461.320.682.54
COVID-19-related factors
Household income loss
 Nonerefrefrefref
 Partial4.731.3616.461.570.406.120.890.441.801.460.673.18
 Complete2.420.4413.182.810.5913.400.900.461.741.010.482.12
Concern about future loss of income due to COVID-19
 Norefrefrefref
 Yes1.870.546.501.500.326.951.150.433.131.260.503.14
Food insecurity during COVID-19‡
 Nonerefref
 Chronic stable0.340.120.941.300.712.38
 Increased0.990.611.600.740.431.28
Able to socially distance
 Norefrefrefref
 Yes2.560.857.762.270.925.641.340.822.181.130.751.71
Concerned about becoming infected with COVID-19
 Concernedrefrefrefref
 Very concerned0.170.060.471.000.342.931.250.662.380.980.551.74
 A little/not concerned0.510.131.991.480.415.291.530.475.050.670.192.34
  • Multivariate regression adjusted for all other covariates listed and account for the complex survey design. Values in bold type indicate p<0.05.

  • *Among women not using highly effective contraception at baseline.

  • †Among women using effective/highly effective contraception at baseline.

  • ‡Sample sizes for food insecurity were too small in Burkina Faso to detect differences among women who used contraception at baseline and follow-up.

  • AOR, adjusted odds ratio; ;CI, confidence interval; ref, reference.