An analysis of delays among women accessing second-trimester abortion in the public sector in South Africa.

Constant D, Kluge J, Harries J, Grossman D

Contraception
Apr 2019

Objectives

To identify key delays and associated factors in women's pathway to second-trimester abortion that could inform strategies to increase earlier presentation.

Study design

We performed a secondary analysis using data collected from May 2012 to June 2013 as part of a randomized controlled trial among women having abortion at 13.0–20.0 weeks at a public hospital in South Africa. We used ultrasound and participant interview data to calculate 3 key intervals: (1) conception to suspicion of pregnancy, (2) suspicion to first healthcare visit for abortion, and (3) first healthcare visit to abortion procedure. We compared intervals for women at 13–15.0 weeks versus 15.1–20.0 weeks gestation at abortion using Wilcoxon rank-sum tests and tested for associations between gestational age at key events using multivariable linear regression.

Results

Median (interquartile range[IQR]) durations for the 3 intervals among women at 13–15 weeks (n=93) compared to 15.1–20 weeks (n=63) gestation were: (1) 36 days (IQR 21–53 days) versus 62 days (36–71 days), p<.001; (2) 29 days (IQR 15–46 days) versus 23 days (IQR 11–39 days), p=.64; (3) 14 days (IQR 7–21 days) versus 14 days (IQR 12–21 days), p=.32. Multivariable logistic regression showed marginal associations between gestational age at suspicion of pregnancy and no prior pregnancy (aOR=3.8, 95% CI 1.0–14.6) and living in informal housing (aOR=3.1, 95% CI 1.0–9.1). Gestational age on the day of the abortion procedure was significantly associated with living in informal housing (aOR=3.1, 95% CI 1.4–6.6).

Conclusion

The only differences in delay in obtaining second trimester abortion between South African women having an earlier and later second trimester procedure is due to longer time to suspect pregnancy.

Implications

Interventions to improve early pregnancy recognition should be explored and referral processes should be streamlined to avoid unnecessary delays accessing abortion care and possibly reduce the proportion of abortions performed later in the second trimester in South Africa.