Examples of Methods to Address Underreporting of Induced Abortion: Preceding Birth Technique and Randomized Response Technique

Application of PBT in Ghana A pilot test of PBT to assess its feasibility in estimating stillbirths, miscarriages and induced abortions was conducted in Accra, Ghana (Oliveras et al. 2008). Data were collected during prenatal and maternity visits made from November 2003 through January 2004 at three public and two private clinics. As such, the sample reflects women with access to prenatal and maternity services available in Accra and is not representative of the general population. Staff nurses at the participating clinics collected the data as part of their routine care. Prenatal clients were interviewed at the time of check-in, and delivery clients were interviewed at a time that was convenient for them; all interviews were strictly confidential. All women older than 15 who came to the clinics for services were included. Since the pilot showed that the method worked better in terms of implementation and reporting with prenatal than with delivery clients (Oliveras et al. 2008), the text below describes its application in the recommended setting of prenatal care provision. To ensure that each prenatal client was interviewed just once, clients were only eligible (and interviewed) if they were making their first visit. Women’s background characteristics were collected from both health records and the study questions to allow researchers to assess any personal characteristics that could be associated with abortion. Nurses conducting routine intake interviews incorporated four additional questions that asked about the outcome of a woman’s preceding pregnancy (Table 1, see table at the end of the chapter). First, each woman was asked whether she had been pregnant before. Women who said they had were asked about the outcome of their last pregnancy. If that pregnancy did not end in a live birth, the woman was asked if she had had a stillbirth, a miscarriage or an induced abortion. Women who reported a The purpose of this chapter is to present examples of data collection techniques that minimize or eliminate the underreporting of induced abortion, which is particularly relevant in contexts where abortion is illegal or access is highly restricted. The methodological challenge of measuring induced abortion, which is deliberately practiced clandestinely, has led researchers to adapt estimation methods that can investigate stigmatized behaviors. The techniques described in this chapter are the Preceding Birth Technique (PBT) and the Randomized Response Technique (RRT). These methods, when combined with population and pregnancy data, can yield indicators such as abortion rates and ratios, and when applied to data for multiple years, can yield information about trends in induced abortion. However, they cannot obtain detailed information on the characteristics of women who have induced abortions. The results of RRT in particular are usually used to complement findings from other research techniques.