RESUMEN
Objectives Providing counseling on danger signs of pregnancy complications as part of visits for antenatal care (ANC) can raise expecting women's awareness so that if danger signs occur they can seek assistance in time. The study examines the level of agreement in counseling on danger signs between observation of the provider during the ANC visit and the client's report in the exit interview, and the association of this agreement with the client's level of knowledge on danger signs. Methods The analysis used data from service provision and assessment (SPA) surveys in Haiti, Malawi, and Senegal. Agreement between the observation and client's report was measured by Cohen's kappa and percent agreement. Regressions were performed on the number of danger signs the client knew, with the level of agreement on the counseling on danger signs as the main independent variable. Results The study found little agreement between the observation of counseling and the client's report that the counseling occurred, despite the fact that the exit interview with the client was performed immediately following the ANC visit with the provider. The level of positive agreement between observation and client's report was 17% in Haiti, 33% in Malawi, and 23% in Senegal. Clients' overall knowledge of danger signs was low; in all three countries the mean number of danger signs known was 1.5 or less. The regression analysis found that, in order to show a significant increase in knowledge of danger signs, it was important for the client to report that it took place. Conclusions Ideally, there should be 100% positive agreement that counseling occurred. To achieve this level requires raising both the level of counseling on danger signs of pregnancy complications and its quality. While challenges exist, providing counseling that is more client-centered and focuses on the client's needs could improve quality and thus could increase the client's knowledge of danger signs.