Search

International audience; In Madagascar, a country where over 60% of deliveries are not attended by a healthcare professional, late or inadequate responses to complications during childbirth account for a great number of maternal deaths. In this article, we analyse local perceptions of birth-related risks and strategies used to avoid these risks or manage complications of childbirth. We conduct this analysis in light of the social meanings of childbirth and the social expectations placed upon women in a context of socioeconomic vulnerability and a challenged public health system. We conducted two separate studies in the district of Morondava (Menabe region) in June 2014 and March 2015, comprising semi-directive interviews with 111 people (59 mothers, 18 members of their immediate en-tourage and 34 institutional or healthcare stakeholders), and eight focus groups discussions-two with community leaders, and six with fathers. The results show that the social pressure exerted on women to give birth without complications leads them to practices aimed at avoiding a "big" baby including dietary restrictions, physical activity, and refusal of iron supplementation intake. During pregnancy, women are usually accompanied by a traditional birth attendant or matron (reninjaza). Further, they use the public health system by attending antenatal consultations. However, women are reluctant to deliver in a health facility, where the practices of health professionals are in discordance with the social realities of women and local beliefs around childbirth. If complications arise, they are explained by social causes. The parturient woman is only taken to a healthcare facility after carrying out rituals and if the problems do not resolve themselves. These findings support recommendations to reduce the cultural distance between health workers and childbearing women, strengthen the collaborations with reninjazas, and inform women and their decision makers (mother, reninjaza, spouse) about nutrition during pregnancy and signs of complications.

The paper approaches the topic of sampling in the qualitative research design. Particular attention is paid to the composition of the focus groups in two important aspects: the homogeneity of the participants in terms of their positions in socio-occupational hierarchy and the pre-existing relations among the participants in the single group. The later issue is closely and directly intertwined with the former, in the case when the informants are approached in their institutional seting, and the research is conducted in a relatively small community. These problems are discussed in reference to the body of available methodological studies. The example of the research undertaken in aforementioned circumstances is subsequently presented to advocate the homogeneous composition of the focus groups while reconstructing the collective viewpoints of representatives of a particular level of public administrations. Avoidance of power relations within the groups proved profitable and effective. Substantial diferences between the groups were easily identifiable in spite of the pre-existing relations among participants, which were unavoidable in this very case. Artykuł podejmuje temat doboru próby podczas projektowania badań jakościowych. Szczególną uwagę zwraca się tu na kompozycję grup fokusowych ze względu na dwa ważne aspekty: ich homogeniczność pod względem pozycji badanych w społeczno-zawodowej hierarchii oraz wcześniejszych wzajemnych relacji pomiędzy uczestnikami jednej grupy. Ta druga kwestia wiąże się blisko i bezpośrednio z pierwszą w sytuacji, gdy uczestnicy fokusów objęci są analizą w swoim otoczeniu instytucjonalnym, a badanie prowadzone jest w relatywnie małej społeczności lokalnej. Te zagadnienia omówione są w odniesieniu do dostępnej literatury metodologicznej. Następnie zaprezentowany jest przykład badania przeprowadzonego w wyżej wymienionych okolicznościach, by wskazać zalety homogenicznej kompozycji grup fokusowych podczas rekonstruowania zbiorowych postaw przedstawicieli poszczególnych szczebli administracji publicznej. Uniknięcie relacji władz i podległości okazało się owocne i skuteczne. Zasadnicze różnice pomiędzy grupami było łatwo zidentyikować, mimo że w przypadku tego badania nie do uniknięcia była wcześniejsza znajomość uczestników przed rozpoczęciem sesji.