Menu
Menu
Menu
Menu
Background: HIV/AIDS has had a significant impact on maternal and child health in South Africa. It is thus of vital importance to implement interventions to prevent mother-to-child transmission of HIV (PMTCT) as early as possible during pregnancy. Negative interactions between patients and health care providers (HCPs) can be an important barrier to antenatal care, PMTCT use and PMTCT adherence. Research about respectful maternity care has focused more on the patient perspective. We therefore compared the patient and HCP perspectives and reflected on how interactions between HCPs and patients can be improved. Objective: To obtain insights into the attitudes of HCPs in the context of HIV and PMTCT-related care, by studying patient and HCP perceptions of their interactions, in a peri-urban hospital setting in Gauteng province, South Africa. Methods: A qualitative study was conducted in a public tertiary-level hospital. Fourteen semi-structured in-depth interviews were conducted with nurses and doctors in the antenatal clinic and postnatal ward. Thirty-one semi-structured in-depth interviews and two focus group discussions were conducted with HIV positive and negative women on the postnatal ward. Results: HCPs experienced a difficult work environment due to a high workload. This was combined with frustrations when they felt that patients did not take responsibility for their own or their child’s health. They were motivated by the need to help the child. Patients experienced judging comments by HCPs especially towards younger, older and foreign women. They expressed fear to ask questions and self-blame, which in some cases delayed health care seeking. No discrimination or isolation of HIV infected patients was reported by patients and HCPs. Conclusion: We hypothesize that more humane working conditions for obstetric HCPs and a caring, personalised approach to patient management can improve patient-provider interactions and access to respectful care. These are critical to preventing mother-to-child transmission of HIV.
This qualitative study was inspired by Bacchi’s social constructivist approach, namely ‘What is the problem represented to be? (WPR)’ [26]. Bacchi highlights that the representation or definition of a problem differs from one stakeholder (patient) to another (HCP) and that these differences are rooted in personal and professional backgrounds. Consequently, this approach focuses on four important questions, namely: ‘What presuppositions or assumptions underlie the problem definitions of the different stakeholders?’; ‘What dimensions of the problem are left unproblematic in each of these problem definitions?’; ‘What solutions for the problem follow from the problem definition?’; and ‘What impact do these problem definitions and proposed solutions have?’ [26]. Applying the WPR-approach, we chose qualitative methods, which allow open-ended, emerging questions and probing, to develop a holistic understanding of the problem and context, based on patients’ and HCPs’ own words [27–29]. Semi-structured interviews (SSIs) were combined with focus group discussions (FGDs) for triangulation [27] and comprehensive understanding [30]. The study was conducted in a public hospital serving a poor population, in Gauteng province, South Africa. Although mainly peri-urban, the patient population includes patients from rural areas, other South African provinces and other African countries. English is the second language of most patients. A recent study that has just been accepted for publication noted a 22% HIV prevalence amongst women delivering at the hospital (personal communication, Goga A). Patients for the semi-structures interviews (SSIs) were recruited in the postnatal ward and approached by the research assistant and FH. The ward has space for up to 40 patients at a time and, on average, one patient was interviewed per day during 8 weeks of data collection. Patients were eligible, if they met all the following criteria: at least 18 years old; inpatient in postnatal ward; within the first week postpartum and not being in acute pain, emotional distress or discomfort. Both HIV-positive and HIV-negative patients were eligible for inclusion, because in this setting all patients are at risk of HIV infection and need timely access to antenatal care (ANC) to test for HIV. Patients who fitted the eligibility criteria were identified through ward records. Purposive sampling [27] was used so that key groups (such as younger women, older women) of participants were targeted. Patients were selected either with different characteristics compared to the patients who were interviewed before to test emerging theory, or with similar characteristics as previous interviewees until theoretical saturation occurred [27,28]. These characteristics included age, origin, HIV-status and delayed access to maternity care (Table 1). Characteristics of participants. FGD questions were informed by the interviews and were conducted in weeks 3 and 5 of data collection. A mix of purposive and convenience sampling [27] was used to select FGD participants. On the day of the FGD the nurse in charge of the ward helped to identify patients that met eligibility criteria listed under SSIs, and sixteen patients were approached. Fourteen HCPs participated in the SSIs. HCPs who worked in the hospital were eligible if they met all the following criteria: at least 18 years old; working as a professional nurse, staff nurse, auxiliary nurse or medical doctor and working in the antenatal clinic or postnatal ward. Purposive sampling was used [27]: HCPs were targeted based on their age, profession and the location (antenatal clinic or postnatal ward) until theoretical saturation was reached. Participating HCPs were aged between 25 and 58 years and their work experience in maternity care ranged from 9 months to 36 years (Table 1). To bridge cultural and language barriers during the interviews, two research assistants were employed. Both assistants attended a structured training course, conducted by FH, about qualitative data collection using the research protocol and the Qualitative Research Methods Field Guide by Mack et al. [28]. During three consecutive days they were trained in participant recruitment, moderator skills for SSIs and FGDs, and ethical considerations. FH mentored the research assistants who interviewed consenting patients in the postnatal ward. A guide consisting of eleven questions for patients and twelve questions for HCPs was used for the SSIs and FGDs (Appendix). SSIs usually lasted between 30 and 60 minutes and FGDs lasted between 55 and 80 minutes. All interviews were conducted in a private space. Interviews were only voice recorded with participant consent. 35 out of 45 SSIs and both FGDs were voice-recorded. Notes were taken during all interviews. Following each interview, FH and the interviewing assistant held a debriefing. The voice-recorded SSIs and FGDs were transcribed verbatim. One of the research assistants aided with the transcription of voice recordings. Field notes of the remaining ten SSIs and of both FGDs were elaborated in Microsoft Word. All SSIs and FGDs were analysed with thematic analysis [31] using NVivo (11) software for qualitative data analysis. Notes were taken of emerging themes during data collection and analysis. Transcripts and notes of the interviews were first reviewed to familiarize ourselves with the data. Initially, a set of fifteen transcripts was analysed to generate initial codes using an inductive approach [27] by going through each transcript and identifying topics line-by-line. An iterative process was used to organize topics that arose from the transcripts into (sub)categories, for example ‘negligence’ under the subcategory ‘causes of HIV’ and the category ‘HCP experiences’. Coding of the transcripts and the definition of (sub)categories was discussed among the authors. An initial diagram was developed in search for relationships between categories. Using a deductive approach [27], the codes and subcategories were organised into themes based on Bacchi’s WPR-approach. The code ‘negligence’ as a cause of HIV for example, became part of the subtheme ‘patients who do not take responsibilities’ under the theme ‘problems ascribed to the patients’ and the category ‘problem definition’ as part of the HCP perspective [26]. A coding scheme was developed following discussion and was based on emerging themes. The coding scheme was used to code the complete set of SSIs and FGDs in NVivo (11). When needed, new codes and subthemes were added, and the coding scheme was refined until expert agreement on the themes was reached between the authors. The initial diagram was adjusted, guided by Bacchi’s WPR approach [26], illustrating relationships between findings. Resulting themes and subthemes are reported in this article. The South African Medical Research Council Ethics Committee approved the research protocol (Protocol ID: EC008-6/2014), and permission was received from the hospital research committee. All participants provided written informed consent before participating in the study. The two research assistants had a similar cultural background as patients and HCPs to ensure cultural integrity and research validity. The research assistants and FH emphasised that participants were not obligated to answers questions that caused discomfort. All interviews and data analyses were conducted confidentially and in line with the Helsinki Declaration [32].
N/A
