LANGUAGE, VIOLENCE AND STIGMA: BARRIERS TO CARE FOR HIV-POSITIVE PREGNANT WOMEN IN THE DOMENICAN REPUBLIC

The Regional Initiative for the Elimination of Mother-to-Child Transmission of HIV goal is to reduce the prevalence of HIV in the newborn population respectively to 2% by the year 2015. This Initiative is promoted by the World Health Organization (WHO) and the United Nations Children's Fund (UNICEF). Governments from low and middle income countries have adopted and adapted the international directives of the Initiative to their national context in order to respond to international commitments, including the achievement of the Millennium Development Goals (MDGs) 5 and 6, related to maternal health and preventable diseases and infections, respectively. Even though bio-medical innovations, mainly Anti Retroviral Drugs, and well-known best practices in public health implementation have made it feasible to prevent most of the Mother to Child transmission, approximately 2.5 million children worldwide are currently living with HIV and every day 1000 children are born with HIV (Marcos et al., 2012; UNAIDS, 2010). In the Dominican Republic (DR), regardless a 98% coverage of antenatal services and institutional birth with skilled staff (MSP and UNICEF, 2011), there are lingering shortcomings that result in high maternal mortality and child mortality rates (150 per 100,000 and 36 per thousand, respectively) (UNICEF et al., 2013). If skilled attendance on birth is \u201cthe single most important factor in reducing maternal mortality\u201d (WHO, 1999), how could anyone reconcile Dominican Republic\u2019s quasi-universal skilled attendance with a maternal mortality rate that almost doubles the Latin American average? The answer to this paradox, according to UNDP, is the low quality and inequality of health care services (2010). These systemic and institutional weaknesses also affect the Program for the Prevention of Mother to Child Transmission -PMTCT- where the current transmission rate has been estimated in 6% and the rate of full screening for HIV in pregnant women is only 19% (UNICEF, 2012). Consequently, the implementation of the PMTCT in Dominican Republic faces various challenges that range from the disempowerment and vulnerability of the target population (HIV positive pregnant women) to the weak incorporation of the initiative into mainstream prenatal care in order to avoid parallel programs (MSP and UNICEF, 2011). Prevention of transmission from a HIV positive mother to her child requires the completion of a series of consecutive and necessary steps in a continuum of care, also known as the PMTCT cascade (Towle 2009, Marcos et al., 2012). One of the main problems identified in Low and Middle Income Countries (LMICs) is the poor retention and low adherence along this cascade. This phenomenon also called loss to follow-up (LFU) \u201chas been recognized as a major hurdle by PMTCT programs in resource poor settings\u201d (Panditrao et al., 2011). It affects 40% of women enrolled in India, up to 81% in Malawi, 84% in Cote d\u2019Ivoire and more than 70% in South Africa and Zimbawe (Panditrao et al., 2011; Manzi et al., 2005; Painter et al., 2004). Given such high rates of non-adherence to the PMTCT Cascade, the effectiveness of these programs is eroded \u201cnot only because the objective of reducing pediatric HIV transmissions is compromised but also because of the missed opportunity to link HIV-infected women and their partners to further care and support activities\u201d. (Panditrao et al., 2011). This programmatic failure is compounded by the lack of information and academic literature about the institutional or social mechanisms that trigger a higher or lower adherence level and the risk-factors that influence drop-out in PMTCT (Towle, 2009; Panditrao et al., 2011). According to Horne \u201cnonadherence is often a hidden problem: undisclosed by patients and unrecognised by prescribers\u201d (2005). Notwithstanding massive advance in the evolution of health care services and in the presentation and design of medicine, adherence is still an unsolved problem and therefore the research question has been formulated to investigate: Why are adherence and non-adherence triggered in the intended users of the Program for the reduction of mother to child transmission in the Dominican Republic? In low and middle-income countries, where lack of appropriate data and information systems restrain longitudinal monitoring of patients and hinders systematic investigation, the study of adherence is often confined to epidemiological approaches, leaving the interaction between patient and the health system superficially addressed. Most of the existing literature focuses on the demand for HIV health services (patient behavior), ignoring supply (institutional arrangements and organizational culture). With regards to PMTCT, very few exploratory studies have discussed the social, economic and structural barriers and facilitators to full adherence in low and middle income countries (Towle, 2009; Painter et al., 2004; Castro and Farmer, 2005; Campbell, 2003; Parker, 2001; Trickett, 2005, 2004). In order to analyze and unpack this phenomenon a Fuzzy Set Qualitative Comparative Analysis will be used to address the task of improving the understanding of PMTCT program, its mechanisms and the context in which adherence develops. Service level delivery will be analyzed using in depth interviews with 120 HIV positive women that attended the program. Their experiences with the program have been recorded, translated where necessary, and transcribed. All the data has been systematized, coded and analyzed with fuzzy set Qualitative Comparative Analysis looking at how different combinations of conditions concur to adherence and risk for them and for the baby. Policy designers\u2019 opinions and practices have been studied during one year of participating in the National Technical Group for the Elimination of Mother to Child Transmission in the Dominican Republic that integrate members of the Ministry of Health and International Organization such as UNICEF and PAHO. This participation has endorsed the elaboration of a series of research questions and interest of the Ministry of Health opening up the access to the two main maternities in the country and partly their documents. Health care providers and practitioners\u2019 practices have been studied during two years of observations in maternities and this experience has been essential for creation of the interview questionnaire and in the interpretation of the fsQCA solutions. This thesis is developed in five chapters. Chapter 1 presents the insertion of the study in the trend of analysis of Public Policies in LMICs and the state of the art in this academic field. The Chapter starts by describing current trends and challenges in the literature of health public policies in low and middle-income countries and develops by looking at the importance of adopting a locus at the front line service level to study the experience of non-traditional actors: the women involved in the program. Subsequently, the Chapter presents the factors found in literature that are related to the possible barriers faced by the women in the program from a psychological, social, economical and structural perspective. The chapter ends by presenting the most important and known definitions used to define different types of patient commitment to health programs. The second Chapter presents the outcome, each single condition adopted for the model and their operationalization, the hypothesis and the data set. The third Chapter is divided in two parts. The first one presents the analysis with fuzzy set Qualitative Comparative Analysis starting with the results of the Analysis of Necessary Conditions for Positive and Negative Outcome and following with the Analysis of Sufficiency. The second part describes the result of the episodes of violence\u2019s categorization that emerged during in depth interviews with the women interviewed. The Chapter ends by exploring the most relevant topics encountered during the analysis of the transcriptions related to the program and the mechanisms involved in adherence and non-adherence. The last Chapter presents the discussion of the results and applied policy implication namely that the most significant determinants for adherence in the Dominican case are language and HIV Knowledge. Despite the numerous literature regarding education and female employment as strong determinants of adherence, the evidence collected does not support their status as necesary conditions. This final chapter ends by addressing the limitation of the current research and future possible development in the field of adherence to the PMTCT in LMICs