Infant feeding in the context of HIV: a qualitative study of health care workers’ knowledge of recommended infant feeding options in Papua New Guinea

Vertical transmission is the leading source of human immunodeficiency virus (HIV) in children under the age of 15 years [1]. Without intervention 30-45% of infants born to HIV-positive mothers in developing countries will become infected during pregnancy, childbirth and breastfeeding [2]. Interventions to prevent mother-to-child transmission during childbirth and breastfeeding have been shown to reduce infections in infants to less than 5% in low and middle income countries [3–5].

The superiority of breastfeeding over artificial feeding is well documented and exclusive breastfeeding remains one of the most valuable interventions for improving child survival, especially in resource poor settings [6]. The transmission of HIV through breastfeeding was first identified in 1985 [7] and since then the issue of breastfeeding within the context of HIV and prevention of mother to child transmission (PMTCT) has continued to be at the centre of much debate and policy. The World Health Organization (WHO) recommends all mothers, regardless of their HIV status practice exclusive breastfeeding for the first six months of an infant’s life.

In the absence antiretroviral therapy (ART) the risk of mother-to-child transmission of HIV through breastfeeding is between 20-45% [8]. However, with the use of ART by the mother this risk can be reduced to less than 5%, even among infants who are breastfed [9–11], because ART reduces the HIV viral load in the mother’s milk [12]. While breastfeeding is associated with risk of transmission of HIV [13], exclusive breastfeeding for the first six months is associated with a lower risk of HIV transmission when compared to mixed feeding, even without ART [3, 4]. The risk of infants acquiring HIV through breastfeeding, therefore needs to be weighed against the increased risk of death from causes other than HIV, in particular malnutrition and serious illnesses such as diarrhoea, among non-breastfed infants [14].

In order to continue encouraging and promoting breastfeeding while at the same time reducing the risk of HIV transmission, global infant feeding guidelines relating to HIV-exposed infants have been developed and subsequently amended as knowledge and understanding of HIV and its biomedical management has advanced [15–20]. These guidelines are particularly relevant for HIV-positive mothers in low income countries [21]. While clearly favouring breastfeeding, the guidelines note that in some situations replacement feeding may be more appropriate, providing it is “acceptable, feasible, affordable, sustainable and safe”. If these provisions cannot be met, exclusive breastfeeding in the first few months of life (currently the first six months) is recommended [17, 19]. While earlier WHO guidelines (2001) recommended abrupt cessation of breastfeeding at six months [17], their most recent guidelines (2009) support the continuation of breastfeeding up to twelve months, after six months of exclusive breastfeeding [19], along with the introduction of other foods and fluids. Additionally, at this time, breastfeeding should only be stopped if an alternative nutritionally adequate diet can be provided [1]. Review of the international literature highlights difficulties for health care workers conveying best practice recommendations for infant feeding, especially for HIV-positive mothers [22–24]. A health care workers’ lack of up-to-date evidence and informed knowledge subsequently impacts on a mothers’ ability to provide safe infant feeding to their children [25, 26], with women often not adhering to best infant feeding practices to reduce HIV transmission [23, 27].

This study was designed in 2008, when Papua New Guinea (PNG) was classified as having a generalized HIV epidemic [28]. Since that time estimates have been progressively revised downwards to less than 1% among adults aged 15–49 years [28]. But HIV infection rates in PNG are still among the highest in the Asia Pacific Region, and unlike most other countries of the region, transmission is predominantly heterosexual. Overall antenatal clinic prevalence in 2010 was 0.5% [29]; HIV prevalence among pregnant women varies across the country with the highest rates in Enga (1.4%), Western Highlands (1.3%) and the National Capital District (1.2%) [29].

Initiated through the National Department of Health in 2004, PMTCT services are now available in all provinces in PNG, however coverage remains low [30]. In PNG only 60% of women attend antenatal clinic [31], therefore many women are not screened for HIV and if positive, prescribed ART (prophylaxis and treatment) for either herself or for preventing mother-to-child transmission (ART prophylaxis). Antenatal clinic provider initiated HIV counseling and testing services currently reach only 24% of pregnant women [32]. In 2009 only 12.2% of the estimated 2,150 HIV-positive pregnant women were recorded as having received ART (prophylaxis and treatment) to reduce the risk of mother-to-child transmission [30].

PNG has a culture of breastfeeding, however it is not one of exclusive breastfeeding. Infants are often mix-fed with foods and fluid other than breast milk from as early as a few weeks old [33, 34]. This practice poses significant risks to the infant, especially HIV exposed infants. Improving rates of exclusive breastfeeding to six months is recognized in PNG as crucial to achieving better nutrition throughout childhood [35] and to improved infant outcome regardless of HIV status.

Reflecting the WHO, UNICEF and UNAIDS joint statement (1997) and with evidence from other resource poor settings [12, 36] the Government of Papua New Guinea developed the first country-specific guidelines relating to HIV and infant feeding in early 2009. Prior to this PNG followed the 2001 WHO guidelines [17]. These guidelines complement national guidelines relating to infant feeding (2009) [35], which seek to protect, promote, and support breastfeeding for all mothers regardless of their HIV status. The guidelines explicitly recommend exclusive breastfeeding for infants of HIV-infected mothers for the first six months of life; replacement feeding is only supported if it is acceptable, feasible, affordable, sustainable and safe [37]. Throughout this six month period of exclusive breastfeeding the mother must continue to receive ART [35]. Reiterated within these guidelines is the Baby Feeding Supplies (Control) Act 1977 (amended 1984), an act to limit the availability of feeding bottles and to ensure that those who use them know how to use them safely [38]. Specifically the aim was to prevent the spread of bottle feeding into rural areas [38]. While this legislation is still in place, feeding bottles, teats, and other feeding aids are freely available in pharmacies throughout PNG without such a prescription.

The aim of this paper is to explore health care workers’ knowledge regarding infant feeding options in PNG. Specifically it aims to examine the extent to which their knowledge reflects the current recommendations outlined in the Papua New Guinea National HIV Care and Treatment Guidelines on HIV and infant feeding (2009).

As part of a study investigating women’s and men’s experiences of the Prevention of Mother to Child Transmission (PMTCT) program in PNG, key informants were interviewed in two sites in PNG, Mount Hagen in the Western Highlands Province and Port Moresby in the National Capital District. These sites were identified by a Research Advisory Group meeting in 2009 because they each have high HIV prevalence with well-established PMTCT, HIV and maternal and child health (MCH) services. Antenatal clinic HIV prevalence rates in these sites in 2010 were 1.3% (Western Highlands) and 1.2% (National Capital District) [29]. Informants were purposively selected based upon their involvement in PMTCT and MCH and were drawn from government, non-government and faith based services, reflecting the diversity of such services in PNG.

An interview guide for key informant interviews was developed in parallel with interview guides for women and men engaged in PMTCT programs. The guides were finalized by the research team, piloted and adapted as necessary. The guides were designed to triangulate data from services users and to elicit further details on providing PMTCT services in PNG. All interviews were conducted in a private location in the workplace and lasted between 45 and 90 minutes.

In depth interviews with twenty-eight key informants across the two sites took place between March and November 2011. All but two interviews were conducted in English and transcribed and cross-checked by the lead author. The two interviews conducted in Tok-pisin (a lingua franca of Papua New Guinea) were transcribed verbatim and translated into English.

Key informants included twenty-four health care workers: sixteen nurses/paediatric nurses or midwives and eight senior medical officers working in the field of HIV, obstetrics or pediatrics. Of these, twenty-two were in clinical practice at the time of the interview. The two not in clinical practice were in senior management and advisory positions within the field of MCH/HIV. Seven of the twenty four health care worker informants held positions specifically relating to PMTCT. The remaining four informants included a social worker, a provincial AIDS council coordinator and two People Living with HIV (PLHIV) patient experts.

More than half of the key informants in clinical practice reported having attended a PMTCT training program. The majority of these were nurses or midwives of whom four were PMTCT program managers or coordinators and four were sisters in charge of their respective clinical areas (well baby clinic, labour ward, and antenatal clinic). The remaining informants included senior medical officers, a voluntary counseling and HIV testing (VCT) counselor, a nurse working in a well-baby clinic and a midwife from antenatal clinic.

A code book was developed to thematically analyse the data using NVivo v.9, a qualitative software management program. Codes included those that were inductive that emerged from the data and deductive codes which were used to design the interview guide and drew upon theoretical and empirical findings from other studies. Coding was undertaken by the lead author and cross-checked by the second. To ensure confidentiality, all informants were assigned a pseudonym. For those in clinical practice only their clinical setting is mentioned in this paper, for example well baby clinic.

Ethical approval was gained from the Papua New Guinea Institute of Medical Research Internal Review Board (PNGIMR IRB: 0914), the Medical Research Advisory Committee (MRAC9.25) and the Research Advisory Committee of the National AIDS Council in Papua New Guinea (RES09 010). In Australia ethics approval was gained through the accredited Human Research Ethics committee of University of New South Wales (09208). Written consent to participate in the study was gained from all informants. This research was funded through the Australian Development Research Award Scheme.

Our findings are presented as they relate to the feeding options described in the current 2009 national guidelines (Table 1).

Since the introduction of the PMTCT program in PNG in 2004 health care workers have been exposed to changing advice relating to infant feeding. Specifically, the shift in thinking from abrupt cessation of breastfeeding HIV-exposed infants at four to six months, as advised by WHO in 2001, to the continuation of breastfeeding after the introduction of food and other fluids at six months, as advised in 2009 and in line with recommendations from the WHO.

As part of a wider study we explored health care workers knowledge relating to HIV and infant feeding. These findings highlight that while the majority of informants were familiar with the national policy of exclusive breastfeeding, this was less so in relation to the continuation of breastfeeding after the first six months, by far the most important change. Despite the recommendation from the National Department of Health for dissemination of the 2009 infant feeding guidelines, responses from our informants indicate that more than two years after the policy was released the updated information has not been adequately disseminated. The importance of clarity regarding infant feeding recommendations for HIV-positive mothers and the need to integrate new guidelines to all relevant levels of the health system through a systematic approach is highlighted in other parts of the globe [23–25, 27].

While many informants in our study discussed the importance of providing the necessary information enabling mothers to make an informed choice about their preferred feeding practice, the idea of choice is somewhat nebulous. Although the health care workers say that they give choice, the guidelines actually limit choice. For example, it states in the guidelines that artificial or replacement feeding “is not an option in the PNG context” [35]. Furthermore because health care workers were providing out dated information on best practices for safe feeding, as suggested by their narratives, women’s ability to make an informed choice was compromised. Moreover, as a result of the power dynamics between patients and healthcare workers who are positioned as experts transferring knowledge about infant feeding is never an objective process, as implied by the informants. A similar concern regarding up-to-date health care worker knowledge of safe infant feeding practices is well documented in Southern Africa [23, 24, 26]. The risks associated with dissemination of out dated feeding advice in terms of infant health are not only in terms of HIV but also diarrhoeal diseases, malnutrition and mortality [13]. As long as health care workers continue to disseminate out dated information, women will continue to practice unsafe infant feeding. It is incumbent upon PMTCT service providers to ensure they convey the most up to date information in order for women to protect their infants [23, 25].

None of the health care workers in this study reported that women expressed a lack of trust about their expertise in PMTCT as a result of discussing feeding options, rather than giving directives. This is in contrast to international research, which has shown that health care workers report women doubting their expertise and thus not trusting them because they offer options and not directives [22].

Our study shows that although four feeding options are listed in the guidelines only two were discussed at any length by our informants. Rather than reflecting poor knowledge, this study highlights what is relevant to PNG and its women. Similarly, international literature on HIV and infant feeding reinforce this emphasis [27].

Providing advice on optimal infant feeding in resource poor settings is problematic, especially in relation to HIV transmission. Findings from our study reflect those found elsewhere in identifying that key health care workers are not aware of up to date information relating to infant feeding, especially within the context of HIV. This lack of up-to-date information and knowledge leads to inaccurate information being relayed to women, especially in regard to the importance of exclusive breastfeeding for six months with continuation of breastfeeding at this time with the introduction of other foods and fluids. In PNG there is a need to prioritise the dissemination of the 2009 revised infant feeding guidelines thus ensuring those involved in PMTCT services are able to provide optimal advice to all mothers.