Context matters: a systematic review of neonatal care in humanitarian emergencies

Of the 15 countries with the highest neonatal mortality rates, 13 are characterised by conflict and political instability. Despite well-documented evidence of best practice interventions for neonatal survival, it remains less clear on how these practices are implemented in humanitarian emergency settings. To conduct a systematic review of published and grey literature on the implementation strategies and challenges in addressing neonatal care in humanitarian emergencies. A systematic literature search was conducted in SCOPUS, MEDLINE, Web of Science, CINHAL and Global Health for studies published between 1 January 2003 and 30 June 2018. Additionally, websites of organisations actively working in humanitarian emergencies were searched. Interventions were reviewed against the existing essential newborn care framework according to the standards outlined in the Newborn Health in Humanitarian Settings Field Guide. Twenty-one studies were identified: eight reporting on conflict and refugee settings, nine followed natural disasters and four discussed multiple emergency settings. Few studies addressed all the components of essential newborn care outlined in the field guide regardless of the emergency type. The review of literature demonstrated challenges in addressing essential newborn care identified in all humanitarian settings including the lack of adequate equipment, financing, and trained staff. Implementation strategies identified included quality improvement training for staff, the development of evacuation procedures, integrating with local and government resources and generating spaces in health facilities specifically for newborn care. The requirements and initiatives needed to deliver essential newborn care in humanitarian settings are highly variable and context dependent. Given the diversity of factors needing to be addressed by the field guide, more research should be directed towards the adaptability of the implementation strategies to differing emergency contexts. PROSPERO registration ID: CRD42018098824

the annual rate in reduction in mortality amongst children aged 1-59 months: 2.6% for neonates compared to 3.6% for the 1-59-month age group (UNICEF, WHO, WBG, UN 2018). As a result of the lower annual rate of reduction in mortality, neonatal mortality equates to a 47% share in all under five deaths, increasing from 40% in 1990(UNICEF, WHO, WBG, UN 2018. The top 15 countries with the highest neonatal mortality rates occur in low-and middle-income countries (LMIC) ( Table 1); these neonatal deaths are attributable to three main preventable causes: preterm birth complications (35%), intrapartum-related complications (24%), and sepsis (15%) (UNICEF, WHO, WBG, UN 2018;WHO, MCEE 2016).

Neonatal health in humanitarian emergencies
Of the 15 countries with the highest NMR, 13 are characterised by fragility, conflict, and violence (The World Bank Group 2017). Humanitarian emergencies are the result of armed conflict, natural disasters, food insecurity, and other crises that affect large populations (Humanitarian Coalition 2018). These emergencies are often characterised by excessive mortality, insecurity, and large population displacement, both within and across countries (Humanitarian Coalition 2018). In these contexts, the provision of essential newborn care has proven particularly challenging Save the Children, UNICEF 2015;UNICEF, WHO 2017).
Programmes to provide newborn care in humanitarian emergencies have been informed by two main field guides that provide technical and programmatic guidance. The first and earliest guide developed in 1999 is the Inter-Agency Field Manual on Reproductive Health in Humanitarian Settings developed by the Inter-Agency Working Group on Reproductive Health in Crisis (IAWG 2010). This field guide addresses newborn care in tandem with maternal care as a component of the Minimum Initial Services Package (MISP). The MISP outlines priorities in addressing the reproductive health needs of a population at the onset of an emergency with an emphasis on primary health care facility and hospital levels of care. The five components of the MISP are the identification of an agent to lead the implementation, prevention of sexual violence, reduction in HIV transmission, prevention of maternal and infant mortality, and the integration of reproductive health services into primary care (IAWG 2010). Whilst the provision of critical newborn services is a part of the MISP, evaluations of the MISP and improved neonatal data has resulted in a call for greater focus on specific interventions to address the disproportionately high numbers of newborn deaths (Casey 2015;Save the Children, UNICEF 2015).
In response to this, and the emerging evidence base on effective newborn care interventions, a second companion guide was developed: Newborn Health in Humanitarian Settings Field Guide (Save the Children, UNICEF 2015). This field guide is a compilation and summary of the World Health Organization (WHO) standards of care for neonatal health with additional guidance and indicators on how to provide neonatal services in humanitarian settings (UNICEF, WHO 2017). Informed by strategies of the Every Newborn Action Plan (UNICEF, WHO 2014), the field guide outlines several critical components of Essential Newborn Care (ENC) addressing household, facility and hospital levels of care. The components of ENC represent evidence-based interventions for the basic care required for a newborn in any setting; they include thermal care, infection prevention, initiation of breathing, feeding support, delayed umbilical cord clamping, monitoring, and postnatal care (Save the Children, UNICEF 2015; UNICEF, WHO 2014). Two thirds of newborn deaths are preventable with the adequate provision of ENC at birth and during the first month of life (Lawn et al. 2014).

Research agenda
Whilst these field guides exist (IAWG 2010

Methods
This review follows the PRISMA guidelines for the accurate reporting of systematic reviews and as per these guidelines the protocol for this systematic review has been registered with PROSPERO (ID: CRD42018098824). Ethical approval was not required for this review as data is publicly available.

Screening, selection and extraction
The search strategy was applied to all databases and the citations were imported into Covidence software, which allowed access by both reviewers. Following duplicate removal, title and abstract screening was conducted through the inclusion and exclusion criteria. Subsequently, full-text screening of selected publications was conducted to ensure they met the inclusion criteria. All screening was performed by the first reviewer with a second reviewer consulted for areas of uncertainty. Data was extracted through an Excel template that highlighted key variables for each of the studies (country and type of emergency; population; level of care, i.e. household, community, and hospital; ENC component/s identified; strengths and challenges; outcomes; study type; limitations and quality assessment). The final included studies underwent critical appraisal to determine the risk of bias, overall quality and strength in which conclusions could be drawn. For this review, National Institute for Health Care Excellence (NICE) and Effective Public Health Practice Project (EPHPP) were used to assess the quality of qualitative and quantitative studies respectively. For systematic reviews and mixed methods studies, Critical Appraisal Skills Programme (CASP) checklists were utilised.

Results
The search strategy yielded 1108 papers after duplicates were removed. Following title and abstract screening, 144 papers met the inclusion criteria and were submitted for full-text review; 20 papers met the inclusion criteria. Of the 124 papers excluded, 16 were excluded because they were logistical and field guides. These field guides and the 20 included papers' references were screened for additional papers, which yielded one additional paper (Fig. 1). Overall, 21 studies were included in the analysis (ten qualitative, three quantitative, six mixed methods, and two systematic reviews).

Study characteristics
Sixteen of the 21 included studies were in LMIC and the remaining five described high-income countries (Fig. 2).  (Casillo et al. 2016) and the USA respectively Espiritu et al. 2014;Orlando et al. 2008). Finally, four articles described multiple settings and emergency types (Casey 2015;Casey et al. 2015;Gopalan et al. 2017;Lam et al. 2012). Table 2 describes a review of all the studies including information gathered from the analysis and quality appraisal. Country context and type of emergency play a significant role in influencing the timing, level of preparedness, governance, and community factors that impact on service delivery during a humanitarian emergency. This review will comment on the specific humanitarian emergency types that were addressed in the included studies and how this impacted on neonatal care. Following this, the review will address the overall implementation strategies and challenges in providing ENC in humanitarian settings.

Earthquakes
Five studies reported on the complexities of newborn care following earthquakes: three in LMIC (Amibor 2013;Ayota et al. 2013;DeYoung et al. 2018) and two in highincome settings (Bengin et al. 2010;Iwata et al. 2017). Four out of the five included studies commented only on one ENC component, breastfeeding, and discussed the complexities of supporting breastfeeding women. Strategies presented in these studies included developing a space for women to breastfeed during the crisis (Amibor 2013;Ayota et al. 2013;DeYoung et al. 2018) and the promotion of accurate breastfeeding information to reduce misconceptions about breastfeeding and unregulated formula distribution (Ayota et al. 2013;Bengin et al. 2010;DeYoung et al. 2018). The remaining study by Iwata et al. (Iwata et al. 2017) was a descriptive editorial of a neonatal intensive care unit evacuation following an earthquake in Japan.
Following an earthquake in Haiti, Ayota et al. (Ayota et al. 2013) outlined the effectiveness of establishing 193 baby tents over five cities that reached 180,399 infant-mother pairs and 53,503 pregnant women. Over a 29-month period, the baby tents promoted safe breastfeeding environments and allowed for the registration, assessment, and referral of sick newborns and were an avenue to give culturally appropriate health promotional messages and psychosocial support (Ayota et al. 2013). A qualitative study by DeYoung et al. (DeYoung et al. 2018) in Nepal of a similar baby tent programme identified facilitators and barriers to breastfeeding for Nepalese women post-disaster. The study showed that the baby tents generated a sense of community amongst women, providing greater comfort for new mothers and pregnant women (DeYoung et al. 2018). Additionally the study revealed that the Nepalese women felt a sense of abandonment and worried for their safety after the baby tents were discontinued as humanitarian groups left (DeYoung et al. 2018).
An editorial by Bengin et al. (Bengin et al. 2010) in China additionally described newborn feeding patterns               following an earthquake event. The editorial commented on how an increased number of caesarean deliveries from 62 to 87%, to reduce uncertainty around the timing of birth, resulted in delayed breastfeeding with only 14.8% of post-partum mothers initiating breastfeeding within the hour (Bengin et al. 2010). Delayed breastfeeding can raise health concerns, as breast milk is vital for the growth and development of the immune system and the provision of basic nutrition for neonates (UNICEF, WHO 2014). All three studies on breastfeeding highlighted common misconceptions about breastfeeding including the inability to produce milk due to stress and the negative impact of unregulated infant formula (Ayota et al. 2013;Bengin et al. 2010;DeYoung et al. 2018).
A secondary data analysis study in Haiti conducted by Amibor (Amibor 2013) assessed the child and maternal health situation pre-and post-earthquake. Indicators specifically focused on the implementation of MISP, which was not established until four months after the earthquake. This led to an uncoordinated response from multiple humanitarian organisations and the government with insufficient and strained resources for newborn care (Amibor 2013).

Hurricanes
The timing of an event and level of preparedness can lead to a coordinated pre-disaster response. This is a unique characteristic of meteorological events (e.g. hurricanes and typhoons) and in some cases geophysical events (e.g. earthquakes and tsunamis), whereby pre-warning of the event can lead to greater governance and coordination of the response (WHO 2008). In the hurricane and earthquake events described by four included studies, the provision of ENC was defined by pre-disaster preparation responses, prioritising ENC components such as respiratory support, and the establishment of an integrated stepped-up emergency response plan (Iwata et al. 2017;Bernard and Mathews 2008;Espiritu et al. 2014;Orlando et al. 2008;Casey et al. 2015;Gopalan et al. 2017;Lam et al. 2012;Lawn et al. 2016;Froen et al. 2016;Culver et al. 2017;WHO 2020;Ban 2015;WHO 2008;Wise and Darmstadt 2015a).
All four descriptive studies commented on the evacuation of hospital neonatal intensive care units (NICU) during hurricane events in the United States of America Espiritu et al. 2014;Orlando et al. 2008) and an earthquake event in Japan (Iwata et al. 2017). All studies cited similar challenges to addressing neonatal care including supply shortages such as oxygen tanks and thermoregulation warming pads, as well as difficulties locating receiving hospitals due to the breakdown of Information Communications Technology (i.e. lack of internet and phone services). Two studies described the prioritisation of newborns based on immaturity, respiratory support, and vascular access which determined how far and when the newborn was to be evacuated, and allowed greater efficiency in resource allocation (Iwata et al. 2017;Espiritu et al. 2014). NICUs described in hurricane events in America noted predisaster responses leading up to the hurricane's landfall including discharging stable newborns, relocating high-risk patients to other hospitals, increasing staffing capacity, and re-checking equipment and supplies Espiritu et al. 2014;Orlando et al. 2008). All studies identified a hospital level stepped-up emergency response plan and highlighted the need to develop a multi-sector integrated response plan, both regionally and nationally, to assist in the swiftness of evacuation procedures for future incidents (Orlando et al. 2008).

Refugee settings
In contrast to natural disaster settings, the ENC provision in refugee settings described in the included studies primarily depended on three factors: integration with local host-country health systems for the establishment of referral pathways, the number of trained staff to deliver ENC, and whether refugees were based in camp versus non-camp settings (Turner et al. 2013;Bouchghoul et al. 2015;Krause et al. 2015;Casey et al. 2015).
A cross-sectional mixed methods study by Casey et al. ) explored the quality and availability of services for camp-and non-camp-based refugees and found that all health workers at the facilities lacked training in newborn infection management and supplies for newborn resuscitation. In Burkina Faso, one out of four camp health facilities (25%) and two out of 21 non-camp health facilities (10%) had basic ENC available . Non-camp health facilities also lacked functioning referral systems, identifying an advantage for the coordination of neonatal care within a camp setting . The advantage for neonatal care co-ordination in the camp setting was further highlighted in the evaluation of reproductive health services in Jordanian refugee camps whereby reproductive services in urban areas were lagging behind those coordinated in the camp setting, which resulted in higher neonatal mortality rates outside the camp compared to within the refugee camp (Krause et al. 2015).
An observational study of an obstetric unit in a Syrian refugee camp in Jordan by Bouchghoul et al. (Bouchghoul et al. 2015) identified the need for partnerships and integration into local host-country health systems. The study acknowledged difficulties with following up high-risk pregnancies and sick newborns that were being referred to an outside camp hospital, because there was no communication between these two facilities (Bouchghoul et al. 2015).

Protracted conflict
All seven studies that described protracted conflict events in LMIC identified the reliance on task shifting and community health worker (CHW) networks to deliver ENC (Ahamadani et al. 2014;Akseer et al. 2016;Khan et al. 2012;Sami et al. 2017;Hynes et al. 2017;Gopalan et al. 2017;Lam et al. 2012). Khan et al. (Khan et al. 2012) commented on how increased government policy and funding into CHW programmes for child and maternal health in Pakistan has helped to address workforce shortage and increase coverage of ENC in low wealth quintiles and hard to reach areas. The authors reported that CHW programmes were particularly important in Pakistan because foreign assistance for ENC were only concentrated in a few provinces and large or capital cities (Khan et al. 2012). Two studies identified cultural perceptions about motherhood as challenges for health workers when promoting newborn care. In studies conducted in South Sudan (Sami et al. 2017) and Pakistan (Khan et al. 2012), mothers were particularly opposed to kangaroo mother care interventions for thermal care due to religious traditions and family cultural practices. Particularly in the South Sudan study by Sami et al. (Sami et al. 2017) that interviewed health workers post an ENC training intervention, the health workers found it difficult to convince mothers to allow for delayed cord clamping because it was incorrectly perceived that the blood still moving in the cord could result in mother-to-child transmission of an infection.

Implementation strategies for addressing ENC
The primary strategies for addressing ENC to improve neonatal survival included the provision of separate areas in health facilities for neonatal care, training specific health worker teams to address neonatal care, increasing government financing, and up skilling/task shifting to community health workers (CHW) with training in ENC. Quality improvement training for health workers was found to be the most effective with a quasi-experimental study in The Democratic Republic of the Congo reporting a 2.4-fold increase in coverage of ENC (OR = 2.44 95% C.I. 1.28-4.66) after a -12 day enhanced ENC training intervention compared to health facilities who received standard training in neonatal care (Hynes et al. 2017). Similar success occurred in South Sudan (Sami et al. 2017) and after a typhoon in the Philippines (Casillo et al. 2016) with pre-/post-interviews and assessments showing increased awareness and confidence of identifying and carrying out ENC, and increased facility equipment in both settings.
A retrospective descriptive study at the Maela refugee camp on the Thailand-Myanmar boarder identified that the creation of separate newborn areas in the health facility, employing a specific staffing team for newborn care during labour and delivery, creating locally appropriate standardised newborn guidelines, and conducting annual retraining led to a reduction of the NMR by 51% over five years: 21.8 to 10.7 per 1000 live births (Turner et al. 2013). The creation of a dedicated team to address newborn care around the time of birth was similarly shown to increase the likelihood of newborn survival in Jordanian refugee camps (Bouchghoul et al. 2015).
Leveraging CHW programmes was identified in several studies as being critical to delivering ENC to mothers in communities. The presence and training of CHW in ENC increased the possibility of a mother having a facility-based birth, strengthened the connection to formal services, and provided continuity of care, especially settings characterised by protracted or prolonged conflict (Casey 2015;Khan et al. 2012;Sami et al. 2017;Gopalan et al. 2017).

Challenges for addressing ENC
All 21 studies identified two main challenges in addressing neonatal care regardless of emergency type and location: lack of appropriate and adequate equipment for ENC and limited staff knowledge of ENC components. In the 16 LMIC studies included, the initiation of breastfeeding was often noted as the only intervention routinely provided for neonatal care. In these settings, neonatal care was cited as being difficult to establish because of the perceived additional time and technical capacity required to deliver ENC (Casey 2015;Sami et al. 2017).
A systematic review evaluating reproductive health programmes in humanitarian emergencies by Casey (Casey 2015) identified similar challenges across all LMIC settings including the need to train health workers, procure equipment for ENC, and update competency based training particularly for emergency newborn and obstetric care. Comparable challenges were also recognised in the included studies in Iraq (Ahamadani et al. 2014), South Sudan (Sami et al. 2017), Jordan (Bouchghoul et al. 2015Krause et al. 2015), and The Democratic Republic of the Congo (Hynes et al. 2017). Additionally, these studies identified inequitable coverage of skilled birth attendants and access to health facilities across wealth quintiles, which impacted on the delivery of ENC.
The lack of knowledge of ENC components also extended to humanitarian workers. Lam et al. (Lam et al. 2012) utilised a web-based survey to understand newborn care practices for 56 humanitarian workers across 27 organisations involved in humanitarian emergencies. Almost two thirds (62.5%) of survey respondents reported having policies and guidelines on maternal health, and only 36.7% reported having training and guidelines for newborn care (Lam et al. 2012). In addition, none of the surveyed organisations implemented all ENC interventions with some components more commonly utilised over others. For example, the promotion of the immediate drying of the newborn (80.4%), resuscitation (62.5%), breastfeeding promotion within 1 h (87.5%), and kangaroo mother care (73.2%) contrast with the levels of provision of umbilical cord disinfectant (48.2%) and newborn care kit provision (39.3%) (Lam et al. 2012). Furthermore, 91.8% of staff surveyed expressed the desire for training on the management of neonatal complications (Lam et al. 2012).

Discussion
To the authors' knowledge, this is the first review to systematically synthesise and appraise current literature on strategies to support the implementation of ENC interventions in a range of humanitarian emergency settings. The 21 articles included in this review were analysed against ENC components as they form the standard protocol for care as outlined in the Newborn Care in Humanitarian Settings Field Guide (Save the Children, UNICEF 2015).

Limitations
There are several limitations associated with this review. The first being the small number of articles that met the inclusion criteria, with the majority being of weak to moderate quality. Critical appraisal of each of the articles identified 12 out of the 21 included articles lacked methodological rigour including small sample sizes and, publication and recall bias, which resulted in weak to moderate quality ratings. Further to this, many humanitarian organisations do not tend to publish the results of programmes implemented in humanitarian settings, and so the articles included in this review are not representative of the breadth of ENC programmes implemented. Secondly, this review was limited by the strict search and inclusion criteria to the neonatal period only, which could result in a number of articles missed due to neonatal and maternal health historically placed together in earlier humanitarian programming under the MISP. Neonatal services are a component in a continuum of care for sexual and reproductive health, and there are synergies in care required for the mother and baby during the peripartum period. It is notable however that across the included studies there were gaps in ENC in settings where maternal health services were in place, which supports the need for greater attention to those strategies designed to scale up ENC interventions. Limiting the inclusion criteria to the neonatal period may also have excluded studies reporting on programmes designed to prevent stillbirths; only four out of the 21 included studies mentioned strategies to prevent stillbirths (Ahamadani et al. 2014;Khan et al. 2012;Turner et al. 2013;Bouchghoul et al. 2015). Furthermore, our review of the literature was constrained by the fact that just four studies included neonatal mortality as an outcome measure in their results. There is a risk that without reported outcome measures the promising strategies to improve neonatal care described in this review may result in increased crude coverage of ENC interventions without good evidence that this will translate into improved outcomes (Marsh et al. 2020). Finally, the inclusion of English only articles could have impacted on the number of articles included in this review.

Challenges in addressing newborn care in humanitarian settings
In humanitarian emergencies, the disruption of health services and systems place women and neonates at a greater vulnerability to excessive mortality and morbidity compared to non-emergency settings. LMICs are often disproportionately affected by humanitarian emergencies, with the occurrence of an emergency often exacerbating an already under-resourced health system within the country (Culver et al. 2017). Addressing the existing gaps in the provision of ENC interventions in humanitarian settings provides an opportunity for significant gains in newborn survival. These gaps in ENC provision are not exclusive to humanitarian settings and are also well documented across many LMICs (The Lancet Neonatal Survival Steering Team 2005). However, this review has identified particular challenges of delivering ENC within differing humanitarian emergency contexts.

Gaps in the delivery of ENC
The findings in this review reflect diverse settings both geographically and by the type of emergency experienced. In almost all cases, lack of awareness and preparedness to deliver ENC components, equipment, and trained health workers were reported to be the main challenges in delivering ENC, impacting negatively on neonatal outcomes. The majority of studies identified the initiation of breastfeeding as the only routinely implemented intervention for newborn care, reflecting limitations in the quality, breadth, and consistency in the delivery of a comprehensive package of ENC interventions. Across all humanitarian emergency settings reported in our review, it was clear that all components of ENC needed to be translated into practice and indicators for neonatal care routinely monitored. It is encouraging to see the recent release of guidelines by the WHO that bring a stronger focus on the provision of quality care in fragile and conflict affected settings, and it is hoped this will encourage more implementation research on how to meet this gap between knowledge and practice in humanitarian settings (WHO 2020). Incorporating neonatal care into country emergency response assessments, intervention packages, and monitoring indicators is a part of several national milestones in the Reaching Every Newborn National Milestones Report (UNICEF, WHO 2017). These milestones provide programmatic guidance and hold countries accountable to incorporating neonatal care in national plans as outlined in the Sustainable Development Goals and The Global Strategy, 2016-2030(Ban 2015. Furthermore, this review highlights that additional guidance and action is needed for disaster-prone areas, particularly in LMICs, to incorporate these neonatal indicators into humanitarian emergency response plans to ensure that every newborn is counted.

Context
The most widely used humanitarian emergency typology categorises emergencies by time and the level of response required at each phase of the emergency: predisaster, acute, long term, and recovery and mitigation (WHO 2008). The Reproductive Health in Humanitarian Settings Field Guide identifies the immediate responses that must take place in order to set up the MISP in the acute emergency response phase; however, this does not take into consideration different types and phases of the emergency being experienced. Similarly, the Newborn Health in Humanitarian Settings Field Guide does not address differing humanitarian emergency types and phases and how this might influence delivery and implementation of ENC. It is evident from the studies included in this review that the requirements and ability to address neonatal care alters in different emergency contexts, for instance, addressing newborn care in hurricane-prone America versus refugee camps in South Sudan.
Outside of the emergency typology, political and governance structures also play a significant contextual role in the ability to respond effectively to an emergency. To exemplify this, Wise and Darmstadt (Wise and Darmstadt 2015a) utilised Worldwide Governance Indicators to develop composite measures of political instability and governance within a country and established an inverse relationship between NMR and political stability (r=−0.55), and government effectiveness (r=−0.77). Whilst the relationship is complex, the study recognises the contribution of political instability and poor governance in shaping global NMR (Wise and Darmstadt 2015a).
The included studies in this review on earthquakes in Haiti and Nepal demonstrate how the effectiveness of a humanitarian response can deteriorate rapidly in the face of a weak health system and poor governance (Amibor 2013; Ayota et al. 2013;DeYoung et al. 2018). There is limited discussion and evaluation of strategies required to address political and governance requirements of implementing a national maternal and neonatal service in areas of crisis presented in the papers included in this review. Financing, supply chain infrastructure, policy, adequate staffing, and establishing a coordinated response system between national and international partners are all key aspects to be considered when responding to any crisis (IAWG 2010;Casey 2015). Adequate health service provision in areas of conflict or protracted crisis is in no doubt complex, but in many circumstances not impossible. Experiences of reducing maternal-to-child transmission of HIV in Zimbabwe, conducting mass child polio immunisation campaigns in Afghanistan, and Pakistan's lady health worker programme have in put forth the message that even in areas labelled as "fragile states" effective services can still be provided (Wise and Darmstadt 2015b;Hafeez et al. 2011).
Equitable coverage of maternal and newborn health services and quality health infrastructure is a common barrier for LMICs, especially those experiencing a crisis. Akseer et al. (Akseer et al. 2016) highlighted these issues in the protracted conflict setting of Afghanistan whereby promising public health policy at the national level was restricted by subnational disparities. This can be particularly prevalent for crises that are characterised by large population movement across borders, where refugee populations are not granted the same protections and access to services for example the Syrian refugee crisis in Turkey (Ekmekci 2017).

Localisation
The concept of localisation is a growing area of reform in the recent debate for improving the effectiveness of humanitarian aid at the 2016 World Humanitarian Summit. By localising a humanitarian response, international actors recognise, support, and strengthen leadership by local and national leaders in order to better address the affected populations, increase local accountability, and promote sustainable outcomes, such as increased training and capacity of local humanitarian responders (OECD 2017). Experiences of the baby tent programmes in Haiti (Ayota et al. 2013) andNepal (DeYoung et al. 2018) highlighted the need for a localised response to strengthen local capacity and integration of programmes within existing health systems to ensure the continuation of programmes once international humanitarian actors leave.