Human Resources for Health in Conflict Affected Settings: A Scoping Review of Primary Peer Reviewed Publications 2016–2022

Background: Conflict has devastating effects on health systems, especially on healthcare workers (HCWs) working in under-resourced and hostile environments. However, little evidence is available on how policy-makers, often together with development partners, can optimize the organization of the health workforce and support HCWs to deliver accessible and trustworthy health services in conflict-affected settings (CAS). Methods: A scoping review was conducted to review recent evidence (2016-2022) on human resources for health (HRH) in CAS, and critically discuss HRH challenges in these settings. Thirty-six studies were included in the review and results were presented using an adapted version of the health labour market (HLM) framework. Results: Evidence from CAS highlights that conflict causes specific constraints in both the education sector and in the HLM, and deepens any existing disconnect between those sectors. Parallel and inadequate education and performance management systems, attacks on health facilities, and increased workload and stress, amongst other factors, affect HCW motivation, performance, distribution, and attrition. Short-term, narrowly focused policy-making undermines the long-term sustainability and resilience of the health workforce in CAS, and also contributes to the limited and narrow available research base. Conclusion: While HRH and workforce issues in CAS include those found in many other low- and middle-income countries (LMICs), an additional set of challenges for HCWs, governance dynamics and institutional constraints in CAS ‘multiply’ negative effects on the health workforce. HRH policies, programmes and interventions must be aligned with the political and broader societal context, including the stage, severity and other dynamics of conflict. During conflict, it is important to try to monitor in- and outflow of HCWs and provide HCWs the support they need at local level or through remote measures. The post-conflict situation may present opportunities for improvement in HRH, but a clear understanding of political economy dynamics is required to better act on any such a window of opportunity.


Background
Globally, the number of conflict-affected settings (CAS) is rising.In 2017, the number of people living in proximity of conflict (defined as within 60 km of at least 25 conflict-related deaths) was 220 million: double compared to the number a decade ago.The Word Bank estimates that by 2030, twothirds of the global extreme poor will be living in fragile and conflict-affected situations. 1 Conflict has devastating effects on health systems.It results in destruction of health infrastructure, loss of healthcare workers (HCWs), and weakening of health governance at all levels.At the same time, demand for health services increases. 2,3This poses a serious threat to achieving universal health coverage.
The health workforce is the engine of any health system, as all activities and programmes have to be adopted or adapted through them. 4Insecurity and direct attacks lead HCWs to leave or avoid working in CAS.Consequently, small numbers of health workers are left in CAS, working in difficult conditions with often very little support. 5,6The COVID-19 pandemic further aggravated this situation for many HCW.
Responding to crises in human resources of health (HRH), including in CAS, requires a systems approach. 7Increasing the availability of HCWs is often difficult in low-income countries and particularly in CAS.Maintaining or enhancing HCWs' competence, responsiveness, and productivity is often also constrained, for example because of disconnection from social and professional support systems, limited supplies and equipment, increased workload and stress. 8espite the increasing availability of tools and guidelines on HRH, only a few resources are available for how policy-makers can optimize the organization of the health workforce and support HCWs to deliver accessible and trustworthy health services in CAS. 5,9Health systems research (including on First, studies from the period 2016-2019 were selected from the systematic review of Bou-Karroum et al. 5 This systematic review comprehensively mapped the evidence base on HCWs in conflict and post-conflict settings, yet did not itself thematically analyse the content of these papers.Out of a total of 304 studies, 20 studies were included for further assessment.Second, PubMed and Google Scholar searches were conducted for papers from the period of 2016 -August 2022.These searches combined terms (in titles and abstracts) on intervention characteristics with the relevant countries (Table 1 and Supplementary file 1).Assessment of study titles (and where unclear, abstracts) resulted in 85 studies being included for further assessment.Third, 7 studies were suggested, and 3 included, through expert consultation and reviewer suggestion.Bibliographies of included studies were checked during extraction to identify additional sources.
One hundred eight studies were divided among three researchers for abstract reading.Studies that included a focus on all three of the publication characteristics (Table 1) were included.In case of doubt, the full study was read.A second reading was done by another researcher if the article was to be included.This led to 36 studies being included in this scoping review (Figure 1).
Included studies were read, extracted and charted using an Excel data extraction form.We used the HLM framework as the basis for the thematic analysis of papers and presentation of results.The choice for this framework, as well as inclusion of two specific adaptations to this model to better reflect the situation in conflict affected settings are described in Box 1 and Figure 2.
Four researchers extracted relevant excerpts from the results sections of the included studies and sorted these according to key elements of the adapted HLM framework (Box 2).Quality assessment was not part of this scoping review.

Collating, Summarising, and Reporting Results
Once extracted, discussions between the researchers took place on the findings for each of these main result areas of the adapted framework (see Box 2), which were summarized in narratives per results area.A first draft was discussed with key experts in HRH to further refine the focus of the results and strengthen the discussion.

Results
Overview of Included Studies Details of the 36 included studies are shown in Table 2.The most reported contexts were Syria (n = 11), followed by the Democratic Republic of Congo (DRC) (n = 6).Most studies were qualitative (n = 29).They focused on a variety of HCWs.The findings of these studies are summarised in Table 3.

Education Sector
In CAS, production or supply of HCWs coming from education and training is often disrupted.Disruptions or inefficiencies in the education pathway have adverse effects on the number, quality and recognition of (future) HCWs in CAS.
Conflict can bring disruptions to production of HCWs as teaching staff flee, educational institutes seize operations or even become direct focus of attacks.As a response to disruptions, new institutes (often private and/or non-aligned with the government) were founded in several contexts including Libya, Syria, and DRC. A few studies reported about the accreditation of (medical) education, which often lacked quality 17,23,30 or was different in parts on the country controlled by non-state actors. 15,24Development partners partly and temporarily filled gaps regarding training of HCWs, but these initiatives were also not harmonized in terms of focus, content and quality of training. 24ike qualified health workers, students and their teachers face serious challenges in CAS.Barnett-Vanes et al reported that students in Iraq were concerned that conflict caused them to have a low level of clinical competence due to perceived low quality of education, mental exhaustion, and fears for personal safety.The same study indicated that the majority First developed by Sousa et al, the HLM framework takes a comprehensive view of the education sector and HLM dynamics (such as inflows and outflows of HCWs), including market failures, and links these to potential policy interventions. 13An important reason for us to choose this framework is its particular strength in visualizing how market disruptions within these sectors can compound issues in other parts of the framework affect workforce outcomes for years to come.This is especially relevant in CAS, as conflict causes specific disruptions across markets that further add to many difficulties already experienced in more stable LMICs.
We have chosen to further develop two elements of the HLM framework to better portray the devastating effects conflict has on the health workforce: Emphasis on Dynamics of Conflict Current versions of the HLM framework do consider 'economy, population, and broader societal drivers' as influencing the education sector and labour market dynamics.However, the dynamics of conflict in CAS can have such far-reaching consequences for individual HCWs and the workforce that we added this specification in the framework.In our review, it becomes clear that conflict causes significant challenges for individual HCWs (including insecurity, displacement, and mental health) as well as shocks to institutions and functioning of health systems.

Addition of Focus on Health Workforce Governance
While the HLM framework clearly addresses various entry points for policies, the authors felt that especially in CAS the underlying capacity and willingness of governments to fulfil key governance functions related to the health workforce cannot be taken for granted.As the review shows, in certain context the main coordinating function may fall to a non-state actor like ISIS in Syria and Iraq 14 or ethnic health organizations in Myanmar. 15Moreover, in certain settings where governments are unable or unwilling to reach, development partners and/or civil society actors can take a bigger role. 16To highlight that an ideal stewardship role from a government cannot be presumed in CAS, we have added an additional box to highlight governance dynamics.
Abbreviations: HRH, human resources for health; CAS, conflictaffected settings; HCWs, healthcare workers; ISIS, Islamic State of Iraq and Syria; HLM, health labour market; LMICs, low-and middle-income countries.
Box 1.The Health Labour Market Framework

Education Sector
The organisation, accessibility, and quality of pre-service education that affect production or supply of HCWs.

Inflows and Outflows
The mobility of HCWs to enter and exit the labour market, and the ability of employing organisations to develop and implement policies that can shape and direct that mobility.

Distribution
The (mal)distribution of the health workforce within a country, taking into account key factors such as urban/rural, level, cadre composition, age, and gender.

Employment and Remuneration
HCWs' varying employment conditions, both official and de facto, including contractual conditions and remuneration (salary, stipends, and other).

Performance
Various factors mediating HCWs' (in)ability to perform, including challenges pertaining to in-service training, supervision, performance management and task divisions.

Contextual Factors Affecting HCWs in CAS
Contextual challenges particular to CAS that impact the capacity and willingness of HCWs to operate as part of the workforce, including security-related issues, economic pressures and IDP movements.Changing dynamics around gender, ethnicity, race, class, ideology and other group dynamics can impact HCWs and patients.

Healthcare Workforce Governance
The processes that structure the roles and responsibilities of health systems stakeholders which influence if and how policies are enacted to achieve HCW outcomes (including administration of the workforce, regulation and accreditation, and policies to set and maintain strategic direction of the HCW).These processes are shaped by the interactions and relationships of a network of stakeholders with varying interests and power to influence governance outcomes.
Box 2. Key Elements of the Adapted Health Labour Market Framework Used to Structure This Review of medical students surveyed intended to leave the country after graduation. 23In Syria, some medical students also had to transfer to other universities in Syria, because of the risk of arrest by the Government of Syria if they were thought to oppose the government or to have studied at 'opposition institutions.' 24ouad et al reported that the content of existing curricula and training was inadequate for preparing HCWs to deliver quality of care in times of conflict. 27The medical training in Syria did not include specialisation in trauma management, intensive care, or emergency medicine, which resulted in HCWs needing to acquire these knowledge and skills on   48 Afghanistan To discuss the motivations of Afghani women to become CHWs, their status in the community and within the health system, the threatening situations under which they operate, and the challenges they face as working women in a deeply patriarchal society within a conflict zone Qualitative; case study approach CHWs Performance and motivation; contextual factors Tang and Zhao (2019) 15

Myanmar
To audit the health systems and their performance of ethnic health organizations in selected ethnic-controlled, government-controlled or mixed-controlled areas and identify key challenges faced by the ethnic health systems in achieving universal health coverage  the job. 27,30The quality of education was compromised by inadequate infrastructure, issues with recruitment and retention of qualified professors, and limited guidance and mentorship provided to interns.Al-Areibi reported that the operation of educational facilities in Libya was highly constrained by, amongst other factors, a lack of physical space, insufficient library resources and inadequate laboratory equipment. 17Barnett-Vanes et al described that in Iraq staff recruitment and retention became a serious issue, resulting in missed classes, study delay and perceived lower quality. 23ooter et al reported that in high intensity conflict settings like in Syria, interns and other HCWs in training often had to take on more responsibilities than could normally be expected, whilst at the same time receiving less guidance, supervision and mentorship.This had implications for standards of care and caused significant psychological pressure and increased physical workload. 29Some adaptations to curricula and teaching methods were reported, such as the inclusion of focus on gender based violence in the midwifery curricula in the DRC 39 and distance learning being provided by certain institutions in Syria. 24bour Market Dynamics

Inflows and Outflows
Studies describe specific several conflict related dynamics that impact the mobility of HCWs to enter and exit the labour market, and the ability of employing organisations to develop and implement policies that can shape and direct that mobility.Disconnects and imbalances between in-and outflows of HCWs in the labour market lead to an inadequate and/or unstable pool of HCWs.The review found that most studies indicated limited

Inflows and outflows
• Limited recruitment and deployment of HCWs was caused by insufficient numbers of graduates or HCWs to meet demand, and limited motivation of potential or current HCWs to work in areas with high insecurity and workload • When it proved not possible to recruit specific cadres or types of HCWs, a common policy response was recruitment and deployment of CHWs • Conflict often led to the (forced) outflow of HCWs, both in terms of migration to more secure areas within or outside the country and leaving the profession all together • Besides insecurity, attrition was also caused by poor living and working conditions and in some cases, tension between locals, regional workers and expatriates • High workload and mental health issues among HCWs caused attrition Distribution • Maldistribution of HCWs was associated with attractiveness of working areas (urban versus rural, stable versus instable/ in conflict)

Employment and remuneration
• HCWs needed and/or requested more remuneration and other benefits than they were offered • There was some evidence about unequal payment of HCWs with similar functions, leading to demotivation and reduced performance Performance • In-service training, supervision and continuous professional development opportunities were disrupted, leading to inequity in access to these job motivators among HCWs and, potentially, patient safety concerns • HCWs often took on tasks and roles that they were not trained for, leading to risks for quality of care and patient and HCW safety

Contextual factors affecting HCWs in CAS
• Security challenges required adjustments in the organization of and trust in healthcare, posing a challenging environment for HCWs • Female health workers are more vulnerable in conflict areas • Mental health impacts on HCWs can be a significant issue

HRH governance
• There were significant gaps and fragmentation in leadership, HRH governance and regulation at national and regional levels • Conflict can intensify corruption, which can demotivate HCWs and compromise quality of care • Development partners' short timelines of support and specific focus on certain activities or cadres, which might not be in line with government's priorities, can disrupt HCWF composition over the medium to long term • HRH policy development capacity was low, and implementation (and financing) of existing policies was constrained • Non-state conflict actors, such as insurgent groups or rebel movements, are an actor group unique to CAS and bring with them serious ramifications for the health workforce • A few studies reporting on Syria showed that professional associations or expatriate groups took on governance functions in non-government-controlled areas Abbreviations: HRH, human resources for health; HCWs, healthcare workers; CHWs, community health workers; CAS, conflict-affected settings; HLM, health labour market; HCWF, healthcare workforce.
recruitment and deployment of HCWs was caused by insufficient numbers of graduates or HCWs to meet demand.Other identified factors include the limited motivation of potential or current HCWs to work in areas with high insecurity and high workload.Lar et al described that heightened intercommunal tensions in Nigeria made communities more suspicious of community health workers (CHWs) from outside the main ethno-religious group, which narrowed the potential scope of people to be deployed. 38hen it proved not possible to recruit specific cadres or types of HCWs, a common policy response was recruitment and deployment of CHWs.CHWs programmes can be effective when CHWs receive appropriate training, supervision and other support.However CHWs, like in many low-and middle-income countries (LMICs), were found to have poor job security and be under-remunerated. 19,48onflict often led to the (forced) outflow of HCWs, both in terms of migration or internal displacement to more secure areas within or outside the country and leaving the profession or health sector employment all together.Fear for being targeted professionally as well as general decreasing security for health workers and their families is reported on in literature and is a key driver for both of these outflows. 18,35,49tudies from Syria reported that many health workers either fled or voluntarily emigrated, and that whilst this attrition affected the whole of Syria, the situation was very different between government-controlled areas and nongovernmentcontrolled areas. 30Other reported factors causing attrition were economic meltdown caused by conflict 30 and better opportunities abroad. 44,49A key issue to understand the volume of outflows is inadequate data monitoring HCW outflow. 22eliance on external humanitarian HCWs was noted in some studies to fill gaps.Tensions between locals, regional workers and expatriates lead to HCWs leaving In South Sudan, HCWs reported unfavourable government policies towards foreign workers and short-term employment contracts due to stringent donor funding requirements. 25

Distribution
The review indicated that maldistribution of HCWs was associated with attractiveness of working areas (urban versus rural, stable versus unstable/in conflict).In Afghanistan, reliance on CHWs in remote and more conflict prone areas meant they were effectively the main health workers accessible to certain communities. 48Some studies described differences in distribution in terms of the gender of HCWs.A study in DRC found that less than one third of nurses in rural health districts were female, while in the urban health district they made up 61% of the nurses. 22This can be problematic, particularly in settings where certain cadres are underrepresented by one of the genders or where communities have a preference to receive care by one gender.

Employment and Remuneration
The studies that reported about pay or incentives in CAS showed that HCWs often received little and irregular income, and that this was worse in rural and unstable areas than in urban and more stable areas. 21,22Risk allowances were officially in place in some settings like the DRC, but often remained unpaid. 21,22Health workers reported a need to supplement income through user fees in DRC 21 or through over prescription of certain drugs in Myanmar. 36A few studies reported on issues related to staff within the same institution being employed on different contracts, terms and conditions. 21,45One study in Yemen described that HCWs actively shifted between contracts to find better pay -with one study specifically noting the distorting role of humanitarian organisations, who come in in offering higher pay.This leads to tensions between different workers on different types of contracts. 49One study focusing on experiences of frontline workers in Burkina Faso highlighted that it was generally considered an advantage to be employed in the public sector compared to the private sector, because it was perceived to be a more secure position and it included a pension system. 47n Syria, at the beginning of the conflict some doctors would remain officially employed and paid by the government, but volunteer in underground hospitals. 34 study among local aid workers in South Sudan reported that they may be influenced to continue working in the humanitarian and development sectors, if there is increasing availability of competitive salaries and benefits such as medical insurance and paid time off, training and education opportunities, career progression, performance feedback, effective communication, and staff empowerment opportunities. 25

Performance
This review found significant issues with performance of HCWs in CAS, with issues around in-service training, supervision, performance management and task divisions worsening labour market inefficiencies and negatively impacting retention and motivation.
We found that in-service training, supervision and continuous professional development opportunities were disrupted in conflict settings, leading to inequity in access to these job motivators among HCWs and, potentially, patient safety concerns.In CAR, refresher training was organized only irregularly for CHWs, and major discrepancies were found between districts in relation to supervision practices. 28n a study of surgical teams in DRC, all interviewees highlighted the need for continuous medical education while underscoring inequity in access to training.Some participants highlighted the increased risk of errors due to the lack of 'supervision' of inexperienced trainees in surgical care. 37Nongovernmental organizations (NGOs) sometimes filled the gap by providing in-service trainings, with evidence from South Sudan indicating that this led to variations in content and duration of training. 16everal studies reported increased workload for HCWs remaining in position, often aggravated by the double impact of fewer colleagues to treat more and more complex cases (such as trauma cases after an attack, or increased levels of communicable diseases due to an influx of internally displaced persons [IDPs]). 30,32HCWs often took on tasks and roles that they were not trained for, leading to risks for quality of care and patient and HCW safety.An example was related to junior doctors being hurriedly employed in full service in Syria, and various specialists having to retrain in active conflict areas where HCWs were forced to re-specialise on triage and trauma. 27In Syria, insufficient numbers of HCWs led to an increased focus on roles of non-physician HCWs and skill substitution.Bdaiwi et al reported that specialised dialysis nurses in Syria had been taking the roles of renal specialists: this was done with the support of the single remaining renal physician and a team of expatriate renal physicians providing training and advice. 24Despite all challenges facing them, studies widely report the importance of intrinsic factors such as humanitarianism and medical ethics as a strong motivating factor for HCWs to remain in position. 29,34ntextual Factors Affecting HCWs in CAS This review identified many contextual challenges particular to CAS that impact the capacity and willingness of HCWs to operate as part of the workforce, including security-related issues, economic pressures and IDP movements.Changing dynamics around gender, ethnicity, race, class, ideology and other group dynamics impact HCWs and their patient relationships -with female HCWs notably facing many additional challenges in CAS.
1,43,44,49 In Syria this included persecution for treating protestors, 27,30 systematic attacks and bombings on health facilities in opposition held areas (including strategies such as chemical attacks and 'double-tap' attacks meant to specifically target onrushing aid personnel). 26,48Fouad et al interpreted the many ways in which health and HCWs become active targets of conflict actors as a pattern of 'weaponisation of health.' 30,50 Security challenges required adjustments in the organization of and trust in healthcare, posing a challenging environment for HCWs.In Syria, the threats to HCWs and patients were so severe that they necessitated a re-organisation of healthcare infrastructure and staff.For example, various wards of hospitals were split among different (fortified) positions, and hospitals were placed underground to reduce risks of targeting and minimise potential for all services being interrupted. 29,30,36ield hospitals in homes, schools, basements, mosques, and caves were established to treat casualties before transporting patients to permanent medical facilities. 30Mowafi et al reported on the problem of transferring patients in need of specialty care between hospitals in active war zones in Syria. 43nsecurity and curfews could also force HCWs to change their working hours, with evidence from the DRC indicating shorter opening hours to avoid the need for HCWs to travel home at night. 18In Yemen and Nigeria, offering services in areas other than CHWs' and supervisors' village of residence, became a problem. 38,41In Afghanistan, safety on the roads was a problem for particularly female CHWs. 45In DRC, ethnic grouping on either side of the conflict generated patient distrust of HCWs or between each other, which sometimes compromised HCWs' attitudes towards particular ethnic groups. 37Furthermore, for the predominantly female lay mental health workers in DRC, gender-based violence was a fear, as well as one of the biggest reasons for secondary trauma. 33he enormous stress and impact on mental health of HCWs is described in many studies. 14,29,33,36The stress of providing care was partly caused by the scarcity of HCWs, leading to heavy workload, an inability to "switch off " from work, and a feeling of helplessness because of the number of people they could not assist. 29Moreover, delayed or non-payment of salaries and increasingly expensive cost of living could complicate HCWs home situation. 14,48Psychological assistants (lay mental health workers) in DRC were susceptible to secondary trauma, as they may have been exposed, or were afraid to get exposed, to traumatic events their clients had experienced. 33Footer et al reported on the stress of health workers, after being threatened being to prioritise the government forces, at the expense of the civilian population. 29Provision of care to protestors was criminalized in Syria, and suspected sympathetic HCWs could be interrogated, disappeared or killed. 30,36In ISIS (Islamic State of Iraq and Syria) occupied Mosul (Iraq), the constant presence of morality police in clinics created a climate of fear and stress for HCWs, as even minor infractions of staff could be punished severely. 14althcare Workforce Governance Available evidence suggests significant gaps and fragmentation in leadership and coordination, as well as institutional fragility at national and sub-national levels.In Syria, the de-facto autonomy of the Kurdish and rebel held areas spurred a fragmentation of HRH governance functions like regional healthcare workforce planning and a breakdown in coordination between decentralised structures. 24,27In Iraq, HCWs described the lack of flexibility of ministerial departments to adjust expectations for public health programme implementation to the realities in overstretched hospitals, resulting in tensions with patients and delays in critical care. 32Several studies highlighted that inter-sectoral coordination was weak.Of particular note was the often dysfunctional relationship between the Ministry of Health and the Ministry of (higher) Education, 17,45 that have shared responsibility for health sector education and/ or accreditation.In South Sudan, implementation of CHW programmes was fragmented due to lack of coordination and regulatory frameworks. 16In the DRC, Labat and Sharma reported on increased corruption, organized crime (where people do not dare to take action because of their own safety), a lack of accountability of hospital management towards national health authorities, and patronage in human resource management (where people supportive of authorities are at higher levels). 37In Nigeria, patent and proprietary medicine vendors -small owner operated drug retailers -provided services that went far beyond the role that was outlined by the pharmacy council of Nigeria. 19Co-ordination between the national level government and United Nations agencies and other development partners on health, including HRH, can be limited and uneven in CAS, as was the case in Nigeria. 19In South Sudan, a large influx of donor money in the context of inadequate funding resulted in overdependence on donors. 16n the most unstable (humanitarian) settings, the de-facto absence of government capacity in the health sector has created a scenario in which international non-governmental organizations (INGOs), faith-based organisations and nonprofits have taken on increased responsibilities.Risk of parallel services was identified in some papers, as was donor rigidity in approach when more flexibility is required. 19,24,44,46he few studies that reported on HRH policy development showed capacity was low, and implementation (and financing) of existing policies was constrained.In DRC, the Ministry of Public Health developed national strategies to attract and retain different categories of HCWs.However, many of these strategies were either not implemented or only poorly implemented as Provincial Health Offices had limited control over the recruitment and deployment of health workers, which may be partly attributed to the HRH information system being unreliable.This resulted in inequitable distribution of HCWs across the districts. 21,22Policy implementation gaps were also reported in a study from Burkina Faso, 47 where implementation of a national-level policy on free prenatal consultations and provisions for obstetric emergencies was launched before the technical management and communication plan was ready.Not only did this prove problematic for attaining the policy goals, it also put HCWs in a difficult position where they were expected to deliver on the promises of the policy to patients, without the funding and resources to provide services.Amodu et al reported on low budgetary provisions for health in conflict affected areas of Northern Nigeria and highlighted a gap between budget allocation and actual release of the budget in CAS.Funding for IDP-related (health) concerns was lower than expected and insufficient to carry out the numerous activities that implementing agencies were saddled up with.The study identified links to "[broader] issues with trust, equity, and transparency in the government and health system especially where it concerns funding management." 19 In CAS where non-state actors, such as insurgent groups or rebel movements, were in control, studies indicated their interest in also controlling the health workforce.Bdaiwi et al report that in Syria, both opposition-held and ISIS-held areas saw significant focus on attracting/retaining HCWs within a provisional health system to boost their legitimacy. 24However at the same time there are clear examples from ISIS distrusting and using harsh measures to control HCWs. 14,36In Myanmar's ethnic regions ethnic health organisations (EHOs) emerged as a parallel, overlapping health systems model where national and more localised ethnic public models coexisted under a sustained cease-fire.In these regions, EHOs were the de-facto actors organising the formal health system, though other parallel structures also existed: some facilities from the central level ministry, INGOs and private providers.While the EHOs were trusted more than central ministry-led facilities, big concerns about quality issues persisted with regard to training of HCWs, poor financial protection mechanisms, resources and regulation. 15In Syria professional associations or expatriate groups took on governance functions in nongovernment-controlled areas.Bdaiwi et al highlighted the role of the Syrian Board of Medical Specialists, (a body composed of both in-country HCWs and diaspora specialists who had aligned themselves with the Syrian interim government), who actively took on roles beyond those normally taken by professional boards, eg, the coordination and certification of education programmes in the rebel held territories. 24he dominant presence of non-state actors in Syria, such as NGOs/donors/humanitarians/diaspora has led in some instances to a focus on empowering HCWs, enabling them to become potential advocates for peace and using healthcare as a potential avenue for peacebuilding after conflict, as well as emphasising the responsibility of governments to guarantee health worker safety. 29,30

Discussion
The evidence from the different settings and countries presented in this review highlights that dynamics of conflict both worsens 'common' HRH challenges found in LMICs and creates a unique set of additional challenges.The combination of these factors deepens existing disconnects between the education sector and HLM, leading to further weakened HRH in CAS.Using Syria as an example, the country already experienced a number of key HRH issues prior to the civil war, including maldistribution, high turnover of skilled staff and inadequate numbers of nurses. 51Once the civil war erupted, these existing issues were further exacerbated by HCWs fleeing the country/region, being arrested or even killed.Direct additional effects included non-qualified HCWs shifting to trauma care, need for concealment of facilities and fragmentation of the health system into several regionalised systems. 27,34So while some symptoms of disconnects in the education sector and HLM bear close resemblance to those in many other LMICs, we argue that conflict 'multiplies' negative workforce outcomes in CAS through more, and more complex, challenges for HCWs, governance dynamics and institutional constraints.In the next sections we discuss some of the key implications of our findings for HRH policymaking in CAS.
HRH policies, programmes and interventions are embedded in, and have an impact on, the political and broader societal context. 52,53Therefore, HRH policies must take into account the context, including the stage, severity and other dynamics of conflict.Depending on the dynamics of conflict, this highlights the vulnerability of the health workforce as visibly associated to the state or opposition (requiring protection), and the responsibility of governments to provide access to services to all its citizens (avoiding maldistribution of HCWs between favoured and non-favoured regions).Moreover, the effects of health workforce reforms on larger reconciliation and mid-to-long term reconstruction efforts needs to be considered.As little discussion has taken place on what HRH policy options work 'best' in CAS in general, or in specific CAS, feasibility assessments and a policy dialogue on how to best spend available resources are necessary first steps.
Long-term HRH planning efforts in CAS are constrained by a lack of (financial and human) resources, instable institutions with limited capacity (and sometimes willingness) to implement, relatively strong but unpredictable donor influence, data limitations and by the uncertainties and disconnections between (systems in) geographical areas created by ongoing conflict.While not explicitly discussed in the studies included in this review, health is often not a priority of governments in CAS 54 and countries' constrained economic situations and increased population movements make long-term planning difficult.Filling immediate service gaps as a short-term approach to health systems relief appears to be a common trait of HRH policies and programmes in CAS.While this makes sense given the contextual challenges and urgency of the situation, it results in a lack of focus on sustaining quality of care and building and sustaining community trust.Short-term or too narrowly focused policymaking likely undermines the long-term sustainability and resilience of the health workforce, because HRH policy decisions (especially those focused on training and deploying new cadres, changing skills mix and distribution) have longterm impact and are not easily corrected.
Many conflict-affected countries are formally ruled by contested, repressive or mistrusted governments.This in itself has major negative implications for governance, institutions, and the effective development and implementation of HRH policies and developments.In addition, it is not a simple "before and after" dichotomy.Many of these countries had governance issues before the conflict erupted (in some cases being causes of conflict).In these contexts, state withdrawal from a sector or region may actually open space and potential for innovation and improvement, for instance in terms of job descriptions, training programmes and task-shifting, where these had become ossified.One emerging issue from the literature is the potential to increase 'value for money' of HRH investments through strengthening financing and decision-making authority at sub-national levels to facilitate local training, recruitment, and support to HCWs. 55artly as a result of less central-level control this review found parallel systems in the education sector as well as in deployment and management of the health workforce in CAS -for instance, the lack of central planning and proliferation of commercial training institutions is likely to contribute to imbalances in the types of cadres trained and affect health workforce composition for many years.The policy implications for how governments should deal with the existence of a health workforce controlled by nonstate actors are not straightforward.On the one hand, tacit support (eg, through continuation of certification or access to training) may strengthen the legitimacy of non-state actors through increased capacity to provide services to the population under their control.However on the other hand, active obstruction of HCWs undermines the obligation of the government to protect the right to health of all of its citizens.In the post-conflict period, attention should be paid to harmonizing the regulatory divergence or duplication and promoting integration those HCWs who did not have the 'right' qualifications or allegiance. 56A failure to adequately approach this may have implications for community trust in, and access to, services.
Labour markets are dynamic in all countries and contexts, with changing patterns of HCW mobility being an everpresent factor that must be accounted for.However, in CAS, labour markets are likely to be more unstable and fractured.The evidence on HCW outflow within and out of CAS highlights the need for improved monitoring to understand variations in HCW mobility patterns between areas regarded as relatively more or less safe.In addition to geographic mobility, there can be sectoral mobility, with HCWs moving to NGO employment if this is perceived to be more attractive. 57A secondary issue related to outmigration is the potential to harness capacity support from HCW diaspora working in other countries. 58The overall understanding of the issue of internal and international flows of HRH in CAS is hampered by a lack of accurate, complete and up-to-date data on mobility patterns.This is not unique to CAS, but there is a risk of assumption, rather than analysis, directing policy focus.Individual country contexts, in terms of ease of travel and recognition of qualifications will be amongst the determining factors.In addition, the broader context of migration, its size and complexity will impact on health system and HRH demands.Somalia, for example, has been identified as a "country of origin, destination, transit and return for a large number of people moving across the Horn of Africa region and beyond, " with a 2016 report suggesting that remittances from the diaspora accounted for 35 per cent of gross domestic product in 2012, and were among the highest in the world. 59he evidence emphasises the need to address retention challenges with a co-ordinated policy response, using a bundle of different policy interventions.A systematic assessment is required to pinpoint which mix of interventions is likely to be most effective at supporting staff retention.The recently updated World Health Organization (WHO) Retention guidelines outline that community support in assuring security, housing and other livelihoods can improve HCWF retention. 60Though little comprehensive evidence is available on this in CAS, the role of community protection to support HCWs to practice safely and remain in CAS seems important to focus on in policy development.Moreover, trust of communities in HCWs is an especially important focus for ensuring access to care 61 -despite severe challenges, civil society driven attempts to rebuild structures in places like Syria can give hope. 29he evidence highlights that the combination of underresourcing, widespread societal trauma, and the (fear for) security threats facing facilities exacts a heavy toll on HCWs, highlighting the need for investments in mental health support services. 62In addition, the expansion of active monitoring of security incidents involving HCWs and patients' needs to be supported to ensure perpetrators of violence are held to account. 63Particular attention must be paid to HCWs from different genders, ethnicities, ages and other intersecting characteristics that may increase vulnerability to violence. 64urther research on this is needed.
6][67] However, this review shows that there is little documented recent evidence on how technology could support HCWs in CAS, which could partly be because of problems with infrastructure and connectivity.
Limitations of available workforce data constrain monitoring, planning and policy-making.Whilst this is not unique to CAS, the disruption and fragmentation of existing HRH information systems in CAS should be addressed.Where feasible, labour market analysis, underpinned by surveys of HCWs should be undertaken. 68The provision of digital HRH information systems is often part of donor-led post conflict re-building, and there is a need to ensure that the choice of systems is driven by user needs and policy priorities, and that local capacity to sustain the system is considered, with an emphasis on local "user" involvement during development, and on training of local operatives.
Finally, investments in the health workforce will not be effective if not accompanied by investments in other elements of the health system.Effective governance is required to effectively implement the best human resources for health interventions and to co-ordinate across the other health system elements.The literature on health systems recovery after conflict often highlights that the immediate post-conflict period provides a unique opportunity for the government and development partners to increase efforts to rebuild health systems within the context of larger peace building efforts.The length, nature and challenges of a specific conflict -in addition to the specifics of the process of conflict resolution -all impact the likelihood and timing of such a window of opportunity. 54The implication for policy and strategy development processes is that this postconflict opportunity cannot simply be presumed, and a clear understanding of political economy dynamics is required to better identify -and increase the chances of acting ona window of opportunity. 69This is also an area that needs further research.

Reflection on Available Evidence
A key finding of the review is the relative lack broad based and policy relevant scientific evidence.In our sample studies, the HLM elements most reported on were related to individual performance or HCWs experiences, with governance aspects being least reported on.However, research gaps exist related to all elements of the HLM in CAS.The most important research gap is an absence of evidence taking a 'long-term and broadbased view' on the HLM dynamics in CAS.Most studies take a narrower entry point, and have a shorter horizon for their examination, such as a specific aspect of retention of one cadre in a localised area or region.There is a dearth of studies that look across the range of financial incentives available.Most studies take a narrow and incomplete focus, although a comprehensive approach is essential to investigate both the causes (at individual and system level) of the fragmentation and variability in income levels and the consequences of this fragmentation.In addition, the broader HLM dynamics and the connections between different labour market characteristics are usually not considered.Political economy analysis and related approaches could help to build a better understanding of power dynamics between stakeholders influencing these dynamics -and contextualise current market failures.Furthermore, studies that focus on HCWs in CAS usually focus on those workers who have stayed, not those who have left.The evidence available may therefore not fully represent considerations relevant for all HCWs, which in turn has implications for developing effective HRH policies.Moreover, there is a scarcity of evidence adequately considering gender and equity considerations in the health workforce in general, 64 but this is even more so the case for CAS.In addition, this review did not yield evidence on the impact of COVID-19, but the implications of the pandemic on HRH in CAS are important to consider.Some initial evidence describes enormous destabilising effects of the pandemic on health services in Yemen, where a perfect storm of intensifying conflict, a cholera epidemic linked to lack of services and COVID-19 threaten systems collapse. 70Whether COVID-19 becomes a new dimension of fragility or a test of the resilience of a health workforce already under pressure is something which needs further exploration.Future research could also consider including (quality assessed) grey literature.

Limitations
Our scoping review only considered English language, primary research peer reviewed publications -our goal was to specifically map scientific evidence but future reviews could consider further inclusion and quality assessment of grey literature.There are a wide range of CAS contexts, each presenting unique dynamics of conflict and their interaction with the socio-cultural context and labour market.This necessarily complicates relevance of findings across settings.

Conclusion
Against a backdrop of growing conflict across the world, this scoping review has identified progress in generating evidence on HRH aspects of health and care systems in CAS.It has also identified continued gaps in knowledge, which limit the scope to identify and implement evidence-based policies.Whilst the HRH challenges may resemble those in many other LMICs, we argue that the unique set of societal drivers of conflict, governance dynamics and institutional constraints in CAS 'multiply' negative affects to the health workforce.Moreover, active conflict brings a set of additional HRH challenges, including the targeting of HCWs by combatants and effects of widespread societal trauma on the mental health of HCWs.HRH policies, programmes and interventions must be aligned with the political and broader societal context if they are to be successful.The post-conflict situation may present opportunities for improvement in HRH, but a clear understanding of political economy dynamics is required to better act on any such a window of opportunity.Specialist) and Aarushi Bhatnagar (Economist).Moreover, the team would like to thank Jo Raven and Tim Martineau from Liverpool School of Tropical Medicine (LSTM) for their critical reflections.

Figure 2 .
Figure 2. Health Labour Market Framework, Adapted to Emphasise Dynamics of Conflict and Health Workforce Governance.Abbreviations: CAS, conflict-affected settings; NSAs, non-state actors.

Table 1 .
Publication Characteristics and Criteria Afghanistan, Libya, Somalia, Syrian Arab Republic[Syria], Burkina Fa-so, Cameroon, CAR, Chad, DRC, Iraq, Mali, Mozambique, Myanmar, Niger, Nigeria, South Sudan, Yemen Dimensions of conflict Humanitarian, crisis, emergency, fragile and conflict affected [states/settings/situations/countries], post-conflict, conflict Exclusion criteria Lack of specific reporting on dimensions of conflict when discussing HRH issues Abbreviations: HRH, human resources for health; DRC, Democratic Republic of Congo; CAR, Central African Republic.

Author, Years and Citation Country Study Objective Study Methods Types of HCWs Main Reported Elements of the Adapted HLM Framework
SyriaTo explore the complex challenges health workers face in providing care when healthcare services and personnel are themselves subjected to violence and forced to operate in the midst of ongoing human rights violations and war crimes Qualitative: interviews with health workers Various cadres Labour market dynamics: inflows and outflows, performance and motivation; contextual factors; HRH governance Fouad et al (2017) 30 Syria To focus on four analytical themes: attacks on health-care facilities and targeting of health workers as part of a broader pattern of systematic violations of international humanitarian law, the attrition of health workers, the challenges facing health workers in different areas, and the evolving roles of health workers Mixed methods: scoping review, expert consultations, testimonials of health workers Various cadres Education sector, labour market dynamics: inflows and outflows, performance and motivation; contextual factors; HRH governance Hamid et al (2020) 31 Syria To explore the impact of the provision of care of forcibly displaced Syrian MHPs to Syrian clients in the community, given shared experiences and backgrounds with clients Qualitative: in-depth interviews with forcibly displaced Syrian MHPs across two cities in Turkey, Istanbul and Gaziantep, the latter being a city 97 km north of Aleppo, Syria

Table 2 .
Continued NigeriaTo explore the perceptions, experience, and performance of CHWs in the context of task sharing within communities in Bassa, Plateau State in Northcentral Nigeria and potentially to illuminate workforce dynamics in similar situations where regions experience conflict Qualitative; 18 interviews with CHVs and facility heads in rural communities

Table 2 .
Continued The aim of the study is to examine the specific role of the Global Fund in strengthening HRH in the EMR.(1)What are the levels and composition of Global Fund investments in HRH in EMR countries?(2)Whattypes of HRH activities have been supported by these investments?(3)In what ways have these investments contributed to health system strengthening in some of these countries?
HRH, human resources for health; HCWs, healthcare workers; HLM, health labour market; DRC, Democratic Republic of Congo; CHW, community health worker; UN, United Nations; NCQAA, National Center for Quality Assurance and Accreditation; CAR, Central African Republic; MHPs, mental health professionals; ICCM, integrated community case management; NGO, Non-Governmental Organization; EMR, Eastern Mediterranean Region; ISIS, Islamic State of Iraq and Syria; RMNCAH+N, reproductive, maternal, new-born, child and adolescent health and nutrition; CHVs, community health volunteers; MISP, Minimum Initial Service Package for sexual and reproductive health in crisis settings; HPF, Health Pooled Fund; IDIs, in depth interviews; CAS, conflict-affected settings.

Table 3 .
Summary of Main Review Findings Conflict resulted in a higher demand for HCWs and sometimes in increased numbers of schools and establishment of new universities, which often provided sub-standard training • A lack of, outdated or parallel accreditation mechanisms led to divergence in quality of services delivered by HCWs and non-recognition of trained HCWs • The content of existing curricula or training could, in certain cases, be inadequate for preparing HCWs to deliver quality of care in times of conflict • The quality of education was compromised by inadequate infrastructure and limited guidance and mentorship of interns