Health governance approaches in NWS

The consultations with experts revealed three distinct approaches to health governance in NWS. In Idlib governorate and western Aleppo, the governance structure follows a bottom-up approach, encompassing 20 sub-districts (Fig. 1, green). In northern Aleppo, two primary approaches were identified: a top-down approach in 5 sub-districts (Fig. 1, blue) and a hybrid approach in 10 sub-districts (Fig. 1, orange). However, the hybrid approach is more aligned with the top-down model, operating under the strategic plan of the Turkish health authorities, as noted by participants in the consultations.

Fig. 1

Different types of health governance approaches in northwest Syria

Quantitative resultsAverage scale of HSLI, sub-indices and indicators across the governance models

Table 4 illustrates that HSLI ranks highest in the bottom-up governance model, followed by the hybrid model, and is lowest in the top-down model. These differences across governance models are statistically significant (Appendix 2). This trend is similarly observed in the four sub-indices. Statistically significant differences exist across all governance models except for the difference between hybrid and top-down for the Act of Consent sub-index and between bottom-up and hybrid for the View of Performance sub-index (Appendix 2). The View of Performance sub-index has the highest average score across all three governance models, while the View of Justification sub-index has the lowest average score. The average score of substitutive indicators is better than that of constitutive indicators in all governance models.

Table 4 Average scalea of HSLI, its sub-indicesb and indicatorsc across the governance models

Moreover, the analysis reveals no statistically significant differences observed across the three governance models concerning four indicators (Appendix 2). These indicators include ‘compliance’ of the Act of Consent sub-index, ‘reliability’ of the View of Performance sub-index, and ‘approval’ and ‘participation’ within the Legality sub-index. The remaining 14 indicators within the bottom-up governance model exhibit statistically higher values compared to those within the top-down governance model. Conversely, the differences between these indicators within the hybrid governance model, in comparison to both the bottom-up and top-down governance models, do not reach statistical significance (Appendix 2).

Percentage distribution of HSLI and sub-indices levels across the governance models

Figure 2 presents the percentage distribution of HSLI and its sub-indices across three governance models, categorized as ‘below average,’ ‘average,’ or ‘above average.’ The HSLI results show that the percentage of ‘above average’ is highest in the bottom-up governance model (63.6%), followed by the hybrid model (34.1%), while the top-down governance model exhibits the lowest percentage (20.4%). These differences between percentages are substantial and statistically significant. The ‘average’ level of HSLI is the lowest across the three governance models; however, the differences in ‘average’ levels among the different governance models are statistically insignificant.

Fig. 2

Percentage distribution of HSLI* level and its sub-indices by governance models**, * HSLI and its sub-indices are continuous variables ranging from 1 to 5, with 1 indicating very poor and 5 signifying very good. Thus, we consider any value below 3 as ‘below average’, while values above 3 are classified as ‘above average’, and a value of 3 is assigned as ‘average’. ** We used a two-sample test of proportions to check for statistical significance between each pair of percentages within the same scale category across the three governance models (Appendix 2)

In terms of HSLI sub-indices, the bottom-up model consistently exhibits statistically significant higher proportions of the ‘above average’ across all sub-indices compared to the other two governance models. Conversely, the top-down governance model consistently shows statistically significant higher proportions of scores in the ‘below average’ across all sub-indices, except for the Performance sub-index. It is worth noting that the Performance sub-index shows relatively low ‘below average’ scores across all governance models, with insignificant differences observed between them.

Percentage distribution of legitimacy indicators across the governance models

Table 5 provides a detailed percentage distribution of the indicators within the four sub-indices across the three governance models. These indicators are categorised as very bad, bad, average, good, very good, don’t want to answer, don’t know. Analysing the table highlights crucial patterns and disparities between governance models, complementing the insights derived from the analysis of the average scale in Table 4. For instance, when examining the View of Justification sub-index, which includes transparency and equity in addition to other indicators, the bottom-up model consistently demonstrates relatively higher proportions of scores in the ‘good’ and ‘v. good’ categories across these indicators. These high scores could suggest that bottom-up governance structures foster greater transparency, equity, impartiality, and understanding in public health services decision-making compared to more centralized governance models.

Table 5 The percentage of people’s perceptions of legitimacy indicators

Similarly, in the Act of Consent sub-index, which includes compliance with guidelines and delegation of authority, the bottom-up model again reveals higher proportions of scores in the ‘good’ and ‘v. good’ categories for these indicators. This could also indicate that the decentralized governance model of public health services may enable more effective compliance and better delegation than hybrid and top-down governance approaches.

In terms of Views of Performance sub-index, the highest percentage of ‘good’ and ‘v. good’ is also seen in the bottom-up governed areas regarding all indicators. However, the top-down model demonstrates relatively better scores in the indicators of this sub-index compared to the top-down indicators of other sub-indices. Within the Legality sub-index, and across all indicators, the bottom-up model consistently demonstrates the highest percentage of indicators classified as ‘good’ and ‘v. good’. In contrast, the top-down model generally displays the lowest percentage of indicators classified as ‘good’ and ‘v. good’.

Qualitative insights and discussion

From the beginning of our study, we assumed that the international response in the aftermath of the earthquake used the existing structures and governance models and did not change or replace them. This assumption was confirmed by consultancies with local experts. They confirmed that although the response to the earthquake enhanced collaboration between different actors, it did not change governance models.

Views of justification

The ‘bottom-up’ model scores are higher across all aspects of the ‘views of justifications’ compared to ‘top-down’ and ‘hybrid’ models. These results suggest that a community-based governance structure, as the ‘bottom-up’ model was described in the expert panel, is more attuned to the local context and specific needs and values.

The lack of transparency in all different models aligns with previous research by Alaref et al., which reported poor transparency in the governance system in NWS. The most significant deficiency was ‘weakness in the internal system or operational model and project design,’ followed by ‘the lack of resources and poor sustainability planning’ and ‘the lack of legitimacy.’ [68] Our participants in the expert panel mentioned more reasons, in areas heavily reliant on aid, medical staff supported by NGOs earn much more than the average person. As a result, financial data is kept private to avoid clashes with local communities. Additionally, sharing health-related information in opposition-controlled areas is risky, as the central government has criminalised health practitioners and systematically targeted health facilities. In regions with poor security, sharing essential information could pose a threat, making institutions vulnerable to armed robbery.

Equity scores below average in all areas. The impact of conflict on health equity is not well-documented, especially in conflict-affected fragile states. Factors such as displacement, gender inequity, and financial barriers can affect health equity [99, 100]. According to participants in the expert panel, the distribution of health services in NWS tends to depend more on institutional systems than on people's needs. For example, since the military campaign by the Syrian and Russian armies started in NWS in 2019 [101], which led to displacing a million people [102] and destroyed more than 60 health facilities [103], more than 200,000 people in Al-Zawya Mountain have remained without facilities. All the attempts by the local health authorities and NGOs to provide health services in this area have failed because of the routine targeting by the Russian and Syrian armies just as they open. Additionally, there is a gap observed in general between people's needs and health and humanitarian aid [45, 104].

Regarding corruption, one of the main challenges in the ‘bottom-up’ area is the weakness of the rule of law and the monopoly of law enforcement by the SSG's courts. Health institutions tend to handle corruption issues internally to avoid dealing with these courts and to comply with red lines drawn by donors regarding interactions with the SSG, according to the expert panel. However, the incorruptibility indicator in the ‘bottom-up’ area is significantly better than that of the ‘top-down’ and ‘hybrid’ areas controlled by the SIG and Turkish authorities. Participants in the expert panel also noted that the system of courts that emerged in northern Aleppo has no authority over Turkish health facilities. Conversely, it is unclear to the public if there is an active internal mechanism to combat corruption in these areas.

Therefore, the ‘bottom-up’ model exhibits acceptable internal regulatory and customary mechanisms to control corruption, with community oversight being more transparent than in other models. However, evidence from a 2012 bribery experiment by Serra comparing top-down and bottom-up accountability indicated that a “combined” accountability system might be highly effective at reducing corruption, even in environments with weak institutions where the likelihood of formal punishment and fines is low [105].

The impartiality score in the ‘bottom-up’ area (4.07) is significant, especially considering the Idlib population of 3 million, 65% of whom are IDPs. More than a million people live in camps. Concerns have been raised about the hostility of local communities and discrimination against IDPs in other sectors. Several studies have investigated the ethical dilemmas faced by medical personnel in overseas humanitarian military operations, including issues of impartiality [106]. However, studies on the impartiality of local medical staff in conflict areas are rare.

According to the expert panel, several factors contribute to this positive result:

1) Entire communities, including their Health Care Workforces (HCWs), were forcibly displaced by the Syrian government from several governorates, such as Daraa, Ghouta, Homs, and Hama, to NWS. These HCWs integrated with the local health sector and became part of the service delivery process. 2) The nature of health system governance, which involves the participation of all health facilities in the elections of the General Assembly and then the Board of Trustees of the IHD, allows for the inclusion of diverse HCW backgrounds in the decision-making process. 3) The majority of the population shares similar political positions regarding the ongoing war in Syria against the Syrian government. And 4) There are no significant ethnic and religious differences, as most people are from the same ethnic and religious background. In contrast, the ‘top-down’ approach, while involving displaced HCWs in providing health services, does not include them in the decision-making process, which is concentrated in the hands of Turkish health officials. Additionally, there are allegations of discrimination against Kurdish people in northern Aleppo by Turkish authorities and the military groups that control the area.

The respect for traditions indicator scores higher in all models compared to other justification values because most HCWs are from the same community or understand its traditions and values.

Acts of consent

The overall score of acts of consent is better in the bottom-up governance compared to top-down and hybrid governance. Compliance with HCWs’ orders and advice is above average, with no significant differences among governance approaches. This result demonstrates trust in HCWs who gain significant practical experience in dealing with ‘war medicine,’ especially trauma cases. This result is in line with previous research by Ekzayez et al., which mentioned good compliance to health authorities in Idlib during the response to COVID-19 due to many reasons, including gained experience from prior health emergencies, local-level coordination, community engagement, local health leadership, and the role of diaspora medical networks—using knowledge networks and eHealth tools to great effect [51]. However, the expert panel stressed that the HCWs still need a higher level of professional training to handle such emergencies.

When the contribution of society, including the medical community, in selecting medical leaders increases, even if this is imperfect, people feel more confident in the ability of these leaders to express their interests and act on it, according to a participant in the expert panel. This is a possible explanation for the relatively high delegation score in the ‘bottom-up’ approach in Idlib compared to the ‘top-down’ approach in northern Aleppo. In the ‘top-down’ area, people do not have the right to delegate or remove authorisation by any known mechanism, so the score was below average. However, delegating Turkish health authorities is part of a big dilemma in which Turkish authorities play a role in political negotiation as a partner and sometimes a representative of the opposition. In political science, the extent of delegation in the case of crises and resource constraints is typically influenced by the need for specialisation and efficiency of decision-making of responding bodies, the level of trust, the desire for centralized control, and the situational context [107,108,109].

Views of performance

When evaluating the views of performance of various health systems in NWS in the aftermath of the earthquake, the overall scores of the different health systems were above average.

In the immediate emergency response phase, particularly in the initial days following the earthquake, the ‘bottom-up’ approach exhibited an advantageous capability to swiftly mobilize and deliver urgent health services, outperforming other health systems in terms of the speed of the health response and the quality of health services. This result is in line with previous research by Alzoubi & Alkhalil et al., which emphasised the significance of civil society and bottom-up networks in responding to compound crises compared to governmental institutions and top-down structures [31]. On the other hand, the ‘hybrid’ health system demonstrated a relatively better capacity for maintaining the sustained provision of health services (availability).

By comparing the strengths of these various approaches during different phases of the earthquake response, it is evident that the ‘bottom-up’ system was able to promptly assess immediate healthcare needs, mobilize resources, and establish direct engagement with local communities, which played a crucial role in the rapid deployment of humanitarian health assistance [31]. However, our data shows that such flexibility in decision-making capacity may fall short of ensuring the durability of health services over the long run. Achieving sustained effectiveness necessitates a more comprehensive needs assessment and coordination among responders, including national and international NGOs, as well as adopting long-term policies and procedures that often require a more robust institutional capacity and harmony at the national and sub-national levels. Such requirements are found better in ‘hybrid’ health systems, where the sustainability of health services depends on more centralised management of resources and assessment of long-term needs at the macro level.

Compared to other systems, the ‘top-down’ approach, found in the health governance structures in Azaz, Afrin and al-Bab, for instance, has notably lower scores concerning responsiveness and quality. According to the experts, the internal patient’s referral in the aftermath of the earthquake was clearly from the ‘top-down’ area to the ‘bottom-up’ area in Idlib, especially for advanced specialised services. This could be attributed to the overreliance of the ‘top-down’ structures on the Turkish health authorities, their main conduit of support, which were institutionally overburdened and overwhelmed by the profound impact of the earthquake on the southern provinces of Turkey, thus impeding the efficacy of the health services provided in the aforementioned regions.

The only indicator in the ‘top-down’ approach overtaking that in the ‘bottom-up’ is the ‘availability’ of health services. According to the expert panel, this is because of the significant hospitals built by Turkey in northern Aleppo. Additionally, transferring patients with complicated diseases, including cancers, for treatment in Turkey is also much easier compared to the ‘bottom-up’ area because the ‘top-down’ area applies the Turkish health system.

Globally, national and international frameworks like the 2005 Hyogo Framework for Action by the United Nations have been established to reduce the impact of natural disasters [110]. While a top-down approach is helpful for governments, a bottom-up approach focusing on individual and community responsibility could be even more effective, potentially saving more lives [111].

Views of legality

Most respondents indicate a level of coordination between health and humanitarian responders that surpasses the average in areas where the ‘bottom-up’ approach to health governance is implemented, followed by the hybrid system. This can also be linked to the responsive coordination role played by the Health Cluster in Gaziantep in the ‘bottom-up’ area compared to the Turkish health authorities’ role in both ‘top-down’ and ‘hybrid areas’ [31].

The consistently lower scores assigned to all health systems in terms of involving the local community in decision-making processes during the initial response phase (participation), coupled with the deficiencies in community-based accountability of responding bodies, indicate persistent inadequacies in good governance capabilities and institutional capacity beyond the immediate provision of health services. Most respondents expressed dissatisfaction with the local health actors’ level of community-based consultations and their transparency in sharing progress and financial reports in an accessible manner. WHO emphasises community participation as a core element in enhancing primary health care, integrated health services and diminishing health disparities [112,113,114]. However, despite the growing interest in participation, the evidence linking participation directly to better health remains weak, which creates barriers to gaining full support from governments, funding agencies and health professionals to enhance this concept [72].

All models are below average in terms of accountability. These results align with previous research by Alaref et al., which mentioned poor accountability of the health governance system in the area [68]. However, the ‘Bottom-up’ system has a slightly higher score than the other two systems due to the direct engagement with local communities and their representatives, albeit restricted, in the decision-making mechanisms. Three types of accountabilities were mentioned by Brinkerhoff in 2004, including financial, performance and political, with three purposes: reducing abuse, assuring compliance with procedures and standards, and improving performance [115]. However, criminal accountability is another significant type in conflict zones due to some parties’ involvement in attacking health facilities and medical personnel [116]. The last type is beyond the scope of this paper.

Notably, the public perception of the health systems’ external connectivity in all studied areas, such as the abilities to coordinate with international bodies such as the WHO, and to liaise with and secure funding from external donors, both international donor agencies and the diaspora, surpassed the average, with the ‘hybrid’ system receiving the highest score, closely followed by the ‘bottom-up’ system, and the ‘top-down’ approach, respectively.

The slightly higher rating of the ‘hybrid’ health system regarding ‘external connectivity’ could be due to the profound impact of the earthquake on particular localities within these areas, such as Genderes and its surroundings, which prompted the Syrian diaspora to organise highly effective fundraising campaigns within the first week of the earthquake. According to our respondents, most of these funds were directed towards addressing the specific needs of these affected areas. Additionally, diaspora organisations played a significant role in bridging the gap between donors and local actors, understanding urgent local needs in the aftermath of the earthquake and the first response.

Overall legitimacy

To assess the overall legitimacy under different governance models, two key factors need to be considered: the Health System Legitimacy Index (HLSI) scores and the percentage distribution of the index across the ‘below average,’ ‘average,’ and ‘above average’ scales in each governance area. The findings highlight the advantage of the ‘bottom-up’ model in conflict zones, where it is perceived as a more legitimate model.

When examining the HLSI scores, the ‘bottom-up,’ ‘hybrid,’ and ‘top-down’ models scored 3.18, 2.92, and 2.69, respectively. Although the ‘bottom-up’ model outperformed the others, all models scored around the average. This indicates that perceptions regarding health system legitimacy were slightly ‘above average’ in areas where the ‘bottom-up’ model was implemented. However, this slight advantage in a highly volatile and unstable region is considered somewhat muddled.

The superiority of substitutive indicators average over constitutive indicators across all governance models illustrates that people perceive public health authorities' performance as better than the procedures they adopt. This result aligns with Alzoubi & Alkhalil’s findings that undocumented, tacit governance developed during the conflict, relying on health personnel's experience and collective memory, created efficient responses to disasters [31].