Introduction
The serious public health impact of human influenza infection in Cambodia is increasingly apparent. Highly pathogenic avian influenza is now considered enzootic among poultry in the country, with a total of 34 human cases of H5N1 infection confirmed as of early July 2013, of which 28 have been fatal [
1,
2]. At 82%, the case fatality rate for H5N1 in Cambodia is second only to Indonesia’s (83%) in the list of countries with sustained cases [
2]. Situated within Southeast Asia, a region recognized as playing a central role in the circulation and evolution of influenza viruses [
3,
4], Cambodia is at risk for the emergence and spread of novel strains with pandemic potential. Strategies to contain and mitigate future influenza pandemics in low- and middle-income countries are critically important, both because many of these countries, like Cambodia, are in regions with very high risk of emerging infections and because health systems in these areas are weaker compared to those in high-income countries [
5‐
7]. Developing countries are likely to be at particular risk in the event of future influenza pandemics, with one study predicting that they will bear the burden of more than 90% of influenza-related deaths globally [
8]. Within developing countries, too, poor populations are disproportionately affected by the health impacts of pandemics, in part due to decreased access to efficiently functioning healthcare systems; this vulnerability was illustrated in the higher rates of infection and greater morbidity and mortality among poor communities during the H1N1 pandemic of 2009 [
9].
The H1N1-2009 pandemic also demonstrated the limitations of health system pandemic response capacities, even in developed countries [
10,
11]. While much emphasis has been placed on strengthening surveillance and outbreak response activities throughout South East Asia and globally, there is increasing recognition of the need to employ resource mobilization and redistribution strategies in pandemic mitigation planning (that is, activities aimed to reduce morbidity and mortality in the event of a pandemic) [
12] and to evaluate critically existing and future investments in heath system resources. This has proven true in Cambodia, where semi-structured interviews conducted with members of the pandemic influenza national coordinating committees, performed in response to the 2009 H1N1 swine flu pandemic, revealed a need for data to inform health system investment allocation for maximum efficiency in pandemic mitigation strategies [
13]; the research described here was performed in response to this identified need. Although resource-constrained and largely dependent on external donors for health sector financing, Cambodia has proven its ability to successfully increase investment in health care in the recent past, with an explicit goal of achieving more efficient and equitable distribution of health care resources. The country has made remarkable achievements in rebuilding its national health system since its systematic destruction under the Khmer Rouge regime in the 1970s, including the introduction of health system reforms in the 1990s focused on delivering primary health care nationwide through a district-based system. Since the development of this system, important advancements have been made in the health status of Cambodians, including improvements in life expectancy and infant mortality [
14]. The country is now divided into 77 Operational Districts (ODs) on the basis of population density (100,000 – 200,000 people per OD), each supporting a network of referral hospitals, health centers and health posts that provide basic health care at the local level [
15]. In order to strengthen pandemic influenza mitigation in Cambodia, a clearer understanding of current healthcare resource distribution at the OD (as opposed to Province) level is required.
This report builds upon the findings of a previous research project, AsiaFluCap [
16]. The AsiaFluCap Project analyzed how health system resources can be most effectively and efficiently deployed across selected Asian countries (Cambodia, Indonesia, Lao PDR, Taiwan, Thailand and Vietnam) in the event of a pandemic. Using a mathematical transmission model to simulate a mild to moderate pandemic influenza scenario, gaps in health care resources key for responding to pandemic influenza were estimated within and compared across countries, with calculation of potential mortality burden associated with those gaps. The results of these analyses, carried out at the Province level in all study countries, revealed high variability in resource capacities both between and within countries, with substantial avoidable mortality attributed to severe shortages in mechanical ventilators [
12]. A follow-up analysis, focusing only on study countries in the Greater Mekong Subregion (Cambodia, Lao PDR, Thailand and Vietnam) mapped and analyzed key health care resource distributions at the Province level in order to identify relative resource shortages, mismatches, clustering, and inequalities in distribution [
17]. Again, a high degree of heterogeneity in resource distribution was noted. In Cambodia, the availability of key resources was found to be particularly low and unequally distributed compared to the other study countries, suggesting that a pandemic influenza here could have a disproportionately high public health impact.
The current analysis constitutes a sub-component of an overarching project, the CamFlu Project, which draws on current health service resource distribution patterns in Cambodia to model scenarios for cost-effective options to improve pandemic response capacity and evaluate the public health impact of scaling up resources for influenza pandemic response. Thus, the present study aims to further explore the challenges involved in efficiently and effectively distributing key health system resources in Cambodia for pandemic influenza mitigation, specifically by refining previous distribution analyses from the Province to the OD level and investigating associations with a socioeconomic indicator, the Predicted Family Poverty Rate (PPR).
Discussion
The results presented here reveal notable heterogeneity in distribution of health resources considered critical for pandemic influenza response across Cambodia. This diversity in distribution is related to a number of variables, including resource type (human resources vs. medical supplies), clinical setting (hospital vs. NHMF), geography (including proximity to the major urban center, Phnom Penh), and poverty level. The results build and expand on previously published data on pandemic health resource distribution in Cambodia [
12,
17] by focusing the analysis at the OD rather than Province level, allowing better appreciation of varying patterns in resource distribution. Additionally, the introduction of a measure of socioeconomic status, the PPR, allows for novel analysis of resource distribution by poverty quintile, an important consideration given the higher vulnerability of poor communities to pandemic influenza.
In the event of a pandemic, inpatient beds can be redistributed to affected areas with insufficient resources relatively quickly, particularly among ODs within a given Province. Indeed, the Cambodian Ministry of Health anticipates the need for redistribution of certain resources, such as isolation beds and personal protective equipment (PPE), from areas of high concentration to areas experiencing a pandemic-related surge in demand. Detailed knowledge of existing resource density by OD is, thus, essential for rapid and efficient redistribution in the event of a pandemic.
In terms of doctors and nurses, there may be benefit to temporarily deploying medical personnel most heavily concentrated in Phnom Penh to remote ODs in the event of a pandemic influenza, assuming that affected individuals are unable to voluntarily travel to the capital to seek care (either due to distance or illness severity). The fact that the distribution of doctor and nurses by OD exhibited markedly different patterns, particularly by PPR, is worth noting and raises the question of whether enhancing clinical capacity of nurses to serve as physician extenders in areas of the country where doctors are scarce might also be beneficial in a resource-constrained setting. Indeed, this may present an opportunity to improve pandemic preparedness and prevention through maximizing population health, as nurses represent core workforce capacity for basic primary care provision and many public health education activities. At the same time, increased investment in health care workers (both doctors and nurses) in relatively undersupplied ODs is prudent. In order to achieve its articulated goals of improving baseline health in three identified priority areas – reducing newborn, child and maternal morbidity and mortality while improving reproductive health; reducing morbidity and mortality attributable to communicable diseases, including HIV/AIDS, tuberculosis, and malaria; and reducing the burden of non-communicable and other diseases – the Ministry of Health has acknowledged the importance of improving health service delivery at the OD level, including through more effective human resource planning and management [
14,
25]. Focusing future government health spending on incentivizing health care worker deployment in underserved areas and providing the necessary durable medical equipment to support such an influx in workers would not only strengthen future pandemic mitigation strategies but would aid the country in achieving the goals of the Health Sector Strategic Plan 2008–2015 (HSP) as well. Indeed, health service delivery and human resources for health are two of five strategic areas identified in the HSP, and the Cambodian Ministry of Health has explicitly linked successful implementation of this plan with more equitable access and health outcomes for all Cambodians [
25].
The availability of oseltamivir reported here must be interpreted within the regional context in which Cambodia is embedded. As a member of ASEAN since 1999, Cambodia received an initial national stockpile of antivirals (as well as PPE) from the Japan-ASEAN Antiviral Stockpile for ASEAN Countries and is eligible to draw on additional stores from the central stockpile storage site in Singapore in the event of a pandemic. Cambodia was the site of a test exercise in early 2007 to determine adequacy of protocols for redistributing antivirals and PPE from the central stockpile, suggesting that such a strategy could efficiently augment Cambodia’s existing supplies of these important resources [
26].
For resources that require the presence of trained medical personnel and reliable electricity to be clinically useful, such as ventilators, it will likely prove most efficient to refer or transport patients to areas where these resources are known to be concentrated. Given the low absolute numbers of ventilators available in Cambodia and the logistical challenges (i.e. requirements for attendant electricity and personnel) involved in mobilizing ventilators either domestically or from other countries in the region, it is likely that the capacity of even the most well-supplied areas would be rapidly overwhelmed during a pandemic by patients seeking care at referral hospitals.
It is important to emphasize that the data presented here reflect only the distribution of selected resources provided through the public health care system in Cambodia and do not account for complementary resources available in the private and non-government organization (NGO) sectors. Given the historical development of the Cambodian health care system over the past few decades, with substantial reliance on donor investment and NGO health programming activities in addition to an intentional strategy of decentralized health system management and public-private partnerships, a comprehensive assessment of resource distribution requires complementary data on resources available in these other sectors. Indeed, an estimated 50-60% of all health treatments are provided by the private sector and 15-20% by practitioners outside the traditional medical sector (i.e. by drug vendors, religious leaders, and traditional healers), compared with approximately 20% through government facilities [
27,
28]. Thus, it may be appropriate, for instance, for the Ministry of Health to focus public health resources in areas where private health care services and NGO-supported services have traditionally been absent. Additionally, this analysis does not include any measures of health resource accessibility or health outcomes, as the data required for such an analysis were beyond the scope of this study. For these reasons, we cannot draw conclusions about the appropriateness or equity of resource distribution from these data; indeed, it may be fully intentional (and clinically reasonable) to concentrate health care workers in areas of high population density and numerous referral hospitals, such as Phnom Penh, and equally logical to maximize the utility of scarce resources such as ventilators by locating them in facilities where they can be reliably maintained and accessed.
Conclusions
There is substantial variation in health care resources important for pandemic influenza response across Cambodia. This diversity is related to a number of factors, including geographic location, resource type, clinical setting, and poverty rates at the OD level. Absolute availability of specific key resources also plays a role, as resources that are scarce nationwide are concentrated in selected areas. The implications of these findings vary by resource type, as the strategies for distribution and improved accessibility must take into account clinical appropriateness as well as resources available from outside the country. Further assessment of the feasibility and appropriateness of resource distribution options requires a more comprehensive understanding of complementary health resources available in the private and NGO sectors. In conjunction with such data, the resource distributions presented here will prove useful for strategic health resource investment around the country, both for future pandemic mitigation and to strengthen the baseline functioning of the Cambodian health care system.
Understanding the existing distribution of selected healthcare resources important for treating severe influenza is a necessary condition for ensuring equity in future pandemic mitigation planning in Cambodia. Health-related equity is a broad concept that encompasses health service financing, health service delivery, and health outcomes. Although there are numerous definitions for health equity, most authors agree that achieving equity in health requires minimizing avoidable inequities (i.e. disparities between groups of people as defined by gender, race or ethnicity, socioeconomic status, or geographic location) in health determinants and outcomes by ensuring equal access to a minimum standard of healthcare [
15,
29‐
31]. The conditions that affect access are numerous, and include availability of effective health services (including infrastructure, staff, medications, equipment) and funding mechanisms [
15]. As there were insufficient available data to link the availability of inpatient beds, doctors, nurses, ventilators and oseltamivir to patient access patterns or health outcomes, we cannot make any claims regarding the equity of the distributions reported here. However, as resource availability is an essential component of health service delivery, we believe this analysis represents an important contribution toward future equity analyses of pandemic mitigation planning in Cambodia. Specifically, it draws attention to differences in selected resource distribution across the country and thus lays the foundation for further study of potential inequities by socioeconomic status and geographic location. Given the disproportionately high impact of pandemic influenza on the health of poor populations, further exploration of vertical equity (i.e. differential access for unequal need) in pandemic mitigation planning in Cambodia is of vital importance.
Acknowledgements
We sincerely thank all collaborators and respondents who assisted the AsiaFluCap and CamFlu Projects in collecting health system resource information, including the AsiaFluCap Consortium. We are grateful to the CamFlu project partners, which include: CelAgrid, Cambodia; Ministry of Health, Cambodia; Hamburg University of Applied Sciences; and London School of Hygiene and Tropical Medicine. Ms. Leah Holzworth provided invaluable assistance with GIS analysis and Mr. Tapley Jordanwood, on behalf of the University Research Co., provided access to data that greatly facilitated the socioeconomic analysis. The project was supported by Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH on behalf of the German Federal Ministry for Economic Cooperation and Development (BMZ). GIZ played no role in project design; data collection, analysis, or interpretation; writing of the manuscript; or decision to submit the manuscript for publication.
Competing interests
SUSK: The author declares no competing interests.
JWR: JWR has worked on an unrelated project funded by Hoffman La Roche, the maker of oseltamivir.
TD: The author declares no competing interests.
IC: The author declares no competing interests.
KB: The author declares no competing interests.
ST: The author declares no competing interests.
RC: RC has held grants unrelated to this research from Hoffman La-Roche, the maker of oseltamivir.
Authors’ contributions
SSK carried out the socioeconomic analysis and drafted the manuscript. JWR participated in study coordination and data analysis and helped draft the manuscript. TD coordinated data analysis and helped draft the manuscript. IC performed the GIS analysis and drafted corresponding sections of the manuscript. KB and ST facilitated data collection and participated in resource characterization. RC conceived, designed and coordinated the study. All authors read and approved the final manuscript.