Balancing science and political economy: Tobacco control and global health

Background: Global tobacco control is a major public health issue, as smoking-related disease burden remains high worldwide. The World Bank and the World Health Organization (WHO) are the driving forces in global tobacco control. However, little research has focused on their development, financing, decision-making, and accountability structures. Methods: We used two strategies to identify the development and structure of global tobacco control initiatives. First, we reviewed the published literature through electronic databases. Second, we conducted grey literature searching. Results: We identified four periods in the Bank’s involvement in global tobacco control, from creation of the evidence base in the 1990s to the implementation of tax reforms. We identified three phases in the WHO’s efforts, from its early recognition of the link between tobacco and health risks in the 1970s to its implementation of the Framework Convention on Tobacco Control. Both organisations are financed by a handful of private philanthropies, and face similar risks for effective tobacco control: reduced accountability and resource mobilisation, poor decision-making authority due to specific donor influence, and difficulty in monitoring and evaluation. Conclusions: Continued attention should be paid not only to the primary health-related outcomes of tobacco use, but also to the decision-making and financing structures to promote tobacco control activities.


Introduction
Tobacco use kills seven million people worldwide annually and tobacco-related disorders cause a substantial burden, accountable for approximately 150 million disability-adjusted life years 1,2 . Since the 1990s, evidence-based measures to combat tobacco use have been identified and implemented 3 . At the global level, two international organisations, the World Bank and the World Health Organization (WHO), have become the most influential agencies in tobacco control. The Bank was the driving force in creating evidence on tobacco economics 4 , and used its longstanding relationships with finance ministries to implement tax reforms in low-and middle-income countries 3 . The WHO, the only normative agency in global health, used that evidence and negotiated toward the enactment of the Framework Convention on Tobacco Control (FCTC). At the same time as tobacco control became a cornerstone of the global health agenda, the past two decades also gave rise to philanthropic agencies, notably the Bill and Melinda Gates Foundation and Bloomberg Philanthropies. The Bank and WHO, given their resource limitations from governmental sources, have accepted this private support in pursuit of the tobacco control agenda.
The shift toward public-private partnerships for global tobacco control raises the risk of "Trojan multilateralism" 5 , in which a handful of actors play significant roles in the decision-making processes of multilateral organisations. Safeguarding the Bank and WHO's multilateral mandate and ability to make equitable decisions requires close attention to their governance 6 . Analysis of global health projects, therefore, should include a broader scope of work, from primary health outcomes (e.g., mortality or the disease burden) to financing, decisionmaking, and monitoring structures 7 . However, researchers have provided limited insight into the development, current decisionmaking, and financing structures of global tobacco control projects by the Bank and WHO 8,9 . The particular focus of previous research has been on creating epidemiological datasets and analysing FCTC negotiations 2,10 .
In this paper, we review published and grey literature on the Bank and WHO's tobacco control policies. We use this literature to comparatively examine the development, current decisionmaking, and financial structures of each institution. We then identify major opportunities and challenges facing these two institutions in tobacco control priority setting and effective resource allocation.

Methods
We used the following strategies to identify the development and structure of tobacco control established by the WHO and the Bank. First, we reviewed the published literature (Supplementary File 1). We primarily focused on the governance and financing structures of each organisation. Therefore, literature that focused on general cost-effective measures and epidemiological studies were excluded. Second, we searched for research and project reports through the websites of the Bank (Projects and Operations database, Development Topics database, Documents and Reports database, digitalised archive holdings, annual trust funds reports, and Finances trust fund paid-in contributions database), WHO, Organization for Economic Co-operation and Development (OECD), Gates Foundation, Bloomberg Philanthropies, International Monetary Fund (IMF), and Institute for Health Metrics and Evaluation (IHME).

Tobacco control and the World Bank
Evidence of tobacco's danger first became available in 1950, with the publication of three research papers, which reported a correlation between tobacco consumption and lung cancer 11 . The initial evidence was followed by reports from the Royal College of Physicians of the United Kingdom (1962) 12 and the United States Surgeon General (1964) 13 . This accumulated evidence also paved the way for the first international tobacco control event, the World Conference on Smoking and Health, in 1967. However, it took several decades for the Bank to initiate tobacco control projects worldwide. We identified four major periods in the Bank's involvement in global tobacco control ( Figure 1).

Phase I: The politicisation of tobacco control (1950s-1991).
Starting in the 1950s, the Bank extensively supported tobacco farmers, as part of an effort to strengthen economies by increasing specific countries' exports of expensive crops 14 .
According to the Bank's Projects & Operations database, the Bank committed USD 32 million in the 1970s to four projects aimed at enhancing tobacco production in Tanzania, Zambia, and Uganda [15][16][17][18] . However, in 1991, the Bank adopted an Operational Policy prohibiting it to "lend directly for, invest in, or guarantee investments or loans for tobacco production, processing, or marketing" (See World Bank Operations Manual).
This reversal of tobacco lending policies was largely due to the efforts of a handful of the Bank staff in the early 1990s. Howard Barnum, a then senior economist, played a key role in demonstrating the cost-effectiveness of tobacco control 19 . A tension existed between the humanitarian concept of health as a human right, particularly following the Alma Ata declaration of 1978 20 , and the Bank's economically-driven, politicised development approach to health 21 . To prioritize investment in tobacco control for health purposes, therefore, the Bank needed to demonstrate its cost-effectiveness. Senior economist Howard Barnum advanced a particularly compelling argument that market efficiency could not be applied to tobacco because of its addictiveness and consumers' lack of knowledge of its dangers 19 . (1991-2005). The Bank then emerged as a "knowledge bank", which could create and marshal expertise on development topics 22 . In the early 1990s, the Bank still saw tobacco as a health-versus-economics issue; the 1993 World Development Report argued for the importance of tobacco taxation, but simultaneously urged caution in applying tax reforms in countries that relied on tobacco exports 23 .

Phase II: Creation of the knowledge bank
With the accumulation of evidence on tobacco's dangers, the Bank began to make tobacco a political agenda in the mid-1990s. The 1997 Health, Nutrition and Population (HNP) Strategy Paper stated that tobacco control required a cost-effective approach, such as taxation or price measures. Based on this paper, researchers in the Bank's HNP sector, led by Prabhat Jha, presented the idea of tobacco control to the HNP board in 1998 24 . They proposed potential control activities, including the production of an analytic report, construction of partnerships with other organisations, and the support for the WHO's efforts. In this period, President James Wolfensohn also proposed a Comprehensive Development Framework, in which he noted that development required a holistic approach; health was central to the development; and tobacco control was a key development agenda 25 .
There was opposition to this control agenda inside the Bank. Some argued that the policy was against the idea of trade liberalisation. One needs to note that the Uruguay Round was concluded around this time, which enhanced trade liberalization and led to the establishment of the World Trade Organization. Thanks to Barnum's efforts, however, the Bank had already adopted its 1991 operational policy, which forbade its investment in tobacco production. Despite struggling with this opposition, the Bank continued its effort to create the evidence for tobacco control. Such efforts were crystallised in the publication of two reports on the economics of tobacco-related deaths 4 and its evidence-base in developing countries 26 .
This second phase was also characterised by a close collaboration with the WHO. Gro Harlem Brundtland was elected as the WHO's Director-General, and initiated the Tobacco Free Initiative in 1998. The WHO's leadership enhanced inter-agency efforts in tobacco control, including the Bank's tobacco control projects. (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013). The Bank provided the WHO with the evidencebase on tobacco's harm and the cost-effectiveness of tobacco control, which became a driving force toward the enactment of the FCTC in 2005. However, after its enactment, the Bank entered The World Bank's tobacco control programmes can be captured in four phases, from its support to tobacco production in the 1950s to the implementation of tax reforms in the 2010s. The WHO recognised the tobacco's harm early, and its efforts are divided into three phases. a slow phase in tobacco control. Some researchers speculate it was because of Bank bureaucracy; taxation and health were in different silos and the two policy groups could not collaborate with each other effectively on implementing tax reforms (See Devex article on The World Bank and Tobacco Taxes). Another possible explanation is the lack of leadership for tobacco control. Following President James D. Wolfensohn's retirement in 2005, the new President Paul Wolfowitz (1995)(1996)(1997), showed little interest in global health 27 .

Phase III: Proliferation of new actors in tobacco control
In this phase, new actors started to work on tobacco control globally, including the Gates Foundation and Bloomberg Philanthropies. In 2006, the Bloomberg Philanthropies initiated the Bloomberg Initiative to Reduce the Tobacco Use in developing countries, based on the success in tobacco control in New York City. In 2008, this initiative went into the second phase, and was joined by the Gates Foundation (See Gates Foundation press release).

Phase IV: Global Tobacco Control Programme (2013-). The
Bank launched multiple country-based tobacco control projects that were initiated around 2013 (Development Topic database). In 2015, the Bank officially initiated the Global Tobacco Control Programme, which enhanced collaborations between the Bank's HNP Global Practice and its Global Taxation Team. Such efforts were summarised in its major 2017 report, Tobacco Tax Reform: At the Crossroads of Health and Development 3 .
The major driver for renewed efforts in tobacco control at the Bank was the election of President Jim Yong Kim and effective domestic resource mobilisation for health. President Kim introduced a new model of Global Practices, and tobacco control was placed in the HNP Global Practice, which fostered interactions between technical staff working in different regions. Also, in order to implement tax reforms in target countries, the Bank needed to establish buy-ins from country experts and leaders. The Bank therefore began to collaborate closely on tax reforms with experts from the Ministries of Finance in these countries. The results from such collaborations have been published in multiple Bank country reports 3 .
The Global Tobacco Control Programme is financed through the Tobacco Control Programme Multi-Donor Trust Fund, supported by the Gates Foundation and the Bloomberg Philanthropies ( Figure 2). Bloomberg Philanthropies and the Gates Foundation each committed USD 5 million to the three-year (December 2014 to 2017) fund. It is a bank-executed trust fund; this type of trust fund typically supports advisory services and the development of the Bank's knowledge agenda 28 . According to the trust fund agreement, the supported projects have a narrow objective to assist selected countries in implementing tobacco tax reforms 29,30 . These objectives are accomplished through four major tobacco control activities: providing ministries with policy advice and technical assistance; establishing knowledge exchange systems; helping to build capacity and promote tobacco control as a priority; and coordinating the programme with appropriate partnership. The fund is governed by a Consultative Group, which is chaired by the Bank, and includes representatives of the donors, the Gates Foundation and Bloomberg Philanthropies. This Consultative Group selects country programmes. Tobacco control and the WHO The development of tobacco control frameworks at the Bank and WHO shared several features: both organisations faced internal objections, identified collaborations within the UN agencies as a key to success, and had strong leadership, from Wolfenson and Brundtland, respectively. In contrast to the Bank, however, the WHO recognised the scientific evidence early on, and subsequently initiated its control efforts in the 1970s. (Figure 1) Phase I: Tobacco control as a primary healthcare issue . In 1970, the WHO adopted its first tobacco resolution, which urged the Director-General to consider establishing an expert group on tobacco control. The resolution also highlighted the importance of educating younger generations on tobacco's harm 31 . Such efforts were advanced through two landmark reports: Smoking and Its Effects on Health in 1975, and Controlling the Smoking Epidemic in 1979. The 1979 report called for action 32 , based on which the first permanent programme "Tobacco or Health" was created in 1980 33 . Although it started small, this programme contributed to the preparation of technical reports 34 and lead to the inception of the World No Tobacco Day 35 to raise awareness among member states. This period coincided with the organisation's programming goal "Health for All," which aimed to secure the health of all people through primary healthcare. Director-General Halfdan Mahler integrated tobacco control in a primary healthcare agenda 36 .

Phase II: Toward enactment of the FCTC (1993-2005).
In the 1980s, some researchers expressed their support for an international law to regulate the consumption of tobacco products 37 . However, it was in the 1990s when the WHO developed a realistic idea for a global tobacco control facility. The original idea, discussed by lawyers Ruth Roemer and Allyn Taylor in 1993, was to use the WHO's constitutional authorities 38 . Despite opposition inside and outside of the WHO 38 , the World Health Assembly adopted a resolution in 1995 to develop an international framework on tobacco control. After Brundtland's election in 1998, the idea gained political support, and the Tobacco Free Initiative (TFI) was launched 39 .
Tobacco control was different from other health problems by nature, due to active industry lobbying. The WHO's noteworthy strategy was to gain support from the entire UN system 39 . Brundtland requested that other organisations shift the leadership role in tobacco control to the WHO. The WHO became the coordinator of the UN and Bretton Woods systems, leading to the creation of an ad hoc Inter-Agency Task Force in 1999. In that year's World Health Assembly, member states agreed to start a formal negotiation toward the proposed framework 40 . Beyond paving the way to the FCTC in 2005, Brundtland's move created a momentum across the UN system. For example, in the Inter-Agency Task Force meeting, the UN linked tobacco to the eight Millennium Development Goals (MDGs) 41 , which pushed tobacco control into a global development agenda 42 .

Phase III: Implementation of the FCTC (2005-).
After the enactment of the FCTC, the WHO implemented tobacco control projects globally. The FCTC serves as a negotiation entity for resource allocation and decision-making, whereas the TFI conducted technical support for national or regional tobacco control activities, with the support of Bloomberg Philanthropies.
The biannual Conference of the Parties (COP) is the FCTC's decision-making entity. The COP is attended by the Parties (181 countries as of March 2018), and discusses its budget and programmes proposed by the Secretariat. The main sources of the funding (Figure 3a) are the Voluntary Assessed Contributions (VAC) (Figure 3b) by the Parties and extra-budgetary funds. Therefore, the FCTC is essentially financed by WHO trust funds: resources mobilised voluntarily from donors, and held apart from the core budget 28 . Sources of extra-budgetary funds include governmental The TFI supports national and regional tobacco control programmes. With the financing and close collaboration of Bloomberg Philanthropies, the WHO's TFI works with countries as part of the Bloomberg Initiative to Reduce Tobacco Use. The exact amount of funding is not available on public websites, but the initiative has provided grants to implement tobacco control with a focus on high-burden countries. The WHO, however, does not play any role in the grant's selection process (Figure 4) (See WHO page on TFI).
Challenges and opportunities of the World Bank and WHO Financing, decision-making, and accountability of tobacco control. A comparative analysis of the tobacco control policies by the Bank and WHO illustrates similarities and differences in their financing, decision making, and accountability (Table 1) Second, in terms of decision-making processes, the FCTC is the only entity that is not influenced by private philanthropies. The FCTC holds biannual COPs, in which the Parties (sovereign states) have the voting power. In contrast, the WHO's TFI is not involved in the selection process of country projects supported by Bloomberg Philanthropies' tobacco control grants.
Also, the Bank holds a Consultative Group, which includes the representatives from the Bank, the Gates Foundation, and Bloomberg Philanthropies.
Third, the Bank is accountable mainly to its two donors funding tobacco control, and detailed progress reports and financial disclosures are available only for these donors on its secure websites 29,30 . The WHO, on the other hand, is accountable mostly to its Parties, but most of the documents from the COP are disclosed on its website. Such detailed financing data are not available for the WHO's TFI.

Challenges and opportunities.
What are the opportunities that the two organizations created for tobacco control? The FCTC provides an international platform for financial and legal support to tobacco control activities. Through the FCTC, Parties can negotiate and support global efforts, and civil societies can participate in the conferences as observers. The Bank has also created the evidence on tobacco control, and implemented tax reforms globally. Based on this evidence base, attention is now being paid to effective implementations of tobacco control measures. Finally, funding has increased for tobacco control globally, reaching from USD 7.9 million in 1990 to USD 99.8 million in 2015 ( Figure 5) (See Viz Hub page on financing global health). There is a growing diversity in channels of global financial assistance, such as nation states, philanthropies, and NGOs. Such coherence toward tobacco control was not observed in the early 1990s, when the amount of funding was limited and mostly channelled by the UN organisations.   As leaders in tobacco control, the two organisations should consider three issues: their institutional segmented nature, the proliferation of new actors, and monitoring and evaluation.
First, tobacco control will continue to be challenged by the segmented nature of global health institutions and government ministries as well as global regime complexity, i.e., conflict between trade and health law 45 . An initiative can easily be classified into one sector, to which other sectors pay little attention. Some argued that this occurred in the late 2000s to early 2010s in the Bank, during which tobacco control was implemented only by health experts, and did not collaborate effectively with fiscal policy groups to implement tax reforms globally (See Devex article on The World Bank and Tobacco Taxes). Such miscommunication across sectors could occur in national governments as well. The Bank has made enormous efforts to build national capacities in implementing tobacco taxation since 2013, but the existing bureaucracies across ministries should be scrutinised; implementation, monitoring and evaluation of taxation measures require collaborations that go beyond the authority of finance and health ministries.
In addition, tobacco industries are undertaking a growing number of tobacco control litigations under trade law 46 . Tackling such litigations requires close collaboration between health and legal practitioners, which poses a particular challenge for the FCTC, since it cannot legally force the Parties to adhere to the provisions 47 .
Second, the proliferation of donors, such as the Gates Foundation, is a double-edged sword. Tobacco control programmes are not the exception to the shift in global health governance, in which multiple actors -including nation states, philanthropies, NGOs, and industries -are involved in decision-making.
The involvement of private philanthropies poses risks, such as the alignment of objectives of implementation bodies and their own, and the use of material incentives 5 . For example, in 2010 the Gates Foundation terminated a tobacco control grant to a Canadian agency, the International Development Research Centre, because the leadership of the organisation also directed Canadian tobacco industries (see Gates Foundation press release). Also, the Bank's efforts are currently implemented with a narrow mandate that matches the philanthropies' objectives: tax reforms in targeted countries. Similarly, the WHO's TFI operates with a particular focus on high-burden countries without having any governmental authorities in the grant's selection process. Such an approach possesses an inherent risk of "Trojan multilateralism", the ability of a few agents (such as specific states or industry representatives) to have an undue influence on decision-making at multilateral organisations 4 .
Lastly, monitoring and evaluation have become increasingly difficult due to the trust funds' inadequate transparency 28 . Tracking development assistance for health (DAH) allows researchers to analyse the trend of priorities set by donors and implementing agencies, and the lack of a comprehensive database to track development assistance for trust funds has been a major concern.
Although the Bank sometimes releases annual trust reports, they are not available for recent years, and its Trust Fund Directory is out of date. Furthermore, its World Bank Finances database's trust funds datasets generally only include trust fund commitments after 2005, and were last updated in 2013-2014 28 . Datasets available from other institutions such as the IHME, AidFlows (a collaboration between the Bank, OECD, and regional development banks), and the OECD also do not capture the entire picture of the trust funds. For example, our literature search identified at least seven trust funds for tobacco control projects that were channelled by the Bank (Table 2), but no databases systematically included all these activities. Moreover, while the IHME and AidFlows allow researchers to specifically search for Gates Foundation commitments, Bloomberg Philanthropies' contributions were either unavailable (AidFlows) or included in NGO totals (IHME). Given that the Bloomberg Philanthropies has invested approximately USD 1 billion in tobacco control (see Bloomberg page on tobacco control), this is a concerning trend. The health financing landscape for tobacco control may look significantly different, if such activities are properly reflected and private foundation commitments are clearly identifiable.

Conclusions
Despite the small beginnings in the 1970s, the past two decades have seen the crystallisation of global efforts in tobacco control by the Bank and the WHO. This has resulted in major progress through the establishment of the FCTC and the implementation of evidence-based tobacco control policies with a particular focus on low-and middle-income countries.

Competing interests
A senior member of the World Bank is on our project's advisory board.

Grant information
This work was supported by the Wellcome Trust [106635] and the Nakayama Foundation for Human Science.
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
What is more, it would be interesting to say more about the exact implications of regime complexity for the political role and influence of tobacco companies. The key question is, as Alter and Meunier put it, whether and how international regime complexity impacts "decision-making and political strategies, as well as empower some actors and interest groups." That is, regime complexity provides internationally operating firms with ample opportunities to engage in forum shopping and, as such, target those institutions that are most favourably disposed towards their policy preferences. Research has shown that tobacco companies use the WTO Dispute Settlement System and Investor-State Dispute Settlement Mechanisms within Bilateral Investment Treaties to challenge domestic tobacco control policies like the introduction of plain packaging . Further research is needed on how business-government relations are affected by the trans-nationalisation of tobacco firms and the institutional complexity of tobacco control governance.
Finally and related, the authors briefly mention the tension between trade and health law and this issue could be unpacked further as well. During the FCTC negotiations a compromise was reached whereby no explicit reference was made to the relationship between tobacco control and international trade law. As a result, doubts have been raised as to the likely outcomes of disputes involving a conflict between trade and health concerns . In a recent ruling in a dispute concerning tobacco products within the WTO, where several countries complained about the adoption of plain packaging (PP) requirements by Australia, the WTO Panel ruled in favour of the defendant (Australia) and argued that the public health measure taken (i.e. the introduction of plain packaging) was in fact not inconsistent with trade law (see: On current smoking patterns, about 1 billion smokers will be killed by their addiction in the 21 century, as opposed to "only" 0.1 billion in the 20 century. Reductions in premature deaths from tobacco consumption are possible in every country, particularly with significant increases in tobacco excise taxation . Tobacco control involves ministries of finance, as well as vested interests in agriculture, commerce and trade, it is unusual from most health challenges that are normally within the purview of health ministries. Thus, understanding the global response to the tobacco epidemic is crucial, and given the multitude of sectors, is complex. Devi Sridhar and her team have made an important contribution to understanding the global architecture for tobacco control . In this review, Mukaigawara and colleagues review the major periods of global tobacco control that started first with attention by the World Health Organization (WHO) and then the World Bank, and more recently by private philanthropies.
Their review is serious, scholarly and sobering. It points to the obvious disconnect between the huge, and largely avoidable burden from tobacco, and its response. Tobacco taxation is, arguably, the single most cost-effective intervention for health worldwide . Yet it continues to be vastly underused. Why? The obvious reason is the enormous profitability for tobacco dealers, who earn approximately $50 billion of profit every year from cigarette sales, or roughly about $10,000 per tobacco death . The second reason, articulated nicely in this review, is that the international architecture for tobacco control has been less than optimal. The alignment of financial and development interests with tobacco control demands a much stronger role for the World Bank and the International Monetary Fund, as key counterparts to ministries of Finance.
The World Bank made substantial progress during its phase when it produced the widely-influential "Curbing the Epidemic" report ( ; full disclosure-I led that report, and organized its funding and implementation while a Senior Health Specialist at the World Bank). Since then, the track record has st th 1 2 3 1 4 implementation while a Senior Health Specialist at the World Bank). Since then, the track record has been uneven. However in the last three years, Patricio Marquez and team have substantially reinvigorated the Bank's response and produced an important report that calls for a major scale of big, fast tax increases so as to change future smoking behaviour.
WHO estimates that over the last few years, about 100 countries raised excise taxes, and that the proportion of the world now covered by some form of tobacco control has increased substantially . This is welcome news, but marked increases in adult cessation (which is the only practicable way to avoid much of the tobacco deaths prior to 2050) would need large tax hikes paired with strategies to prevent downward substitution.
What then for the future of global tobacco control? First, as with most health successes, country ownership is key to fund and to support tobacco control. Here, the recent evidence that strongly links tobacco excise increases to reductions in poverty (mostly from avoided treatment cost for tobacco-attributable diseases; ) is an important milestone that may increase country ownership. Second, the forthcoming Tax Force on Fiscal Policies for Health, chaired by Larry Summers and Michael Bloomberg might provide important spur for governments to signals future price hike expectations to smokers. Indeed, France was able to halve consumption per adult in about 15 years versus the 30 years it took to halve consumption in the United States by using large excise taxes and announcing these in advance to the public . Today French ex-smoking prevalence among middle age adults is well above the European average. Mukaigawara and colleagues warn, appropriately, about the undue reliance on private philanthropic funding for tobacco control. A small criticism of an otherwise fine paper is that they did not spend sufficient time discussing research innovations for tobacco control (such as better and more local epidemiological and economic research) or new tools (such as is now appearing with use of plain packaging; ). Indeed, a useful trajectory of philanthropic research funding would be to move away from implementation to a much deeper investment in creating new tools, and evidence-driven networks that could support countries to raise taxes substantially.
Large tax hikes might well avoid about 200 million deaths over the next few decades and are the only practicable way to achieve the United Nations global goals of reducing deaths from chronic disease by 2030 . 6. WHO report on the global tobacco epidemic 2017.
. 2017. World Health Organization 7. Global Tobacco Economics Consortium: The health, poverty, and financial consequences of a cigarette price increase among 500 million male smokers in 13 middle income countries: compartmental model study.
. 2018; : k1162 BMJ 361 PubMed Abstract Is the work clearly and accurately presented and does it cite the current literature? Yes Is the study design appropriate and is the work technically sound? Yes

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?

Not applicable
Are all the source data underlying the results available to ensure full reproducibility? Partly

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Referee Expertise: Epidemiology, economics, global mortality I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.