Relatively recently, global institutions such as the World Health Organization (WHO) and the World Bank have made the link between macroeconomics and health [2,3]. Analysis of poverty around the world highlights those countries that are 'very unlikely' to meet the World Bank groups' millennium development goals (MDGs) [4]. These MDGs include the combat of HIV/AIDS, malaria and other diseases, improvement in maternal health, achievement of universal primary education, promotion of gender equality and empowerment of women, reduction in child mortality and the eradication of extreme poverty and hunger.

Some of the shared societal structures underpinning economic growth and health are the absence of violence, government effectiveness, the rule of law, lack of corruption and the ability to select a government. Even though all of these are clearly desirable the relative weight of each societal structure necessary for a strong nation state is debatable [5]. The risk of infectious disease is determined not only by pathogens and the response of the patient but also by powerful societal forces that override individual knowledge and choice [6]. Paul Farmer has coined the phrase 'structural violence' that reflects the limit of life choices, particularly of women, by racism, sexism, political violence, and grinding poverty.

The 2004 World Health Report discusses the challenges of tackling the HIV pandemic [7]. In the African continent, HIV is implicated in poor economic performance and falling gross domestic product (GDP). Within this document it describes the wide range of international support garnered to meet this challenge. However, even though the requirement of local and national government co-operation is stressed within this document, it does not elaborate on the massive heterogeneity inherent within this mandatory component.

The 2002 HIV prevalence estimates were obtained for each country. HIV prevalence is the percentage of adults aged between 15 and 49 years of age infected with HIV. One hundred and forty nine of the 199 countries / regions cited by the World Bank paper had published UNAIDS 2002 HIV prevalence estimates. Those countries / regions excluded from analysis due to the lack of HIV prevalence data are listed in Table ​Table3.3. Those countries given a UNAIDS HIV prevalence estimate of < 0.1% were given a set value of 0.05%.

In addition to separate analysis of each governance dimension, an average governance figure was obtained based on the assumption that each governance dimension was of equal importance. The null hypothesis was tested by measuring association between ranked governance and HIV prevalence data across the whole spectrum of countries (Kendall tau test, two tailed). Statistical analysis was performed using SPSS version 12.

Other health and economic data have been included in the Tables to illustrate the development needs of countries. The most recent maternal mortality data from the World Health Organization (WHO) data are available from the year 2000 [10]. The mean maternal mortality ratio (MMR) is the number of mothers who die per 100,000 live births. The number of physicians available per 100,000 inhabitants from 1990–2003 were obtained from United Nations [11]. Access to improved drinking water in 2002, expressed as a percentage, was obtained from WHO / UNICEF [12]. Life expectancy and the GDP per capita in US dollars corrected for purchaser power parity (GDP-PPP) were obtained from the Central Intelligence Agency World Factbook 2002 [13]. The GINI index data (1994–2001) from the United Nations are quoted to give an indication of the equity of income and resource distribution for each country [14]. A value of zero on the GINI index indicates fully equitable distribution of income and resources.

Non-parametric analysis of association between governance and HIV prevalence is shown in Table ​Table4.4. The negative correlations indicate that HIV prevalence falls as the governance improves for each governance dimension and mean governance. The three most influential dimensions of governance were government effectiveness, the rule of law and corruption. All correlations were significant thus rejecting the null hypothesis.

The dataset of 149 nations has been divided into three groups that represent the lowest (n = 50), middle (n = 50) and the highest (n = 49) governance ranking positions for each governance dimension and mean governance in Tables ​Tables5,5, ​,66 and ​and77 respectively. When these nations were divided into three groups, the median (range) HIV prevalence estimates remained constant at 0.7% (0.05 – 33.7%) and 0.75% (0.05% – 33.4%) for the lower and middle mean governance groups respectively despite improvements in other health and economic indices. The median HIV prevalence estimates in the higher mean governance group was 0.2% (0.05 – 38.8%).

The HIV prevalence is generally low in higher governance group however this figure conceals differences found within specific population groups. For example in the USA, HIV prevalence amongst African American women is almost twenty three times that in whites [16]. Whilst in the UK, the prevalence of HIV amongst men who have sex with men (MSM) within London in 2001 was 100 times the national average [17]. The disparity in HIV prevalence amongst 'at risk' groups in the UK and US highlight the general difficulty of using the UNAIDS country HIV prevalence estimates. The quality of surveillance methods has been discussed and graded by UNAIDS surveillance teams, and it is clear that some HIV prevalence estimates are inaccurate [18].

Most sentinel surveillance methods use antenatal screening due to the ease of patient access and the benefits of provision of anti-retroviral treatment (ART) to prevent mother to child transmission. This surveillance strategy, though easier to implement, does not sample high risk groups such as MSM, intravenous drug users and commercial sex workers and thus underestimates the true HIV prevalence figure for the country. Within the 2002 UNAIDS HIV prevalence estimates used in this analysis there are at least four grades of surveillance. As highlighted in Table ​Table33 some countries / regions were not included in formal UNAIDS surveillance and then there are countries where a UNAIDS report was commissioned yet no HIV prevalence estimate was provided. Reasons were not given as to why certain countries did not have a report commissioned. One fifth of the countries with UNAIDS reports quote a HIV prevalence estimate less than 0.1% and yet other health and economic indices would predict that this is an optimistic figure. Finally, there are those countries that report a HIV prevalence estimate greater than 0.1% which is complicated to varying degrees by the observer bias described above. Additional file 1 tabulates the UNAIDS HIV prevalence estimates for each of the 149 countries included in this analysis.

The governance dataset by Kaufmann et al in 2002 is the first global assessment of societal structures. These authors point out the variability inherent in collecting subjective material and, like UNAIDS, state the need to improve the quality of the data collected in subsequent analyses. Ideally sources would be able to freely report on each nation however, the extent of data available decreases in those nations with poorer governance. The relatively large margins of error within the governance data make direct cross-country comparisons difficult to interpret. Due to the variability of both the governance and HIV prevalence dataset the whole spectrum of data was chosen and subjected to non-parametric ranking analysis. The null hypothesis 'HIV prevalence is not associated with governance' is rejected for each dimension of governance with variations in the relative importance of different governance dimensions. Previously, Fareed Zakaria [19] has argued that democracy is less important in the development of a strong nation than the rule of law, corruption and political stability. The correlation coefficient of the voice and accountability dimension of governance with HIV prevalence was the lowest in this analysis somewhat supporting this contention.

Those countries in the lowest governance ranking group of governance are defined by poverty, ineffective health care systems, elevated HIV prevalence and significant international debt. The elevated HIV prevalence in many of these vulnerable countries was predicted more than a decade ago following the analysis of health, economic and human rights data [20]. Historically, international support has focused on short-term 'vertical' disease control strategies to tackle healthcare problems [21]. Long-term, 'horizontal' capacity building strategies are vital if HIV / AIDS is going to be effective managed in nations with limited healthcare infrastructure [22]. It has been shown in a number of resource poor settings that the provision of voluntary counselling and testing (VCT) for HIV is facilitated by the provision of free primary care services and ART [23]. The provision of effective primary care support to pregnant women is the most effective way to provide VCT services for HIV and thereby identify HIV positive mothers, prevent mother to child transmission and facilitate VCT of their partner(s). Like surveillance, this strategy though relatively effective fails to test and treat vulnerable 'high risk' groups within the population.

The poor are those most at risk of infectious disease. The role of poverty as a risk factor for disease has been clear for over 300 years [24]. Health and wealth are inextricably linked. All who become chronically ill enter a negative cycle of limited horizons. Indeed, what is true for the individual is equally true for the nation state. The effect of HIV on economic under performance and negative growth is testament to this. It is vital that essential healthcare is free, so that those that catch treatable infectious diseases are allowed to live. Encouragingly there are a few positive examples in resource poor countries, such as Uganda, Senegal and Cuba, where leadership, good communication and support of civil society have made a difference in their respective HIV epidemics. There are however many countries within this group of vulnerable nations that need the bulk of international healthcare and financial institution commitment in order to address their devastating healthcare challenges.

One example is Nigeria which is the most populous nation in Africa that has been a democracy since 1999. Despite its vast resource wealth, this country has suffered from repeated religious and ethnic conflicts that have compromised its development [25]. Only recently has the civilian government made HIV / AIDS its top priority and initiated some selective treatment programs. This change followed pressure from civil society and the military, in this latter group HIV prevalence is estimated to be 20% [26]. In contrast, South Africa is 115 mean governance ranking points higher than Nigeria, has had universal franchise for a decade and the per capita wealth is almost ten times greater than Nigeria yet the HIV prevalence is four times greater in South Africa. Some of the multiple factors that help explain the HIV prevalence in South Africa include: an earlier HIV epidemic, migration from high HIV prevalence neighbours, the violence and inequality of the Apartheid era and government inaction over the last decade. The government of South Africa has recently responded to national and international civil society pressure and has promised to provide ART free to all patients with advanced HIV disease [27].

Cuba and Haiti are islands with a similar population size and GDP-PPP per capita yet the HIV prevalence estimates are 0.05% and 6.1% respectively. HIV is thought to have entered Haiti from the USA via the sex trade in the early 1980s. The main exposure risk for Cuban nationals was from military and healthcare worker interaction with sub-Saharan Africa. Cuba was one of the first countries in the Americas to launch a nationwide HIV policy to contain transmission and care for those people living with HIV / AIDS [28]. Healthcare in Cuba is provided free to its citizens by the state and there is strong political commitment supporting health as well as national and international HIV / AIDS action. In contrast, there have been 33 coups in Haiti in the last two centuries of independence. Political instability in addition to other governance factors have been attributed to the lack of development of a responsive healthcare system [29].

As governance improves fewer women die in childbirth, more physicians exist per population, there is better access to improved water and life expectancy is longer. In addition with improvements in governance there is more GDP-PPP per capita, more equitable distribution of income and greater investment in health and education compared to the military. Interestingly, the median HIV prevalence estimates does not change between the lower and middle third groups of mean governance ranking despite a step-wise improvement in all other indices. The majority of middle ranking economies can be found within the middle governance group. Many of these countries have large populations where the HIV epidemics are set to explode. Three nations in this group who have the economic and technological power to halt their respective epidemics are Russia, China and India.

Russia has over 3 million intravenous drug users and relatively expensive ART that help to fuel the HIV epidemic [30]. The collapse of the USSR produced significant strain on the health of the people [31]. Life expectancy in Russia fell 9 years following its transition to a market economy and there has been a significant rise in 'social diseases' of Tuberculosis (TB), HIV and Hepatitis. Intravenous drug use accounts for approximately 80% of those infected with HIV however recently a new phase of the epidemic has developed that is driven by sexual transmission [32]. It is only since 2003 that there has been an increase in leadership and commitment at higher political levels to combat HIV and AIDS.

UNAIDS reported in 2002 that the number of overall infections in China increased 30% since 1998, with over 1 million people infected with HIV [33]. It is feared that China may soon experience an explosive and widespread HIV epidemic. Intravenous drug use and the sharing of contaminated needles in the south and north-west of China was one mechanism of initial transmission the other was unsafe practices among paid blood donors. Unsafe blood collections in the 1990s led to the appearance of HIV and subsequent AIDS deaths in Chinas central provinces. In response to this the Chinese authorities have recently announced that they are providing free ART in central provinces [34].

The first main focus of HIV in India was Mumbai where there is a large commercial sex work industry and the HIV prevalence reported amongst these workers is 50%. It is expected that HIV will become the largest cause of adult mortality in India in the coming decade. Despite the government making HIV its national topmost priority, any attempt to address the problem is hampered by its fractured health care infrastructure, poor literacy figures and widespread poverty [35]. At the end of 2003 the Indian government began providing free ART in eight government hospitals with the plan to expand it to a total of 25 centres [36].

The aim of this paper was to attempt to dissect out the role of governance in the HIV pandemic. It is not possible yet to determine if the relationship seen represents correlation or causation. Even though this first analysis alludes to causation, for those 149 countries with UNAIDS HIV prevalence data, the relationship will become clearer over time when it is possible to compare nations that appear similar today. Currently Brazil and India have equivalent overall governance and HIV prevalence estimates at 0.8% and 0.7% respectively. However when other health and economic indices are examined it is clear that India invest less than Brazil in health and education, has one quarter the number of physicians and double the MMR. The GDP-PPP per capita is three times greater in Brazil but it is more equitably distributed in India which is likely to contribute to equivalent life expectancy seen in both countries. India and Brazil are the main producers of generic ART. However Brazil, unlike India, has consistently provided strong political support for HIV / AIDS patients after the end of the military dictatorship in 1990. In 1996, the Brazilian government guaranteed by national law the permanent allocation of financial resources and universal access to care, including ART [37]. The current disparities between India and Brazils HIV treatment policy predicts that the Indian epidemic will progress more rapidly and is likely to impact on its development. All ten countries selected for discussion are summarised in Table ​Table88.

HIV / AIDS control in Russia, China and India will only be possible if they follow the example set by Brazil. International institutions need to support national civil society groups within these nations to focus the attention and resources of their respective governments for progressive healthcare changes. The global plan to stop TB outlines the possibilities and challenges that will be faced treating chronic illness, such as HIV [38]. It is pertinent that international health and financial institutions work together to influence change so that robust healthcare networks and responsive government are developed in order to apply best healthcare and economic practice.