The production of consumption: addressing the impact of mining on tuberculosis

Mining for coal, diamond, gold and other minerals has been associated with some of the greatest disasters in occupational history, with recent mine collapses in Chile, West Virginia, and China capturing headlines. The environmental impact of mining, along with the use of gold and “conflict diamonds” to fund proxy armies, has also been the subject of novels and documentaries.

But from a public health perspective, the dangers of mines are not isolated to dramatic explosions, trapped miners and guerilla wars; the epidemiology of mining, in fact, suggests that the occupation carries with it a unique set of secondary effects on the rest of the population. In this week’s post, we look at the evidence suggesting that mining amplifies infectious disease epidemics (especially tuberculosis, TB) on a regional and worldwide scale. We look at how the use of statistics and simulation models can enable us to investigate the broad effects of mining on a wider population, as well as compare alternative policy options to control the epidemic effects of mining on TB.

Mining and the risk of tuberculosis

The mining industry is among southern Africa’s largest employers, particularly in the Republic of South Africa, where one of every ten employed men (at least 500,000 men) mines for gold, diamonds or other minerals. The South African government reports the incidence of tuberculosis to be as high as 7,000 cases per 100,000 miners, about ten times higher than among the general population. Miners also have a four-fold greater odds of dying from TB than the general population – but why?

A few key environmental and social factors are to blame. Exposure to silica dust increases the risk of pulmonary tuberculosis, particularly among gold miners who drill through hard rock. Miners with the scarring lesions characteristic of silicosis–about 18% to 31% of gold miners in Botswana and South Africa–have about a three-fold increased risk of pulmonary tuberculosis compared with those without silicosis. Living and working conditions are also a cause for concern. Mine shafts themselves are crowded and poorly-ventilated, but so are hostels where over a dozen men can share a small room. These conditions are highly conducive to infection; the rate of recurrent tuberculosis in a recent South African prospective cohort of 600 miners was about 8 per 100 person-years (as opposed to half of that rate or less in the general population), with 69% of recurrent cases attributable to reinfection rather than relapse.

But occupational and environmental risks on mines apply to the mining sector in wealthy countries, just as in poor ones. What makes mining in southern Africa so dangerous that tuberculosis rates are far higher among African miners than in miners in the US or Europe? While regulations are weaker in southern Africa, the companies owning the mines and determining typical occupational conditions are multi-national corporations. The problem is not simply one of differential occupational hazards, but also of the social context for transmission and the interaction between miners and the rest of the population. An extensive migration system throughout southern Africa was constructed over a century ago to facilitate the movement of workers to mines. The system, which until the early 1990’s prohibited black workers from settling permanently in “whites-only” areas, created patterns of circular migration conducive to the spread of tuberculosis both on the mines and to rural areas from which men migrated. Shantytowns developed around hostels, with alcoholism and prostitution proliferating around many. This corresponded to the spread of sexually-transmitted diseases.

HIV has rapidly spread among miners and their partners since the 1980’s. The dramatic rise in HIV prevalence among miners (upwards of 30% in some cohorts) has been thought to be responsible for a subsequent increase in tuberculosis incidence among them. According to one industry study, nearly one-third of new mineworkers without HIV will become infected within the first eighteen months of employment. HIV increases the likelihood that a person infected with tuberculosis will progress to active disease, shortens survival times among individuals with both HIV and TB, and increases the likelihood of atypical tuberculosis manifestations that can be difficult to diagnose. HIV- tuberculosis “co-infection” is particularly problematic for miners: HIV and silicosis multiplicatively increase the risk of tuberculosis, so tuberculosis incidence among HIV-positive silicotic miners is about 15 times higher than among HIV- negative miners without silicosis. Migrants moving between their homes and the mines usually do not have continuous access to treatment, risking poor patient outcomes as well as the development and subsequent transmission of drug-resistant forms of disease.

Since the relaxation of rules restricting movement under apartheid, miners are able to travel more frequently between mines and their home communities. In South Africa today, roughly 230,000 men migrate each year from other countries for mining jobs. Over 50,000 men travel to South African mines from Lesotho (a country entirely contained within the Republic of South Africa), and 60% of them return home at least once per month; these individuals would normally travel home only once or twice per year in prior decades. The circular migration pattern not only exposes people in low prevalence areas to migrants with a higher prevalence of HIV and tuberculosis, but also prevents continuity of care, adherence support, and consistent access to diagnostic facilities for migrating miners. In a cross-sectional study of over 28,000 South African goldminers, 18% of 425 tuberculosis cases acquired multi-drug resistance, and a further 9% had already-resistant tuberculosis strains transmitted to them (primary resistance). Over 13% of cases had previously failed therapy. Since August 2007, one-quarter of new multi-drug and extensively drug resistant tuberculosis cases (MDR and XDR TB) in Lesotho were among miners or former miners.

The wave of new tuberculosis infections related to HIV is also being accompanied by significant secondary transmission of tuberculosis to HIV-uninfected persons. In this context of migration and secondary transmission, the number of mines in a population correlates strongly to the overall population’s tuberculosis incidence (see the Figure below); how do we know that this correlation really implies causation, and isn’t just a spurious correlation because mines are commonly found in poor populations also affected by HIV?

Measuring the epidemic impact of mining

Behind this question is another common question in epidemiology: how do you tell if an industry is having a population-wide impact on public health? There are two common approaches to answering the question. One is to conduct a statistical study of how much risk is “attributable” to that industry, regressing the population-wide rates of disease against some measure of the exposure of the population to that industry, correcting for “confounding” factors and, if possible, doing a time-series analysis to see if there’s a dose-dependent and sequential relationship that might imply causation. The second is to create a simulation model (a mathematical model of the epidemic) that tries to capture how disease spreads through a population, conditional on the different risks faced by the industry’s workers and others exposed to those workers in the community. When we tried both approaches to investigate the impact of mining on tuberculosis epidemics in southern Africa, the results were striking:

Using data on mining and TB for 44 countries in southern Africa, we first investigated the relationship between tuberculosis rates and mining, correcting for HIV and other potential confounders such as poverty and urbanization. We found that an increase in mining production by one standard deviation corresponded to a 33% higher TB incidence in the overall population (not just among the miners) – about 760,000 more cases of TB. By comparison, a 1% rise in HIV corresponded to a 4 to 5% rise in TB incidence.

Because gold mining in particular is strongly associated with silicosis, we looked separately at the impact of gold mines, and found that a 1 standard deviation rise in the number of gold mines was associated with a 49% higher TB incidence. Mining also was significantly statistically related to the number of HIV-TB coinfections; it appeared that mining did not just correspond to higher rates of TB because mining increased HIV risks, but also that there was an independent and direct effect of mining on TB infection rates, particularly in gold mines were silicosis is likely to be a key risk factor. This also helps explain why TB continues to be common even in southern African nations that have maintained low levels of HIV and high levels of TB case detection and treatment; the mining activity seems to be sustaining transmission of TB in spite of these factors.

As with any statistical analysis, the results could arise from confounding or by “ecological fallacy”—the fact that population-level associations don’t always correspond to individual-level mechanisms of disease. But these findings were consistent with individual-level empirical research on silicosis, HIV and tuberculosis, and the results were the same when we checked them using other indicators of mining activity, such as the number of mines in a given population (a “robustness check”).

What to do?

We next created a mathematical model to investigate how best to control TB among the general population when the population is influenced by TB among miners. The relationship between TB among miners and the rest of the population is analogous to many other TB outbreak scenarios: a very vulnerable population is at high risk of infection, but migrates into and back out of the rest of the community, where a “secondary epidemic” occurs. This is similar to the relationship between inmates in Russian prisons and the broader population in the former Soviet Union where we found that the resurgence of TB epidemics can be explained by the rate of incarceration.

The standard approach to TB control in the world has been the break the “transmission chain”: that is, if we can find and treat enough people with active, infectious tuberculosis, the effective treatment will suppress their infectiousness and they won’t infect others. The World Health Organization had previously set goals of 70% case detection and 85% treatment success (that is, find at least 70% of cases and successfully treat at least 85% of them) to break epidemics of TB. These calculations were based on mathematical models that assumed “homogeneity”—that people in the population mix around like tapioca in pudding, and all share relatively the same risks. But Russian prisons and African mines present a very different problem: a small subset of the population that is at dramatically higher risk than everyone else, and face social barriers to treatment that aren’t accounted for in these general models.

It turns out that if you incorporate these “institutional amplifiers” (prisons or mines) into a model of TB epidemics, you find that the WHO’s targets are likely to be totally inadequate. We found that even very high rates of case detection and treatment success—beyond those achieved in nearly any country—were still insufficient to counteract the high rate of infection and secondary epidemics caused by these institutional amplifiers. Furthermore, the “amplifiers” often expose people who have a high risk of getting inadequate treatment and dropping out of care, as they are released from prison or lost from therapy when they migrate back home from mines. So the risk that they’ll develop drug resistant forms of TB goes up in proportion to the percent of the population that’s at risk for being exposed to these amplifier settings (see the Figure below, where MDR TB refers to multi-drug resistant TB; we used a stochastic model here to estimate the percent of time that a new drug-resistant strain will propagate successfully and emerge into an epidemic in the overall population after it starts out in a single person in the amplifier). So to really tackle these kinds of epidemics, treatment alone isn’t sufficient. It’s necessary to actually decrease people’s exposure to the risky environment and prevent infections in the first place—a dramatic shift from current global TB control policy.

Most of the ideas about what to do with the mining-TB problem have come out of the Aids and Rights Alliance of Southern Africa (ARASA), an advocacy group for patients. Having interviewed the major stakeholders from government regulators to mine workers to corporate directors, the group came up with a core set of recommendations about what to do next with the threat of TB among miners. The list included three main sets of recommendations that we assembled in a broad political analysis:

(1) Healthcare recommendations: First, because mines provide tertiary care for accidents and injuries but weak primary care, the mining companies needed to set up and monitor facilities where mine workers can be adequately treated to prevent drug-resistant TB from propagating, and evaluate and improve standards of living and safety on mines, particularly in barracks and mining shafts where the risk of infection is greatest. They also needed to address continuity of care as migrants travel back and forth across borders, providing a standardized set of patient charts with TB treatment histories for doctors to review, coordinating with clinics across the borders to continue therapy during migration, and working with the mineworkers’ unions to monitor adherence.

(2) Recruitment: Contracts for the mine workers appear to contain major loopholes, and are signed by workers with limited literacy. The workers’ unions need to supervise and translate some of the contract language for workers, and push to include health benefits and basic protections for contract employees (the seasonal workers who are not always included in formal payrolls). Providing family housing is a mitigating step to controlling HIV, so that male miners aren’t far from their wives as they migrate back-and-forth from far rural regions.

(3) Compensation: There are a number of formal and informal barriers to miners’ access to compensation during sickness, ranging from bullying to a massively-bureaucratic process for making claims (in some cases, requiring tissue samples to be extracted and analyzed by a company pathologist). Avoiding these unnecessary obstacles and empowering the government agency charged to oversee the processing of claims could help miners facing critical healthcare costs that current limit their ability to pursue or continue medical treatment.

Of course, the broadest reform angle is to diminish dangerous work in the mining sector and allow for other opportunities to engage in safer work. But how likely are these and other initiatives to be undertaken? Thanks to the work of several activists and public health physicians, progress looks likely. The studies on mining and TB hit The New York Times. And at a subsequent meeting of top officials, South Africa’s Minister of Health began a speech declaring: “If TB and HIV are a snake in Southern Africa, the head of the snake is here in South Africa. People come from all over SADC [Southern African Development Community] to work in our mines and export TB and HIV, along with their earnings. If we want to kill a snake, we need to hit it on its head.” Subsequently, a parliamentary committee on health opened investigation into the mining sector and its impact on TB.

For the first time, the mining sector is being made to undergo rigorous data collection on disease burden and treatment outcomes, as well as answer directly to public authorities on the matter. As ARASA and others monitor the situation, the choice seems clear: whether to continue deregulating an industry that has been strongly associated with major epidemics throughout the population, or to think creatively of an alternative African reality that doesn’t involve the traumas and risks of migration and mining.

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s