Intended for healthcare professionals

Education And Debate

Pitfalls of tuberculosis programmes in prisons

BMJ 1997; 315 doi: (Published 29 November 1997) Cite this as: BMJ 1997;315:1447
  1. Hernán Reyesa, medical coordinator for detention related activities,
  2. Rudi Coninx, medical coordinator for traininga
  1. a International Committee of the Red Cross, Geneva


    Among its other activities, the International Committee of the Red Cross visits prisoners in countries all over the world, essentially in countries at war or affected by conflict. As part of its work aimed at ensuring that prisoners receive adequate care, it has had to deal with the issue of tuberculosis.

    Recent experience in countries of the former Soviet Union has given us an insight into how complicated the treatment of tuberculosis can be in prisons. There are pitfalls that must be avoided if the disease is to be treated in accordance with the directly observed treatment, short course (DOTS) strategy drawn up by the World Health Organisation and the International Union against Tuberculosis and Lung Disease.1

    Directly observed therapy is designed to ensure, by means of direct observation, that patients actually take their full course of treatment. Prisons are, however, particularly difficult environments for applying such a strategy. Prevalencies five to ten times the national average are not uncommon and can be up to 50 times the reported national average.2 3 Tuberculosis may be a, or even the, major cause of death in prisons in developing countries, with mortality rates as high as 24%.4 In the case of tuberculosis, it is better to do nothing than to do something badly—and failure to complete courses of treatment can have disastrous results, leading to the development of multidrug resistant strains of Mycobacterium tuberculosis.

    The problems described here apply essentially to countries where problems of low income are accompanied by a high prevalence of the disease. Prisons are full beyond capacity, with prisoners from impoverished unhealthy backgrounds living in an even unhealthier environment. Prison health services suffer serious shortcomings, and the internal violence of prisons also has its influence. The setting is perfect for tuberculosis to develop and thrive.

    Summary points

    An emerging issue in the care of prisoners in poor countries is treatment of tuberculosis

    Prisoners and prisons present difficulties for healthcare staff trying to implement directly observed short course therapy for tuberculosis

    These difficulties include poor conditions in the prisons, overcrowding, demoralised and underfunded prison health services, and the unofficial power structures that flourish in all prisons

    It is better not to implement tuberculosis treatment at all than to do so if courses of treatment cannot be completed since this simply encourages the development of multidrug resistant tuberculosis that is then transmitted from inside the prison to outside, as prisoners have contact with their families and are released

    This is an unacceptable situation. As prison commissioner Alexander Paterson said in the 1930s: “Men are sent to prison as punishment, not for punishment.”5 Contracting tuberculosis in prison is most certainly not part of a prisoner's sentence.

    Prisons are bad for tuberculosis

    Most prisoners come from underprivileged sectors of the general population. High risk factors for the disease, such as malnutrition, poor hygiene, inadequate living conditions, and generally poor health conditions are all present in their normal environment. These people are more likely to have contracted tuberculosis before their arrival in prison.

    Prisons as neglected institutions

    Health care in prisons is usually the responsibility of the ministry in charge of prisons and almost never the ministry of health. This explains why national statistics on tuberculosis seldom include figures on the prison population. Prisons are never regarded as a priority for health care. Even when tuberculosis inside prisons is recognised as a specific prison health hazard, often nothing is done about it.

    Prison health services have a duty to identify prisoners with the disease and to treat them. Prison health not being a priority, budget allocations are usually hopelessly insufficient and there is a general lack of interest in the issue. Diagnostic facilities for tuberculosis are often inadequate and not based on (recommended) sputum microscopy. Medical records may be scanty outside prison hospitals. Prison medical staff are poorly paid, poorly trained, and therefore often poorly motivated. The inefficiency of the medical services may also lead to delay in diagnosis and thereby late case finding, making treatment of patients more difficult. The overall result is in many cases therapeutic chaos.

    Medical services in remand prisons may be reluctant to diagnose and hence have to start treatment of a chronic disease such as tuberculosis for prisoners who may well be released. This should not be an issue, as any such patients should be referred to a national tuberculosis treatment programme outside the correctional system.

    Prison as a factor for contagion

    Prisons are usually overcrowded, sometimes extremely so, with poor hygiene and inadequate ventilation creating ideal conditions for airborne contagion. (Even where there are windows, prisoners in countries with cold winters may never open them because of lack of heating.)

    There is often no medical screening on admission, and prisoners contagious for tuberculosis may be put into a crowded cell with others. In the prisons of low income countries, malnourishment has a deleterious effect on prisoners who are most prone to disease in these ideal conditions for contagion: the weakest, the unhealthiest, the underdogs.

    Prisons as violent societies

    Prisoner culture obviously varies between countries, and prison populations are anything but homogeneous, even within a single establishment. One common denominator in all prisons, however, is the existence of power structures parallel to the official administration. In many cases this unofficial hierarchy is more powerful than the official authority, and prison administrations often condone these parallel systems as they help to maintain order. The type of prison society that results will of course depend on many factors (see box).

    Power structures in prisons in the former Soviet Union

    In the countries that made up the former Soviet Union a caste system exists in establishments for sentenced prisoners. Inmates are stratified into four groups. The “bosses” (blatniye) are the upper caste—professional criminals. Then comes the silent majority of “blokes” (muzhiki), non-professional criminals just in to serve their time, with no power. The third caste is the “collaborators” (kozly), who are shunned because they work for the prison administration. At the bottom are the “untouchables” (petukhi), the despised members of the prison society: homosexuals, sex offenders, outcasts from the other groups, and anyone who has contravened the unofficial laws of the prison hierarchy.

    The rules and laws of this unofficial hierarchy have direct implications for the management of tuberculosis, in terms of unfair selection of patients for treatment and of trafficking of medicines.

    Prisoners understand that tuberculosis is a dread disease, having seen fellow inmates die for lack of treatment. Newly introduced medicines, such as rifampicin, will be very desirable. The bosses will want to get hold of the pills for their monetary value. This will automatically create an incentive to be included in tuberculosis treatment programmes. Influential prisoners may thus try to get on a tuberculosis treatment programme whether they have tuberculosis or not. Poorly paid prison doctors may turn a blind eye to exchanges of sputum, after taking bribes from wealthy prisoners. They may even put pressure on laboratory technicians to find bacilli in negative sputum samples. In addition, patients lower down the hierarchy receiving medicines within a tuberculosis treatment programme may be pressurised by the bosses to deliver up their pills.

    Internal hierarchies can influence tuberculosis programmes in other ways. In prison hospital wards it is unrealistic to try to mix prisoners of different status in the same room. Any doctor who tries to put underdog patients in a room with other inmates will find that during the night the prisoners have sorted themselves out again according to the unofficial hierarchy.

    Moreover, individual prisoners in poor countries will also try to hoard pills for their own use. They may sell the medicines to the guards, give them to their relatives during family visits, use them as currency for gambling, or use them to pay their debts, rather than having to use more distasteful methods of obtaining ready cash (H Reyes, Corrections Health Service Conference, Sydney 1997).

    Education of patients, so essential normally, is often hopeless in prisons. Prisoners have more immediate worries than the dangers of not receiving a full course of treatment. Others may want to take their medicines but are prevented from doing so. The public health argument relating to the danger of creating multidrug resistance will probably have no impact on inmates incarcerated for long terms in a violent world.

    For all these reasons medical teams working in prisons have a hard time if they want to comply with the strict recommendations of directly observed therapy. Nurses will be faced every day with new tricks invented by prisoners to avoid taking all their pills (see box). Prisoners who have actually taken the treatment may try to substitute the sputum of an infectious patient for their own, so they can stay on the programme and continue to receive better food and more medicines.


    The International Committee of the Red Cross participated in a tuberculosis programme for prisoners in Peru in 1986, treating them with the same regime used by the national tuberculosis programme. The patients were both common law and security (Shining Path guerillas) prisoners. The guerillas turned out to be very cooperative: as soon as their internal, quasimilitary, hierarchy had understood the reasons for complying with treatment, clear orders were given to all members of the group, and their compliance rate was close to 100%.

    Among common law prisoners, however, the story was very different. Either individually, or under coercion from gang bosses, these patients invented stratagems so as not to swallow their pills and smuggle them back to their cells, from where they entered the commercial prison circuit. This turned out to be impossible to control, as local staff were often also under pressure to look the other way. It was here that we realised for the first time that management of a tuberculosis programme for prisoners was much more difficult than for outside populations.

    Conversely, and paradoxically, there may be disincentives for staying on a programme. This happens if a prisoner thinks that showing signs of still active tuberculosis somehow hinders his release. These patients will try to present negative sputums they have obtained from other prisoners. If their ploy works, and they are taken off the programme, they become automatic defaulters.

    Tuberculosis is bad for prisons

    The uncontrolled spread of tuberculosis is also bad for a prison. Apart from the medical implications for the patients themselves, other prisoners soon realise that the disease cannot be controlled, which could result in serious rioting and other security problems.

    If patients receive incomplete doses of medicines the conditions for the development of multidrug resistant strains of tuberculosis bacilli will be present. The results of the Red Cross study in Azerbaijan show that the problem of multidrug resistance already exists among prisoners entering the tuberculosis programme (R Coninx et al, unpublished data).

    Contrary to popular belief, prisons are not hermetically sealed institutions. Unlike their hosts, the prisoners, the tuberculosis bacilli thriving in a prison will not be contained by the prison bars. If untreated, infectious prisoners may infect their spouses and children during family visits, as well as prison guards. Moreover, if prisoners infectious with tuberculosis are freed, not having received a full course of treatment, and are not referred to an outside tuberculosis centre for follow up, this will also create a risk for transmission outside (see box).


    In January 1995 the Red Cross started supporting a tuberculosis programme in six prisons in Ethiopia. The prisoners, all common law, were mostly from impoverished, high risk backgrounds for tuberculosis. The prison environment, with severe overcrowding, poor hygiene and ventilation, was ideal for contagion. Directly observed therapy was implemented inside the prisons, supervision ensured by expatriate nurses.

    The programme had to be discontinued because of an unacceptably high number of defaulters, both within and without the prison system. Unfortunately the national tuberculosis programme for the general population was unable to provide treatment. Government decrees freed great numbers of prisoners, many of them under treatment for tuberculosis, who disappeared into the countryside. Even if they could have been traced, there was no medical structure to continue treatment. Furthermore, even within the prison system, many prisoners were transferred to other prisons without follow up being possible, or left the programme to “melt back” into the prison population, and medical staff were not able to trace them. The overall defaulter rate thus being high, up to 62% in Addis Ababa prison, it was decided at the end of 1996 to discontinue the programme after having completed full course treatment for all prisoners still on medication.

    The possibility that multidrug resistant strains of M tuberculosis that develop within a malfunctioning prison tuberculosis treatment programme will eventually spread to the outside community is not merely theoretical. In the Russian Federation, for example, there is evidence from tuberculosis control programmes in the community that a high proportion of patients have served time in prisons, and that having been in jail is a major risk factor for the development of multidrug resistant strains of M tuberculosis (A Khomenko, personal communication; Médecins sans Frontières, personal communication).

    For all these reasons, prisons must be included in national tuberculosis programmes; otherwise a major breeding ground for the disease will be overlooked, with possibly disastrous consequences for both prisoners and the general community.

    Could prisons be good for tuberculosis?

    If tuberculosis management is properly planned and implemented, and if all the adverse factors relating to the prison environment are effectively countered, there might be ways in which prisons could be an ideal environment for treatment. Tuberculosis might in this sense even be “good” for prisons, if the dangers inherent in the disease made health officials aware of the public health time bomb inside prisons. If concern for prison health and adequate funding resulted from tuberculosis, this would be achieved.


    Knowledge of the many pitfalls associated with tuberculosis programmes in prisons is essential if directly observed therapy is to succeed in curbing the disease in poor countries. Special measures must be taken to implement a prison tuberculosis programme. Prison health care should be integrated into the health priorities of ministries of health. Sufficient funding for prison medical facilities and decent salaries for staff should be provided to ensure adequate care. Prison conditions must be improved and the crucial issue of overcrowding addressed urgently. Neglecting to take such measures may result in disaster, as prisons are not only breeding grounds for the disease but also sources of transmission to the outside and could lead to an increase in the prevalence of tuberculosis in the general population. Finally, if the issue of erratic and incomplete treatments is not addressed urgently the incidence of multidrug resistant tuberculosis could rise in those countries that can least afford expensive second line treatments.

    The Baku Declaration and its origins

    The healthcare system in the former Soviet Union is in chaos following the collapse of the USSR in 1991. Assessements in the transcaucasian republics report nothing less than a collapse of national tuberculosis programmes, a paralysis of tuberculosis dispensaries, and a shortage of both diagnostic supplies and medicines.

    Diagnostic and treatment practices in the former Soviet Union do not follow the recommended WHO guidelines. Too many diagnoses are still done by radiology instead of sputum smear microscopy. Treatment regimens vary widely, and drug supplies are often interrupted.

    In most republics a bleak picture of multidrug resistance (resistance to both rifampicin and isoniazid) has started to emerge: up to 4% in Armenia, 7.3% in Russia (Ivanovo Oblast), 11.7% in Estonia, and 22.1% in Lithunania. Rates of resistance in patients who have received treatment, generally considered to reflect recent case management, reach 19.2% in Estonia, 54.5% in Latvia, and 27.3% in Russia.

    Throughout the world rates of tuberculosis in prisons are 10–50 times higher than in the general population. Even though HIV does not play a major role in the current epidemics in the former Soviet Union, mortality rates in prison are high and related to late diagnosis and erratic treatments. In some prisons mortality from tuberculosis is as high as 24%, and tuberculosis is the commonest cause of death, accounting for 50-80% of deaths.

    This situation goes largely unreported, because prisoners are not included in department of health statistics. The World Health Organisation, for example, estimates that there are 3550 cases of tuberculosis in Azerbaijan, but the International Committee of the Red Cross estimates that at least another 700 cases in prisons go unreported. In one study in Baku, Azerbaijan, 89% of the patients whose sputum did not convert after they had received first line antituberculosis treatment were found to have drug resistant strains of M tuberculosis. In consecutive patients admitted to the tuberculosis programme 24% of the strains were multidrug resistant.

    A workshop held in Baku, Azerbaijan, in July 1997 by the WHO and the International Committee of the Red Cross together with prison health authorities from Russia and Georgia and with medical non-governmental organisations working in the region highlighted this situation and led to the adoption of the Baku declaration.

    Baku Declaration

    We, the participants at the Baku Tuberculosis in Prisons meeting,

    Recognising that tuberculosis has become a major health threat to prisoners, and

    Observing that often incurable multidrug resistant forms of tuberculosis are increasing in prisons, and

    Further observing that the spread of HIV within prisons increases the risk of death from tuberculosis, and

    Noting that tuberculosis in prisons easily spreads into the community form infectious prisons and infectious prison staff, and

    Acknowledging that adequately funded and staffed prison health services are essential to address the problem of tuberculosis in prisons

    call upon

    • Governments, ministries of justice and interior and state security and health to work together towards providing prisoners with adequate health care and the means to cure tuberculosis, and Prison health services to implement DOTS (directly observed treatment, short course), and

    • Ministries of health to strengthen national tuberculosis programmes through the DOTS strategy

    and warn

    That if there is no response to our call for action, incurable tuberculosis will increase death among prisoners and their families and prison staff and their communities.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.