Intended for healthcare professionals


Drug resistant tuberculosis in prisons in Azerbaijan: case study

BMJ 1998; 316 doi: (Published 09 May 1998) Cite this as: BMJ 1998;316:1423
  1. R Coninx (rconinx{at}, medical coordinator for training and evaluationa,
  2. G E Pfyffer, headb,
  3. C Mathieu, programme officer in chargec,
  4. D Savina, nursec,
  5. M Debacker, laboratory technicianc,
  6. F Jafarov, head of tuberculosis wardd,
  7. I Jabrailov, doctor for tuberculosis programmed,
  8. A Ismailov, programme officerc,
  9. F Mirzoev, programme officerc,
  10. R de Haller, chairman of health commissiona,
  11. F Portaels, head of mycobacteriologye
  1. a International Committee of the Red Cross, CH-1202 Geneva, Switzerland
  2. b Swiss National Centre for Mycobacteria, Department of Medical Microbiology, University of Zurich, CH-8028 Zurich, Switzerland
  3. c International Committee of the Red Cross, Tuberculosis Program, Baku, Azerbaijan
  4. d Central Penitentiary Hospital, Baku, Azerbaijan
  5. e Institute of Tropical Medicine, Department of Microbiology, Antwerp, Belgium
  1. Correspondence to: Dr Coninx
  • Accepted 5 February 1998


Objectives: To document the existence of drug resistance in a tuberculosis treatment programme that adheres strictly to the DOTS principles (directly observed treatment, short course) and to determine the extent of drug resistance in a prison setting in one of the republics of the former Soviet Union.

Design: Case study.

Setting: Central Penitentiary Hospital in Baku, the referral centre for tuberculosis patients from all prisons in Azerbaijan.

Subjects: Prisoners with tuberculosis: 28 selected patients not responding clinically or bacteriologically to the standard treatment (group 1) and 38 consecutive patients at admission to the programme (group 2).

Main outcome measures: Drug resistance of Mycobacterium tuberculosis strains grown from sputum.

Results: All the non-responding patients (group 1) had strains resistant to at least one drug. 25 (89%) of the non-responding patients and nine (24%) of the consecutive patients had M tuberculosis strains resistant to both rifampicin and isoniazid. A further 17 patients in group 2 had strains resistant to one or more first line drugs.

Conclusions: Drug resistant M tuberculosis strains are common in prisons in Azerbaijan. Tuberculosis problems tend to be worse in prisons, but prisoners and former prisoners may have an important role in the transmission of tuberculosis, particularly of drug resistant forms, in the community. National programmes to control tuberculosis will have to take into account and address the problems in prisons to ensure their success.

Key messages

  • Tuberculosis is an important problem in prisons in Azerbaijan

  • Multidrug resistant tuberculosis was common and an important cause of non-response to standard treatment

  • National tuberculosis control programmes must include prisons and take account of drug resistance

  • Unless WHO recommended treatment protocols are followed the problem of multidrug resistant tuberculosis may result in untreatable tuberculosis which will spread to the general community


Tuberculosis is an increasing problem in Azerbaijan. 1 2 In prisons this problem is compounded by overcrowding, poor general health, high prevalence of risk groups, late case finding, and incomplete treatments. As a result prevalences in prisons in Azerbaijan were almost 50 times higher than the country's average and the mortality was as high as 24% in 1994.3 The common practice of self treatment with drugs provided by family members often results in treatment at insufficient doses and with multiple interruptions. The incidence of HIV infection in the general population is low,4 and there is no evidence that HIV plays an important part in the current epidemic of tuberculosis in prisons. Drug resistance is known to exist in the countries of the former Soviet Union, but exact data have rarely been presented. 5 6 No data are available for Azerbaijan.

The International Committee of the Red Cross has implemented a tuberculosis control programme in the Central Penitentiary Hospital in Baku, Azerbaijan, which is the country's only treatment centre for patients with tuberculosis in prison. The programme started in June 1995 and used a short course of directly observed treatment (DOTS) as recommended by the World Health Organisation7 and the International Union Against Tuberculosis and Lung Diseases.8 The objective of this study was to document the existence and the extent of multidrug resistance.

Subjects and methods

We analysed sputum from two groups of patients in whom Mycobacterium tuberculosis was identified. Group 1 comprised 28 patients who did not respond, either clinically or bacteriologically, after a minimum of eight weeks to the fully supervised WHO treatment regimen. Group 2 comprised 38 patients consecutively enrolled over four weeks from whom sputum was taken before the start of the treatment. The two groups were mutually exclusive. One sputum sample was taken from each patient. We supervised collection of sputum to avoid irregularities.

Treatment protocols

We included only patients whose sentence was long enough to allow a full course of treatment. Patients received the WHO recommended DOTS regimen in the recommended doses.9 For new cases, defined as having never been treated or treated for less than a month, treatment comprised daily doses of rifampicin, isoniazid, pyrazinamide, and streptomycin for two months followed by rifampicin and isoniazid thrice weekly for four months. For failure and relapse cases initial treatment was daily doses of rifampicin, isoniazid, pyrazinamide, and ethambutol for three months plus streptomycin for the first two months followed by rifampicin, isoniazid, and ethambutol thrice weekly for five months. Failure of treatment was defined as patients who had stopped treatment after having taken drugs for at least one month. Relapse cases were patients who had previously completed a full treatment course. Medicines were bought from the International Dispensary Association in Amsterdam and imported to ensure compliance with international standards. Drug intake was directly supervised by our staff, who observed the patients swallowing the medicines. This allowed us to achieve a compliance rate of 94%.

Microbiological analyses

Microscopy —Smears were stained with the Ziehl-Neelsen technique.10

Culture —Sputum specimens were collected in the morning, stored at 4°C, and taken in batches to the laboratories for processing. They were decontaminated by the Petroff method or by N-acetyl-L-cysteine and sodium hydroxide,10 inoculated on to solid media, and incubated at 36°C for eight weeks in 6% carbon dioxide. In addition, 0.5 ml of the sediment was cultivated by the radiometric BACTEC 460 TB technique.11

Identification of M tuberculosis —Routine biochemical methods10 and Accuprobe culture confirmation kits (Gen-Probe, San Diego, CA, USA) were used to identify the isolates.

Susceptibility testing Susceptibility to first line drugs (isoniazid, rifampicin, ethambutol, and pyrazinamide) and to streptomycin was tested with the radiometric BACTEC 460 TB method (isoniazid 0.2 µg/ml; rifampicin 2.0 µg/ml; ethambutol 2.5 µg/ml; pyrazinamide 100 µg/ml; streptomycin 2 µg/ml) as well as by the proportion method on solid medium. Resistance was defined according standard protocols.12


M tuberculosis was isolated from all sputum specimens (n=66). The table shows the antibiotic resistance pattern of the strains. In group 1, 25 of the strains (89%) proved to be multidrug resistant (resistant to rifampicin and isoniazid). None was fully susceptible. The strains of all new cases and 14 out of 17 failure or relapse cases (82%) were multidrug resistant.

Susceptibility patterns of M tuberculosis strains among non-responding cases (group 1) and consecutive cases (group 2)

View this table:

In group 2, nine strains (24%) were multidrug resistant and only 12 (32%) were fully susceptible. Three (15%) strains in the 20 new cases and six (33%) strains in the 18 failure and relapse cases were multidrug resistant.


Our data show that drug resistance is common in the Central Penitentiary Hospital in Baku and an important factor in non-response to treatment. Although these data cannot be generalised because of the limited number of samples investigated, they have implications for the management of tuberculosis in prisons. Treatment practices in Azerbaijan need to be reviewed as frequent monotherapy and multiple interrupted treatment courses undoubtedly contribute to the widespread multidrug resistance problem.

Our results suggest that the new cases were not really new or that patients became infected with drug resistant strains. We used the WHO definition of new cases, and many patients had taken antituberculous drugs for up to four weeks. It may be more appropriate to restrict the definition of new cases to patients without any history of treatment for tuberculosis.

The epidemiological implications of tuberculosis in prisons in the republics of the former Soviet Union may be more serious than commonly assumed. For example, data from Mariinsk (Siberia) indicate that a third of patients with tuberculosis have been in prison.13 A similar trend is seen in Ivanovo, Russia (A Khomenko, personal communication). These data suggest that prisoners and former prisoners may in future have an important role in the transmission of tuberculosis, particularly of multidrug resistant forms, in the community.

National tuberculosis programmes will have to take into account the problems currently existing in prisons. The extent of tuberculosis in prisons, which is much greater than in the community, is underreported. Statistics on tuberculosis in the republics of the former Soviet Union are provided by the Ministry of Health and do not include data from prisons, which are under the jurisdiction of the Ministry of Justice (or in some countries the Ministry of Interior). Organisations running tuberculosis control programmes in prisons need to determine the presence and extent of multidrug resistance before starting a project.


We thank the staff of Baku prison and the Ministry of Justice in Azerbaijan, whose efforts to tackle the problem of tuberculosis in prisons led to the establishment of the current DOTS programme and who continue to provide all necessary support.

Contributors: RC initiated and coordinated the project, participated in discussion and analysis of the data, wrote the paper, and is the guarantor. GEP guided the laboratory analyses, participated in the discussion and analysis of the data and helped write and revise the paper. CM participated in the design of the project, was in charge of the patient management and sputum collection, and participated in discussion and analysis of the data and writing the paper. DS participated in the patient management, sputum collection, and discussions. MD performed the laboratory examinations in Baku, prepared the sputum specimens, and participated in discussion and analysis of the data. FJ was in charge of patient management, IJ participated in the patient management. AI helped with laboratory examinations and sputum collection. FM helped with patient management and data collection. RdeH initiated the project and participated in the discussion and analysis of the data and in writing the paper. FP participated in the analysis of specimens, discussion and analysis of the data, and writing the paper.

Funding: The Damien Foundation (Brussels) partly funded the laboratory investigations in Antwerp and the Zurich Lung Association those in Zurich.

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
View Abstract