Multidrug resistant tuberculosis in prisons located in former Soviet countries: A systematic review

Background A systematic literature review was performed to investigate the occurrence of multidrug-resistant tuberculosis (MDR TB) in prisons located in countries formerly part of the Soviet Union. Methods A systematic search of published studies reporting MDR TB occurrence in prisons located in former Soviet countries was conducted by probing PubMed and Cumulative Index Nursing and Allied Health Literature for articles that met predetermined inclusion criteria. Results Seventeen studies were identified for systematic review. Studies were conducted in six different countries. Overall, prevalence of MDR TB among prisoners varied greatly between studies. Our findings suggest a high prevalence of MDR TB in prisons of Post-Soviet states with percentages as high as 16 times more than the worldwide prevalence estimated by the WHO in 2014. Conclusion All studies suggested a high prevalence of MDR TB in prison populations in Post-Soviet states.


Introduction
It is estimated that as of 2014, there were 480,000 cases of MDR TB worldwide [1]. In 2015, of the 27 countries worldwide designated as having a high burden for multidrug-resistant tuberculosis, TB, multi drug-resistant, prison, jail, incarcerated, inmate, penitentiary, and the individual names of all the countries that are defined as Post-Soviet. Studies were included if they were original research, published in a peer-reviewed scientific journal, were in the English language, focused on one or more Post-Soviet countries, included prisoners as the study population, reported prevalence of MDR TB among prisoners, and used data that were collected between 1992 and 2015.
A total of 239 potentially eligible studies were initially identified. We subsequently excluded 225 studies for the following reasons: 71 did not focus on a Post-Soviet country, 48 did not have information on MDR TB prevalence in prisons 37 were reviews or guidelines, 30 were reports, 28 did not include prisoners in the study population, 8 included a study period before 1992, 2 were excluded because they were court cases, and 1 was excluded because it used the same data set as a previously included study. Three additional studies that matched all of the inclusion and none of the exclusion criteria were brought in through looking at sources of other review papers. Seventeen studies were included for review. Two researchers (AJ and AMJ) independently assessed the methodological integrity of each included study.
Five studies reported adjusted odds ratios (AOR) using multivariate analysis [25,26,30,33,37]. While one study in Russia reported that those with MDR TB had a 4 times greater odds of being a prisoner [30], a study done in Georgia found no association between MDR TB and prisoners [33]. The same study in Georgia found that older age was associated with decreased odds of MDR TB, while another study in Georgia found the odds of having MDR TB increased with age [25,33].
Prisoners with MDR TB had greater odds of being overweight or obese (body mass index [BMI] > 25kg/m 2 ) compared to those with drug-susceptible TB. However, the same study found individuals with drug-susceptible TB were at decreased odds of being overweight or obese. No association between MDR TB and a BMI under 20 kg/m 2 was reported, but prisoners with drug-susceptible TB had increased odds of having a BMI under 20 kg/m 2 compared to those with MDR TB [25].
The only study that examined recreational drug users found an association between illicit drug use and MDR TB. However, those with MDR TB were at decreased odds of having HIV [26].
Every study that looked at current or previous TB treatment as a risk factor found significant associations between treatment and MDR TB. Two studies found that those with MDR TB had 2.7 times greater odds of having had previous treatment [25,26]. Another 2 studies found that those with MDR TB had over 5 times greater odds of having had treatment failure  [30,37]. One study that followed a cohort of 18 prisoners through TB treatment found that 94% (17) developed MDR TB [38]. The major findings for each study are presented in Table 2.

Discussion
The goal of this study was to investigate the prevalence of MDR TB among prisoners of Post-Soviet states. We used systematic review to analyze studies and extrapolate results. Our findings suggest a high prevalence of multidrug-resistance in prisons of Post-Soviet states with percentages as high as 16 times higher than the worldwide prevalence estimated by the WHO in 2014 [1,36]. While there were no common risk factors among studies reporting adjusted odds ratios [25,26,30,33,37], all studies suggested a high prevalence of MDR TB in prison populations in Post-Soviet states. Within the Soviet Union, even after the advent of antibiotics for TB, the standard procedures for treatment were diagnosis by X-rays, admittance to a sanitarium, isolation, and occasionally surgery [42]. This treatment was, in fact, effective as TB mortality rates in the Soviet Union decreased from 400 per 100,000 persons in the 1910s to 17.3 per 100,000 men and 1.9 per 100,000 women in 1990 [43,44].
In 1991, the Soviet Union dissolved and the entire interconnected health, economic, and prison system collapsed with it. Suddenly millions of people found themselves living in poverty in newly formed nations [45]. Between 1991 and 1998, TB incidence in most Post-Soviet states more than doubled [46]. Additionally, during this time period, both crime and incarceration rates increased in the region as arrests tripled between 1988 and 1995 [20]. The gulag prison work camps of the former Soviet Union were repurposed to serve as prisons for the newly formed counties in which they were located. These prisons were not only over-crowded but also characterized by inadequate ventilation and malnutrition [20]. Such conditions resulted in high transmission of TB [7][8][9][10]. Between 1991 and 1997 the mortality rate in Russian prisons more than doubled with approximately half of these deaths being attributed to TB [47] and by 1999, one-third of all TB cases in Russia were in prisons [48].
Following the collapse in 1991, many of the newly formed nations started to implement the direct observed therapy short course (DOTS) system as recommended by the WHO. However, among physicians of the Post-Soviet states there was strong opposition to this move as they felt it threatened their livelihoods [49]. Conflicting instructions on how to best treat TB during the early 1990s likely did little to help the situation as isoniazid, rifampin and other TB medications started to enter the market. Partly because of stigmas associated with undergoing DOTS, TB medications took on value and were sold in open air markets, without prescription from pharmacists trying to make ends meet, and traded among inmates in prisons [49,50].
By 1998, it was estimated 20% of all TB cases in Russian prisons were multidrug-resistant [51]. Our findings suggest the prevalence of MDR TB in Russian prisons has since increased, and in some cases, more than doubled [26][27][28]30,31,35,37,38,40,41]. In Azerbaijan, Moldova, and Kazakhstan the proportion of TB among prisoners that was multidrug-resistant was found to be around 52%, 65% and 81% respectively [29,32,36]. Georgia and Kyrgyzstan, although similar in terms of incarceration rates to Azerbaijan, Moldova and Kazakhstan, had the lowest prevalence of multidrug-resistance with a reported prison prevalence of 18.1% and 26.8%, respectively [19,33,39].
• Prisoners were nearly twice as likely to have MDR TB compared to the general population (Risk Ratio = 1.9 [95% CI 1.1 to 3.2]).
• Among the study population, those with MDR TB had 0.7 times lower odds of having an HIV infection compared to those who had drug susceptible TB.
• Among the study population, those with MDR TB had 2.8 times greater odds of currently taking medications for TB compared to those with drug susceptible TB.
• Among the study population, those with MDR TB had 3.0 times greater odds of being recreational drug users compared to those with drug susceptible TB. Among the countries in this study, lower incarceration rates were not associated with lower MDR TB. The studies from Russia, which had one of the highest rates of incarceration in the world at the beginning of each study period, had a more than 20% lower mean prevalence than Moldova which had one of the lowest rates of incarceration among all the countries in this study [19,[26][27][28][30][31][32]35,37,38,40,41], Georgia, which had one of the highest rates of incarceration and one of the highest percentages of pre-trial detention in the world [19], had the lowest prevalence of MDR TB in prisons of Post-Soviet states [25,33,34].
The use of FDC has been suggested to account for low MDR TB prevalence in Sub-Saharan Africa [52,53]. However, no information on FDC usage was mentioned in any of the studies [25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41]. Only one study suggested the implementation of FDC as a possible solution for decreasing MDR TB prevalence rate in the Samara Oblast of Russia. 26 Data on nutrition and daily caloric intake of prisoners in Post-Soviet states has not been found, however several studies have suggested that nutritional status is fairly standard among prisoners in much of the Post-Soviet region [20,21]. There also seems to be conflicting data regarding prison capacity and overcrowding, making it difficult to assess these findings [19,54].
Earlier studies have suggested MDR TB is inversely associated with GDP [55,56]. Our findings, however, did not show such a relationship. Regions with some of the highest GDPs also had the highest rates of MDR TB; whereas, regions with some of the lowest GDPs also had the lowest rates of MDR TB [25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41]57,58]. In fact, among the 3 Georgian studies conducted in 1997, 2001 and 2009, MDR TB seems to be increasing during this period even though during the same period the GDP almost tripled [25,33,34,58]. This may result from better detection and reporting in Post-Soviet states [59,60]. However, Georgia, in comparison to other Post-Soviet states included for review, has a relatively low rate of multidrug-resistance. Further investigation is needed to elucidate possible reasons why.
It is important for future research to focus on risk factors so that this epidemic could be better understood and addressed. With conflicting data and small sample sizes of publications, more studies are needed to create a more comprehensive landscape of MDR TB in Post-Soviet prisons. Notably, none of the studies reported a previous history of TB infection among prisoners. Many included previously treated cases, and indeed most of these did have MDR TB [25][26][27][28][30][31][32][33][34][35][36][37]41], but studies failed to differentiate between treatment failure cases and those that successfully completed treatment and then were re-infected. Information on the length of prison terms were lacking. Only one study looked at this potential association and found that prisoners who had MDR TB were at decreased odds of having spent a prison term between 2-4 years [25]. This may suggest prison terms less than 2 years and more than 4 years could correlate with MDR TB but further investigation is needed.
Malnutrition has been considered a problem in Post-Soviet prisons [20,21], but only one study included BMI and it found that being overweight was associated with MDR TB [25]. While being overweight is not an indicator of proper nutrition, this link needs to be investigated further. It is not clear if diets varying among countries could have an effect as nutrition has been shown to be a correlated with TB [13][14][15][16][17].
While the link between TB and HIV in prisons has been well established [58][59][60][61][62][63], the link between MDR TB and HIV is less clear. Studies have found conflicting results and one review focusing on prisons worldwide found that overall there was no significant association, although the relationship was variable based on region [64,65]. The HIV epidemic in the Post-Soviet states has been fueled by injection drug use [62,66,67]. Injection drug use has also been speculated to be associated with increased likelihood of becoming a prisoner as well as acquiring MDR TB [68], but only one of our studies looked at it as a factor [26]. It found that injection drug use was associated with MDR TB but also that those with MDR TB were at decreased odds of having HIV compared to those without MDR TB [26]. The other three studies we included in the review that looked at HIV in prisons found that all MDR TB patients were HIV negative [27,39,40]. As the current sample size of studies is too small to suggest an association, more studies are needed to examine the pathway between injection drug use, MDR TB, and HIV in Post-Soviet prisons.

Limitations
Our review did encounter some limitations. Because our inclusion criteria included studies published from 1992 to 2015, studies published after this timeframe would not have been identified. Additionally, although PubMed and CINAHL are extensive databases for health research, limiting our search to these two databases may have resulted in the exclusion of any potential studies not cataloged in either PubMed or CINAHIL.