Prevalence of pulmonary tuberculosis and associated factors among prisoners in Western Oromia, Ethiopia: A cross-sectional study

Background Prisoners are a disproportionately at high risk for tuberculosis. This is because; prisons represent dynamic communities where at-risk groups congregate. It increases the transmission rate because of overcrowding and living together with infected individuals. This study was done to determine the prevalence of pulmonary tuberculosis and associated factors among prisoners of Western Oromia, Ethiopia in 2017. Methods A cross-sectional study was conducted among prisoners who have a history of cough for two weeks or more. Data were collected from 270 participants and sputum sample was collected from 249 prisoners and analyzed in GeneXpert for having pulmonary tuberculosis. Logistic regression analysis was used to identify factors associated with the development of pulmonary tuberculosis among prisoners. Results The overall prevalence among suspected cases was (15.6%; 95% CI (11.5, 20)) which makes the point prevalence of pulmonary tuberculosis were 744 per 100,000 of prisoners. Prisoners who had history of cigarette smoking before imprisonment (AOR=3.55; 95% CI (1.29, 9.78)), contacted with known TB patient (AOR=5.63; 95% CI (2.19, 14.41)), share prison cell with TB patients (AOR=3.51; 95% CI (1.34, 9.19)) and Body Mass Index <18.5kg/m2 (AOR=8.87; 95% CI (3.23, 24.37)) were more likely to have pulmonary tuberculosis. Conclusion A higher prevalence of pulmonary tuberculosis was observed among prisoners in the three prisons of Wollega Zones. To avert this problem, screening of prisoners should be done at the entry and separation of inmates with symptoms of tuberculosis should be done.

1 Introduction 2 Tuberculosis (TB) is an airborne infectious disease caused by bacillus mycobacterium tuberculosis. It 3 typically affects the lungs (pulmonary Tuberculosis), but it can also affect other sites (extra pulmonary 4 Tuberculosis). The disease spreads in the air when sick people expel the bacteria while talking, coughing, singing, 5 sneezing and spitting [1].

6
About 95% of all cases and 99 % of deaths occur in low and middle-income countries. Twenty-two 7 countries contribute to 99% of the world's TB burden. Besides this, household costs of TB are substantial estimates 8 suggest that tuberculosis costs the average patient three or three months of lost earnings, which can represent up to 9 30 percent of annual household income [2, 3].

10
TB mostly affects socially marginalized and other poor high-risk groups such as prisoners, intravenous 11 drug users, migrants, and poor socio-economic groups. Prison inmates constitute a high risk-group for tuberculosis 12 (TB) in both developing and industrialized countries [4]. In terms of cases, the best estimates for 2015 are that there 13 were 10.4 million new TB cases (including 1.2 million among HIV-positive people), of which 5.9 million were 14 among men, 3.5 million among women and 1.0 million among children. Overall, 90% of cases were adults and 10% 15 of children [5].  where TB infection and transmission are higher [6].

3
In addition to overcrowding, long prison stays, low Body Mass Index (BMI), previous TB treatment, loss 4 of appetite, poor nutrition, and HIV infection have been documented as risk factors for TB [7,8]. Because of these, 5 prison is also specifically taken as a risky place for the transmission of TB in low and middle-income countries. Its 6 prevalence is estimated to be about ten to a hundredfold in the prison than the general population [9, 10].

7
Studies that indicated the prevalence of TB in the prison have reported higher prevalence in Ethiopia. The 8 study was done in Southern Ethiopia, Gamo Gofa prison, which indicated about 3.2% prevalence of TB among 9 inmates. Another study was done in the Hadiya Zone; south Ethiopia also showed a three times higher prevalence of 10 TB than in the general population [11,12]. But there is limited evidence regarding the TB prevalence in Western

12
This study generated information on the current prevalence of TB among prisoners in Wollega Zones, 13 western Ethiopia. The study was also aimed to identify factors associated with TB infection in prisons. This will 14 help TB programmers and authorities in re-designing the existing programs for TB prevention in prisons. Thus,

15
there are needs to accurately define specific factors driving TB and the prevalence of the problem among prisoners 16 of Wollega Zones.

24
Participants of the study were selected based on their history of cough of two weeks and above before data 1 Beside this, PTB patients who were taking anti-TB treatment were included in the study. HIV positive inmates with 2 cough of any duration were also included in this study. However, prisoners who were unable to produce sputum and 3 unable to communicate during data collection were excluded. Then, all prisoners with cough of two weeks or more 4 were screened for having pulmonary TB through mass screening strategy.

5
Data were collected by using interviewer administered questionnaire that was initially adapted from 6 previous studies [14]. The questionnaire was prepared in English and translated into 'Afaan Oromo' (regional 7 working language) for data collection. It has sections on socio-demographics, imprisonment history and morbidity 8 history of prisoners. Data was collected by six nurses and three laboratory technologists. One nurse and one 9 laboratory technologist supervised the overall data collection process.

10
Before data collection, prisons' health committees and health professionals working in prisons were 11 oriented on the purpose of the research. Then, all inmates were registered with support of the health committee 12 members in each prison.

Sputum Collection and transportation 14
A single sputum specimen is recommended for GeneXpert. The sputum was collected outside the 15 laboratory or clinics in a well-ventilated space. The prisoners were instructed to produce sputum through coughing 16 to get sputum from lower respiratory organs. Respondents were also asked to wash their mouth using clean water 17 before the sputum was taken.

18
Then, they were asked to inhale deeply 2-3 times and breathe out strongly each time cough deeply from 19 chest to produce sputum. Sputum specimens were collected in "Falcon tube" with 30-50 ml capacity. This was 20 because it was translucent and has walls that allow easy labeling. Then, the sputum was transported using triple 21 package to hospital for laboratory to be evaluated. Appropriate procedure of using the GeneXpert test was followed 22 during laboratory test [15].

Weight and height measurements 24
Body weight was determined to the nearest 0.1kg on weight scale and height was measured to the nearest

15
Based on this, variables with p-value <0.25 were entered to multivariable logistic regression to identify the factors 16 that affect pulmonary tuberculosis. Odds ratio and corresponding 95% confidence intervals were used to quantify 17 the degrees of association between independent variables and PTB status. All associations with p-value ≤0.05 were 18 considered as being statistically significant.

19
Operational definitions 20 TB Positive screen: identifying a person with symptoms and findings consistent with tuberculosis involves 21 screening of patients with particular attention to cough of two weeks or more duration.

22
Suggestive symptoms: that help to identify presumptive TB include fever, night sweating, and weight loss.
1 New cases of TB: those patients who have never been diagnosed for TB before but diagnosed by GeneXpert during 2 data collection.

3
Existing TB case: refers to a patient who had been diagnosed for TB and receiving anti-TB drugs during data 4 collection.

5
Malnutrition: a person with Body Mass Index (BMI) less than18.5kg/m2.   Dambi Dollo Hospital to get permission for the study and material support. Before data collection written consent 16 was taken from study participants. Consent was obtained from a parent or guardian on behalf of any participants 17 under the age of 18. Confidentiality was insured and maintained by the investigators and data collectors.

5
This study showed that the prevalence of Pulmonary TB in prison is about 744 cases per 100,000 prisoners.

6
Furthermore, the magnitude of TB among suspects was 15.6 % (95% CI (11.5, 20)). This finding indicated higher 7 prevalence when compared to the southern Ethiopia, Gamo Gofa which has prevalence 623 per 100,000 prisoners. It 8 is also higher than the study done in Eastern Ethiopia (9%) and North Gondar prison (10.4%).

1
The observed prevalence is 7 times higher than in general population. New cases of TB were also observed 2 during data collection which indicated undiagnosed TB in the prisons which are source of infection for other 3 prisoners. Because of this, there is a need for regular screening of TB during intake on routine basis. The use of 4 GeneXpert may also contribute for better case detection due to that GeneXpert is more accurate and reliable than 5 sputum smear microscopy in predicting pulmonary TB [17].

6
Smoking was significant risk factor of PTB in the current study. This finding is similar with the study 7 conducted in Tanzania   Availability of data and materials 2 All the data and materials are available with the authors.

Competing interests 4
The authors declare that they have no competing interest.

5
Authors' contributions 6 KE has been involved in conception, writing the study protocol, formulating the study design, and training of data 7 collectors, data entry, analysis and interpretation of data. ZD participated in design, interpretation of data, reviewing 8 intellectual content; supervise overall process of the process and manuscript preparation. BE was part of data quality 9 check, providing important comments, supervise overall process and review manuscript.

12
Hospital and West Wollega Zone Health Office for material support. We are also grateful to the data collectors,

13
supervisors and study participants for their voluntarily participation. 14 15 16