BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from
Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a multicenter case control study
ForJournal: peerBMJ Open review only Manuscript ID: bmjopen-2015-009058
Article Type: Research
Date Submitted by the Author: 11-Jun-2015
Complete List of Authors: Alemu, Yihun Mulugeta; University , Epidemiology Aweke, Werku; University, Epidemiology wilder-smith, Annelies; Heidelberg University,Institute of Public Health, International Health
Primary Subject Global health Heading:
Secondary Subject Heading: Epidemiology
Keywords: EPIDEMIOLOGY, PUBLIC HEALTH, Tuberculosis < INFECTIOUS DISEASES
http://bmjopen.bmj.com/ http://bmjopen.bmj.com/
on September 24, 2021 by guest. Protected copyright. on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a 6 7 multicenter case control study 8 9 Correspondence author; Name: Yihun Mulugeta, Alemu E: mail: [email protected] 10 11 Yihun Mulugeta Alemu 1, 2*, Worku Awoke2, Wilder Smith Annelies1 12 13 14 1Institute of Public Health, Heidelberg University, Germany 15 For peer review only 16 2School of Public Health, College of Medicine and Health Science, Bahir Dar University, 17 18 Ethiopia 19 20 21 3Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore 22 23 Determinants of tuberculosis in HIV infected adults 24 25 Key words: Determinants, tuberculosis, HIV, adults, multi-center 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 1 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 Summary Boxes 5 6 What is already known in this subject? 7 8 9 Previous studies to investigate the link have identified independent determinants for TB as 10 clinical (highly activated antiretroviral therapy, lower level of CD4 cell count, cotrimoxazole 11 12 preventive therapy and isoniazid preventive therapy) and life style (alcohol consumption). 13 14 However, determinants for TB among HIV infected adults are not well described in resource 15 For peer review only 16 limited settings. 17 What this study adds? 18 19 This study identified independent determinants for tuberculosis (highly activated antiretroviral 20 21 22 therapy, lower level of CD4 cell count, cotrimoxazole preventive therapy, isoniazid preventive 23 24 therapy and alcohol consumption) in HIV infected adults at resource limited settings. In addition 25 26 the study identified independent life style determinants for tuberculosis (Khat chewing and 27 28 29 tobacco smoking) in HIV infected adults. The independent association of life style variables 30 31 (Khat chewing and tobacco smoking) have never been identified by any published study in this 32
33 types of source population (HIV infected adults). Determinants for TB among HV infected adult http://bmjopen.bmj.com/ 34 35 36 are complex and encompass a wide range of factors. This study will contribute to a more 37 38 comprehensive & multidisciplinary approach for prevention and control of TB among HIV 39 40 infected adults in resource limited settings. 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 2 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 Abstract 5 6 Background: 7 8 Tuberculosis presents a major infectious diseases burden in resource limited settings. HIV is 9 10 driving the tuberculosis epidemic in many countries. Determinants for tuberculosis among HIV 11 12 13 infected patients is not well described in resource limited settings. 14 15 Objective: For peer review only 16 17 The objective of this study was to identify determinants for tuberculosis among HIV infected 18 19 adults in Northwest Ethiopia. 20 21 22 Methods: 23 24 A multicenter case control study was conducted in three hospitals and ten health centers. Cases 25 26 were HIV infected adults diagnosed with active tuberculosis and controls were HIV infected 27 adults without active tuberculosis. We recruited 150 cases and 296 controls. Interviewers 28 29 administered questionnaires and case report forms were used to systematically collect the data. 30 31 Results: 32
33 Being smokers (AOR (adjusted odds ratio) =5.47; 95%CI: 2.26, 13.22), presence of tuberculosis http://bmjopen.bmj.com/ 34 35 patient in the family (AOR=2.66; 95%CI: 1.25, 5.66), alcohol drinker (AOR=2.49; 95%CI: 1.29, 36 37 4.8) and chewing khat (AOR=2.22; 95%CI: 1.11, 4.41) were independent determinants for 38 tuberculosis. Highly active antiretroviral therapy (HAART) (AOR=0.25; 95%CI: 0.13, 0.51) had 39 40 a protective benefit against tuberculosis. 41 on September 24, 2021 by guest. Protected copyright. 42 Conclusion: 43 44 HIV infected adults who have substance use (tobacco smoking, khat chewing and alcohol) 45 46 should be prioritized for tuberculosis screening. This study reaffirmed that initiation of HAART 47 is one of the best strategies to reduce tuberculosis occurrence in HIV infected adults. Increased 48 49 coverage of isoniazid and cotrimoxazole preventive therapy have protective benefit against 50 51 active tuberculosis. Contact tracing of household contacts for tuberculosis should be intensified. 52 53 54 55 56 57 3 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 ARTICLE SUMMARY 4 5 Strength and limitation of the study 6 7 This is the first multicenter case control study in Northwest Ethiopia that identify determinants 8 9 for tuberculosis in HIV infected adults. 10 11 The study identified determinants for tuberculosis that will be important to prioritize TB 12 screening, treatment, prevention and control. 13 14 This study reaffirmed that initiation of HAART is one of the best strategies to reduce 15 For peer review only 16 tuberculosis among HIV infected adults in resource limited settings. 17 18 As we used a retrospective case control approach; temporal relationship could not be established; 19 the study design could not proof causation. 20 21 Ascertainment of tuberculosis patient among HIV infected adults might lack sensitivity and 22 23 specificity. 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 4 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 INTRODUCTION 4 5 The advent of HIV was a massive setback for the prevention and control of tuberculosis (TB) 6 7 1.TB is the leading cause of morbidity and mortality among people living with HIV (PLWHIV) 8 2 9 .There is a strong synergy between TB and HIV infection in HIV high burden countries 10 3 11 particularly in resource limited settings where the impact of both diseases are more significant . 12 One third of the world’s population is infected with TB 2. In 2013, an estimated 9.0 million 13 14 people developed TB; 1.1 million was HIV infected 4. Ethiopia ranks as the seventh TB burden 15 For5 peer review only 16 country in the world . 17 18 In resource limited settings the health care systems are overwhelmed by preventive, therapeutic 19 and diagnostic challenges of "HIV TB syndemic". In Ethiopia; the dual epidemics drained 20 21 resources and overburdened the already very limited health work force1. HIV increases the 22 6 23 lifetime risk for developing TB . However, HIV is not the only factor determines TB; various 24 25 other determinants contribute to TB HIV co infection. Studies to investigate the link have 26 identified determinants for TB as socio demographic7,8, clinical9,10, life style11,12 and 27 28 environmental13. However, determinants for TB among HIV infected adults are not well 29 30 described in resource limited settings. This study assessed determinants for TB among HIV 31 32 infected adults in Northwest Ethiopia.
33 http://bmjopen.bmj.com/ 34 METHODS 35 Study design 36 37 We conducted a multicenter case control study from May 12 to June 5, 2014 in Northwest 38 39 Ethiopia. All eligible governmental health institutions in ten provinces were included. Those 40 41 were Bahir Dar, Dangella, Kossober, Fnote Selam, Bure, Dur Betie, Deber Tabor, Wereta, Addis on September 24, 2021 by guest. Protected copyright. 42 Zemen and Nefas Mewcha. Three hospitals and ten health centers were included. 43 44 Participants 45 46 Cases were HIV infected adults diagnosed with active TB and on TB treatment during the data 47 48 collection period. Controls were HIV infected adults without active TB. The diagnosis of TB was 49 50 made according to the national guidelines for TB HIV. In Ethiopia, active TB diagnostic 51 guidelines include bacteriological methods; among those sputum microscopy is the mainstay 52 53 diagnostic tool. Culture, molecular, histopathological and radiological examinations are other 54 1 55 diagnostic methods for active TB . 56 57 5 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Sample size determination 4 5 The sample size was calculated using Epi Info version 7 software. Proportion of low CD4 cell 6 7 count 5.9% among the controls and 13.9% among cases 13. The level of significance 5%, power 8 9 of 80% and a case to control ratio of 1:2. By using the two proportion formula, the calculated 10 11 sample size was 144 for cases and 287 for controls; adding 10% for none response, the resulting 12 sample size was 474 (158 cases and 316 controls). The sample size was calculated for exposure 13 14 status of different variables: CPT, IPT, CD4 and hemoglobin. We took the largest sample among 15 For peer review only 16 these exposure variables. 17 18 Sampling procedures 19 Governmental hospitals and health centers in Northwest Ethiopia were assessed whether they 20 21 had adequate number of patients. Three hospitals and ten health centers in West Gojjam, Awie 22 23 and South Gondar Zone were found to be eligible and included in the study purposely to obtain 24 25 adequate sample. For the cases, HIV infected adults (age greater than 15 years) diagnosed with 26 active TB and for controls, HIV infected adults (age greater than 15 years) without active TB 27 28 were identified. Study subjects unable to give informed consent and suspected but unconfirmed 29 30 TB were excluded. All TB HIV co infected patients attending HIV care clinic and who were 31 32 taking TB treatment were included. Controls were sampled by systematic random sampling with
33 http://bmjopen.bmj.com/ 34 a sampling interval of five. 35 Data collection and analysis 36 37 Data were collected from two sources. Trained nurses who were in charge of HIV care clinic 38 39 conducted face to face interview with study participants by using structured questionnaires that 40 41 were prepared in English translated to Amharic and back translated and pretested for consistency on September 24, 2021 by guest. Protected copyright. 42 and easily understandable. The primary data were collected to assess: socio demographic 43 44 variables (age, sex, educational and marital status), host related variables (cigarette smoking, 45 46 khat chewing and alcohol consumption) and environmental determinants (floor the house, latrine 47 48 and separate kitchen). Trained nurses also collected data from patient’s record to assess: clinical 49 50 variables (HAART, CD4 cell count, CPT and IPT). Data collection procedures were supervised 51 by medical officers. 52 53 Data were entered into Epi info version 7 and analysis was done with SPSS version 20. 54 55 Frequencies and proportions were used to describe the study subjects in relation to relevant 56 57 6 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 variables. Odds ratio with its 95% confidence interval and p value were used to measure strength 4 5 of association and identify statistical significance result. Logistic regression models were applied 6 7 to assess the relationship between determinants and TB. Identification of confounding variables 8 9 by logistic regression models was applied. Multivariable analysis (backward stepwise) was used. 10 11 Hosmer Lemeshow test was applied and the fit of the model was checked. A poor fit of the 12 model if the significance P value is less than 0.05 and good fit greater than 0.05. Here in this 13 14 study the model was adequately fits the data hence P value was 0.34. 15 For peer review only 16 Ethical considerations 17 18 Ethical approval from Heidelberg ethical commission was obtained. Additionally, ethical 19 approval from the country of research, Amhara regional research and ethical core process, Bahir 20 21 Dar, Ethiopia was received. Written permission to conduct the study was obtained from each 22 23 health institution involved in the study. Since there were illiterate participants the data collectors 24 25 inform each study participant about the informed consent sheet. Informed oral consent was 26 obtained from each study participant. This consent procedure was approved by the ethical 27 28 commission. The data collectors document the participant’s consent on the informed consent 29 30 sheet. 31 32 RESULTS
33 http://bmjopen.bmj.com/ 34 Socio demographic and host related factors of HIV infected adults: 35 From the total of 474 eligible participants, 446 study participants responded and consented to 36 37 participate in the study; 150 cases (33.6%) and 296 controls (66.4%); making the overall 38 39 response rate 94.09 % ( 94.93% for cases and 93.67 % for controls). 40 41 The mean and inter quartile range (IQR) for the age of cases were 33.4 and 27 39 years, on September 24, 2021 by guest. Protected copyright. 42 respectively. The corresponding values for controls were 34.03 and 28 39 years. A higher 43 44 percentage of women was observed in both groups; 81 (54%) in cases and 188 (63.5%) in 45 46 controls (Table 1). 47 48 49 50 51 52 53 54 55 56 57 7 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Table 1. Socio demographic and host related factors of HIV infected adults in Northwest 4 5 Ethiopia, 2014 6 7 Variables Case n (%) Control n (%) Total n (%) 8 9 Sex 10 11 Male 69(46) 108(36.5) 177(39.7) 12 13 Female 81(54) 188(63.5) 269(60.3) 14 15 Education For peer review only 16 17 No formal education 46(30.7) 120(40.5) 166(37.2) 18 Primary 48(32) 78(26.4) 126(28.3) 19 20 Secondary 42(28) 60(20.2) 102(22.9) 21 22 Tertiary 14(9.3) 38(12.8) 52(11.6) 23 24 Marital status 25 26 Married 53(35.3) 141(47.6) 194(43.5) 27 Divorce / widowed 50(33.3) 98(33.1) 148(33.2) 28 29 Never married 47(31.3) 57(19.3) 104(23.3) 30 31 Smoking 32
33 Yes 25(16.7) 12(4) 37(8.3) http://bmjopen.bmj.com/ 34 35 No 125(83.3) 284(96) 409(91.7) 36 37 Khat chewing 38 Yes 44(29.3) 37(12.5) 39 40 No 106(70.7) 259(87.5) 365(82) 41 on September 24, 2021 by guest. Protected copyright. 42 Alcohol 43 44 Yes 49(32.7) 49(16.6) 98(22) 45 46 No 101(67.3) 247(83.4) 348(78) 47 Previous history of TB 48 49 Yes 41(27.3) 50(16.9) 91(20.4) 50 51 No 109(72.7) 246(83,1) 355(79.6) 52 53 54 55 56 57 8 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Bivariate Analysis: 4 5 The bivariate analysis showed that male patients (COR (crude odds ratio) =1.48; 95% CI: 0.99, 6 7 2.12; P value = 0.053) were marginally associated with tuberculosis compared to female 8 9 patients. Married patients (COR=0.45; 95%CI: 0.27, 0.75) were less likely to develop TB 10 11 compared with never married. Cases were more likely to be smokers (COR= 4.73; 95% CI: 2.3, 12 9.72), Khat chewer (COR=2.9; 95%CI: 1.77, 4.75) and alcohol drinker (COR= 2.44; 1.54, 3.86). 13 14 Study participants who had a history of TB (COR=1.85; 95%CI: 1.15, 2.96) were more likely to 15 For peer review only 16 develop TB. But educational status was not a determinant for TB (Table 2). 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 9 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Table 2: Socio demographic and host related determinants for developing tuberculosis among 4 5 HIV infected adults in Northwest Ethiopia, 2014 6 7 Variables Cases n (%) Controls n (%) COR 95%CI P value 8 9 Sex 10 11 Male 69(46) 108(36.5) 1.48 0.99, 2.12 0.053 12 13 Female 81(54) 188(63.5) 1 14 15 Education For peer review only 16 17 No formal education 46(30.7) 120(40.5) 1.04 0.51, 2.09 0.912 18 Primary 48(32) 78(26.4) 1.67 0.82, 3.39 0.157 19 20 Secondary 42(28) 60(20.2) 1.90 0.9,3.93 0.084 21 22 Tertiary 14(9.3) 38(12.8) 1 23 24 Marital status 25 26 Married 53(35.3) 141(47.6) 0.45 0.27, 0.75 0.002* 27 Divorce / widowed 50(33.3) 98(33.1) 0.61 0.37, 1.03 0.068 28 29 Never married 47(31.3) 57(19.3) 1 30 31 Smoking 32
33 Yes 25(16.7) 12(4) 4.73 2.30, 9.72 < 0.0001* http://bmjopen.bmj.com/ 34 35 No 125(83.3) 284(96) 1 36 37 Khat chewing 38 Yes 44(29.3) 37(12.5) 2.9 1.77, 4.75 < 0.0001* 39 40 No 106(70.7) 259(87.5) 1 41 on September 24, 2021 by guest. Protected copyright. 42 Alcohol 43 44 Yes 49(32.7) 49(16.6) 2.44 1.54, 3.86 < 0.0001* 45 46 No 101(67.3) 247(83.4) 1 47 Previous history of TB 48 49 Yes 41(27.3) 50(16.9) 1.85 1.15, 2.96 0.01* 50 51 No 109(72.7) 246(83,1) 1 52 53 54 55 56 57 10 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 HAART (highly active antiretroviral therapy) (COR=0.33; 95%CI: 0.19, 0.56), cotrimoxazole 4 5 preventive therapy (CPT) (COR=0.35 95%CI: 0.22, 0.53) and isoniazid preventive therapy (IPT) 6 7 (COR=0.30 95%CI: 0.18, 0.50) had protective benefit against TB. HIV infected adults with a 8 9 separate kitchen in the house (COR=0. 61; 95%CI; 0.40, 0.93) were less likely to develop TB. 10 11 HIV infected adults with presence of a TB patient in the family (COR=1. 99; 95%CI: 1.11, 3.57) 12 were more likely to develop TB. But floor of house and presence of latrine were not 13 14 determinants for developing TB (Table 3). 15 For peer review only 16 Table 3: Clinical and environmental determinants for developing tuberculosis among HIV 17 18 infected adults in Northwest Ethiopia, 2014 19 20 Variables Cases n (%) Controls n (%) COR 95%CI P value 21 22 HAART 23 24 Yes 114(76) 268(90.5) 0.33 0.19, 0.56 < 0.0001* 25 26 No 36(24) 28(9.5) 1 27 28 CPT 29 Yes 83(55.3) 231(78) 0.35 0.22, 0.53 < 0.0001* 30 31 No 67(44.7) 65(22) 1 32
33 INH prophylaxis http://bmjopen.bmj.com/ 34 35 Yes 22(14.7) 107(36.1) 0.30 0.18, 0.50 < 0.0001* 36 37 No 128(85.3) 189(63.9) 1 38 CD4 cell count (cells/ µl) 39 40 <=200 87(58) 61(20.6) 5.53 2.96, 10.35 < 0.0001* 41 on September 24, 2021 by guest. Protected copyright. 42 200 – 500 46(30.7) 169(57.1) 1.05 0.56, 1.97 0.86 43 44 >=500 17(11.3) 66 (22.3) 1 45 46 TB patient in the family 47 48 Yes 25(16.7) 27(9.1) 1.99 1.11, 3.57 0.021* 49 No 125(83.3) 269(90.9) 1 50 51 Separate kitchen 52 53 Yes 96(64) 220(74.3) 0.61 0.40, 0.93 0.024* 54 55 No 54(36) 76(25.7) 1 56 57 11 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Floor of the house 4 5 Mud / Soil 123(82) 240(81) 1.06 0.64, 1.76 0.814 6 7 Cement 27(18) 56(19) 1 8 9 Latrine 10 11 Yes 136(90.7) 267(90.2) 1.05 0.53, 2.06 0.875 12 13 No 14(9.3) 29(9.8) 1 14 Multivariable analysis: 15 For peer review only 16 In order to identify independent determinants for TB multiple logistic regression model was 17 18 used. After adjusting for possible confounders some variables were remained in the multivariable 19 20 model: chewing khat (AOR=2.22; 95%CI: 1.11, 4.41), being a smoker (AOR=5.47; 95%CI: 21 2.26, 13.22) and alcohol drinker (AOR=2.49; 1.29, 4.80) were independently determine 22 23 increased TB occurrence. However, HAART (AOR=0.25; 95%CI: 0.12, 0.51), CPT (AOR=0. 24 25 32; 95% CI: 0.19, 0.52) and IPT (AOR=0.22; 95%CI: 0.11, 0.42) had independent protective 26 27 benefit against TB. Subjects with separate kitchens in the house (AOR=0.48; 95%CI: 0.28, 0.83) 28 29 were less likely to develop TB. Patients who live with a TB patient in the household (AOR=2.66; 30 95%CI: 1.25, 5.66) were more likely to develop TB. Patients whose CD4 cell count < 200 31 32 cells/µl (AOR= 7.22; 95%CI: 3.39, 15.37) were more likely to develop TB. But sex, marital
33 http://bmjopen.bmj.com/ 34 status and a previous history of TB were not remained in the multivariable model (Table 4). 35 36 Table 4: Factors independently associated with tuberculosis among HIV infected adults in 37 38 Northwest Ethiopia, 2014 39 40 Variable COR(95%CI) P value AOR(95%CI) P value 41 on September 24, 2021 by guest. Protected copyright. 42 HAART 43 44 Yes 0.33(0.19, 0.56) < 0.0001 0.25(0.13, 0.51) < 0.0001 45 46 No 1 1 47 CPT 48 49 Yes 0.35(0.22, 0.53) < 0.0001 0.32(0.19, 0.55) < 0.0001 50 51 No 1 1 52 53 INH prophylaxis 54 55 Yes 0.30(0.18, 0.50) < 0.0001 0.22(0.11, 0.41) < 0.0001 56 57 12 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 No 1 1 4 5 Smoking 6 7 Yes 4.73(2.30, 9.72) < 0.0001 5.47(2.26, 13.22) < 0.0001 8 9 No 1 1 10 11 Khat 12 13 Yes 2.9(1.77, 4.75) <0.001 2.22(1.11, 4.41) 0.023 14 No 1 1 15 For peer review only 16 Alcohol 17 18 Yes 2.44(1.54, 3.86) <0.0001 2.49(1.29, 4.80) 0.006 19 20 No 1 1 21 22 TB patient in the family 23 Yes 1.99(1.11, 3.57) 0.021 2.66(1.25,5.66) 0.011 24 25 No 1 1 26 27 Separate kitchen 28 29 Yes 0.61(0.40, 0.93) 0.024 0.48(0.28, 0.83) 0.010 30 31 No 1 1 32 CD4 cell count (cells/ µl) 33 http://bmjopen.bmj.com/ 34 < 200 cells/ µl 5.53(2.96, 10.35) < 0.0001 7.22(3.39, 15.37) < 0.0001 35 36 200 – 500 cells/ µl 1.05(0.56,1.97) 0.86 1.31(0.63, 2.74) 0.461 37 38 >= 500 cells/ µl 1 1 39 40 DISCUSSION 41 on September 24, 2021 by guest. Protected copyright. 42 This study was shown that HIV positive adults on HAART, CPT, IPT and household members 43 with separate kitchen were less likely to develop TB. But determinants such as smoking, alcohol 44 45 consumption, Khat chewing, presence of TB patient in the family and CD4 count less than 200 46 47 cells were increased TB occurrence. 48 49 In this study HIV positive males were more likely to develop TB compared to women. Men are 50 more susceptible to tuberculosis infection due to genetic makeup of sex chromosome and 51 52 metabolic phenomena14. But the variable male sex did not remain in the multivariable model. 53 54 This could be due to higher prevalence of female in the source population. 55 56 57 13 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 In this study level of education was not a determinant for TB occurrence. It was consistent with 4 5 literature7. This might explain increased level of school training not bring successful behavioral 6 7 change to TB prevention and control in this source population. 8 9 In this study married HIV positive adults were less likely to develop TB compared to never 10 8 11 married. This was consistent with a study . Increased marriage prevalence observed in the 12 higher income class of the society15. Tuberculosis and poverty are interrelated 16. In this study 13 14 marriage was not remained in the final multivariable model. This could be due to increased rate 15 For peer review only 16 of divorce that reduced the proportion of married in the source population. 17 18 The statistically significant association of low CD4 count and TB in HIV patients was consistent 19 with literature7. Also our finding IPT had a protective benefit against TB was consistent with 20 21 published study17. Isoniazid reduces the mycobacterium load and reduce progression of latent 22 9 23 bacilli to active tuberculosis . The increase mycobacterial load coupled with progressive 24 18 25 impairment of mycobacterium specific T cell response . 26 In this study participants on HAART were less likely to develop TB. This was consistent with 27 28 publication10. This could be due to strong association between increasing HAART coverage and 29 30 decreased viral load11. Low level of viral load was linked with reduced occurrence of active 31 19 32 TB .
33 http://bmjopen.bmj.com/ 34 In this study patients on cotrimoxazole preventive therapy were associated with protective 35 13 36 benefit against tuberculosis. This was consistent with literature elsewhere . Cotrimoxazole 37 38 preventive therapy has beneficial effects on enhancing CD4 count as well as reduction of viral 39 40 20 41 load . on September 24, 2021 by guest. Protected copyright. 42 43 Cigarette smoking induces immune impairment and damage the ciliary clearance21,22. In this 44 45 study smoking was an independent determinants for TB. It was consistent with a previous 46 23 47 study . 48 Khat chewing is associated with immune modulation12. No published study has detected 49 50 independent association between khat chewing and TB among HIV infected adults. In this study 51 52 Khat chewing was an independent determinants for TB. 53 54 55 56 57 14 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 In our study subjects who consumed alcohol were more likely to develop TB consistent with 4 5 previous studies24 26. Alcohol can modulate the immune response. Long term alcohol consumers 6 7 have been shown to have impaired immunity 27. 8 9 In this study HIV infected adults who had past history of tuberculosis were more likely to 10 11 develop TB. But a past medical history of TB was shown to be a determinant for TB in bi 12 variable model only. This might be explained by the lower TB case detection rate in Ethiopia2. 13 14 The established tuberculosis cases detection mechanisms overlooked some tuberculosis patients. 15 For peer review only 16 A proportion of tuberculosis patient might remain undiagnosed i.e. died with other means of 17 18 diseases or died without visiting health care; the detected proportion of reoccurrence might be 19 low. Therefore the prevalence of reoccurrence was underestimated. 20 21 This study has the following potential limitations: as we used a retrospective case control 22 23 approach temporal relationship cannot be established; the study design could not proof causation. 24 25 Ascertainment of tuberculosis patients from HIV infected adults might lack sensitivity and 26 specificity. 27 28 CONCLUSION 29 30 HIV infected adults who have substance use (tobacco smoking, khat chewing and alcohol) 31 32 should be prioritized for tuberculosis screening. This study reaffirmed that initiation of HAART
33 http://bmjopen.bmj.com/ 34 is one of the best strategies to reduce tuberculosis occurrence in HIV infected adults. Increased 35 coverage of isoniazid and cotrimoxazole preventive therapy has protective benefit against active 36 37 tuberculosis. Contact tracing of household contacts for tuberculosis should be intensified. 38 39 Data sharing statement 40 41 The data of this research report is available. The authors are agreed to share the data. There is no on September 24, 2021 by guest. Protected copyright. 42 additional unpublished data. You may contact through [email protected] when 43 44 needed. 45 46 Acknowledgement 47 48 The authors gratefully acknowledge the Amhara Regional Health Bureau Research Core Process 49 50 Team, hospitals and health centers, data collectors and study participants involved in the study. 51 Author Contributions 52 53 54 55 56 57 15 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Conceived and designed the experiments: YMA WA AWS. Performed the experiments: YMA 4 5 WA AWS. Analyzed the data: YMA WA AWS. Contributed materials/analysis tools: YMA WA 6 7 AWS. Wrote the paper: YMA WA AWS. 8 9 10 11 References 12 1. Federal Ministry of Health Ethiopia (2013) Guide lines for clinical and program management of 13 14 TB, TB/HIV and leprosy. Ministry of Health, 2013. 15 For peer review only 16 2. World Health Organization (2013) Global tuberculosis report. WHO, 2013 17 18 3. Kwan CK, Ernst JD: HIV and tuberculosis: a deadly human syndemic. Clinical microbiology 19 reviews 24:351 376, 2011. 20 21 4. World Health Organization (2014) Global tuberculosis report. WHO, 2014. 22 23 5. United States Agency for International Development (2013) Report on the global AIDS epidemic. 24 25 USAID, 2013. 26 6. Pawlowski A, Jansson M, Sköld M, et al: Tuberculosis and HIV Co Infection. PLoS Pathog 27 28 8:e1002464, 2012. 29 30 7. Batista J, de Albuquerque Mde F, Maruza M, et al: Incidence and risk factors for tuberculosis in 31 32 people living with HIV: cohort from HIV referral health centers in Recife, Brazil. PloS one
33 http://bmjopen.bmj.com/ 34 8:e63916, 2013. 35 8. M. Tekkel, M. Rahu, H M. Loit AB: Socio economic status and duration of TB symptoms in 36 37 males treated at the Mazovian Treatment Centre of Tuberculosis and Lung Diseases in Otwock. 38 39 International Journal of Tuberculosis and Lung Disease 11/2002; 6:887 894., 2002. 40 41 9. Wilkinson D: Drugs for preventing tuberculosis in HIV infected persons. Cochrane Database Syst on September 24, 2021 by guest. Protected copyright. 42 Rev 4, 2000. 43 44 10. Suthar AB, Lawn SD, del Amo J, et al: Antiretroviral Therapy for Prevention of Tuberculosis in 45 46 Adults with HIV: A Systematic Review and Meta Analysis. PLoS Med 9:e1001270, 2012. 47 48 11. Montaner JS, Lima VD, Barrios R, et al: Association of highly active antiretroviral therapy 49 50 coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a 51 population based study. Lancet 376:532 539, 2010. 52 53 12. Alvi A, Rizwan M, Sunosi RA, et al: Does khat chewing increases the risk of Mycobacterium 54 55 tuberculosis infection by macrophage immune modulation? Med Hypotheses 82:667 669, 2014. 56 57 16 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 13. Kibret KT, Yalew AW, Belaineh BG, et al: Determinant factors associated with occurrence of 4 5 tuberculosis among adult people living with HIV after antiretroviral treatment initiation in Addis 6 7 Ababa, Ethiopia: a case control study. PLoS One 8:e64488, 2013. 8 9 14. Neyrolles O, Quintana Murci L: Sexual Inequality in Tuberculosis. PLoS Med 6:e1000199, 10 11 2009. 12 15. Albrecht DE, Albrecht CM: The Implications of Economic Structure for Marriage Prevalence. 13 14 The Open Sociology Journal, 2008. 15 For peer review only 16 16. Zammarchi L, Bartalesi F, Bartoloni A: Tuberculosis in tropical areas and immigrants. Mediterr J 17 18 Hematol Infect Dis 6:e2014043, 2014. 19 17. Golub JE, Saraceni V, Cavalcante SC, et al: The impact of antiretroviral therapy and isoniazid 20 21 preventive therapy on tuberculosis incidence in HIV infected patients in Rio de Janeiro, Brazil. 22 23 Aids 21:1441 1448, 2007. 24 25 18. Day CL, Moshi ND, Abrahams DA, et al: Patients with tuberculosis disease have 26 Mycobacterium tuberculosis specific CD8 T cells with a pro apoptotic phenotype and impaired 27 28 proliferative capacity, which is not restored following treatment. PLoS One 9:e94949, 2014. 29 30 19. Moreno S, Jarrin I, Iribarren JA, et al: Incidence and risk factors for tuberculosis in HIV positive 31 32 subjects by HAART status. Int J Tuberc Lung Dis 12:1393 1400, 2008.
33 http://bmjopen.bmj.com/ 34 20. Mermin J, Lule J, Ekwaru JP, et al: Effect of co trimoxazole prophylaxis on morbidity, mortality, 35 CD4 cell count, and viral load in HIV infection in rural Uganda. Lancet 364:1428 1434, 2004. 36 37 21. Sopori ML, Kozak W: Immunomodulatory effects of cigarette smoke. J Neuroimmunol 83:148 38 39 156, 1998. 40 41 22. Den Boon S, van Lill SWP, Borgdorff M, et al: Association between smoking and tuberculosis on September 24, 2021 by guest. Protected copyright. 42 infection: a population survey in a high tuberculosis incidence area. Thorax 60:555 557, 2005. 43 44 23. Leung CC, Yew WW, Chan CK, et al: Smoking and tuberculosis in Hong Kong. Int J Tuberc 45 46 Lung Dis 7:980 986, 2003. 47 48 24. Rabirad N, Mohammad Nejad E, Hadizadeh MR, et al: The Prevalence of Tb in HIV Patients and 49 50 Risk Factor With Frequent Referral (Iran, 2009 10). Iran Red Crescent Med J 15:58 61, 2013. 51 25. Gajalakshmi V, Peto R: Smoking, drinking and incident tuberculosis in rural India: population 52 53 based case control study. Int J Epidemiol 38:1018 1025, 2009. 54 55 56 57 17 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 26. Rehm J, Samokhvalov AV, Neuman MG, et al: The association between alcohol use, alcohol use 4 5 disorders and tuberculosis (TB). A systematic review. BMC Public Health 9:450, 2009. 6 7 27. Szabo G: Alcohol's contribution to compromised immunity. Alcohol Health Res World 21:30 41, 8 9 1997. 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 18 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 STROBE Statement—checklist of items that should be included in reports of observational studies 3 4 Item 5 No Recommendation 6 Title and abstract 7 1 (a) Indicate the study’s design with a commonly used term in the title or the abstract 8 (b) Provide in the abstract an informative and balanced summary of what was done 9 and what was found 10 Introduction 11 12 Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 13 Objectives 3 State specific objectives, including any prespecified hypotheses 14 Methods 15 For peer review only 16 Study design 4 Present key elements of study design early in the paper 17 Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, 18 exposure, follow-up, and data collection 19 Participants 6 (a) Cohort study—Give the eligibility criteria, and the sources and methods of 20 21 selection of participants. Describe methods of follow-up 22 Case-control study—Give the eligibility criteria, and the sources and methods of 23 case ascertainment and control selection. Give the rationale for the choice of cases 24 and controls 25 26 Cross-sectional study—Give the eligibility criteria, and the sources and methods of 27 selection of participants 28 (b) Cohort study—For matched studies, give matching criteria and number of 29 exposed and unexposed 30 Case-control study—For matched studies, give matching criteria and the number of 31 32 controls per case
33 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect http://bmjopen.bmj.com/ 34 modifiers. Give diagnostic criteria, if applicable 35 Data sources/ 8* For each variable of interest, give sources of data and details of methods of 36 37 measurement assessment (measurement). Describe comparability of assessment methods if there 38 is more than one group 39 Bias 9 Describe any efforts to address potential sources of bias 40 Study size 10 Explain how the study size was arrived at 41 on September 24, 2021 by guest. Protected copyright. 42 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, 43 describe which groupings were chosen and why 44 Statistical methods 12 (a) Describe all statistical methods, including those used to control for confounding 45 (b) Describe any methods used to examine subgroups and interactions 46 (c) Explain how missing data were addressed 47 48 (d) Cohort study—If applicable, explain how loss to follow-up was addressed 49 Case-control study—If applicable, explain how matching of cases and controls was 50 addressed 51 Cross-sectional study—If applicable, describe analytical methods taking account of 52 53 sampling strategy 54 (e) Describe any sensitivity analyses 55 Continued on next page 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml1 Page 21 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Results 4 5 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, 6 examined for eligibility, confirmed eligible, included in the study, completing follow-up, and 7 analysed 8 (b) Give reasons for non-participation at each stage 9 (c) Consider use of a flow diagram 10 11 Descriptive 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and information 12 data on exposures and potential confounders 13 (b) Indicate number of participants with missing data for each variable of interest 14 (c) Cohort study—Summarise follow-up time (eg, average and total amount) 15 For peer review only 16 Outcome data 15* Cohort study—Report numbers of outcome events or summary measures over time 17 Case-control study—Report numbers in each exposure category, or summary measures of 18 exposure 19 Cross-sectional study—Report numbers of outcome events or summary measures 20 21 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their 22 precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and 23 why they were included 24 (b) Report category boundaries when continuous variables were categorized 25 26 (c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful 27 time period 28 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and sensitivity 29 analyses 30 31 Discussion 32 Key results 18 Summarise key results with reference to study objectives
33 Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or imprecision. http://bmjopen.bmj.com/ 34 Discuss both direction and magnitude of any potential bias 35 36 Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, multiplicity 37 of analyses, results from similar studies, and other relevant evidence 38 Generalisability 21 Discuss the generalisability (external validity) of the study results 39 Other information 40 41 Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, on September 24, 2021 by guest. Protected copyright. 42 for the original study on which the present article is based 43 44 *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and 45 46 unexposed groups in cohort and cross-sectional studies. 47 48 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 49 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 50 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 51 52 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 53 available at www.strobe-statement.org. 54 55 Author notice: This checklist is a very good guide to prepare an observational research report. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml2 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from
Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a multicenter case control study
ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2015-009058.R1
Article Type: Research
Date Submitted by the Author: 25-Sep-2015
Complete List of Authors: Alemu, Yihun Mulugeta; Heidelberg University,Institute of Public Health; Bahir Dar University , School of Public Health Awoke, Worku; Bahir Dar University, School of Public Health Annelies, Wilder-Smith; Heidelberg University, Institute of Public Health; Lee Kong Chian School of Medicine , Nanyang Technological University
Primary Subject Global health Heading:
Secondary Subject Heading: Epidemiology, Global health
PUBLIC HEALTH, Epidemiology < THORACIC MEDICINE, Clinical trials < Keywords: THERAPEUTICS
http://bmjopen.bmj.com/
on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a 6 7 multicenter case control study 8 9 10 11 1, 2* 2 1,3 12 Yihun Mulugeta Alemu , Worku Awoke , Wilder Smith Annelies 13 14 15 For peer review only 16 17 18 1Institute of Public Health, Heidelberg University, Germany 19 20 21 2School of Public Health, College of Medicine and Health Science, Bahir Dar University, 22 Ethiopia 23 24 25 3Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore 26 27 28 29 30 Corresponding author; Yihun Mulugeta, Alemu E: mail: [email protected] 31 32
33 http://bmjopen.bmj.com/ 34 35 Determinants of tuberculosis in HIV infected adults 36 37 Key words: Determinants, tuberculosis, HIV, adults, multi-center 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 1 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Summary Boxes 4 5 What is already known in this subject? 6 7 Previous studies have identified independent determinants for TB as clinical (highly activated 8 9 antiretroviral therapy, lower level of CD4 cell count, cotrimoxazole preventive therapy and 10 11 isoniazid preventive therapy) and life style (alcohol consumption). However; among HIV 12 infected adults, determinants for TB are not well described in resource limited settings. 13 14 What this study adds? 15 For peer review only 16 This study identified independent determinants for tuberculosis (highly activated antiretroviral 17 18 therapy, lower level of CD4 cell count, cotrimoxazole preventive therapy, isoniazid preventive 19 therapy and alcohol consumption) in HIV infected adults in resource limited settings. In addition 20 21 the study identified independent life style determinants for tuberculosis (Khat chewing and 22 23 tobacco smoking) in HIV infected adults. Independent life style determinants (Khat chewing and 24 25 tobacco smoking) have been rarely identified in this types of source population (HIV infected 26 adults). Determinants for TB among HIV infected adult are complex and encompass a wide 27 28 range of issues. This study will contribute to a more comprehensive & multidisciplinary 29 30 approach for prevention and control of TB among HIV infected adults in resource limited 31 32 settings.
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 2 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 Abstract 5 6 Objective: The objective of this study was to identify determinants for tuberculosis among 7 8 HIV infected adults in Northwest Ethiopia. 9 10 11 Design: Case control study. 12 13 14 Setting: Three hospitals and ten health centers in Northwest Ethiopia. 15 For peer review only 16 17 18 Participants: A total of 446 study participants consented to participate in the study (150 cases 19 20 21 and 296 controls). Cases were HIV infected adults diagnosed with active tuberculosis and 22 23 controls were HIV infected adults without active tuberculosis. 24 25 26 27 Main outcome measure: The link between tuberculosis and determinants was assessed by 28 29 using logistic regression. Determinants were categorized as sociodemographic, host related, 30 31 32 clinical and environmental.
33 http://bmjopen.bmj.com/ 34 35 Results: Smoking had an Adjusted Odds Ratio (AOR) of 5.47 (95%CI: 2.26, 13.22), presence 36 37 of tuberculosis patient in the family an AOR of 2.66 (95%CI: 1.25, 5.66), alcohol consumption 38 an AOR of 2.49 (95%CI: 1.29, 4.8) and chewing khat an AOR of 2.22 (95% CI: 1.11, 4.41) were 39 40 independent determinants for increased occurrence of tuberculosis. Highly active antiretroviral 41 on September 24, 2021 by guest. Protected copyright. 42 therapy (HAART) an AOR of 0.25 (95%CI: 0.13, 0.51), isoniazid (INH) therapy an AOR of 0.22 43 44 (95% CI: 0.11, 0.42) and cotrimoxazole preventive therapy an AOR of 0.32 (95% CI: 0.19, 0.52) 45 had protective benefit against tuberculosis. 46 47 Conclusion: HIV infected adults with substance abuse (tobacco smoking, khat chewing and 48 49 alcohol) should be prioritized for tuberculosis screening. This study reaffirmed that HAART and 50 51 INH therapy are some of the best strategies to reduce tuberculosis occurrence in HIV infected 52 53 adults. These findings provide impetus to intensify contact tracing in TB household contacts. 54 55 56 57 3 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 ARTICLE SUMMARY 4 5 Strength and limitation of the study 6 7 This is the first multicenter case control study in Northwest Ethiopia that investigates 8 9 determinants for tuberculosis in HIV infected adults. 10 11 The study identified determinants for tuberculosis that will be important to prioritize TB 12 screening, treatment, prevention and control. 13 14 This study reaffirmed that HAART and isoniazid therapy are some of the best strategies to 15 For peer review only 16 reduce tuberculosis among HIV infected adults in resource limited settings. 17 18 As we used a retrospective case control approach; temporal relationship could not be established; 19 the study design could not proof causation. 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 4 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 INTRODUCTION 4 5 The advent of HIV was a massive setback for the prevention and control of tuberculosis (TB).1 6 7 TB is the leading cause of morbidity and mortality among people living with HIV (PLWHIV).2 8 9 There is a strong synergy between TB and HIV infection in HIV high burden countries 10 3 11 particularly in resource limited settings where the impacts of both diseases are more significant. 12 One third of the world’s population is infected with TB.2 In 2013, an estimated 9.0 million 13 14 people developed TB; 1.1 million were HIV infected.4 Ethiopia ranks as the seventh TB burden 15 For5 peer review only 16 country in the world. 17 18 In resource limited settings the health care systems are overwhelmed by preventive, therapeutic 19 and diagnostic challenges of the "HIV TB syndemic". In Ethiopia; the dual epidemics drained 20 21 resources and overburdened the already very limited health work force.1 HIV is increases the life 22 6 23 time risk of developing TB. However, HIV is not the only determinant for developing TB; 24 25 various other determinants contribute to TB HIV co infection. Previous studies have identified 26 socio demographic,7,8 clinical,9,10 life style11,12 and environmental13 determinants for TB. 27 28 However, among HIV infected adults, the determinants for TB are still not well described 29 30 particularly in a resource limited setting. Determinants for tuberculosis among HIV positive 31 32 adults vary from one setting to another. Therefore, a context specific study in a high burdened
33 http://bmjopen.bmj.com/ 34 TB/HIV region is indicated. This study aims to assess the determinants for TB among HIV 35 infected adults in Northwest Ethiopia. 36 37 METHODS 38 39 Study design 40 41 We conducted a multicenter case control study from May 12 to June 5, 2014 in Northwest on September 24, 2021 by guest. Protected copyright. 42 Ethiopia. All governmental health institutions in ten provinces were included (i.e. Bahir Dar, 43 44 Dangella, Kossober, Fnote Selam, Bure, Dur Betie, Deber Tabor, Wereta, Addis Zemen and 45 46 Nefas Mewcha). 47 48 Participants 49 50 Cases were HIV infected adults diagnosed with active TB and on TB treatment during the data 51 collection period. The diagnosis of active TB was based on one or more of TB diagnostic 52 53 investigations (sputum microscopy, x ray, histopathology, culture or molecular). Controls were 54 55 HIV infected adults without TB. All controls were actively screened for TB to rule out TB. The 56 57 5 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 screening procedure included history taking for four symptoms at each visit (cough, weight loss, 4 5 night sweating or fever). If the patient presented with any one of these symptoms, this patient 6 7 was a suspect TB case and was investigated according to the national guidelines for TB HIV1. In 8 9 hospitals, sputum microscopy, histopathology and radiology examinations are used as the TB 10 11 diagnostic tools. In health centers, sputum microscopy is the main stay TB diagnostic tool and 12 patients will be sent for other diagnostic evaluation at hospitals and private institutions where the 13 14 diagnostic tools are available. 15 For peer review only 16 17 Sample size calculation 18 The sample size was calculated using Epi Info version 7 software. The study was designed to 19 20 have 80% statistical power with level of significance at 5% and a case to control ratio of 1:2. 21 22 Assuming the proportion of low CD4 cell count were 5.9% among the controls and 13.9% 23 13 24 among cases by using the two proportions formula, the calculated sample size was 144 for 25 cases and 287 for controls. Allowing for 10% of non response, the resulting sample size was 474 26 27 (158 cases and 316 controls). The sample size was calculated for exposure status of different 28 29 variables: cotrimoxazole preventive therapy (CPT), isoniazid preventive therapy (IPT) and CD4. 30 31 We took the largest sample among these exposure variables. 32 Sampling procedures
33 http://bmjopen.bmj.com/ 34 A total of fifty two health institutions are providing HIV care service in 44 districts of Awie, 35 36 West Gojjam and South Gondar Zones.14 Ten districts were randomly selected and all health 37 38 institutions in the ten districts were included in the study. Three hospitals and ten health centers 39 40 were included. HIV infected adults diagnosed with confirmed active TB were recruited as cases 41 while those without active TB were recruited as controls. Study subjects who were unable to give on September 24, 2021 by guest. Protected copyright. 42 43 informed consent and subjects with suspected but unconfirmed TB were excluded. All TB HIV 44 45 co infected patients attending HIV care clinic and those who were taking TB treatment were 46 47 included. Controls were allotted based on the number of cases available in each facility with the 48 control to case ratio of 2:1 and sampled by systematic random sampling with a sampling interval 49 50 of five. 51 52 Data collection and analysis 53 54 Data were collected from two sources. Trained nurses who were in charge of HIV care clinic 55 56 conducted face to face interview with study participants by using structured questionnaires that 57 6 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 were prepared in English, translated to Amharic and back translated and pretested for 4 5 consistency and ease of understanding. The primary data were collected to assess: socio 6 7 demographic variables (age, sex, educational and marital status), host related variables (cigarette 8 9 smoking, khat chewing and alcohol consumption). A person with substance abuse (chewing 10 11 Kahat, alcohol consumption or smoking) was defined as an individual who is currently using the 12 substance or has a history of substance use regularly. Environmental determinants included the 13 14 presence of flooring, latrine and separate kitchen in the household. Trained nurses also collected 15 For peer review only 16 data from patient’s records to assess: clinical variables (HAART, CD4 cell count, CPT and IPT). 17 18 Data collection procedures were supervised by medical officers. 19 Data were entered into Epi info version 7 and analysis was done with SPSS version 20. 20 21 Frequencies and proportions were used to describe the study subjects in relation to the studied 22 23 variables. Odds ratio with its 95% confidence interval and p value were used to measure strength 24 25 of association and identify statistical significance result. Logistic regression models were applied 26 to assess the relationship between determinants and TB. Identification of confounding variables 27 28 by logistic regression models was applied. Multivariable analysis (backward stepwise) was used. 29 30 Hosmer Lemeshow test was applied and the fit of the model was checked. A poor fit of the 31 32 model if the significance p value is less than 0.05 and good fit greater than 0.05. Here in this
33 http://bmjopen.bmj.com/ 34 study the model adequately fits the data and the p value was 0.34. 35 Ethical considerations 36 37 Ethical approval from the Heidelberg Ethics Commission was obtained. Additionally, ethical 38 39 approval from the country of research, Amhara Regional Research and Ethical Core Process, 40 41 Bahir Dar, Ethiopia was granted. Written permission to conduct the study was obtained from on September 24, 2021 by guest. Protected copyright. 42 each health institution involved in the study. Since there were illiterate participants the data 43 44 collectors informed each study participant about the informed consent sheet. Informed oral 45 46 consent was obtained from each study participant. The data collectors documented the 47 48 participant’s consent on the informed consent sheet. This consent procedure was approved by the 49 50 ethical commission. 51 52 53 54 55 56 57 7 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 6 7 8 9 RESULTS 10 11 Socio demographic and host related determinants of HIV infected adults: 12 A total of 446 subjects (150 cases=33.6%, 296 controls=66.4%) out of 474 eligible participants 13 14 responded and consented to participate in this study, resulting in an overall response rate of 15 For peer review only 16 94.1% (94.9% for cases and 93.7% for controls). 17 18 The median age and its interquartile range (IQR) of cases were 32 years (IQR: 27 39 years) 19 while the corresponding value for controls were 33 years (IQR: 28 39 years). A higher 20 21 percentage of women was observed in both groups; 81(54%) in cases and 188(63.5%) in controls 22 23 (Table1). 24 25 Table 1. Socio demographic and host related determinants of HIV infected adults in Northwest 26 Ethiopia, 2014 27 28 Variables Case n (%) Control n (%) Total n (%) 29 30 Sex 31 32 Male 69(46) 108(36.5) 177(39.7)
33 http://bmjopen.bmj.com/ 34 Female 81(54) 188(63.5) 269(60.3) 35 36 Education 37 No formal education 46(30.7) 120(40.5) 166(37.2) 38 39 Primary 48(32) 78(26.4) 126(28.3) 40 41 Secondary 42(28) 60(20.2) 102(22.9) on September 24, 2021 by guest. Protected copyright. 42 43 Tertiary 14(9.3) 38(12.8) 52(11.6) 44 45 Marital status 46 Married 53(35.3) 141(47.6) 194(43.5) 47 48 Divorce / widowed 50(33.3) 98(33.1) 148(33.2) 49 50 Never married 47(31.3) 57(19.3) 104(23.3) 51 52 Smoking 53 54 Yes 25(16.7) 12(4) 37(8.3) 55 56 57 8 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 No 125(83.3) 284(96) 409(91.7) 4 5 Khat chewing 6 7 Yes 44(29.3) 37(12.5) 81(18) 8 9 No 106(70.7) 259(87.5) 365(82) 10 11 Alcohol 12 13 Yes 49(32.7) 49(16.6) 98(22) 14 No 101(67.3) 247(83.4) 348(78) 15 For peer review only 16 Previous history of TB 17 18 Yes 41(27.3) 50(16.9) 91(20.4) 19 20 No 109(72.7) 246(83.1) 355(79.6) 21 22 23 Bivariate Analysis: 24 25 The bivariate analysis showed that male patients (COR (crude odds ratio) =1.48; 95% CI: 0.99, 26 27 2.12; p value = 0.053) were more likely to develop tuberculosis compared to female patients. 28 29 Married patients (COR=0.45; 95%CI: 0.27, 0.75) were less likely to develop TB compared with 30 31 never married. Cases were more likely to be smokers (COR= 4.73; 95% CI: 2.3, 9.72), Khat 32 chewer (COR=2.9; 95%CI: 1.77, 4.75) and alcohol drinker (COR= 2.44; 1.54, 3.86). Study
33 http://bmjopen.bmj.com/ 34 participants who had a history of TB (COR=1.85; 95%CI: 1.15, 2.96) were more likely to 35 36 develop TB. But educational status was not a determinant for TB (Table 2). 37 38 Table 2: Socio demographic and host related determinants for developing tuberculosis among 39 HIV infected adults in Northwest Ethiopia, 2014 40 41 Variables Cases n (%) Controls n (%) COR 95%CI P value on September 24, 2021 by guest. Protected copyright. 42 43 Sex 44 45 Male 69(46) 108(36.5) 1.48 0.99, 2.12 0.053 46 47 Female 81(54) 188(63.5) 1 48 Education 49 50 No formal education 46(30.7) 120(40.5) 1.04 0.51, 2.09 0.912 51 52 Primary 48(32) 78(26.4) 1.67 0.82, 3.39 0.157 53 54 Secondary 42(28) 60(20.2) 1.90 0.9,3.93 0.084 55 56 57 9 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Tertiary 14(9.3) 38(12.8) 1 4 5 Marital status 6 7 Married 53(35.3) 141(47.6) 0.45 0.27, 0.75 0.002* 8 9 Divorce / widowed 50(33.3) 98(33.1) 0.61 0.37, 1.03 0.068 10 11 Never married 47(31.3) 57(19.3) 1 12 13 Smoking 14 Yes 25(16.7) 12(4) 4.73 2.30, 9.72 < 0.0001* 15 For peer review only 16 No 125(83.3) 284(96) 1 17 18 Khat chewing 19 20 Yes 44(29.3) 37(12.5) 2.9 1.77, 4.75 < 0.0001* 21 22 No 106(70.7) 259(87.5) 1 23 Alcohol 24 25 Yes 49(32.7) 49(16.6) 2.44 1.54, 3.86 < 0.0001* 26 27 No 101(67.3) 247(83.4) 1 28 29 Previous history of TB 30 31 Yes 41(27.3) 50(16.9) 1.85 1.15, 2.96 0.01* 32 No 109(72.7) 246(83,1) 1
33 http://bmjopen.bmj.com/ 34 35 36 HAART (COR=0.33; 95%CI: 0.19, 0.56), CPT (COR=0.35 95%CI: 0.22, 0.53) and IPT 37 38 (COR=0.30 95%CI: 0.18, 0.50) had protective benefit against TB. Subjects with separate kitchen 39 40 in the house hold (COR=0.61; 95%CI; 0.40, 0.93) were less likely to develop TB. HIV infected 41 adults with presence of a TB patient in the family (COR=1. 99; 95%CI: 1.11, 3.57) were more on September 24, 2021 by guest. Protected copyright. 42 43 likely to develop TB. But flooring and presence of latrine were not determinants for developing 44 45 TB (Table 3). 46 47 Table 3: Clinical and environmental determinants for developing tuberculosis among HIV 48 49 infected adults in Northwest Ethiopia, 2014 50 Variables Cases n (%) Controls n (%) COR 95%CI P value 51 52 HAART 53 54 Yes 114(76) 268(90.5) 0.33 0.19, 0.56 < 0.0001* 55 56 57 10 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 No 36(24) 28(9.5) 1 4 5 CPT 6 7 Yes 83(55.3) 231(78) 0.35 0.22, 0.53 < 0.0001* 8 9 No 67(44.7) 65(22) 1 10 11 INH prophylaxis 12 13 Yes 22(14.7) 107(36.1) 0.30 0.18, 0.50 < 0.0001* 14 No 128(85.3) 189(63.9) 1 15 For peer review only 16 CD4 cell count (cells/ µl) 17 18 <=200 87(58) 61(20.6) 5.53 2.96, 10.35 < 0.0001* 19 20 200 – 500 46(30.7) 169(57.1) 1.05 0.56, 1.97 0.86 21 22 >=500 17(11.3) 66 (22.3) 1 23 TB patient in the family 24 25 Yes 25(16.7) 27(9.1) 1.99 1.11, 3.57 0.021* 26 27 No 125(83.3) 269(90.9) 1 28 29 Separate kitchen 30 31 Yes 96(64) 220(74.3) 0.61 0.40, 0.93 0.024* 32 No 54(36) 76(25.7) 1
33 http://bmjopen.bmj.com/ 34 Floor of the house 35 36 Mud / Soil 123(82) 240(81) 1.06 0.64, 1.76 0.814 37 38 Cement 27(18) 56(19) 1 39 40 Latrine 41 on September 24, 2021 by guest. Protected copyright. 42 Yes 136(90.7) 267(90.2) 1.05 0.53, 2.06 0.875 43 No 14(9.3) 29(9.8) 1 44 45 46 47 Multivariable analysis: 48 49 To identify independent determinants for TB, a multivariable logistic regression model was used. 50 51 After adjusting for possible confounders, some variables were remained in the multivariable 52 model: chewing khat (AOR=2.22; 95%CI: 1.11, 4.41), being a smoker (AOR=5.47; 95%CI: 53 54 2.26, 13.22) and alcohol drinker (AOR=2.49; 1.29, 4.80) were independent determinants for 55 56 increased TB occurrence. However, HAART (AOR=0.25; 95%CI: 0.12, 0.51), CPT (AOR=0. 57 11 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 32; 95% CI: 0.19, 0.52) and IPT (AOR=0.22; 95%CI: 0.11, 0.42) had an independent protective 4 5 benefit effect against TB. Subjects with separate kitchen in the house hold (AOR=0.48; 95%CI: 6 7 0.28, 0.83) was less likely to develop TB. Patients who live with a TB patient in the household 8 9 (AOR=2.66; 95%CI: 1.25, 5.66) were more likely to develop TB. Patients whose CD4 cell count 10 11 was < 200 cells/µl (AOR= 7.22; 95%CI: 3.39, 15.37) were more likely to develop TB. But sex, 12 marital status and a previous history of TB were not remained in the multivariable model (Table 13 14 4). 15 For peer review only 16 Table 4: Independent determinants for tuberculosis among HIV infected adults in 17 18 Northwest Ethiopia, 2014 19 Variables COR(95%CI) P value AOR(95%CI) P value 20 21 HAART 22 23 Yes 0.33(0.19, 0.56) < 0.0001 0.25(0.13, 0.51) < 0.0001 24 25 No 1 1 26 27 CPT 28 29 Yes 0.35(0.22, 0.53) < 0.0001 0.32(0.19, 0.55) < 0.0001 30 No 1 1 31 32 INH prophylaxis
33 http://bmjopen.bmj.com/ 34 Yes 0.30(0.18, 0.50) < 0.0001 0.22(0.11, 0.41) < 0.0001 35 36 No 1 1 37 38 Smoking 39 Yes 4.73(2.30, 9.72) < 0.0001 5.47(2.26, 13.22) < 0.0001 40 41 No 1 1 on September 24, 2021 by guest. Protected copyright. 42 43 Khat 44 45 Yes 2.9(1.77, 4.75) <0.001 2.22(1.11, 4.41) 0.023 46 47 No 1 1 48 49 Alcohol 50 Yes 2.44(1.54, 3.86) <0.0001 2.49(1.29, 4.80) 0.006 51 52 No 1 1 53 54 TB patient in the family 55 56 57 12 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Yes 1.99(1.11, 3.57) 0.021 2.66(1.25,5.66) 0.011 4 5 No 1 1 6 7 Separate kitchen 8 9 Yes 0.61(0.40, 0.93) 0.024 0.48(0.28, 0.83) 0.010 10 11 No 1 1 12 13 CD4 cell count (cells/ µl) 14 < 200 cells/ µl 5.53(2.96, 10.35) < 0.0001 7.22(3.39, 15.37) < 0.0001 15 For peer review only 16 200 – 500 cells/ µl 1.05(0.56,1.97) 0.86 1.31(0.63, 2.74) 0.461 17 18 >= 500 cells/ µl 1 1 19 20 21 22 DISCUSSION 23 This study showed that HIV positive adults on HAART, CPT and IPT were less likely to develop 24 25 TB. But smoking, alcohol consumption, Khat chewing, presence of TB patient in the family and 26 27 CD4 count less than 200 cells/µl were independent determinants for increased TB occurrence. 28 29 When compared with female HIV positive adults, we found that male HIV positive adults were 30 more likely to develop TB. Men are reported to be more susceptible to tuberculosis infection.15 31 32 But the variable male sex did not remain in the multivariable model.
33 http://bmjopen.bmj.com/ 34 Similar to the finding of other study, this study also showed that level of education was not a 35 7 36 determinant for TB occurrence. Hence, increasing level of school training may not necessarily 37 38 bring successful behavioral change to TB prevention and control in this source population. 39 Married HIV positive adults were also less likely to develop TB compared to those who were 40 41 8 never married; consistent with a study. Previous studies have shown that increased marriage on September 24, 2021 by guest. Protected copyright. 42 16 43 prevalence is observed in higher income society and that tuberculosis and poverty are 44 17 45 interrelated. 46 HIV positive patients with low CD4 count were more likely to develop TB. It was consistent 47 48 with the literature.7 In addition, our findings indicating that IPT had a protective benefit against 49 50 TB was also consistent with a previous research report.18 IPT reduces the mycobacterium load 51 9 52 and reduces the progression of latent TB to active tuberculosis. But increase mycobacterial load 53 19 54 is coupled with progressive impairment of mycobacterium specific T cell response. 55 56 57 13 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Study participants on HAART were less likely to develop TB. It was consistent with 4 5 publication.10 This could be due to strong relation between increasing HAART coverage and 6 7 decreased viral load.11 Lowering of viral load was also linked with reduced occurrence of active 8 20 9 TB. 10 11 We found that cotrimoxazole preventive therapy has protective benefit against tuberculosis. This 12 was consistent with literature elsewhere.13 Cotrimoxazole preventive therapy has beneficial 13 14 effects on enhancing CD4 count as well as reduction of viral load.21 15 For peer review only 22,23 16 Cigarette smoking induces immune impairment and damage the ciliary clearance. In this 17 18 study smoking was an independent determinants for TB. It was consistent with a previous 19 study.24 20 21 Khat chewing is linked with immune modulation.12 Among HIV infected adults, no published 22 23 study has detected khat chewing as an independent determinant for TB. In this study Khat 24 25 chewing was an independent determinant for TB. 26 Our findings showed that study subjects who consumed alcohol were more likely to develop TB. 27 28 It was consistent with previous studies.25 27 Alcohol can modulate the immune response. Long 29 30 term alcohol consumers have been shown to have impaired immunity.28 31 32 Past history of tuberculosis was a determinant for increased TB occurrence. But a past medical
33 http://bmjopen.bmj.com/ 34 history of TB was shown to be a determinant for TB in the bi variable analysis only. This may be 35 explained by the lower TB case detection rate in Ethiopia.2 The established tuberculosis case 36 37 detection mechanisms overlooked some tuberculosis patients. A proportion of tuberculosis 38 39 patient may remain undiagnosed i.e. died with other means of diseases or died without visiting 40 41 health care; the detected proportion of reoccurrence might be low. Therefore the prevalence of on September 24, 2021 by guest. Protected copyright. 42 reoccurrence was underestimated. 43 44 This study has the following potential limitations: as we used a retrospective case control 45 46 approach; temporal relationship could not be established and the study design could not proof 47 48 causation. Tuberculosis has a possibility to spread to another people who is not a household 49 50 member. But this study was considered, only family members were close tuberculosis contacts. 51 Information bias related to substance use (chewing chat, smoking or alcohol consumption) might 52 53 have also affected the accuracy of information. 54 55 CONCLUSION 56 57 14 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 HIV infected adults with substance abuse (tobacco smoking, khat chewing and alcohol) should 4 5 be prioritized for tuberculosis screening. This study reaffirmed that HAART and INH therapy are 6 7 some of the best strategies to reduce tuberculosis occurrence in HIV infected adults. These 8 9 findings provide impetus to intensify contact tracing in TB household contacts. 10 11 Data sharing statement 12 No additional data available. 13 14 Competing interests 15 For peer review only 16 No, there are no competing interests. 17 18 Acknowledgement 19 The authors gratefully acknowledge the Amhara Regional Health Bureau Research Core Process 20 21 Team, hospitals and health centers, data collectors, supervisors and study participants involved in 22 23 the study. 24 25 Author Contributions 26 Conceived and designed the experiments: YMA WA AWS. Performed the experiments: YMA 27 28 WA AWS. Analyzed the data: YMA WA AWS. Contributed materials/analysis tools: YMA WA 29 30 AWS. Wrote the paper: YMA WA AWS. 31 32 Funding: PAGEL (Partnerships for the health sector in developing countries offers German
33 http://bmjopen.bmj.com/ 34 institutions of higher education many different opportunities for international cooperation related 35 to the health sector in developing countries) funded this article. The points expressed in this 36 37 article are the responsibility of authors, do not reflect the view of PAGEL. The funding source 38 39 had no role in design, analysis, wrote paper, manuscript preparation or decision for publication. 40 41 References on September 24, 2021 by guest. Protected copyright. 42 1. Federal Ministry of Health Ethiopia (2013) Guide lines for clinical and program management of 43 44 TB, TB/HIV and leprosy. Ministry of Health, 2013. 45 46 2. World Health Organization (2013) Global tuberculosis report. WHO, 2013. 47 48 3. Kwan CK, Ernst JD: HIV and tuberculosis: a deadly human syndemic. Clinical microbiology 49 50 reviews 24:351 376, 2011. 51 4. United States Agency for International Development (2013) Report on the global AIDS epidemic. 52 53 USAID, 2013. 54 55 5. World Health Organization (2014) Global tuberculosis report. WHO, 2014. 56 57 15 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 6. Pawlowski A, Jansson M, Sköld M, et al: Tuberculosis and HIV Co Infection. PLoS Pathog 4 5 8:e1002464, 2012. 6 7 7. Batista J, de Albuquerque Mde F, Maruza M, et al: Incidence and risk factors for tuberculosis in 8 9 people living with HIV: cohort from HIV referral health centers in Recife, Brazil. PloS one 10 11 8:e63916, 2013. 12 8. M. Tekkel, M. Rahu, H M. Loit AB: Socio economic status and duration of TB symptoms in 13 14 males treated at the Mazovian Treatment Centre of Tuberculosis and Lung Diseases in Otwock. 15 For peer review only 16 International Journal of Tuberculosis and Lung Disease 11/2002; 6:887 894., 2002. 17 18 9. Wilkinson D: Drugs for preventing tuberculosis in HIV infected persons. Cochrane Database Syst 19 Rev 4, 2000. 20 21 10. Suthar AB, Lawn SD, del Amo J, et al: Antiretroviral Therapy for Prevention of Tuberculosis in 22 23 Adults with HIV: A Systematic Review and Meta Analysis. PLoS Med 9:e1001270, 2012. 24 25 11. Montaner JS, Lima VD, Barrios R, et al: Association of highly active antiretroviral therapy 26 coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a 27 28 population based study. Lancet 376:532 539, 2010. 29 30 12. Alvi A, Rizwan M, Sunosi RA, et al: Does khat chewing increases the risk of Mycobacterium 31 32 tuberculosis infection by macrophage immune modulation? Med Hypotheses 82:667 669, 2014.
33 http://bmjopen.bmj.com/ 34 13. Kibret KT, Yalew AW, Belaineh BG, et al: Determinant factors associated with occurrence of 35 tuberculosis among adult people living with HIV after antiretroviral treatment initiation in Addis 36 37 Ababa, Ethiopia: a case control study. PLoS One 8:e64488, 2013. 38 39 14. Amhara Regional Health Bureau (2013) Amhara Health Profile 2013. Amhara Health Bureau, 40 41 Amhara, Ethiopia, 2013. on September 24, 2021 by guest. Protected copyright. 42 15. Neyrolles O, Quintana Murci L: Sexual Inequality in Tuberculosis. PLoS Med 6:e1000199, 43 44 2009. 45 46 16. Albrecht DE, Albrecht CM: The Implications of Economic Structure for Marriage Prevalence. 47 48 The Open Sociology Journal, 2008. 49 50 17. Zammarchi L, Bartalesi F, Bartoloni A: Tuberculosis in tropical areas and immigrants. Mediterr J 51 Hematol Infect Dis 6:e2014043, 2014. 52 53 54 55 56 57 16 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 18. Golub JE, Saraceni V, Cavalcante SC, et al: The impact of antiretroviral therapy and isoniazid 4 5 preventive therapy on tuberculosis incidence in HIV infected patients in Rio de Janeiro, Brazil. 6 7 Aids 21:1441 1448, 2007. 8 9 19. Day CL, Moshi ND, Abrahams DA, et al: Patients with tuberculosis disease have 10 11 Mycobacterium tuberculosis specific CD8 T cells with a pro apoptotic phenotype and impaired 12 proliferative capacity, which is not restored following treatment. PLoS One 9:e94949, 2014. 13 14 20. Moreno S, Jarrin I, Iribarren JA, et al: Incidence and risk factors for tuberculosis in HIV positive 15 For peer review only 16 subjects by HAART status. Int J Tuberc Lung Dis 12:1393 1400, 2008. 17 18 21. Mermin J, Lule J, Ekwaru JP, et al: Effect of co trimoxazole prophylaxis on morbidity, mortality, 19 CD4 cell count, and viral load in HIV infection in rural Uganda. Lancet 364:1428 1434, 2004. 20 21 22. Sopori ML, Kozak W: Immunomodulatory effects of cigarette smoke. J Neuroimmunol 83:148 22 23 156, 1998. 24 25 23. Den Boon S, van Lill SWP, Borgdorff M, et al: Association between smoking and tuberculosis 26 infection: a population survey in a high tuberculosis incidence area. Thorax 60:555 557, 2005. 27 28 24. Leung CC, Yew WW, Chan CK, et al: Smoking and tuberculosis in Hong Kong. Int J Tuberc 29 30 Lung Dis 7:980 986, 2003. 31 32 25. Rabirad N, Mohammad Nejad E, Hadizadeh MR, et al: The Prevalence of Tb in HIV Patients and
33 http://bmjopen.bmj.com/ 34 Risk Factor With Frequent Referral (Iran, 2009 10). Iran Red Crescent Med J 15:58 61, 2013. 35 26. Gajalakshmi V, Peto R: Smoking, drinking and incident tuberculosis in rural India: population 36 37 based case control study. Int J Epidemiol 38:1018 1025, 2009. 38 39 27. Rehm J, Samokhvalov AV, Neuman MG, et al: The association between alcohol use, alcohol use 40 41 disorders and tuberculosis (TB). A systematic review. BMC Public Health 9:450, 2009. on September 24, 2021 by guest. Protected copyright. 42 28. Szabo G: Alcohol's contribution to compromised immunity. Alcohol Health Res World 21:30 41, 43 44 1997. 45 46 47 48 49 50 51 52 53 54 55 56 57 17 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 STROBE Statement—checklist of items that should be included in reports of observational studies 3 4 5 Item No Recommendation 6 7 Title and abstract 1 Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a 8 multicenter case control study 9 Accepted: Background: 10 Tuberculosis presents a major infectious disease burden in resource limited settings. 11 12 HIV is driving the tuberculosis epidemic in many countries. Among HIV infected 13 patients in resource limited settings, determinants for tuberculosis are not well 14 described. 15 For Accepted:peer Objective: review only 16 The objective of this study was to identify determinants for tuberculosis among HIV 17 18 infected adults in Northwest Ethiopia. 19 Methods: 20 A multicenter case control study was conducted in three hospitals and ten health 21 centers. Cases were HIV infected adults diagnosed with active tuberculosis and 22 controls were HIV infected adults without active tuberculosis. We recruited 158 cases 23 24 and 316 controls. Interviewers administered questionnaires and case report forms were 25 used to systematically collect the data. 26 Results: 27 A total of 446 study participants consented to participate in the study (150 cases and 28 296 controls). Being smoker (AOR (adjusted odds ratio) =5.47; 95%CI: 2.26, 13.22), 29 30 presence of tuberculosis patient in the family (AOR=2.66; 95%CI: 1.25, 5.66), alcohol 31 drinker (AOR=2.49; 95%CI: 1.29, 4.8) and chewing khat (AOR=2.22; 95%CI: 1.11, 32 4.41) were independent determinants for tuberculosis. Highly active antiretroviral
33 therapy (HAART) (AOR=0.25; 95%CI: 0.13, 0.51) had a protective benefit against http://bmjopen.bmj.com/ 34 35 tuberculosis. 36 Conclusion: 37 HIV infected adults who use substance (tobacco smoking, khat chewing and alcohol) 38 should be prioritized for tuberculosis screening. This study reaffirmed that initiation of 39 HAART is one of the best strategies to reduce tuberculosis occurrence in HIV infected 40 41 adults. Increased coverage of isoniazid and cotrimoxazole preventive therapy has a on September 24, 2021 by guest. Protected copyright. 42 protective benefit against active tuberculosis. Contact tracing of household contacts 43 for tuberculosis should be intensified. 44 45 46 Introduction 47 Background/rationale 2 INTRODUCTION 48 The advent of HIV was a massive setback for the prevention and control of 49 1 tuberculosis (TB). TB is the leading cause of morbidity and mortality among people 50 2 51 living with HIV (PLWHIV). There is a strong synergy between TB and HIV 52 infection in HIV high burden countries particularly in resource limited settings where 53 the impacts of both diseases are more significant.3 54 One third of the world’s population is infected with TB.2 In 2013, an estimated 9.0 55 million people developed TB; 1.1 million were HIV infected.4 Ethiopia ranks as the 56 5 57 seventh TB burden country in the world. 58 In resource limited settings the health care systems are overwhelmed by preventive, 59 therapeutic and diagnostic challenges of the "HIV TB syndemic". In Ethiopia; the 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml1 Page 19 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 dual epidemics drained resources and overburdened the already very limited health 3 work force.1 HIV is increases the life time risk of developing TB.6 However, HIV is 4 not the only determinant for developing TB; various other determinants contribute to 5 TB HIV co infection. Previous studies have identified socio demographic,7,8 6 9,10 11,12 13 7 clinical, life style and environmental determinants for TB. However, among 8 HIV infected adults, the determinants for TB are still not well described particularly in 9 a resource limited setting. Determinants for tuberculosis among HIV positive adults 10 vary from one setting to another. Therefore, a context specific study in a high 11 12 burdened TB/HIV region is indicated. 13 Objectives 3 This study aims to assess the determinants for TB among HIV infected adults in 14 Northwest Ethiopia. 15 For peer review only 16 Methods 17 Study design 4 We conducted a multicenter case control study from May 12 to June 5, 2014 in 18 Northwest Ethiopia. 19 Setting 5 All governmental health institutions in ten provinces were included (i.e. Bahir Dar, 20 Dangella, Kossober, Fnote Selam, Bure, Dur Betie, Deber Tabor, Wereta, Addis 21 22 Zemen and Nefas Mewcha). 23 Participants 6 Cases were HIV infected adults diagnosed with active TB and on TB treatment during 24 the data collection period. The diagnosis of active TB was based on one or more of TB 25 diagnostic investigations (sputum microscopy, x ray, histopathology, culture or 26 molecular). Controls were HIV infected adults without TB. All controls were actively 27 28 screened for TB to rule out TB. The screening procedure included history taking for 29 four symptoms at each visit (cough, weight loss, night sweating or fever). If the 30 patient presented with any one of these symptoms, this patient was a suspect TB case 31 and was investigated according to the national guidelines for TB HIV1. In hospitals, 32 sputum microscopy, histopathology and radiology examinations are used as the TB
33 http://bmjopen.bmj.com/ 34 diagnostic tools. In health centers, sputum microscopy is the main stay TB diagnostic 35 tool and patients will be sent for other diagnostic evaluation at hospitals and private 36 institutions where the diagnostic tools are available. 37 Variables 7 TB was the outcome variable. Diagnosis of TB stated above. The determinants of TB 38 39 were socio demographic variables (age, sex, educational and marital status), host 40 related variables (cigarette smoking, khat chewing and alcohol consumption), 41
environmental determinants included the presence of flooring, latrine and separate on September 24, 2021 by guest. Protected copyright. 42 kitchen in the household and clinical variables (HAART, CD4 cell count, CPT and 43 IPT) 44 45 Data sources/ 8* Data were collected from two sources. Trained nurses who were in charge of HIV care 46 measurement clinic conducted face to face interview with study participants by using structured 47 questionnaires that were prepared in English, translated to Amharic and back 48 translated and pretested for consistency and ease of understanding. The primary data 49 50 were collected to assess: socio demographic variables (age, sex, educational and 51 marital status), host related variables (cigarette smoking, khat chewing and alcohol 52 consumption). A person with substance use (chewing Kahat, alcohol consumption or 53 smoking) was defined as an individual who is currently using the substance or has a 54 history of substance use. Environmental determinants included the presence of 55 56 flooring, latrine and separate kitchen in the household. Trained nurses also collected 57 data from patient’s records to assess: clinical variables (HAART, CD4 cell count, CPT 58 and IPT). Data collection procedures were supervised by medical officers. 59 Bias 9 Identification of confounding variables by logistic regression models was applied. 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml2 BMJ Open Page 20 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 Study size 10 The sample size was calculated using Epi Info version 7 software. The study was 3 designed to have 80% statistical power with level of significance at 5% and a case to 4 control ratio of 1:2. Assuming the proportion of low CD4 cell count were 5.9% among 5 the controls and 13.9% among cases13 by using the two proportions formula, the 6 7 calculated sample size was 144 for cases and 287 for controls. Allowing for 10% of 8 non response, the resulting sample size was 474 (158 cases and 316 controls). The 9 sample size was calculated for exposure status of different variables: cotrimoxazole 10 preventive therapy (CPT), isoniazid preventive therapy (IPT) and CD4. We took the 11 12 largest sample among these exposure variables. 13 Quantitative variables 11 TB was the outcome variable; Yes or No response. Logistic regression models were 14 applied to assess the relationship between exposure variable and TB. 15 Statistical methodsFor 12 Datapeer were entered intoreview Epi info version 7 and onlyanalysis was done with SPSS version 16 20. Frequencies and proportions were used to describe the study subjects in relation to 17 18 the studied variables. Odds ratio with its 95% confidence interval and p value were 19 used to measure strength of association and identify statistical significance result. 20 Logistic regression models were applied to assess the relationship between 21 determinants and TB. Identification of confounding variables by logistic regression 22 23 models was applied. Multivariable analysis (backward stepwise) was used. Hosmer 24 Lemeshow test was applied and the fit of the model was checked. A poor fit of the 25 model if the significance p value is less than 0.05 and good fit greater than 0.05. Here 26 in this study the model adequately fits the data and the p value was 0.34. 27 Continued on next page 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml3 Page 21 of 22 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Results 4 5 Participants 13* A total of 446 subjects (150 cases=33.6%, 296 controls=66.4%) out of 474 eligible participants 6 responded and consented to participate in this study, resulting in an overall response rate of 7 94.1% (94.9% for cases and 93.7% for controls). 8 Participants were unable respond mainly because of taking time for interview, lack of 9 willingness for interview 10 11 Descriptive 14* The median age and its interquartile range (IQR) of cases were 32 years (IQR: 27 39 years) 12 data while the corresponding value for controls were 33 years (IQR: 28 39 years). A higher 13 percentage of women was observed in both groups; 81(54%) in cases and 188(63.5%) in 14 controls. 15 For peer review only 16 17 Outcome data 15* The bivariate analysis showed that male patients (COR (crude odds ratio) =1.48; 95% CI: 18 0.99, 2.12; p value = 0.053) were more likely to develop tuberculosis compared to female 19 patients. Married patients (COR=0.45; 95%CI: 0.27, 0.75) were less likely to develop TB 20 compared with never married. Cases were more likely to be smokers (COR= 4.73; 95% CI: 21 22 2.3, 9.72), Khat chewer (COR=2.9; 95%CI: 1.77, 4.75) and alcohol drinker (COR= 2.44; 1.54, 23 3.86). Study participants who had a history of TB (COR=1.85; 95%CI: 1.15, 2.96) were more 24 likely to develop TB. But educational status was not a determinant for TB. 25 HAART (COR=0.33; 95%CI: 0.19, 0.56), CPT (COR=0.35 95%CI: 0.22, 0.53) and IPT 26 (COR=0.30 95%CI: 0.18, 0.50) had protective benefit against TB. Subjects with separate 27 28 kitchen in the house hold (COR=0.61; 95%CI; 0.40, 0.93) were less likely to develop TB. HIV 29 infected adults with presence of a TB patient in the family (COR=1. 99; 95%CI: 1.11, 3.57) 30 were more likely to develop TB. But flooring and presence of latrine were not determinants for 31 developing TB. 32 Main results 16 To identify independent determinants for TB, a multivariable logistic regression model was
33 http://bmjopen.bmj.com/ 34 used. After adjusting for possible confounders, some variables were remained in the 35 multivariable model: chewing khat (AOR=2.22; 95%CI: 1.11, 4.41), being a smoker 36 (AOR=5.47; 95%CI: 2.26, 13.22) and alcohol drinker (AOR=2.49; 1.29, 4.80) were 37 independent determinants for increased TB occurrence. However, HAART (AOR=0.25; 38 39 95%CI: 0.12, 0.51), CPT (AOR=0. 32; 95% CI: 0.19, 0.52) and IPT (AOR=0.22; 95%CI: 0.11, 40 0.42) had an independent protective benefit effect against TB. Subjects with separate kitchen 41 in the house hold (AOR=0.48; 95%CI: 0.28, 0.83) was less likely to develop TB. Patients who on September 24, 2021 by guest. Protected copyright. 42 live with a TB patient in the household (AOR=2.66; 95%CI: 1.25, 5.66) were more likely to 43 develop TB. Patients whose CD4 cell count was < 200 cells/µl (AOR= 7.22; 95%CI: 3.39, 44 45 15.37) were more likely to develop TB. But sex, marital status and a previous history of TB 46 were not remained in the multivariable model 47 Other analyses 17 No analyses of subgroups and interactions, and sensitivity analyses 48 49 Discussion 50 Key results 18 Being smoker (AOR (adjusted odds ratio) =5.47; 95%CI: 2.26, 13.22), presence of 51 tuberculosis patient in the family (AOR=2.66; 95%CI: 1.25, 5.66), alcohol drinker 52 (AOR=2.49; 95%CI: 1.29, 4.8) and chewing khat (AOR=2.22; 95%CI: 1.11, 4.41) were 53 independent determinants for tuberculosis. Highly active antiretroviral therapy (HAART) 54 55 (AOR=0.25; 95%CI: 0.13, 0.51) had a protective benefit against tuberculosis. 56 Limitations 19 As we used a retrospective case control approach; temporal relationship could not be 57 established and the study design could not proof causation. Tuberculosis has a possibility to 58 spread to another people who is not a household member. But this study was considered, only 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml4 BMJ Open Page 22 of 22 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 family members were close tuberculosis contacts. Information bias related to substance use 3 (chewing chat, smoking or alcohol consumption) might have also affected the accuracy of 4 information. 5 6 Interpretation 20 HIV infected adults who use substance (tobacco smoking, khat chewing and alcohol) should 7 be prioritized for tuberculosis screening. This study reaffirmed that initiation of HAART is one 8 of the best strategies to reduce tuberculosis occurrence in HIV infected adults. Increased 9 coverage of isoniazid and cotrimoxazole preventive therapy has a protective benefit against 10 active tuberculosis. Contact tracing of household contacts for tuberculosis should be 11 12 intensified. 13 Generalisability 21 This study will be generalized to HIV positive adults in Hence Awie, West Gojjam and South 14 Gonar, Northwest Ethiopia 15 For peer review only 16 Other information 17 Funding 22 PAGEL (Partnerships for the health sector in developing countries offers German institutions 18 of higher education many different opportunities for international cooperation related to the 19 health sector in developing countries) funded this article. The points expressed in this article 20 are the responsibility of authors, do not reflect the view of PAGEL. The funding source had no 21 22 role in design, analysis, wrote paper, manuscript preparation or decision for publication. 23 24 *Give information separately for cases and controls in case control studies and, if applicable, for exposed and 25 unexposed groups in cohort and cross sectional studies. 26
27 28 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 29 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 30 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 31 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 32 available at www.strobe statement.org.
33 http://bmjopen.bmj.com/ 34 35 Author notice: This checklist is a very good guide to prepare an observational research report. 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml5 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from
Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a multicenter case control study
ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2015-009058.R2
Article Type: Research
Date Submitted by the Author: 17-Nov-2015
Complete List of Authors: Alemu, Yihun Mulugeta; Heidelberg University,Institute of Public Health; Bahir Dar University , School of Public Health Awoke, Worku; Bahir Dar University, School of Public Health Annelies, Wilder-Smith; Heidelberg University, Institute of Public Health; Lee Kong Chian School of Medicine , Nanyang Technological University
Primary Subject Global health Heading:
Secondary Subject Heading: Epidemiology, Global health
PUBLIC HEALTH, Epidemiology < THORACIC MEDICINE, Clinical trials < Keywords: THERAPEUTICS
http://bmjopen.bmj.com/
on September 24, 2021 by guest. Protected copyright.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a 6 7 multicenter case control study 8 9 10 11 1, 2* 2 1,3 12 Yihun Mulugeta Alemu , Worku Awoke , Wilder Smith Annelies 13 14 15 For peer review only 16 17 18 1Institute of Public Health, Heidelberg University, Germany 19 20 21 2School of Public Health, College of Medicine and Health Science, Bahir Dar University, 22 Ethiopia 23 24 25 3Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore 26 27 28 29 30 Corresponding author; Yihun Mulugeta, Alemu E: mail: [email protected] 31 32
33 http://bmjopen.bmj.com/ 34 35 Determinants of tuberculosis in HIV infected adults 36 37 Key words: Determinants, tuberculosis, HIV, adults, multi-center 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 1 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 Abstract 5 6 Objective: The objective of this study was to identify determinants for tuberculosis among 7 8 HIV infected adults in Northwest Ethiopia. 9 10 11 Design: Case control study. 12 13 14 Setting: Three hospitals and ten health centers in Northwest Ethiopia. 15 For peer review only 16 17 18 Participants: A total of 446 study participants consented to participate in the study (150 cases 19 20 21 and 296 controls). Cases were HIV infected adults diagnosed with active tuberculosis and 22 23 controls were HIV infected adults without active tuberculosis. 24 25 26 27 Main outcome measure: The link between tuberculosis and determinants was assessed by 28 29 using logistic regression. Determinants were categorized as sociodemographic, host related, 30 31 32 clinical and environmental.
33 http://bmjopen.bmj.com/ 34 35 Results: Smoking had an Adjusted Odds Ratio (AOR) of 5.47 (95%CI: 2.26, 13.22), presence 36 37 of tuberculosis patient in the family an AOR of 2.66 (95%CI: 1.25, 5.66), alcohol consumption 38 an AOR of 2.49 (95%CI: 1.29, 4.8) and chewing khat an AOR of 2.22 (95% CI: 1.11, 4.41) were 39 40 independent determinants for increased occurrence of tuberculosis. Highly active antiretroviral 41 on September 24, 2021 by guest. Protected copyright. 42 therapy (HAART) an AOR of 0.25 (95%CI: 0.13, 0.51), isoniazid preventive therapy (IPT) an 43 44 AOR of 0.22 (95% CI: 0.11, 0.42) and cotrimoxazole preventive therapy an AOR of 0.32 (95% 45 CI: 0.19, 0.52) had protective benefit against tuberculosis. 46 47 Conclusion: HIV infected adults with substance abuse (tobacco smoking, khat chewing and 48 49 alcohol) should be prioritized for tuberculosis screening. This study reaffirmed that HAART and 50 51 IPT are some of the best strategies to reduce tuberculosis occurrence in HIV infected adults. 52 53 These findings provide impetus to intensify contact tracing in TB household contacts. 54 55 56 57 2 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 ARTICLE SUMMARY 4 5 Strength and limitation of the study 6 7 This is the first multicenter case control study in Northwest Ethiopia that investigates 8 9 determinants for tuberculosis in HIV infected adults. 10 11 The study identified determinants for tuberculosis that will be important to prioritize TB 12 screening, treatment, prevention and control. 13 14 This study reaffirmed that HAART and isoniazid therapy are some of the best strategies to 15 For peer review only 16 reduce tuberculosis among HIV infected adults in resource limited settings. 17 18 As we used a retrospective case control approach; temporal relationship could not be established; 19 the study design could not proof causation. 20 21 22 23 24 25 26 27 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 3 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 INTRODUCTION 4 5 The advent of Human Immuno deficiency Virus (HIV) was a massive setback for the prevention 6 7 and control of tuberculosis (TB).1 TB is the leading cause of morbidity and mortality among 8 2 9 people living with HIV. There is a strong synergy between TB and HIV infection in HIV high 10 11 burden countries particularly in resource limited settings where the impacts of both diseases are 12 more significant.3 13 14 One third of the world’s population is infected with TB.2 In 2013, an estimated 9.0 million 15 For peer review4 only 16 people developed TB; 1.1 million were HIV infected. Ethiopia ranks as the seventh TB burden 17 5 18 country in the world. 19 In resource limited settings the health care systems are overwhelmed by preventive, therapeutic 20 21 and diagnostic challenges of the "HIV TB syndemic". In Ethiopia; the dual epidemics drained 22 1 23 resources and overburdened the already very limited health work force. HIV is increases the life 24 6 25 time risk of developing TB. However, HIV is not the only determinant for developing TB; 26 various other determinants contribute to TB HIV co infection. Previous studies have identified 27 28 socio demographic,7,8 clinical,9,10 life style11,12 and environmental13 determinants for TB. 29 30 However, among HIV infected adults, the determinants for TB are still not well described 31 32 particularly in a resource limited setting. Determinants for tuberculosis among HIV positive
33 http://bmjopen.bmj.com/ 34 adults vary from one setting to another. Therefore, a context specific study in a high burdened 35 TB/HIV region is indicated. This study aims to assess the determinants for TB among HIV 36 37 infected adults in Northwest Ethiopia. 38 39 METHODS 40 41 Study design on September 24, 2021 by guest. Protected copyright. 42 We conducted a multicenter case control study from May 12 to June 5, 2014 in Northwest 43 44 Ethiopia. All governmental health institutions in ten districts were included (i.e. Bahir Dar, 45 46 Dangella, Kossober, Fnote Selam, Bure, Dur Betie, Deber Tabor, Wereta, Addis Zemen and 47 48 Nefas Mewcha). 49 50 Participants 51 Cases were HIV infected adults diagnosed with active TB and on TB treatment during the data 52 53 collection period. The diagnosis of active TB was based on one or more of TB diagnostic 54 55 investigations (sputum microscopy, x ray, histopathology, culture or molecular). Controls were 56 57 4 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 HIV infected adults without TB. All controls were actively screened for TB to rule out TB. The 4 5 screening procedure included history taking for four symptoms at each visit (cough, weight loss, 6 7 night sweating or fever). If the patient presented with any one of these symptoms, this patient 8 1 9 was a suspect TB case and was investigated according to the national guidelines for TB HIV . In 10 11 hospitals, sputum microscopy, histopathology and radiology examinations are used as the TB 12 diagnostic tools. In health centers, sputum microscopy is the main stay TB diagnostic tool and 13 14 patients will be sent for other diagnostic evaluation to hospitals and private institutions where the 15 For peer review only 16 diagnostic tools are available. 17 18 Sample size calculation 19 20 The sample size was calculated using Epi Info version 7 software. The study was designed to 21 22 have 80% statistical power with level of significance at 5% and a case to control ratio of 1:2. 23 24 Assuming the proportion of low CD4 cell count were 5.9% among the controls and 13.9% 25 among cases13 by using the two proportions formula, the calculated sample size was 144 for 26 27 cases and 287 for controls. Allowing for 10% of non response, the resulting sample size was 474 28 29 (158 cases and 316 controls). The sample size was calculated for exposure status of different 30 31 variables: cotrimoxazole preventive therapy (CPT), isoniazid preventive therapy (IPT) and CD4. 32 We took the largest sample among these exposure variables.
33 http://bmjopen.bmj.com/ 34 Sampling procedures 35 36 A total of fifty two health institutions are providing HIV care service in 44 districts of Awie, 37 14 38 West Gojjam and South Gondar Zones. Ten districts were randomly selected and all health 39 40 institutions in the ten districts were included in the study. Three hospitals and ten health centers 41 were included. HIV infected adults diagnosed with confirmed active TB were recruited as cases on September 24, 2021 by guest. Protected copyright. 42 43 while those without active TB were recruited as controls. Study subjects who were unable to give 44 45 informed consent and subjects with suspected but unconfirmed TB were excluded. All TB HIV 46 47 co infected patients attending HIV care clinic and those who were taking TB treatment were 48 included. Controls were allotted based on the number of cases available in each facility with the 49 50 control to case ratio of 2:1 and sampled by systematic random sampling with a sampling interval 51 52 of five. 53 54 Data collection and analysis 55 56 57 5 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Data were collected from two sources. Trained nurses who were in charge of HIV care clinic 4 5 conducted face to face interview with study participants by using structured questionnaires that 6 7 were prepared in English, translated to Amharic and back translated and pretested for 8 9 consistency and ease of understanding. The primary data were collected to assess: socio 10 11 demographic variables (age, sex, educational and marital status), host related variables (cigarette 12 smoking, khat chewing and alcohol consumption). A person with substance abuse (chewing 13 14 Kahat, alcohol consumption or smoking) was defined as an individual who is currently using the 15 For peer review only 16 substance or has a history of substance use regularly. Environmental determinants included the 17 18 presence of flooring, latrine and separate kitchen in the household. Trained nurses also collected 19 data from patient’s records to assess: clinical variables (HAART, CD4 cell count, CPT and IPT). 20 21 Data collection procedures were supervised by medical officers. 22 23 Data were entered into Epi info version 7 and analysis was done with SPSS version 20. 24 25 Frequencies and proportions were used to describe the study subjects in relation to the studied 26 variables. Odds ratio with its 95% confidence interval and p value were used to measure strength 27 28 of association and identify statistical significance result. Logistic regression models were applied 29 30 to assess the relationship between determinants and TB. Identification of confounding variables 31 32 by logistic regression models was applied. Multivariable analysis (backward stepwise) was used.
33 http://bmjopen.bmj.com/ 34 Hosmer Lemeshow test was applied and the fit of the model was checked. A poor fit of the 35 model if the significance p value is less than 0.05 and good fit greater than 0.05. Here in this 36 37 study the model adequately fits the data and the p value was 0.34. 38 39 Ethical considerations 40 41 Ethical approval from the Heidelberg Ethics Commission was obtained. Additionally, ethical on September 24, 2021 by guest. Protected copyright. 42 approval from the country of research, Amhara Regional Research and Ethical Core Process, 43 44 Bahir Dar, Ethiopia was granted. Written permission to conduct the study was obtained from 45 46 each health institution involved in the study. Since there were illiterate participants the data 47 48 collectors informed each study participant about the informed consent sheet. Informed oral 49 50 consent was obtained from each study participant. The data collectors documented the 51 participant’s consent on the informed consent sheet. This consent procedure was approved by the 52 53 ethical commission. 54 55 56 57 6 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 RESULTS 12 Socio demographic and host related determinants of HIV infected adults: 13 14 A total of 446 subjects (150 cases=33.6%, 296 controls=66.4%) out of 474 eligible participants 15 For peer review only 16 responded and consented to participate in this study, resulting in an overall response rate of 17 18 94.1% (94.9% for cases and 93.7% for controls). 19 The median age and its interquartile range (IQR) of cases were 32 years (IQR: 27 39 years) 20 21 while the corresponding value for controls were 33 years (IQR: 28 39 years). A higher 22 23 percentage of women was observed in both groups; 81(54%) in cases and 188(63.5%) in controls 24 25 (Table1). 26 Table 1. Socio demographic and host related determinants of HIV infected adults in Northwest 27 28 Ethiopia, 2014 29 30 Variables Case n (%) Control n (%) Total n (%) 31 32 Sex
33 http://bmjopen.bmj.com/ 34 Male 69(46) 108(36.5) 177(39.7) 35 36 Female 81(54) 188(63.5) 269(60.3) 37 Education 38 39 No formal education 46(30.7) 120(40.5) 166(37.2) 40 41
Primary 48(32) 78(26.4) 126(28.3) on September 24, 2021 by guest. Protected copyright. 42 43 Secondary 42(28) 60(20.2) 102(22.9) 44 45 Tertiary 14(9.3) 38(12.8) 52(11.6) 46 Marital status 47 48 Married 53(35.3) 141(47.6) 194(43.5) 49 50 Divorce / widowed 50(33.3) 98(33.1) 148(33.2) 51 52 Never married 47(31.3) 57(19.3) 104(23.3) 53 54 Smoking 55 56 57 7 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Yes 25(16.7) 12(4) 37(8.3) 4 5 No 125(83.3) 284(96) 409(91.7) 6 7 Khat chewing 8 9 Yes 44(29.3) 37(12.5) 81(18) 10 11 No 106(70.7) 259(87.5) 365(82) 12 13 Alcohol 14 Yes 49(32.7) 49(16.6) 98(22) 15 For peer review only 16 No 101(67.3) 247(83.4) 348(78) 17 18 Previous history of TB 19 20 Yes 41(27.3) 50(16.9) 91(20.4) 21 22 No 109(72.7) 246(83.1) 355(79.6) 23 24 25 Bivariate Analysis: 26 27 Although, in bivariate analysis male patients had 1.48 times more likely to have TB compared to 28 29 female patients, the association is not statically significant. Married patients (COR=0.45; 95%CI: 30 31 0.27, 0.75) were less likely to develop TB compared with never married. Cases were more likely 32 to be smokers (COR= 4.73; 95% CI: 2.3, 9.72), Khat chewer (COR=2.9; 95%CI: 1.77, 4.75) and
33 http://bmjopen.bmj.com/ 34 alcohol drinker (COR= 2.44; 1.54, 3.86). Study participants who had a history of TB 35 36 (COR=1.85; 95%CI: 1.15, 2.96) were more likely to develop TB. But educational status was not 37 38 a determinant for TB (Table 2). 39 Table 2: Socio demographic and host related determinants for developing tuberculosis among 40 41 HIV infected adults in Northwest Ethiopia, 2014 on September 24, 2021 by guest. Protected copyright. 42 43 Variables Cases n (%) Controls n (%) COR 95%CI P value 44 45 Sex 46 47 Male 69(46) 108(36.5) 1.48 0.99, 2.12 0.053 48 Female 81(54) 188(63.5) 1 49 50 Education 51 52 No formal education 46(30.7) 120(40.5) 1.04 0.51, 2.09 0.912 53 54 Primary 48(32) 78(26.4) 1.67 0.82, 3.39 0.157 55 56 57 8 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Secondary 42(28) 60(20.2) 1.90 0.9,3.93 0.084 4 5 Tertiary 14(9.3) 38(12.8) 1 6 7 Marital status 8 9 Married 53(35.3) 141(47.6) 0.45 0.27, 0.75 0.002* 10 11 Divorce / widowed 50(33.3) 98(33.1) 0.61 0.37, 1.03 0.068 12 13 Never married 47(31.3) 57(19.3) 1 14 Smoking 15 For peer review only 16 Yes 25(16.7) 12(4) 4.73 2.30, 9.72 < 0.0001* 17 18 No 125(83.3) 284(96) 1 19 20 Khat chewing 21 22 Yes 44(29.3) 37(12.5) 2.9 1.77, 4.75 < 0.0001* 23 No 106(70.7) 259(87.5) 1 24 25 Alcohol 26 27 Yes 49(32.7) 49(16.6) 2.44 1.54, 3.86 < 0.0001* 28 29 No 101(67.3) 247(83.4) 1 30 31 Previous history of TB 32 Yes 41(27.3) 50(16.9) 1.85 1.15, 2.96 0.01*
33 http://bmjopen.bmj.com/ 34 No 109(72.7) 246(83,1) 1 35 36 37 38 HAART (COR=0.33; 95%CI: 0.19, 0.56), CPT (COR=0.35 95%CI: 0.22, 0.53) and IPT 39 40 (COR=0.30 95%CI: 0.18, 0.50) had protective benefit against TB. Subjects with separate kitchen 41 on September 24, 2021 by guest. Protected copyright. 42 in the house hold (COR=0.61; 95%CI; 0.40, 0.93) were less likely to develop TB. HIV infected 43 adults with presence of a TB patient in the family (COR=1. 99; 95%CI: 1.11, 3.57) were more 44 45 likely to develop TB. But flooring and presence of latrine were not determinants for developing 46 47 TB (Table 3). 48 49 Table 3: Clinical and environmental determinants for developing tuberculosis among HIV 50 infected adults in Northwest Ethiopia, 2014 51 52 Variables Cases n (%) Controls n (%) COR 95%CI P value 53 54 HAART 55 56 57 9 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Yes 114(76) 268(90.5) 0.33 0.19, 0.56 < 0.0001* 4 5 No 36(24) 28(9.5) 1 6 7 CPT 8 9 Yes 83(55.3) 231(78) 0.35 0.22, 0.53 < 0.0001* 10 11 No 67(44.7) 65(22) 1 12 13 IPT 14 Yes 22(14.7) 107(36.1) 0.30 0.18, 0.50 < 0.0001* 15 For peer review only 16 No 128(85.3) 189(63.9) 1 17 18 CD4 cell count (cells/ µl) 19 20 <=200 87(58) 61(20.6) 5.53 2.96, 10.35 < 0.0001* 21 22 200 – 500 46(30.7) 169(57.1) 1.05 0.56, 1.97 0.86 23 >=500 17(11.3) 66 (22.3) 1 24 25 TB patient in the family 26 27 Yes 25(16.7) 27(9.1) 1.99 1.11, 3.57 0.021* 28 29 No 125(83.3) 269(90.9) 1 30 31 Separate kitchen 32 Yes 96(64) 220(74.3) 0.61 0.40, 0.93 0.024*
33 http://bmjopen.bmj.com/ 34 No 54(36) 76(25.7) 1 35 36 Floor of the house 37 38 Mud / Soil 123(82) 240(81) 1.06 0.64, 1.76 0.814 39 40 Cement 27(18) 56(19) 1 41 on September 24, 2021 by guest. Protected copyright. 42 Latrine 43 Yes 136(90.7) 267(90.2) 1.05 0.53, 2.06 0.875 44 45 No 14(9.3) 29(9.8) 1 46 47 48 49 Multivariable analysis: 50 51 To identify independent determinants for TB, a multivariable logistic regression model was used. 52 After adjusting for possible confounders, some variables were remained in the multivariable 53 54 model: chewing khat (AOR=2.22; 95%CI: 1.11, 4.41), being a smoker (AOR=5.47; 95%CI: 55 56 2.26, 13.22) and alcohol drinker (AOR=2.49; 1.29, 4.80) were independent determinants for 57 10 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 increased TB occurrence. However, HAART (AOR=0.25; 95%CI: 0.12, 0.51), CPT (AOR=0. 4 5 32; 95% CI: 0.19, 0.52) and IPT (AOR=0.22; 95%CI: 0.11, 0.42) had an independent protective 6 7 benefit effect against TB. Subjects with separate kitchen in the house hold (AOR=0.48; 95%CI: 8 9 0.28, 0.83) was less likely to develop TB. Patients who live with a TB patient in the household 10 11 (AOR=2.66; 95%CI: 1.25, 5.66) were more likely to develop TB. Patients whose CD4 cell count 12 was < 200 cells/µl (AOR= 7.22; 95%CI: 3.39, 15.37) were more likely to develop TB. But sex, 13 14 marital status and a previous history of TB were not remained in the multivariable model (Table 15 For peer review only 16 4). 17 18 Table 4: Independent determinants for tuberculosis among HIV infected adults in 19 Northwest Ethiopia, 2014 20 21 Variables COR(95%CI) P value AOR(95%CI) P value 22 23 HAART 24 25 Yes 0.33(0.19, 0.56) < 0.0001 0.25(0.13, 0.51) < 0.0001 26 27 No 1 1 28 29 CPT 30 Yes 0.35(0.22, 0.53) < 0.0001 0.32(0.19, 0.55) < 0.0001 31 32 No 1 1
33 http://bmjopen.bmj.com/ 34 IPT 35 36 Yes 0.30(0.18, 0.50) < 0.0001 0.22(0.11, 0.41) < 0.0001 37 38 No 1 1 39 Smoking 40 41 Yes 4.73(2.30, 9.72) < 0.0001 5.47(2.26, 13.22) < 0.0001 on September 24, 2021 by guest. Protected copyright. 42 43 No 1 1 44 45 Khat 46 47 Yes 2.9(1.77, 4.75) <0.001 2.22(1.11, 4.41) 0.023 48 No 1 1 49 50 Alcohol 51 52 Yes 2.44(1.54, 3.86) <0.0001 2.49(1.29, 4.80) 0.006 53 54 No 1 1 55 56 57 11 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 TB patient in the family 4 5 Yes 1.99(1.11, 3.57) 0.021 2.66(1.25,5.66) 0.011 6 7 No 1 1 8 9 Separate kitchen 10 11 Yes 0.61(0.40, 0.93) 0.024 0.48(0.28, 0.83) 0.010 12 13 No 1 1 14 CD4 cell count (cells/ µl) 15 For peer review only 16 < 200 cells/ µl 5.53(2.96, 10.35) < 0.0001 7.22(3.39, 15.37) < 0.0001 17 18 200 – 500 cells/ µl 1.05(0.56,1.97) 0.86 1.31(0.63, 2.74) 0.461 19 20 >= 500 cells/ µl 1 1 21 22 23 DISCUSSION 24 25 This study showed that HIV positive adults on HAART, CPT and IPT were less likely to develop 26 27 TB. But smoking, alcohol consumption, Khat chewing, presence of TB patient in the family and 28 29 CD4 count less than 200 cells/µl were independent determinants for increased TB occurrence. 30 31 Similar to the finding of other study, this study also showed that level of education was not a 32 determinant for TB occurrence.7 Hence, increasing level of school training may not necessarily
33 http://bmjopen.bmj.com/ 34 bring successful behavioral change to TB prevention and control in this source population. 35 36 Married HIV positive adults were also less likely to develop TB compared to those who were 37 8 38 never married; consistent with a study. Previous studies have shown that increased marriage 39 prevalence is observed in higher income society15 and that tuberculosis and poverty are 40 41 16 interrelated. on September 24, 2021 by guest. Protected copyright. 42 43 HIV positive patients with low CD4 count were more likely to develop TB. It was consistent 44 7 45 with the previous study. In addition, our findings indicating that IPT had a protective benefit 46 against TB was also consistent with a previous research report.17 IPT reduces the mycobacterium 47 48 load and reduces the progression of latent TB to active tuberculosis.9 But increase mycobacterial 49 50 load is coupled with progressive impairment of mycobacterium specific T cell response.18 51 52 Study participants on HAART were less likely to develop TB. It was consistent with 53 10 54 publication. This could be due to strong relation between increasing HAART coverage and 55 56 57 12 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 decreased viral load.11 Lowering of viral load was also linked with reduced occurrence of active 4 5 TB.19 6 7 We found that cotrimoxazole preventive therapy has protective benefit against tuberculosis. This 8 13 9 was consistent with previous study. Cotrimoxazole preventive therapy has beneficial effects on 10 20 11 enhancing CD4 count as well as reduction of viral load. 12 Cigarette smoking induces immune impairment and damage the ciliary clearance.21,22 In this 13 14 study smoking was an independent determinants for TB. It was consistent with a previous 15 23 For peer review only 16 study. 17 12 18 Khat chewing is linked with immune modulation. Among HIV infected adults, no published 19 study has detected khat chewing as an independent determinant for TB. In this study Khat 20 21 chewing was an independent determinant for TB. 22 23 Our findings showed that study subjects who consumed alcohol were more likely to develop TB. 24 24 26 25 It was consistent with previous studies. Alcohol can modulate the immune response. Long 26 term alcohol consumers have been shown to have impaired immunity.27 27 28 Past history of tuberculosis was a determinant for increased TB occurrence. But a past medical 29 30 history of TB was shown to be a determinant for TB in the bi variable analysis only. This may be 31 2 32 explained by the lower TB case detection rate in Ethiopia. The established tuberculosis case
33 http://bmjopen.bmj.com/ 34 detection mechanisms overlooked some tuberculosis patients. A proportion of tuberculosis 35 patient may remain undiagnosed i.e. died with other means of diseases or died without visiting 36 37 health care; the detected proportion of reoccurrence might be low. Therefore the prevalence of 38 39 reoccurrence was underestimated. 40 41 This study has the following potential limitations: as we used a retrospective case control on September 24, 2021 by guest. Protected copyright. 42 approach; temporal relationship could not be established and the study design could not proof 43 44 causation. Tuberculosis has a possibility to spread to another people who is not a household 45 46 member. But this study was considered, only family members were close tuberculosis contacts. 47 48 Information bias related to substance use (chewing chat, smoking or alcohol consumption) might 49 50 have also affected the accuracy of information. 51 CONCLUSION 52 53 HIV infected adults with substance abuse (tobacco smoking, khat chewing and alcohol) should 54 55 be prioritized for tuberculosis screening. This study reaffirmed that HAART and IPT are some of 56 57 13 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 the best strategies to reduce tuberculosis occurrence in HIV infected adults. These findings 4 5 provide impetus to intensify contact tracing in TB household contacts. 6 7 Data sharing statement 8 9 No additional data available. 10 11 Competing interests 12 No, there are no competing interests. 13 14 Acknowledgement 15 For peer review only 16 The authors gratefully acknowledge the Amhara Regional Health Bureau Research Core Process 17 18 Team, hospitals and health centers, data collectors, supervisors and study participants involved in 19 the study. 20 21 Author Contributions 22 23 Conceived and designed the experiments: YMA WA AWS. Performed the experiments: YMA 24 25 WA AWS. Analyzed the data: YMA WA AWS. Contributed materials/analysis tools: YMA WA 26 AWS. Wrote the paper: YMA WA AWS. 27 28 Funding: PAGEL (Partnerships for the health sector in developing countries offers German 29 30 institutions of higher education many different opportunities for international cooperation related 31 32 to the health sector in developing countries) funded this article. The points expressed in this
33 http://bmjopen.bmj.com/ 34 article are the responsibility of authors, do not reflect the view of PAGEL. The funding source 35 had no role in design, analysis, wrote paper, manuscript preparation or decision for publication. 36 37 References 38 39 1. Federal Ministry of Health Ethiopia (2013) Guide lines for clinical and program management of 40 41 TB, TB/HIV and leprosy. Ministry of Health, 2013. on September 24, 2021 by guest. Protected copyright. 42 2. World Health Organization (2013) Global tuberculosis report. WHO, 2013. 43 44 3. Kwan CK, Ernst JD: HIV and tuberculosis: a deadly human syndemic. Clinical microbiology 45 46 reviews 24:351 376, 2011. 47 48 4. United States Agency for International Development (2013) Report on the global AIDS epidemic. 49 50 USAID, 2013. 51 5. World Health Organization (2014) Global tuberculosis report. WHO, 2014. 52 53 6. Pawlowski A, Jansson M, Sköld M, et al: Tuberculosis and HIV Co Infection. PLoS Pathog 54 55 8:e1002464, 2012. 56 57 14 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 7. Batista J, de Albuquerque Mde F, Maruza M, et al: Incidence and risk factors for tuberculosis in 4 5 people living with HIV: cohort from HIV referral health centers in Recife, Brazil. PloS one 6 7 8:e63916, 2013. 8 9 8. M. Tekkel, M. Rahu, H M. Loit AB: Socio economic status and duration of TB symptoms in 10 11 males treated at the Mazovian Treatment Centre of Tuberculosis and Lung Diseases in Otwock. 12 International Journal of Tuberculosis and Lung Disease 11/2002; 6:887 894., 2002. 13 14 9. Wilkinson D: Drugs for preventing tuberculosis in HIV infected persons. Cochrane Database Syst 15 For peer review only 16 Rev 4, 2000. 17 18 10. Suthar AB, Lawn SD, del Amo J, et al: Antiretroviral Therapy for Prevention of Tuberculosis in 19 Adults with HIV: A Systematic Review and Meta Analysis. PLoS Med 9:e1001270, 2012. 20 21 11. Montaner JS, Lima VD, Barrios R, et al: Association of highly active antiretroviral therapy 22 23 coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a 24 25 population based study. Lancet 376:532 539, 2010. 26 12. Alvi A, Rizwan M, Sunosi RA, et al: Does khat chewing increases the risk of Mycobacterium 27 28 tuberculosis infection by macrophage immune modulation? Med Hypotheses 82:667 669, 2014. 29 30 13. Kibret KT, Yalew AW, Belaineh BG, et al: Determinant factors associated with occurrence of 31 32 tuberculosis among adult people living with HIV after antiretroviral treatment initiation in Addis
33 http://bmjopen.bmj.com/ 34 Ababa, Ethiopia: a case control study. PLoS One 8:e64488, 2013. 35 14. Amhara Regional Health Bureau (2013) Amhara Health Profile 2013. Amhara Health Bureau, 36 37 Amhara, Ethiopia, 2013. 38 39 15. Albrecht DE, Albrecht CM: The Implications of Economic Structure for Marriage Prevalence. 40 41 The Open Sociology Journal, 2008. on September 24, 2021 by guest. Protected copyright. 42 16. Zammarchi L, Bartalesi F, Bartoloni A: Tuberculosis in tropical areas and immigrants. Mediterr J 43 44 Hematol Infect Dis 6:e2014043, 2014. 45 46 17. Golub JE, Saraceni V, Cavalcante SC, et al: The impact of antiretroviral therapy and isoniazid 47 48 preventive therapy on tuberculosis incidence in HIV infected patients in Rio de Janeiro, Brazil. 49 50 Aids 21:1441 1448, 2007. 51 18. Day CL, Moshi ND, Abrahams DA, et al: Patients with tuberculosis disease have 52 53 Mycobacterium tuberculosis specific CD8 T cells with a pro apoptotic phenotype and impaired 54 55 proliferative capacity, which is not restored following treatment. PLoS One 9:e94949, 2014. 56 57 15 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 19. Moreno S, Jarrin I, Iribarren JA, et al: Incidence and risk factors for tuberculosis in HIV positive 4 5 subjects by HAART status. Int J Tuberc Lung Dis 12:1393 1400, 2008. 6 7 20. Mermin J, Lule J, Ekwaru JP, et al: Effect of co trimoxazole prophylaxis on morbidity, mortality, 8 9 CD4 cell count, and viral load in HIV infection in rural Uganda. Lancet 364:1428 1434, 2004. 10 11 21. Sopori ML, Kozak W: Immunomodulatory effects of cigarette smoke. J Neuroimmunol 83:148 12 156, 1998. 13 14 22. Den Boon S, van Lill SWP, Borgdorff M, et al: Association between smoking and tuberculosis 15 For peer review only 16 infection: a population survey in a high tuberculosis incidence area. Thorax 60:555 557, 2005. 17 18 23. Leung CC, Yew WW, Chan CK, et al: Smoking and tuberculosis in Hong Kong. Int J Tuberc 19 Lung Dis 7:980 986, 2003. 20 21 24. Rabirad N, Mohammad Nejad E, Hadizadeh MR, et al: The Prevalence of Tb in HIV Patients and 22 23 Risk Factor With Frequent Referral (Iran, 2009 10). Iran Red Crescent Med J 15:58 61, 2013. 24 25 25. Gajalakshmi V, Peto R: Smoking, drinking and incident tuberculosis in rural India: population 26 based case control study. Int J Epidemiol 38:1018 1025, 2009. 27 28 26. Rehm J, Samokhvalov AV, Neuman MG, et al: The association between alcohol use, alcohol use 29 30 disorders and tuberculosis (TB). A systematic review. BMC Public Health 9:450, 2009. 31 32 27. Szabo G: Alcohol's contribution to compromised immunity. Alcohol Health Res World 21:30 41,
33 http://bmjopen.bmj.com/ 34 1997. 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 16 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 STROBE Statement—checklist of items that should be included in reports of observational studies 3 4 5 Item No Recommendation 6 7 Title and abstract 1 Determinants for tuberculosis in HIV infected adults in Northwest Ethiopia: a 8 multicenter case control study 9 Accepted: Background: 10 Tuberculosis presents a major infectious disease burden in resource limited settings. 11 12 HIV is driving the tuberculosis epidemic in many countries. Among HIV infected 13 patients in resource limited settings, determinants for tuberculosis are not well 14 described. 15 For Accepted:peer Objective: review only 16 The objective of this study was to identify determinants for tuberculosis among HIV 17 18 infected adults in Northwest Ethiopia. 19 Methods: 20 A multicenter case control study was conducted in three hospitals and ten health 21 centers. Cases were HIV infected adults diagnosed with active tuberculosis and 22 controls were HIV infected adults without active tuberculosis. We recruited 158 cases 23 24 and 316 controls. Interviewers administered questionnaires and case report forms were 25 used to systematically collect the data. 26 Results: 27 A total of 446 study participants consented to participate in the study (150 cases and 28 296 controls). Being smoker (AOR (adjusted odds ratio) =5.47; 95%CI: 2.26, 13.22), 29 30 presence of tuberculosis patient in the family (AOR=2.66; 95%CI: 1.25, 5.66), alcohol 31 drinker (AOR=2.49; 95%CI: 1.29, 4.8) and chewing khat (AOR=2.22; 95%CI: 1.11, 32 4.41) were independent determinants for tuberculosis. Highly active antiretroviral
33 therapy (HAART) (AOR=0.25; 95%CI: 0.13, 0.51) had a protective benefit against http://bmjopen.bmj.com/ 34 35 tuberculosis. 36 Conclusion: 37 HIV infected adults who use substance (tobacco smoking, khat chewing and alcohol) 38 should be prioritized for tuberculosis screening. This study reaffirmed that initiation of 39 HAART is one of the best strategies to reduce tuberculosis occurrence in HIV infected 40 41 adults. Increased coverage of isoniazid and cotrimoxazole preventive therapy has a on September 24, 2021 by guest. Protected copyright. 42 protective benefit against active tuberculosis. Contact tracing of household contacts 43 for tuberculosis should be intensified. 44 45 46 Introduction 47 Background/rationale 2 INTRODUCTION 48 The advent of HIV was a massive setback for the prevention and control of 49 1 tuberculosis (TB). TB is the leading cause of morbidity and mortality among people 50 2 51 living with HIV (PLWHIV). There is a strong synergy between TB and HIV 52 infection in HIV high burden countries particularly in resource limited settings where 53 the impacts of both diseases are more significant.3 54 One third of the world’s population is infected with TB.2 In 2013, an estimated 9.0 55 million people developed TB; 1.1 million were HIV infected.4 Ethiopia ranks as the 56 5 57 seventh TB burden country in the world. 58 In resource limited settings the health care systems are overwhelmed by preventive, 59 therapeutic and diagnostic challenges of the "HIV TB syndemic". In Ethiopia; the 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml1 BMJ Open Page 18 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 dual epidemics drained resources and overburdened the already very limited health 3 work force.1 HIV is increases the life time risk of developing TB.6 However, HIV is 4 not the only determinant for developing TB; various other determinants contribute to 5 TB HIV co infection. Previous studies have identified socio demographic,7,8 6 9,10 11,12 13 7 clinical, life style and environmental determinants for TB. However, among 8 HIV infected adults, the determinants for TB are still not well described particularly in 9 a resource limited setting. Determinants for tuberculosis among HIV positive adults 10 vary from one setting to another. Therefore, a context specific study in a high 11 12 burdened TB/HIV region is indicated. 13 Objectives 3 This study aims to assess the determinants for TB among HIV infected adults in 14 Northwest Ethiopia. 15 For peer review only 16 Methods 17 Study design 4 We conducted a multicenter case control study from May 12 to June 5, 2014 in 18 Northwest Ethiopia. 19 Setting 5 All governmental health institutions in ten provinces were included (i.e. Bahir Dar, 20 Dangella, Kossober, Fnote Selam, Bure, Dur Betie, Deber Tabor, Wereta, Addis 21 22 Zemen and Nefas Mewcha). 23 Participants 6 Cases were HIV infected adults diagnosed with active TB and on TB treatment during 24 the data collection period. The diagnosis of active TB was based on one or more of TB 25 diagnostic investigations (sputum microscopy, x ray, histopathology, culture or 26 molecular). Controls were HIV infected adults without TB. All controls were actively 27 28 screened for TB to rule out TB. The screening procedure included history taking for 29 four symptoms at each visit (cough, weight loss, night sweating or fever). If the 30 patient presented with any one of these symptoms, this patient was a suspect TB case 31 and was investigated according to the national guidelines for TB HIV1. In hospitals, 32 sputum microscopy, histopathology and radiology examinations are used as the TB
33 http://bmjopen.bmj.com/ 34 diagnostic tools. In health centers, sputum microscopy is the main stay TB diagnostic 35 tool and patients will be sent for other diagnostic evaluation at hospitals and private 36 institutions where the diagnostic tools are available. 37 Variables 7 TB was the outcome variable. Diagnosis of TB stated above. The determinants of TB 38 39 were socio demographic variables (age, sex, educational and marital status), host 40 related variables (cigarette smoking, khat chewing and alcohol consumption), 41
environmental determinants included the presence of flooring, latrine and separate on September 24, 2021 by guest. Protected copyright. 42 kitchen in the household and clinical variables (HAART, CD4 cell count, CPT and 43 IPT) 44 45 Data sources/ 8* Data were collected from two sources. Trained nurses who were in charge of HIV care 46 measurement clinic conducted face to face interview with study participants by using structured 47 questionnaires that were prepared in English, translated to Amharic and back 48 translated and pretested for consistency and ease of understanding. The primary data 49 50 were collected to assess: socio demographic variables (age, sex, educational and 51 marital status), host related variables (cigarette smoking, khat chewing and alcohol 52 consumption). A person with substance use (chewing Kahat, alcohol consumption or 53 smoking) was defined as an individual who is currently using the substance or has a 54 history of substance use. Environmental determinants included the presence of 55 56 flooring, latrine and separate kitchen in the household. Trained nurses also collected 57 data from patient’s records to assess: clinical variables (HAART, CD4 cell count, CPT 58 and IPT). Data collection procedures were supervised by medical officers. 59 Bias 9 Identification of confounding variables by logistic regression models was applied. 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml2 Page 19 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 Study size 10 The sample size was calculated using Epi Info version 7 software. The study was 3 designed to have 80% statistical power with level of significance at 5% and a case to 4 control ratio of 1:2. Assuming the proportion of low CD4 cell count were 5.9% among 5 the controls and 13.9% among cases13 by using the two proportions formula, the 6 7 calculated sample size was 144 for cases and 287 for controls. Allowing for 10% of 8 non response, the resulting sample size was 474 (158 cases and 316 controls). The 9 sample size was calculated for exposure status of different variables: cotrimoxazole 10 preventive therapy (CPT), isoniazid preventive therapy (IPT) and CD4. We took the 11 12 largest sample among these exposure variables. 13 Quantitative variables 11 TB was the outcome variable; Yes or No response. Logistic regression models were 14 applied to assess the relationship between exposure variable and TB. 15 Statistical methodsFor 12 Datapeer were entered intoreview Epi info version 7 and onlyanalysis was done with SPSS version 16 20. Frequencies and proportions were used to describe the study subjects in relation to 17 18 the studied variables. Odds ratio with its 95% confidence interval and p value were 19 used to measure strength of association and identify statistical significance result. 20 Logistic regression models were applied to assess the relationship between 21 determinants and TB. Identification of confounding variables by logistic regression 22 23 models was applied. Multivariable analysis (backward stepwise) was used. Hosmer 24 Lemeshow test was applied and the fit of the model was checked. A poor fit of the 25 model if the significance p value is less than 0.05 and good fit greater than 0.05. Here 26 in this study the model adequately fits the data and the p value was 0.34. 27 Continued on next page 28 29 30 31 32
33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml3 BMJ Open Page 20 of 21 BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 3 Results 4 5 Participants 13* A total of 446 subjects (150 cases=33.6%, 296 controls=66.4%) out of 474 eligible participants 6 responded and consented to participate in this study, resulting in an overall response rate of 7 94.1% (94.9% for cases and 93.7% for controls). 8 Participants were unable respond mainly because of taking time for interview, lack of 9 willingness for interview 10 11 Descriptive 14* The median age and its interquartile range (IQR) of cases were 32 years (IQR: 27 39 years) 12 data while the corresponding value for controls were 33 years (IQR: 28 39 years). A higher 13 percentage of women was observed in both groups; 81(54%) in cases and 188(63.5%) in 14 controls. 15 For peer review only 16 17 Outcome data 15* The bivariate analysis showed that male patients (COR (crude odds ratio) =1.48; 95% CI: 18 0.99, 2.12; p value = 0.053) were more likely to develop tuberculosis compared to female 19 patients. Married patients (COR=0.45; 95%CI: 0.27, 0.75) were less likely to develop TB 20 compared with never married. Cases were more likely to be smokers (COR= 4.73; 95% CI: 21 22 2.3, 9.72), Khat chewer (COR=2.9; 95%CI: 1.77, 4.75) and alcohol drinker (COR= 2.44; 1.54, 23 3.86). Study participants who had a history of TB (COR=1.85; 95%CI: 1.15, 2.96) were more 24 likely to develop TB. But educational status was not a determinant for TB. 25 HAART (COR=0.33; 95%CI: 0.19, 0.56), CPT (COR=0.35 95%CI: 0.22, 0.53) and IPT 26 (COR=0.30 95%CI: 0.18, 0.50) had protective benefit against TB. Subjects with separate 27 28 kitchen in the house hold (COR=0.61; 95%CI; 0.40, 0.93) were less likely to develop TB. HIV 29 infected adults with presence of a TB patient in the family (COR=1. 99; 95%CI: 1.11, 3.57) 30 were more likely to develop TB. But flooring and presence of latrine were not determinants for 31 developing TB. 32 Main results 16 To identify independent determinants for TB, a multivariable logistic regression model was
33 http://bmjopen.bmj.com/ 34 used. After adjusting for possible confounders, some variables were remained in the 35 multivariable model: chewing khat (AOR=2.22; 95%CI: 1.11, 4.41), being a smoker 36 (AOR=5.47; 95%CI: 2.26, 13.22) and alcohol drinker (AOR=2.49; 1.29, 4.80) were 37 independent determinants for increased TB occurrence. However, HAART (AOR=0.25; 38 39 95%CI: 0.12, 0.51), CPT (AOR=0. 32; 95% CI: 0.19, 0.52) and IPT (AOR=0.22; 95%CI: 0.11, 40 0.42) had an independent protective benefit effect against TB. Subjects with separate kitchen 41 in the house hold (AOR=0.48; 95%CI: 0.28, 0.83) was less likely to develop TB. Patients who on September 24, 2021 by guest. Protected copyright. 42 live with a TB patient in the household (AOR=2.66; 95%CI: 1.25, 5.66) were more likely to 43 develop TB. Patients whose CD4 cell count was < 200 cells/µl (AOR= 7.22; 95%CI: 3.39, 44 45 15.37) were more likely to develop TB. But sex, marital status and a previous history of TB 46 were not remained in the multivariable model 47 Other analyses 17 No analyses of subgroups and interactions, and sensitivity analyses 48 49 Discussion 50 Key results 18 Being smoker (AOR (adjusted odds ratio) =5.47; 95%CI: 2.26, 13.22), presence of 51 tuberculosis patient in the family (AOR=2.66; 95%CI: 1.25, 5.66), alcohol drinker 52 (AOR=2.49; 95%CI: 1.29, 4.8) and chewing khat (AOR=2.22; 95%CI: 1.11, 4.41) were 53 independent determinants for tuberculosis. Highly active antiretroviral therapy (HAART) 54 55 (AOR=0.25; 95%CI: 0.13, 0.51) had a protective benefit against tuberculosis. 56 Limitations 19 As we used a retrospective case control approach; temporal relationship could not be 57 established and the study design could not proof causation. Tuberculosis has a possibility to 58 spread to another people who is not a household member. But this study was considered, only 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml4 Page 21 of 21 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2015-009058 on 15 April 2016. Downloaded from 1 2 family members were close tuberculosis contacts. Information bias related to substance use 3 (chewing chat, smoking or alcohol consumption) might have also affected the accuracy of 4 information. 5 6 Interpretation 20 HIV infected adults who use substance (tobacco smoking, khat chewing and alcohol) should 7 be prioritized for tuberculosis screening. This study reaffirmed that initiation of HAART is one 8 of the best strategies to reduce tuberculosis occurrence in HIV infected adults. Increased 9 coverage of isoniazid and cotrimoxazole preventive therapy has a protective benefit against 10 active tuberculosis. Contact tracing of household contacts for tuberculosis should be 11 12 intensified. 13 Generalisability 21 This study will be generalized to HIV positive adults in Hence Awie, West Gojjam and South 14 Gonar, Northwest Ethiopia 15 For peer review only 16 Other information 17 Funding 22 PAGEL (Partnerships for the health sector in developing countries offers German institutions 18 of higher education many different opportunities for international cooperation related to the 19 health sector in developing countries) funded this article. The points expressed in this article 20 are the responsibility of authors, do not reflect the view of PAGEL. The funding source had no 21 22 role in design, analysis, wrote paper, manuscript preparation or decision for publication. 23 24 *Give information separately for cases and controls in case control studies and, if applicable, for exposed and 25 unexposed groups in cohort and cross sectional studies. 26
27 28 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 29 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 30 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 31 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 32 available at www.strobe statement.org.
33 http://bmjopen.bmj.com/ 34 35 Author notice: This checklist is a very good guide to prepare an observational research report. 36 37 38 39 40 41 on September 24, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml5 Miscellaneous Correction
Alemu YM, Awoke W, Wilder-Smith A. Determinants for tuberculosis in HIV-infected adults in Northwest Ethiopia: a multicentre case–control study. BMJ Open 2016;6: e009058. The first name of the third author was misspelt. The correct spelling is Annelies Wilder-Smith. BMJ Open 2016;4:e009058corr1. doi:10.1136/bmjopen-2015-009058corr1
BMJ Open 2016;4:e009058corr1. doi:10.1136/bmjopen-2015-009058corr1 1