Study area and period
The study conducted in the public health institutions in East Gojjam (Debre Markos Referral Hospital, Debre Markos health center, Dejene health center, Bichena health center, Lummame health center & Yebokela health center). From West Gojjam, Finotselam hospital was included.. The study area was selected according to available number of TB/HIV co-infection in the area and computerized documentation system in the health facilities. The study conducted from March 7 - April 15, 2017 in Northwest, Ethiopia.
Study design: Institutional based unmatched case-control study design was conducted.
Source population: All adult HIV infected adults recorded on ART from public health institutions.
Sampling population: The cases were both active TB and HIV infected adults with TB treatment during the data collection period and registered on TB clinic. Controls were only HIV infected adults and screen negative for TB and follow up and registered on ART at public health institutions.
Eligibility criteria
Inclusion criteria
All adult living with HIV and TB/HIV co-infected individuals with pulmonary tuberculosis, extra Pulmonary tuberculosis and being ≥15 years old and having record on ART and TB clinic in public health institutions in East and West Gojjam Zone were included in the study.
Exclusion criteria
Those patients, who were seriously ill, had a mental problem and with suspected but unconfirmed TB exclude from this study.
Sample size determination
As the investigation was unmatched case-control study, the sample size would require for using two-population proportion formula to achieve statistically significance results. Therefore, a sample was calculated by taking into account the major exposure variables, and using epi-info version 7. Among the exposure variables CD4 count, isoniazed preventive therapy (IPT) and cotrimoxazole preventive therapy (CPT) were evaluated. The variables that were given the largest sample size in this case CD4 count were selected as the main exposure variables which given the most favorable sample size, in this regard and the study were planned to have 80% statistical power with a level of significance at 5% (two-sided) and a case to control ratio of 1:3. Assuming the proportion of low CD4 cell count is 5.6% for the controls and 13.9% for the cases [14] and allowing 5% of non- response rate the resulted sample size were 576.
Sampling procedure
Regarding to cases a base line data was obtained from the governmental hospitals and public health centers. Two hospitals and five health centers in East and West Gojjam were found to be eligible by simple random sampling.
The principal investigator extracts the required data from the ART and TB registries for the identification of cases and controls. All TB–HIV co-infected adults attending HIV care clinics that could diagnose by direct microscopy, culture, Gene expert and fine needle aspiration and confirmed for TB positive. TB-HIV co-infected adults, who would receive TB treatment and fulfill inclusion criteria at the time of data collection, were included as a case in the study. However, controls were adequate to be sampled and confirmed tuberculosis negative according to the guideline diagnosed; lists of controls were prepared using unique identification numbers from records found in ART clinics and selected by simple random sampling technique by using computer-generated random numbers. Medical registration number used to select controls that fulfill inclusion criteria after giving unique identification numbers in increasing order.
Study variables
Dependent variable: Status of tuberculosis among HIV infected adults.
Independent variables
A Socio demographic characteristic includes (sex, age, marital status, residence, level of education, type of occupation and monthly income). Environmental factors (separate kitchen, Family size, number of windows in the house, the types of floor in the house), Host or behavioral related factors (chewing khat, cigarette smoking, drinking alcohol, history of TB in the past, history of asthma, history of TB in the family, & history of diabetic mellitus). Clinical variables (WHO clinical stage, hemoglobin level, IPT prophylaxis, CPT prophylaxis, level of CD4 count, functional status of the patient and level of BMI) were included.
Operational definitions
Substance abuse (chewing khat, alcohol consumption and smoking) defined as an individual who is currently using the substance or has a history of regular substance abuse.
Active tuberculosis is Mycobacterium tuberculosis disease which is in active state in any part of the body as determined by either, a smear microscopy, culture or molecular taken from any source in the person’s body tests positive for tuberculosis and the person has not completed the appropriate prescribed course of medication for active tuberculosis during the study.
Smear positive pulmonary tuberculosis (PTB+) diagnosed if single sputum smear examination positive for Acid Fast Bacilli (AFB) by direct microscopy, culture, Gene expert and laboratory confirmation of HIV infection. Extra-pulmonary tuberculosis diagnosed if one specimen from an extra-pulmonary site culture-positive for mycobacterium tuberculosis or smear positive for AFB or histological or strong clinical evidence consistent with active extra-pulmonary tuberculosis and laboratory confirmation of HIV infection or strong clinical evidence of HIV infection and decision by a clinician to treat with a full course of anti -tuberculosis chemotherapy [14].
Data collection instruments
To determine the predictors of tuberculosis among HIV infected adults, semi-structured questionnaires were used to collect data in study participants from selected public health institutions in East and West Gojjam Zone from primary and secondary sources. The data were collected by face-to-face interview of patients and extracted from ART card and logbooks. Five trained clinical diploma nurses who were not in charge of an HIV and TB care clinic conducted the interview with study participants using questionnaires.
Data quality control
Before data collection period, data collectors trained about the objective of the research. The principal investigators were given the training about the objective of the study and data collection system by using semi-structured questionnaires in a one-day period. The questionnaires would prepare in English, translated into Amharic and back translated into English to check consistency. Pre-test carried out on 16 PLWHIV positive adults in Amanual health center in Machakele woreda to familiarize the interviewer with the instrument and to check the coherence. To keep the quality of data, principal investigator was checking the questionnaires for its completeness in each day.
Data processing and analysis
Data were checked, coded, cleaned and entered using Epi-Data V.3.1computer program and exported to statistical package for social science (SPSS) version 20 for analysis. The whole data were cleaned to minimize data entry errors and observed the unities of the data as well as, inconsistencies were verified using cross tabulation. Descriptive frequency statistics were run to see the general distribution of the data. Frequencies and proportions used to describe the study subjects in relation to the study variables. Multiple logistic regressions were used to review the predictors of TB in HIV infected adults as the most important statistical method of analysis by using backward logistic regression variable selection technique. All explanatory variables that were associated with the outcome variable in bi-variate analysis with p-value of <0.25 [36] were candidate in the initial logistic regression models of multivariate analysis. The crude and adjusted odds ratios together with their corresponding 95% confidence intervals were computed. A P-value < 0.05 considered declaring a result as statistically significant with the outcome variable.