Predictors of Mortality Among Children Co-Infected with Tuberculosis and Human Immunodeficiency Virus in Region, North Ethiopia, Retrospective Follow-Up Study

Predictors of Mortality Among Children Tuberculosis and Human Immunodeficiency Vi rus in Region, North Ethiopia, Retrospective Follow-Up Study. Biomed BJSTR. MS.ID.006252. Abbreviations: AIDS: Acquired Immune Deficiency Syndrome; AHR: Adjusted Risk Ratio; CHR: Crude Risk Ratio; ART: Anti-Retroviral Therapy; CI: Confidence Interval; CPT: Co-Trimoxazole Preventive Therapy; ART: Anti-Retroviral Therapy; HGB: Hemoglobin; HIV: Human Immuno deficiency Virus; EPTB: Extrapulmonary Tuberculosis; PTB: Lung Tuberculosis; IPT: Isoniazid Preventive Therapy; OI: Opportunistic Infections; PLHIV: People Living With HIV; TB: Tuberculosis; TB and HIV: Tuberculosis And Human Immuno deficiency Virus; WHO: preventive therapy (IPT, CPT) and counseling on adherence to ART drugs were crucial interventions to reduce mortality among children co-infected with TB and HIV. Children who have extra-pulmonary tuberculosis and advanced clinical staging (III and IV) need special consideration during treatment.


Tuberculosis (TB) and human immunodeficiency virus (HIV)
co-infection remain a major global and national health problem that requires substantial action to achieve the Sustainable Development Goals (SDG) and the END-TB strategies [1]. Both TB and HIV are the leading causes of death from infectious diseases worldwide [2]. Mycobacterium tuberculosis and HIV co-infection in the human body, potentiate each other and accelerate to death by deteriorating body immunity causing premature death if untreated [3]. Tuberculosis is a major cause of morbidity and mortality in HIV-infected children [4]. In 2015, the World Health Organization (WHO) report showed that nearly 41,000 children died from TB and HIV co-infection. Of which more than 83% were occurred in Africa [5]. Mortality among children co-infected with TB and HIV varied in different settings and fluctuated widely from 6.2% to 36.5% [5][6][7].
In Ethiopia, mortality of children co-infected with TB and HIV was 14% [8] and co-infected children had six times greater death than TB disease alone [9]. Furthermore, more than 1 in 5 TB and HIV coinfected individuals were died [10], but this huge problem was not specifically known in children.

The prevalence of TB and HIV co-infection in children was
under-assured due to the problem of reaching a definitive diagnosis. However, the WHO report showed that HIV prevalence among children with active TB disease ranges from 10 to 60%, depending on the background rates of HIV infection in countries with moderate to high prevalence of TB [11]. The estimated rates of tuberculosis among HIV positive children also had a wide variation, depending on the TB epidemic and the coverage of highly active antiretroviral treatment (HAART) coverage in the area [4].
Data on the survival of TB and HIV co-infection in children are still lacking and the available information is difficult to interpret due to problems with the diagnosis and selection of study populations [4]. In developing countries, including Ethiopia, the management of TB and HIV co-infection in children is very challenging due to the inaccessibility of appropriate formulations of drugs, drug-drug interactions, pill burdens, drug side effects, and poor drug adherence [12][13][14]. This may result in high TB incidence and mortality among HIV-positive children. TB is not only the most commonly reported opportunistic infection [15], but also a major cause of hospital admission and death in HIV infected children [16]. The cause of death is also multifactorial and determined by socio demographic, clinical, laboratory, drug and follow-up related factors [8]. Which are poorly understood. Therefore, studies on mortality and its predictors in TB and HIV co-infection in children are very significant to designate appropriate action according to their ages.

Most of the studies on TB-HIV co-infection focused on adult,
fewer studies on general co-infected population, little is known in pediatrics sub-age group. Still, the problem in children is masked and actions are taken based on findings from studies in the adult population. However, the problem is very alarming in children due to immature immune system and fast deterioration into death [17,18]. A previous study in the comprehensive specialized hospital of Gondar University in Ethiopia lacks a time specification on the TB and HIV co-infection period, rather they prolonged their follow-up after TB was cured. This makes the study more biased.
To some extent, there is better evidence on the incidence and predictors of tuberculosis in HIV-infected children [19,20], but evidence on survival and mortality after co-infection is limited in Ethiopia. Therefore, survival and predictors of mortality among children co-infected with TB and HIV have not been well documented in Ethiopia. Therefore, this study was to try to fill the above gaps by estimating survival and identifying predictors of mortality among children co-infected with TB / HIV in public general hospitals in Mekelle and the southern zone of Tigray region, northern Ethiopia.

Inclusion and Exclusion Criteria
Children infected with TB-HIV co-infected younger than 15 years were included in this study and had follow-up care from January 1/2008 -December 30/2018 in a selected hospital. Children who had missed key information on clinical, immunological, drug information and their outcomes had not been recorded on medical charts were excluded.

Sampling Technique
In the Mekelle and Sothern zones of the Tigray region, five general hospitals were found to provide ART services. These are the general hospitals of Mekelle, Quiha, Maychew, Alamata, and Korem. However, this study used cluster sampling by randomly selecting three hospitals (Mekelle, Alamata, and Maychew). Since

Data Collection and Analysis
Data were collected from medical records (charts) using a data extraction checklist developed from the national HIV intake and follow-up form [22]. The checklist consisted of sociodemographic, Both bivariate and multivariate analysis was computed to determine the association between predictor variables and the outcome variable. These variables that were significantly associated with a p-value of <0.2 in the bivariate analysis were entered into the multivariate analysis. Variables significantly associated with the outcome variable at a p-value <0.05 in the multivariate analysis were considered independent predictors of mortality. Finally, the adjusted hazard ratio with 95% CI and P value was used to measure the significant association between predictors and outcome variable.

Ethical Considerations
The study protocol was evaluated and approved by the Institutional Review Board (IRB) of Mekelle University, a college of health sciences, and then ethical clearance was obtained. A cooperation letter was written to the chief executive managers of each hospital. Since the study was retrospective and document review, it did not cause any risk to the study participants.

Sociodemographic Characteristics
A total of 282 children with co-infected TB and HIV were enrolled in the general hospitals of Mekelle, Alamata, and Maychew.
Of which 29 were excluded from the study due to lost cards or incomplete data. The remaining 253 children co-infected with TB and HIV were included in the study. The median age of the study participants was 8 years with IQR (4-13). One hundred and thirtyone (51.8%) of the children were females ( Table 1).

Education and Follow-Up Related Characteristics
One hundred and ninety-seven (77.9%) of the respondents had taken co-trimoxazole preventive therapy and 145 (57.3%) had also taken isoniazid preventive therapy before developing TB.

Predictors of Mortality Among Children Co-Infected with TB and HIV
Bivariate and multivariate analyzes were used to assess the significant association between exposure variables and the  Table 4).
The risk of death among children with TB and HIV co-infected with underweight was approximately 8 times higher than children with normal weight at baseline (AHR=7.9 (95% CI 1.26, 49.3)).
Children who did not take IPT were approximately 4 times more likely to experience death than children who had taken IPT (AHR=3.69 (95% CI=1. 26, 10.8)). The risk of child death with poor adherence to ART was approximately 4 times higher than children with good adherence to ART (AHR = 3.82 (95% CI: 1.38, 10.54)).
The risk of death among children infected with extrapulmonary TB was also approximately 3 times higher than infected children with pulmonary TB (AHR = 2.9 (95% CI: 1.1, 7.6)). During

Discussion
The study provides information on the overwhelming problem of high mortality and associated predictors among children with TB and HIV coinfected. The mortality rate in this study was 0.17 (95% CI 0.12-0.23) per 1000 child-month observations. The result was lower than the mortality rate reported from a single study conducted in four developing countries (Burkina Faso, Cambodia, Cameroon and Vietnam), which is 0.370 per 1000 child-month observations [23]. The difference may depend on the sample size difference used by the studies.
In this study, mortality was higher in underweight children at baseline. A similar finding was reported from a study conducted in Thailand [24]. This might be the effect of underweight on reducing body metabolic processes resulting in inadequate energy acquisition that increases disease progression, which may end up in death. Furthermore, inadequate weight gain in TB treatment indicates a poor response to treatment [25]. However, stunting and wasting were not significant in this study. This could be due to a higher proportion (90%) of children diagnosed with malnutrition in this study who received treatment for malnutrition. The study also revealed that children who did not take IPT were three times more likely to experience death than children who did take IPT.
This was in line with a study conducted in Gondar, Ethiopia [8].
The possible reason might be that IPT reduces the severity and spread of TB disease. However, CPT was not found to be statistically significant in this study, which was reported as a protective factor for death in a study conducted in Gondar, Ethiopia [8]. This may be because a higher proportion (78%) of our respondents had taken CPT and were unable to make a difference. The number of children who didn't take CPT and died was too few (5.1%). For better survival, HIV positive children should take both CPT and IPT as preventive prophylaxis. In this study, the risk of death among children infected with extrapulmonary TB was three times higher than that of children infected with pulmonary TB. This result was in line with a study conducted in Gondar, Ethiopia [8]. The reason might be that the easy diagnostic technique for EPTB is not available in most of our clinical settings, resulting in delayed initiation of anti-TB treatment leading to rapid disease progression and easy involvement of vital organs.
During follow-up, this study revealed that anemia was associated with higher child death. No previous studies examined anemia during follow-up, but at the beginning of the study, it was identified as a predictor of mortality in studies conducted in Gondar (Ethiopia) and Thailand [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24]. Higher mortality with anemia may be associated with decreased oxygen and nutrient care capacity of the blood, resulting in inadequate oxygen and nutrient supply to vital organs that become synergistic with TB and HIV [8]. In contrast to other studies in Gondar (Ethiopia) [8], Thailand [24], Nigeria [6], Malawi [26], and a single study in four developing countries [23]; WHO staging, CD4 count, and hemoglobin level at baseline were not significantly associated with mortality in this study. The reason might be that unlike these studies, our study assessed the effect of the variables during follow-up time and at baseline. Most of these variables were significantly associated during follow-up, which shows a better effect on the outcome variable than at baseline. This is one of the strengths of this study. Assessing the effect of these variables during follow-up enables us to overlook the more accurate effects of exposure variables on the outcome variable. The study also considered the time of the event, which enables us to consider the contribution of censored cases.

Limitation of the Study
Since the study was a retrospective review of the chart (secondary data), some variables not documented in the child's medical records were missed. A further prospective study is needed to address other important issues not addressed by this study.

Conclusion
The mortality rate of children co-infected with TB and HIV in two zones of the Tigray region was high. Most deaths occurred within the first six months of the follow-up period. Underweight at baseline, IPT non-user, poor ART adherence, extrapulmonary TB, advanced WHO staging during follow-up, advanced/severe immunosuppression status during follow-up, and hemoglobin level < 10mg/dl during follow-up were predictors of increased mortality.
This study is important for planning and decision making by pointing out gaps to make a successful strategy to combat TB and HIV and related consequences to increase the overall effectiveness of therapy in TB and HIV co-infected infected children.

Ethical Approval and Consent to Participate
Ethical approval was obtained from the Institutional Review Board of the College of Health Sciences, Mekelle University. A letter of cooperation from the School of Nursing and a permission letter from hospitals were also secured before data collection. The name of the respondent was not mentioned during data collection.

Availability of Data and Materials
The data set analyzed during the current study is available from the cross-pondering author on reasonable request.