https://orcid.org/0000-0003-1522-7905
Figures
Abstract
Background
HIV-TB co-infection poses a significant public health threat, notably in sub-Saharan Africa including Ethiopia. Despite this public health problem, studies in Ethiopia regarding the mortality of HIV-TB co-infection patients have been inconsistent, and the overall estimate of mortality was not determined. Accordingly, this meta-analysis aims to assess the magnitude of mortality and predictors among HIV-TB co-infected patients in Ethiopia.
Methods
A search of the literature was conducted from three databases (PubMed, Global Index Medicus, and CINHAL), and other sources (Google Scholar, Google, Worldwide Science). All observational studies that reported the mortality of HIV-TB co-infected patients in Ethiopia were included. Joanna Briggs Institute’s (JBI) quality appraisal checklist was used to assess the quality of studies. Effect sizes were pooled using the random effects model. Heterogeneity was assessed using Cochrane Q and I2 test statistics, and the prediction interval was determined. Subgroup analysis was conducted by region. To examine the presence of an influential study, a sensitivity analysis was done. Egger’s test was used to check publication bias. A non-parametric trim and fill analysis was carried out.
Results
A total of 886 studies were identified, using database searches and keywords. Of these, 37 met the criteria for inclusion. The pooled proportion of mortality among HIV-TB co-infected patients was found to be 18.42% (95% CI:14.27–22.57). In the subgroup analysis, the highest mortality was observed in the Tigray region at 31.86% (95% CI: 7.69–56.03), and the lowest mortality was reported in two general studies in Ethiopia 11.95 (95% CI: 4.19–19.00). From the examined 20 predictors, only four predictors such as Anaemia (HR = 2.25, 95% CI: 1.65–3.07), Poor adherence to ART (HR = 2.42, 95% CI: 1.39–4.21), not taking co-trimoxazole preventive therapy (HR = 1.87, 95% CI: 1.28–2.73), and extrapulmonary tuberculosis (HR = 1.23, 95% CI: 1.01–1.51) were significant predictors.
Conclusions
In Ethiopia, 18.42% of HIV-TB co-infected patients died. Anaemia, poor adherence, not taking CPT, and extrapulmonary tuberculosis were found to be significant predictors. Hence, the concerned stakeholders need to expand and strengthen the HIV-TB collaborative services and attention should be given to patients presented with the aforementioned predictors.
Trial registration
This meta-analysis has been registered in PROSPERO with registration number CRD42023466558.
Citation: Wondie WT, Wubneh CA, Legesse BT, Mekonen GB, Zemariam AB, Gebreegziabher ZA, et al. (2025) Mortality and predictors among HIV-TB co-infected patients in Ethiopia: A systematic review and meta-analysis. PLoS ONE 20(1): e0317048. https://doi.org/10.1371/journal.pone.0317048
Editor: Elizabeth S. Mayne, University of Cape Town Faculty of Science, SOUTH AFRICA
Received: February 20, 2024; Accepted: December 19, 2024; Published: January 6, 2025
Copyright: © 2025 Wondie et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All data analyzed and used for this study are available as Supporting information.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AIDS, Acquired Immunodeficiency Syndrome; ART, Anti-Retroviral Therapy; CPT, Co-trimoxazole Preventive Therapy; HIV, human immunodeficiency virus; WHO, World Health Organization
Introduction
Human Immunodeficiency Virus (HIV) and Tuberculosis (TB) infection remain a major public health problem, notably in Sub-Sahara Africa [1]. Patients co-infected with HIV and TB are threefold more likely to die during treatment and continue to suffer from various opportunistic infections [2], which contributes to significant morbidity and mortality [3, 4]. According to the Global TB Report, an estimated 167,000 deaths from HIV-TB occurred in 2022 [5]. Additionally, the World Health Organization (WHO) reported 190, 000 deaths in 2021, with 11% of these deaths occurring in children [6]. This indicates that HIV-TB co-infection accounts for 30% of the total AIDS-related mortality [2, 7].
A global review indicated that HIV-TB co-infection resulted in 2.7 death per 100, 000 individuals, with notably high mortality rates observed in Lesotho and Zambia, at 168 & 53 death per 100,000 patients, respectively [8]. Globally the mortality of HIV-TB co-infected patients is (11%), with the highest mortality reported in Africa (14%) [9]. In Sub-Saharan Africa, 36% of HIV-TB co-infection-related deaths occurred among children [10]. In addition to the aforementioned burden, TB contributes to 24.9% of in-hospital mortality among HIV-infected patients [11], and the co-infection of these diseases imposes a significant financial burden on countries with high rates of HIV-TB [5]. In Ethiopia, the estimated costs are $78 for HIV and $115 for TB per patient [12].
Several studies investigated the risk factors for mortality in HIV-TB co-infected. As the studies showed that the factors were low CD4 count, TB Preventive Therapy, risky behaviors [8, 9, 13–15], Poor nutritional status, not taking Co-trimoxazole Preventive Therapy (CPT), male sex, being female sex worker, functional status (being bedridden and ambulatory) [7, 8], early initiation of Anti-Retroviral Therapy (ART) for TB patients [7, 16], extra-pulmonary tuberculosis, residence, poor adherence, non-disclosure, treatment failure, Multi-Drug Resistance (MDR) TB, opportunistic infection, and WHO stages (stage I, II, III) [7, 13]. To address the dual burden of HIV and TB, WHO has established a global framework for a collaborative strategic program [17], and has recommended a test and treat all strategy since 2016. Additionally, HIV-TB control programs are intensifying their joint efforts to enhance integrated service delivery, case finding, TB preventive therapy, infection control, and the provision of ART and anti-TB medications [2, 7, 8]. Despite these measures, and improved ART coverage in Ethiopia, the rate of HIV-TB co-infection is increasing [6, 7], and the morbidity and mortality associated with this co-infection are not decreasing [7, 18, 19].
Ethiopia is one of the countries with a high burden of HIV-TB co-infection [5], and various individual studies have been conducted in different parts of the country regarding the mortality of co-infected patients. However, the mortality rate associated with HIV-TB co-infection varies across these studies, ranging from 4.4% [20] to 80.45% [21]. These findings are fragmented and do not provide comprehensive and robust conclusions regarding the mortality of co-infected patients at a national level. Current evidence-based findings are necessary to inform health program planners and policy-makers. While there is a general understanding that HIV and TB are interrelated and contribute to increased mortality from one another, to the best of our knowledge, no comprehensive study has estimated the pooled mortality rate of HIV-TB co-infected patients. Therefore, this systematic review and meta-analysis aims to deliver comprehensive findings on the mortality rates and predictors affecting co-infected patients in Ethiopia. The result of this review will provide valuable insights for decision-makers, and program planners in developing effective strategies to reduce mortality among HIV-TB Co-infected patients in Ethiopia. Additionally, it will serve as a baseline data for future researchers.
Methods
Protocol registration and reporting
This systematic review and meta-analysis have been registered in the International Prospective Register of Systematic Reviews (PROSPERO) with registration (ID: CRD42023466558), which is available at https://www.crd.york.ac.uk/prospero/#myprospero. The Preferred Items for Systematic Review and Meta-Analysis (PRISMA) checklist was used for reporting this review [22] (S1 Table).
Databases and search strategy
A literature search through three databases; PubMed/Medline, CINAHL, and Global Index Medicus (GIM) with Medical Subject Heading (MeSH) terms and keywords was done. In addition, Gray literature was searched on Google, Google Scholar, and Worldwide Science. Furthermore, articles from reference lists of included studies, and related reviews were assessed, and retrieved. We searched for articles until November 25, 2023. The search was focused on studies that reported the proportion of mortality and its predictors in Ethiopia. For searching literature the following keywords and phrases were used: “Magnitude”, “Occurrence”, “Prevalence”, “Incidence” “Burden” “Proportion”, “Epidemiology”, “Mortality”, “Death”, “Survival status”, “Survival rate”, “HIV-TB co-infected patients” “HIV-TB co-infected cases”, “HIV-TB co-infection”, “HIV and TB co-infection”, “Tuberculosis/HIV co-infected patients”, “MTB/HIV co-infection”, “Tuberculosis”, “TB”, “Human immunodeficiency virus”, “HIV/AIDS”, “TB/HIV”, “HIV TB co-infected patients”, “Predictors”, “determinants” “Associated factors”, “Risk factors”. The search strings were formed using the “AND” “OR”, and “Asterisk” Boolean operators. Three authors (WTW, BTR, and CAW) systematically searched articles published in the English language (S2 Table).
Eligibility criteria
Studies that fulfilled the following major criteria were considered for inclusion in this systematic review and meta-analysis.
Inclusion criteria
This study included all published observational studies (i.e. cross-sectional, and cohort) conducted in Ethiopia, that reported mortality and/or predictors among HIV-TB co-infected patients. Included studies were in the Population, Exposure, Comparison, and Outcome (PECO) framework (P: HIV-TB coinfected patient, E: HIV-TB coinfection, E: patients with no HIV-TB co-infection and O: mortality among HIV- TB co-infected patients from any other causes). On the other hand, books, meeting reports, and studies that don’t contain the necessary data were excluded (S3 Table).
Screening and data extraction. Based on the pre-specified inclusion criteria, two reviewers (WTW and BTR) screened the titles and abstracts of the studies. They then conducted an independent assessment by reviewing the full text of the studies, and extracted the data from November 26, 2023 to December 26,2023. Any discrepancies between the two reviewers were solved through discussion and common consensus. The following information were extracted: first author’s name, year of publication, study region & design, population, sample size, follow-up period, proportion of mortality, and the hazard ratio for the predictors (S1 File).
Quality assessment
For this review, Endnote version X9 (Thomson Reuters, Philadelphia, USA) software was used to remove duplicate studies. Three investigators (WTW, BTR & CAW) assessed the quality of eligible articles using the Joanna Briggs Institute (JBI) Critical appraisal tools designed for cross-sectional and cohort studies [23]. The cohort checklist includes 11 indicators, while the cross-sectional study consists of 8 indicators. These indicators were converted to a percentage scale, with quality scores graded as high if >80%, medium if between 60 and 80%, and low, if < 60%. Any inconsistencies or discrepancies were resolved through discussion with the fourth and fifth authors (ABZ &BTL) (S4 Table).
Measurement of outcome variables
The present systematic review and meta-analysis had two outcomes, the first was mortality from any cause in patients with HIV-TB co-infected, and the second outcome was predictors of mortality. The pooled hazard ratio (PHR) with 95% confidence intervals was calculated to evaluate these predictors of mortality.
Data synthesis and statistical analysis
Microsoft Excel was used to extract the data and the extracted data were imported into STATA version 18 for analysis. To estimate the proportion of mortality and the effect size of predictors among HIV-TB co-infected patients, a random effect model using the Restricted Maximum Likelihood (REML) method was used [24]. Statistical heterogeneity across studies was assessed using the Cochrane Q-test, tau-squared, and I2 –statistics. I2-statistic of 25, 50, and 75% represent low, moderate, and high heterogeneity respectively [25]. The variation in the proportion of mortality was adjusted through subgroup analysis based on the study characteristics, including study region and age group. Additionally, a sensitivity analysis was conducted to evaluate the impact of each study on the overall estimate. Furthermore, publication bias was assessed through visual inspection of the funnel plot, and Egger’s test [26]. Results of Egger’s test (P-value ≤ 0.05) are considered indicative of publication bias (small study effect). This was addressed using a non-parametric trim and fill analysis with a random effects model [27]. The trim and fill funnel plot analysis was confirmed by a Galbraith plot, and the studies lying inside and outside the central 95% pseudo-CI region agreed with the distributional features in the pooled mortality forest plot. The results were presented using forest plots and tables.
Results
Review processes and findings
In our search, we retrieved a total of 886 studies. Out of these, 807 studies were obtained from three data databases: PubMed (623), Global Index Medicus (159), and CINAHL (25). We removed 123 duplicates, and 684 studies remained in Endnote, which were screened by title and abstract. After this screening, 624 irrelevant studies were excluded. Subsequently, 60 studies were assessed by full-text review. Based on our criteria, 35 studies were excluded and 25 studies were included in the final meta-analysis. In addition to databases, we utilized other relevant sources, such as free web search engines (Google, Google Scholar, Worldwide Science), which yielded 79 full-text articles. From these, 67 irrelevant studies were excluded, and 12 articles were included. In total, 37 studies were ultimately included (Fig 1).
Characteristics of included studies
In the current study, a total of 37 studies [20, 21, 28–62], with a sample size of 12,061 participants were included. Among these, 8 studies were conducted in the Amhara region [29, 32, 33, 36, 46, 52, 54, 61], 7 in the Oromia region [30, 35, 39, 43, 55, 58, 60], 8 in South Nation and Nationality People Region (SNNPR) [20, 34, 38, 41, 45, 51, 53, 62], 5 in Tigray region [21, 31, 37, 56, 59], 3 in Addis Ababa [40, 42, 44], 1 in Dire Dewa [57], 1 in Harari [49], 2 in Harari and Dire Dawa [47, 50], and 2 across various parts of Ethiopia [28, 48]. The studies comprised 27 cohort studies, 9 cross-sectional studies, and 1 longitudinal study. The minimum and maximum sample sizes of the included studies were 58 [42], and 1123 [44] respectively (Table 1)
Quality of included studies
In this study, we found 27 Cohort studies, 9 cross-sectional studies, and 1 longitudinal. The JBI quality appraisal checklist for cohort and cross-sectional studies was utilized to evaluate the quality of the included studies. The result of the quality assessment ranged from 81.81 to 100 for both study designs (S4 Table).
Pooled proportion of mortality
In this study, to estimate the pooled proportion of mortality, 37 studies were included. The death rate varied from 4.4% in a study conducted in SNNPR [20] to 80.45% in a study conducted in Tigray [21]. The present meta-analysis, employing a random effect model using the Restricted Maximum Likelihood (REML) method showed that the pooled proportion of mortality among HIV-TB co.-infected patients in Ethiopia was 18.42% (95% CI: 14.27–22.57). In this study, significant heterogeneity was observed within studies with a Q = 123.16, p-value = 0.00), Tau-squared = 160.41, H = 50.46 I2 = 98.02%, P-value = 0.00. The adjusted prediction interval was found to range from 0 to 80%) (Fig 2).
Publication bias and its handling mechanism
An Egger’s test value β = 7.11, P = 0.001, indicated a significant publication bias, and highly non-normal distribution. A funnel plot confirmed both the bias and the non-normal distribution. Accordingly, the non-parametric trim and fill analysis was done, and three studies were imputed. However, the non-parametric trim and fill analysis failed to resolve the bias because 22 of the 37 studies and the 3 imputed studies remained outside the pseudo 95% CI (Fig 3). A Galbraith plot confirmed the results of the funnel plot. The overall pooled proportion of mortality in the non-parametric trim and fill analysis was 15.66% (95% CI:10.38–20.957).
Handling heterogeneity
In this study, there was high heterogeneity (Q = 123.16, P = 0.00, Tau-squared = 160.41, H2 = 50.46 I2 = 98.02%, P-value = 0.00) between and within the 37 studies, and a forest plot was done using the REML method to show the pooled mortality for the 37 studies (Fig 2). In addition, a Galbraith plot was done, and it showed some studies fell outside the 95% CI, which showed the presence of statistically significant heterogeneity (Fig 4). Although the pooled mortality in the forest plot was 18.42% (95% CI; 14.27–22.57), there was a significant range in mortality values for Ifa 2018, 4.40% (95% CI: 1.21–7.59) [20], to Habtamu et.al. 2021, 80.48% (95% CI: 75.12–85.84) [21]. Due to this high heterogeneity, subgroup and sensitivity analyses were performed by region (Fig 5), and by age group (Fig 6).
Subgroup analysis
Sub-group analysis was performed based on study region, and age group of the study populations. The highest proportion of mortality was reported in Tigray region 31.86% (95% CI: 7.69–56.03), followed by Amhara region 20.73% (95% CI: 15.15–26.30), Addis Ababa 20.36% (95% CI: 1.76–38.97), Oromia 15.19% (95% CI: 10.73–19.65), SNNPR 13.66% (95% CI: 8.98–18.35), Harari and Dire Dawa 13.74% (95% CI: 6.86–20.62), and in studies conducted across various parts of Ethiopia 11.95% (95% CI: 4.91–19.00). In this sub-group analysis, there was variability within groups (P = 0.03), and in the overall pooled estimate (I = 98.02, P-value = 0.00). However, no heterogeneity was found between groups (regions) as indicated by (test of group differences: Q (6) = 7.33, P = 0.30) (Fig 5), (Table 2).
Subgroup analysis was also by age group of the patient. The incorporated studies categorize age groups into pediatrics (under 15 years), adults (15 years and older), and both (studies that include both adults and pediatrics). The highest mortality rate was observed among adults, with a rate of 20.70% (95% CI: 14.29–27.11). In this sub-group analysis, there was significant heterogeneity within groups (P = 0.00), and in the overall pooled estimate (P = 0.00). However, there was no heterogeneity between groups (age groups) as indicated by (the test of group differences: (Qb (2) = 4.39, P = 0.11) (Fig 6).
Sensitivity analysis
A sensitivity analysis was conducted using a random effect model for the overall mortality rate In this analysis, there were no influential studies, as all of the single estimates of the leave-one-out analysis fell within the confidence interval of the pooled proportion of mortality18.42% (95% CI:14.27–22.57). Additionally, the weights in Fig 2 are approximately equal and group around 100/ 37 = 2.70, which is satisfactory, and shows that no study unduly influences the overall weighting. These results indicate that no single study significantly affects the overall magnitude of mortality (S1 Fig).
Predictors of mortality
In this study, the pooled effect size of the 20 independent predictors was assessed. The evaluated predictors include sex, anemia, CD4 count, not taking CPT, not taking Isoniazid Preventive Therapy (IPT), risky behavior, extrapulmonary tuberculosis, residence, poor adherence, non-disclosure, ART failure, Multi-Drug Resistance TB (MDR) TB, presence of opportunistic infection, WHO HIV-stages, commercial sex workers, functional status (bedridden vs. ambulatory). Finally, as shown below, only four predictors anaemia, poor adherence, not taking CPT, and extrapulmonary tuberculosis were found to be significant predictors of mortality in the random effect model.
Anaemia
Among thirty-seven studies, six studies [29, 33, 34, 38, 57, 59] were included to examine the association between anaemia and mortality. The pooled hazard ratio was found to be 2.25 (95% CI: 1.65–3.07). This meta-analysis showed that the hazard of death among anaemic patients was 2.25 times higher compared to non-anaemic patients. The heterogeneity test showed no evidence of variation across studies (I2 = 0.0), (Fig 7).
Poor adherence
Six studies [29, 33, 34, 38, 54, 59] assessed the association between poor adherence to ART and mortality among HIV–TB co-infected patients. The pooled hazard ratio was 2.42(95% CI: 1.39–4.21). Therefore, this meta-analysis revealed that the hazard of death among co-infected individuals with poor adherence was 2.42 higher compared to those with good adherence. For this predictor publication bias was not performed due to a small number of studies (Fig 8).
Not taking CPT
In this review, 10 studies [29, 33, 34, 41, 46, 51, 56–59] reported the association between not taking Co-trimoxazole Prophylaxis Therapy (CPT) and mortality among HIV-TB co-infected patients. The pooled hazard ratio was found to be 1.87 (95% CI: 1.28–2.73). The current meta-analysis showed that patients who were not taking CPT were nearly two times at higher risk of death compared with their counterparts. The Eggers test showed that there was no publication bias (beta = 1.23, P = 0.58) (Fig 9).
Extrapulmonary tuberculosis
In this review, 11 studies [29, 34, 36, 38, 39, 46, 52, 54, 56, 59, 60] reported the association between Extra Pulmonary Tuberculosis (EPT), and mortality among HIV-TB co-infected patients. In this meta-analysis, the pooled hazard ratio was 1.23(95% CI: 1.01–1.51), which means patients who had EPT had a 1.23 times higher hazard of death compared with their counterparts. Likewise, to the above predictors, there was no statistically significant publication bias (Egger test P = 0.244) (Fig 10).
Discussion
This systematic review and meta-analysis aimed to provide and synthesize evidence regarding mortality and its predictors among HIV-TB co-infected patients in Ethiopia. This study showed that the national level magnitude of mortality among HIV-TB co-infected patients was 18.42% (95% CI: 14.27–22.57).
This finding aligns with studies conducted in Sub-Saharan Africa (SSA) (18.1%) [1], Europe and Latin America (19%) [63], and a global systematic review and meta-analysis (15%) [64]. However, this finding was lower than that of a study conducted in Eastern Europe (29%) [63], Brazil (27.4%) [65], and another global systematic review and meta-analysis (33.56%) [66]. This discrepancy may be attributed to geographical variation and the presence of some challenging comorbidities. For example, in those studies, a high number of patients had drug-resistant tuberculosis, whereas in the current study setting the prevalence of multi-drug resistance tuberculosis is low. The other possible justification is the study period, those studies were incorporated studies conducted before the commencement of the test-and-treat strategy, while in the present study, most of the included studies were conducted after the implementation of test and treat all strategy in Ethiopia since 2017, which is a proven strategy in reducing mortality [7, 67].
On the other hand, the finding of the present study was higher than a study conducted in Western Europe (4%) [9, 63], Latin America (11%) [63], United Kingdom (4%), the Caribbean, Central and South America, Central, East, and West Africa (12%) [13], studies in Brazil (13%) [9] & (5.9%) [68], and a global systematic review and meta-analysis (11%) [9]. This discrepancy might be related to limited access to appropriate therapy, delayed diagnosis, and poor awareness of these diseases. In low-income countries like Ethiopia, due to the unavailability/ inadequate supply of diagnostic tools and antiretroviral/anti-TB services, and poor access to healthcare facilities, there is poor management of HIV-TB co-infected patients, and this ultimately leads to increased mortality [69]. On the other hand, delay in ART and Anti-TB initiation resulted in advanced clinical stage, opportunistic infection, and immune reconstitution inflammatory syndrome. Before WHO launched the immediate initiation of ART without clinical or immunological status, there was a delayed initiation of therapy in Ethiopia which escalated the mortality rate [70].
Due to heterogeneity, subgroup analysis was conducted based on regions. Consequently, the highest mortality rate was reported in the Tigray region at 31.86%, followed by the Amhara region 20.73%, while, the lowest mortality rate was observed in studies conducted in all parts of Ethiopia at 11.95%. The possible reason for this discrepancy might be poor awareness about those infectious diseases. Due to this, the co-infected individuals may have poor adherence to medications, and poor access to ART/anti-TB, which may result in increased mortality. Besides, the national HIV/AIDS and TB control program might not be uniformly implemented across different regions of Ethiopia, leading to variation in the survival of co-infected patients. Additionally, TB patients in the Tigray region have poor treatment outcomes [71], which may further elevate the mortality rate in case of co-infection. Therefore, supervision of HIV and TB control services is significantly needed to harmonize the service provision. Moreover, it might be due to demographic factors, like, lifestyle/awareness and socioeconomic factors which may further influence the mortality of co-infected patients. In addition to regions, sub-group analysis was done based on the participants’ age group. Accordingly, the highest proportion of mortality was found among adult patients. This might be due to increased occupational and social interactions in adults. Factors such as unprotected sexual intercourse, injection drug use, and occupational and environmental exposure to HIV, and TB, ultimately resulted in mortality [72]. On the other hand, adults may have other underlying medical problems like diabetes, cardiovascular diseases, etc., which further weaken their immune system, and decrease the survival of HIV-TB co-infected patients.
The present study showed that patients presented with anaemia were more than two times at hazard of death compared with their counterparts. This finding was congruent with systematic review and meta-analysis conducted globally [73], and South Africa [74]. This might be due to both HIV and TB infection compromising the immune system of the patients and affecting lung function, and anaemia exacerbates this problem of oxygen supply, which can result in hypoxia in vital organs such as the brain and heart, this leads to morality [7, 75]. On the other hand, anaemia increases the risk of susceptibility to opportunistic infection and disease progression, which increases the hazard of death [7, 76]. Besides this, in those individuals’ anaemia may be due to nutritional deficiency, which is a significant cause of mortality in co-infected patients [7].
The current study also revealed that patients who had poor adherence to ART were above twofold at hazard of death compared with those individuals with good adherence. A similar finding was reported in a study conducted in Spain [77]. This might be due to, poor adherence could result in reduced viral load suppression and decreased CD4 cell count, which leads to uncontrolled HIV replication, and the patient’s immune system becomes compromised and can’t fight off opportunistic infections, which ends up in mortality. The second reason could be poor adherence causes drug resistance, treatment failure, and susceptibility to opportunistic infection, hence ultimately it results in increased mortality. Thus, healthcare providers should emphasize, and provide counseling about the importance of good adherence to ensure optimal treatment outcomes [7].
In this study, patients who were not receiving co-trimoxazole preventive therapy had twice the hazard of death compared with their counterparts. This increased risk may be due to non-adherence to Co-trimoxazole therapy in individuals co-infected with HIV and TB, which raises the likelihood of opportunistic infections, particularly pneumocystis jirovecii pneumonia, and impairs immune recovery, ultimately leading to mortality [19]. In addition to the above predictors, extrapulmonary tuberculosis was an independent predictor of mortality in HIV-TB co-infected patients. This finding was consistent with a study conducted in South Africa [78]. This is due to the fact, that extrapulmonary tuberculosis further compromises, the immune system’s ability to fight off infections. This leads to uncontrolled dissemination of TB throughout the body and increased viral load, which results in a decreased survival rate. On the other hand, extra pulmonary tuberculosis may involve critical organs such as brain, spine, and meninges which leads to life threatening complications and increased mortality rate [7, 19].
Strength and limitations
This study tried to show the pooled magnitude of mortality among HIV-TB co-infected patients at a national level for the first time. However, our study had the following limitations: the significant heterogeneity in the pooled proportion of mortality of the 37 studies, remained in the subgroup and sensitivity analysis and within study analysis. The funnel plot and Galbraith plot analysis indicated significant bias which was not resolved in the fill and trim procedure in the case of the funnel plot. Nevertheless, the paper has significant value in establishing benchmark values in the case of a country with significant diversity in its independent predictors. Additionally, we could not analyze certain factors, such as nutritional status due to inconsistent measurements.
Conclusion
This study indicated that, at the national level, 18.42% of patients co-infected with HIV and Tuberculosis (TB) died. Anaemia, poor adherence, not receiving CPT, and extrapulmonary tuberculosis were significant predictors of mortality. The high mortality of HIV-TB co-infected patients is an urgent concern. Hence, the FMOH of Ethiopia needs to expand and strengthen the integration of HIV-TB collaborative services and consistent monitoring activities across all health facilities of the country, with critical attention for patients presented with the aforementioned predictors.
Supporting information
S2 Table. Search strategies and terms in different databases.
https://doi.org/10.1371/journal.pone.0317048.s002
(DOCX)
S3 Table. List of all identified studies in the literature search including those excluded studies from the analyses.
https://doi.org/10.1371/journal.pone.0317048.s003
(DOCX)
S1 File. Extracted dataset of the included studies.
https://doi.org/10.1371/journal.pone.0317048.s005
(XLSX)
S1 Fig. Forest plot showing sensitivity analysis of the magnitude of mortality among HIV-TB co-infected patients in Ethiopia.
https://doi.org/10.1371/journal.pone.0317048.s006
(TIF)
Acknowledgments
We authors would like to thank Ambo University for providing us with internet services to access available databases and web pages.
References
- 1. Van Hout M.C. and Hope V.J.B.I.D., Treatment outcomes and antiretroviral uptake in multidrug-resistant tuberculosis and HIV co-infected patients in Sub Saharan Africa: a systematic review and meta-analysis. 2019. 19: p. 1–8.
- 2.
(WHO)., W.H.O. Global HIV program,. Tuberculosis and HIV. 2021,; Available from: https://www.who.int/teams/global-hiv-hepatitis-and-stis-programmes/hiv/treatment/tuberculosis-hiv.
- 3. Getaneh T., et al., The impact of tuberculosis co-infection on virological failure among adults living with HIV in Ethiopia: A systematic review and meta-analysis. J Clin Tuberc Other Mycobact Dis, 2022. 27: p. 100310. pmid:35284661
- 4. Sulochan G., et al., Burden of tuberculosis and hepatitis co-infection among people living with HIV in Nepal: a systematic review and meta-analysis. 2022.
- 5.
World Health Organization(WHO). Global Tuberculosis Report. 2023; Available from: https://reliefweb.int/report/world/global-tuberculosis-report-2023?gad_source=1.
- 6.
Joint United Nation program on HIV/AIDS (UNAIDS). The path that end HIV/AIDS. 2023; Available from: https://thepath.unaids.org/wp-content/themes/unaids2023/assets/files/2023_report.pdf.
- 7.
Ministry of Health Ethiopia (MOH), National Comprehensive HIV Prevention, Care and Treatment Training for Healthcare Providers August 2021: Addis Ababa, Ethiopia.
- 8. Sullivan A. and Nathavitharana R.R., Addressing TB-related mortality in adults living with HIV: a review of the challenges and potential solutions. 2022. 9: p. 20499361221084163.
- 9. Odone A., et al., The impact of antiretroviral therapy on mortality in HIV positive people during tuberculosis treatment: a systematic review and meta-analysis. PLoS One, 2014. 9(11): p. e112017. pmid:25391135
- 10. Dodd P.J., et al., The global burden of tuberculosis mortality in children: a mathematical modelling study. 2017. 5(9): p. e898–e906.
- 11. Ford N., et al., TB as a cause of hospitalization and in‐hospital mortality among people living with HIV worldwide: a systematic review and meta‐analysis. 2016. 19(1): p. 20714. pmid:26765347
- 12. Assebe L.F., et al., Financial burden of HIV and TB among patients in Ethiopia: a cross-sectional survey. 2020. 10(6).
- 13. Humphrey J.M., et al., Mortality Among People With HIV Treated for Tuberculosis Based on Positive, Negative, or No Bacteriologic Test Results for Tuberculosis: The IeDEA Consortium. Open Forum Infectious Diseases, 2020. 7(1). pmid:32010735
- 14. Sari V.H.P., Rahardjo S.S., and Murti B.J.I.J.o.M., Mortality of TB-HIV Co-Infection Patients Based on CD4 Level: Meta-Analysis. 2022. 7(4): p. 456–470.
- 15. Naidoo K., Rampersad S., and Karim S.A., Improving survival with tuberculosis & HIV treatment integration: A mini-review. Indian J Med Res, 2019. 150(2): p. 131–138.
- 16.
Organization W.H., WHO consolidated guidelines on tuberculosis. Module 4: treatment-drug-susceptible tuberculosis treatment. 2022: World Health Organization.
- 17.
World Health Organization(WHO). Tuberculosis and HIV: a framework to address TB/HIV Co-infection in the Western Pacific Region. 2012; Available from: https://www.who.int/publications/i/item/9290610913.
- 18. Zhang J., et al., HIV and tuberculosis co-infection in East Asia and the Pacific from 1990 to 2017: results from the Global Burden of Disease Study 2017. 2019, 2019. 11(9): p. 3822–3835.
- 19.
World Health Organization (WHO). Guidelines for Antiretroviral Therapy & Management of opportunistic infection for Children, adolesecent&Adults living witth HIV 11 December 2023.; Available from: https://www.who.int/timorleste/publications/guidelines-for-antiretroviral-therapy—management-of-opportunistic-infections-for-children—adolescents—adults-living-with-hiv.
- 20. Ifa A.C., Human immunodeficiency virus as a determinant of tuberculosis treatment outcome in tuberculosis patients treated at Arba Minch General Hospital: a five-year retrospective study. HIV & AIDS Review. International Journal of HIV-Related Problems, 2018. 17(2): p. 74–80.
- 21. Alemu H., et al., Modeling Time to Death using the application of Survival model among HIV/TB Co-infected Patients who are under ART Follow-up in Axsum-Hospital, Ethiopia. 2021. 1(2): p. 160–173.
- 22. Page M., et al., & Moher D.(2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. 2021. 372(71): p. 1–9.
- 23.
Sandeep Moola, Z.M., Catalin Tufanaru, Edoardo Aromataris, Kim Sears, Raluca Sfetc, Marian Currie, et al. 7. Systematic reviews of etiology and risk (2020). Aromataris E, Lockwood C, Porritt K, Pilla B, Jordan Z,editors. JBI Manual for Evidence Synthesis 2024; Available from: https://synthesismanual.jbi.global.
- 24. Riley R.D., Higgins J.P., and Deeks J.J., Interpretation of random effects meta-analyses. Bmj, 2011. 342: p. d549. pmid:21310794
- 25. Higgins J.P., et al., Measuring inconsistency in meta-analyses. 2003. 327(7414): p. 557–560. pmid:12958120
- 26. Sterne J.A.C. and Egger M., Funnel plots for detecting bias in meta-analysis: Guidelines on choice of axis. Journal of Clinical Epidemiology, 2001. 54(10): p. 1046–1055. pmid:11576817
- 27. Shi L. and Lin L.J.M., The trim-and-fill method for publication bias: practical guidelines and recommendations based on a large database of meta-analyses. 2019. 98(23): p. e15987. pmid:31169736
- 28. Ali S.A., et al., Outcomes of TB treatment in HIV co-infected TB patients in Ethiopia: a cross-sectional analytic study. BMC Infectious Diseases, 2016. 16: p. 1–9.
- 29. Atalell K.A., Birhan Tebeje N., and Ekubagewargies D.T., Survival and predictors of mortality among children co-infected with tuberculosis and human immunodeficiency virus at University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia. A retrospective follow-up study. PLoS One, 2018. 13(5): p. e0197145. pmid:29787596
- 30. Balcha T.T., et al., Outcome of tuberculosis treatment in HIV-positive adults diagnosed through active versus passive case-finding. Global Health Action, 2015. 8(1): p. -1-N.PAG. pmid:25819037
- 31. Belayneh M., Giday K., and Lemma H., Treatment outcome of human immunodeficiency virus and tuberculosis co-infected patients in public hospitals of eastern and southern zone of Tigray region, Ethiopia. Braz J Infect Dis, 2015. 19(1): p. 47–51. pmid:25467923
- 32. Beyene Y., Geresu B., and Mulu A., Mortality among tuberculosis patients under DOTS programme: a historical cohort study. BMC Public Health, 2016. 16(1): p. 1–6. pmid:27561794
- 33. Chanie E.S., et al., Estimation of lifetime survival and predictors of mortality among TB with HIV co-infected children after test and treat strategies launched in Northwest, Ethiopia, 2021; a multicentre historical follow-up study. PLoS One, 2021. 16(12): p. e0258964. pmid:34932563
- 34. Dawit Z., et al., Incidence and predictors of mortality among children co-infected with tuberculosis and human immunodeficiency virus at public hospitals in Southern Ethiopia. PLoS One, 2021. 16(6): p. e0253449.
- 35. Gebremariam G., et al., Impact of HIV Status on Treatment Outcome of Tuberculosis Patients Registered at Arsi Negele Health Center, Southern Ethiopia: A Six Year Retrospective Study. PLoS One, 2016. 11(4): p. e0153239. pmid:27096159
- 36. Gebreyes D.E., A survival analysis of prognostic determinant factors of time-to-death of HIV/TB co-infected patients under HAART followed-up in a public hospital in Ethiopia. HIV & AIDS Review. International Journal of HIV-Related Problems, 2023. 22(2): p. 110–130.
- 37. Geleso M.G., Modeling the Survival of Tuberculosis Patients in Eastern Zone of Tigray Regional State. Risk Manag Healthc Policy, 2020. 13: p. 473–481. pmid:32581610
- 38. Gemechu J., et al., Predictors of mortality among TB-HIV co-infected children attending anti-retroviral therapy clinics of selected public hospitals in southern, Ethiopia: retrospective cohort study. Archives of Public Health, 2022. 80(1): p. 1–13.
- 39. Gesesew H., et al., Predictors of mortality in a cohort of tuberculosis/HIV co-infected patients in Southwest Ethiopia. Infect Dis Poverty, 2016. 5(1): p. 109. pmid:27915999
- 40. Kassa A., et al., Incidence of tuberculosis and early mortality in a large cohort of HIV infected patients receiving antiretroviral therapy in a tertiary hospital in Addis Ababa, Ethiopia. Trans R Soc Trop Med Hyg, 2012. 106(6): p. 363–70. pmid:22521216
- 41. Lelisho M.E., et al., Modeling Survival Time to Death Among TB and HIV Co-infected Adult Patients: an Institution-Based Retrospective Cohort Study. Journal of Racial & Ethnic Health Disparities, 2023. 10(4): p. 1616–1628. pmid:35697902
- 42. Palme I.B., et al., Impact of human immunodeficiency virus 1 infection on clinical presentation, treatment outcome and survival in a cohort of Ethiopian children with tuberculosis. Pediatr Infect Dis J, 2002. 21(11): p. 1053–61. pmid:12442029
- 43. Reepalu A., et al., Long-term Outcome of Antiretroviral Treatment in Patients With and Without Concomitant Tuberculosis Receiving Health Center-Based Care-Results From a Prospective Cohort Study. Open Forum Infect Dis, 2017. 4(4): p. ofx219. pmid:29226173
- 44. Seyoum E., et al., Increased Mortality in HIV Infected Individuals with Tuberculosis: A Retrospective Cohort Study, Addis Ababa, Ethiopia. HIV AIDS (Auckl), 2022. 14: p. 143–154. pmid:35370424
- 45. Shaweno D. and Worku A., Tuberculosis treatment survival of HIV positive TB patients on directly observed treatment short-course in Southern Ethiopia: a retrospective cohort study. BMC Res Notes, 2012. 5: p. 682. pmid:23234241
- 46. Sileshi B., et al., Predictors of mortality among TB-HIV Co-infected patients being treated for tuberculosis in Northwest Ethiopia: a retrospective cohort study. BMC Infectious Diseases, 2013. 13(1): p. 297–297. pmid:23815342
- 47. Sime T., et al., Effect of active tuberculosis on the survival of HIV-infected adult patients who initiated antiretroviral therapy at public hospitals of Eastern Ethiopia: A retrospective cohort study. PLoS One, 2022. 17(10): p. e0277021. pmid:36315573
- 48. Teklu A.M., et al., Factors Associated with Mortality of TB/HIV Co-infected Patients in Ethiopia. Ethiop J Health Sci, 2017. 27(Suppl 1): p. 29–38. pmid:28465651
- 49. Tola A., et al., Treatment Outcome of Tuberculosis and Associated Factors among TB-HIV Co-Infected Patients at Public Hospitals of Harar Town, Eastern Ethiopia. A five-year retrospective study. BMC Public Health, 2019. 19(1): p. 1–12.
- 50. Weldegebreal F., Mitiku H., and Teklemariam Z., Treatment outcome of tuberculosis among Human Immunodeficiency Virus positive patients in Eastern Ethiopia: a retrospective study. Pan Afr Med J, 2018. 30: p. 32. pmid:30167059
- 51. Wondimu W., Dube L., and Kabeta T., Factors Affecting Survival Rates Among Adult TB/HIV Co-Infected Patients in Mizan Tepi University Teaching Hospital, South West Ethiopia. HIV AIDS (Auckl), 2020. 12: p. 157–164. pmid:32368156
- 52. Adegeh S.C. and Kebede E., Statistical Modeling of Survival of Tuberculosis Infected HIV Patients Treated with Antiretroviral Treatment: A Case of Felege Hiwot Referral Hospital, Bahir Dar, Ethiopia. 2021.
- 53. Alemu M.A., et al., Impact of HIV-AIDS on tuberculosis treatment outcome in Southern Ethiopia–a retrospective cohort study. 2021. 25: p. 100279.
- 54. Birhan H., et al., Statistical analysis on determinant factors associated with time to death of HIV/TB co-infected patients under HAART at Debre Tabor Referral Hospital: an application of accelerated failure time-shared frailty models. 2021: p. 775–787.
- 55. Fekadu G., et al., Impact of HIV status and predictors of successful treatment outcomes among tuberculosis patients: A six-year retrospective cohort study. 2020. 60: p. 531–541.
- 56. Gezae K.E., et al., Predictors of accelerated mortality of TB/HIV co-infected patients on ART in Mekelle, Ethiopia: an 8 years retrospective follow-up study. 2019. 3: p. 2–6.
- 57. Habtegiorgis S.D., et al., TB/HIV Co-Infected Patients’ Mortality Rate and its Predictors in Dire Dawa, Eastern Ethiopia, 2018. 2018.
- 58. Lelisho M.E., et al., Survival rate and predictors of mortality among TB/HIV co-infected adult patients: retrospective cohort study. Sci Rep, 2022. 12(1): p. 18360. pmid:36319734
- 59. Nigussie J., et al., Predictors of Mortality Among Children Co-Infected with Tuberculosis and Human Immunodeficiency Virus in Region, North Ethiopia, Retrospective Follow-Up Study. 2021. 39(1): p. 31026–31034.
- 60. Refera H. and Wencheko E.J.E.j.o.h.d., Survival of HIV-TB co-infected adult patients under ART in Ambo Referral Hospital, Ethiopia. 2013. 27(2): p. 88–93.
- 61. Sinshaw Y., et al., Successful TB treatment outcome and its associated factors among TB/HIV co-infected patients attending Gondar University Referral Hospital, Northwest Ethiopia: an institution based cross-sectional study. 2017. 17(1): p. 1–9. pmid:28178936
- 62. Teshome Kefale A. and Anagaw Y.K.J.I.J.o.G.M., Outcome of tuberculosis treatment and its predictors among HIV infected patients in southwest Ethiopia. 2017: p. 161–169.
- 63. Podlekareva D.N., et al., Tuberculosis-related mortality in people living with HIV in Europe and Latin America: an international cohort study. 2016. 3(3): p. e120–e131. pmid:26939735
- 64. Khan F.A., et al., Treatment of Active Tuberculosis in HIV-Coinfected Patients: A Systematic Review and Meta-Analysis. Clinical Infectious Diseases, 2010. 50(9): p. 1288–1299. pmid:20353364
- 65. Rossetto M., et al., Factors associated with hospitalization and death among TB/HIV co-infected persons in Porto Alegre, Brazil. 2019. 14(1): p. e0209174. pmid:30601842
- 66. Isaakidis P., et al., Treatment outcomes for HIV and MDR-TB co-infected adults and children: systematic review and meta-analysis. 2015. 19(8): p. 969–978. pmid:26162364
- 67. Ethiopia F.J.A.H., National guidelines for comprehensive HIV prevention, care and treatment. 2017. 2(1).
- 68. Domingos M.P., Caiaffa W.T., and Colosimo E.A., Mortality, TB/HIV co-infection, and treatment dropout: predictors of tuberculosis prognosis in Recife, Pernambuco State, Brazil. Cad Saude Publica, 2008. 24(4): p. 887–96. pmid:18392367
- 69. Dlatu N., Longo-Mbenza B., and Apalata T., Models of integration of TB and HIV services and factors associated with perceived quality of TB-HIV integrated service delivery in O. R Tambo District, South Africa. BMC Health Serv Res, 2023. 23(1): p. 804. pmid:37501061
- 70. Tao Y., et al., Prevalence of delayed antiretroviral therapy initiation among people living with HIV: A systematic review and meta-analysis. 2023. 18(10): p. e0286476. pmid:37874794
- 71. Eshetie S., et al., Tuberculosis treatment outcomes in Ethiopia from 2003 to 2016, and impact of HIV co-infection and prior drug exposure: A systematic review and meta-analysis. 2018. 13(3): p. e0194675. pmid:29554144
- 72. Ayobami A.A., et al., RISK FACTORS ASSOCIATED WITH HIV-TB CO-INFECTION AMONG CLINIC ATTENDEES IN DOTS AND ART CENTRES IN IBADAN, NIGERIA. 2022. 2(7): p. 21–38.
- 73. Abioye A.I., et al., Anemia, Iron Status, and HIV: A Systematic Review of the Evidence. Adv Nutr, 2020. 11(5): p. 1334–1363. pmid:32383731
- 74.
Calderwood C.J., et al. Risk Factors for Hospitalization or Death Among Adults With Advanced HIV at Enrollment for Care in South Africa: A Secondary Analysis of the TB Fast Track Trial. in Open Forum Infectious Diseases. 2022. Oxford University Press.
- 75. Redig A.J. and Berliner N.J.H., the American Society of Hematology Education Program Book, Pathogenesis and clinical implications of HIV-related anemia in 2013. 2013. 2013(1): p. 377–381.
- 76. Damanik E.H., et al., ANEMIA, NUMBER OF T CD4 LYMPHOCYTE, OPPORTUNISTIC INFECTIONS AS MORTALITY PREDICTOR FACTORS IN PATIENTS INFECTED OF HUMAN IMMUNODEFICIENCY VIRUS (HIV) IN HAJI ADAM MALIK GENERAL HOSPITAL MEDAN. 2019. 6(10): p. 26–33.
- 77. de Olalla P.G., et al., Impact of adherence and highly active antiretroviral therapy on survival in HIV-infected patients. 2002. 30(1): p. 105–110.
- 78. Mabunda T.E., Ramalivhana N.J., and Dambisya Y.M.J.A.H.S., Mortality associated with tuberculosis/HIV co-infection among patients on TB treatment in the Limpopo province, South Africa. 2014. 14(4): p. 849–854.