Correlates of HIV-TB co-infection and mental health of adults living in countries across sub-Saharan Africa: systematic review and meta-analysis protocol
Correlates of HIV-TB co-infection and mental health of adults living in countries across sub-Saharan Africa: systematic review and meta-analysis protocol
HIV and tuberculosis (TB) are two of the most devastating disease conditions of public health concern globally. A co-infection of the two diseases poses serious health challenges to patients including mental health problems, ranging from mild to severe, with differing outcomes. This systematic review aims to assess the correlates of HIV-TB co-infection and mental health of adults living in sub-Saharan Africa.
We will conduct comprehensive database and non-database searches for studies (published and unpublished). We will search Google Scholar, PubMed, CINAHL, LILACS, JSTOR, Cochrane Library, SCOPUS, PsycINFO, HINARI and African Journals Online from inception to 31 May 2025, without restriction on language. We will also search the grey literature, including conference proceedings, preprint repositories, databases of dissertations, WHO and governmental databases. In the event where there is a need to contact experts and corresponding authors for further information, the review team will do so to enrich the content of the review. At least three reviewers will independently undertake study selection, data extraction and risk of bias assessment using validated tools. We will resolve discrepancies or disagreements through discussion. We will analyse dichotomous data as risk ratio, OR or proportion and continuous data as mean difference with their SD; all estimates will be presented with their 95% CI. Where applicable, we will determine SD from point estimates and the appropriate denominators assuming a binomial distribution. The magnitude of heterogeneity between the included studies will be assessed quantitatively using the index of heterogeneity (I2 statistic). The I2 values of 25%, 50% and 75% will be considered to represent low, moderate and significant heterogeneity. The significance of heterogeneity will be determined by the p value of the I2 statistic, and a p value of <0.05 will be considered as statistically significant. For studies with moderate to significant heterogeneity, the random-effects model will be used to obtain a pooled estimate of the outcome, and if heterogeneity is low, a fixed-effect model will be used.
This systematic review will collate secondary research based on publicly available published and unpublished studies, and no ethical approval is required. However, an eligible study with serious ethical issues will be excluded and the reasons for exclusion documented. The review findings will be shared with key stakeholders, health authorities, agencies involved in the mental health of persons living with HIV-TB (PLHTB) co-infection, social services providers and policy implementers. The findings will be presented at scientific conferences and symposia. The final review report will be in the form of a scientific paper in a high-impact factor peer-reviewed journal.
CRD42024572331.
http://creativecommons.org/licenses/by-nc/4.0/
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
The co-existence of HIV and tuberculosis (TB) is a major public health challenge worldwide.1 Persons Living with HIV (PLHIV) have disproportionately high risk of dying when co-infected with TB even when on recommended treatment.2 About 820 000 of the 10 million people who were infected with TB in 2019 were also living with HIV.3 More than 160 000 PLHIV co-infected with TB died globally in 2022.4
Sub-Saharan Africa (SSA) accounts for the highest incidence and prevalence of HIV-TB co-infection, of which South Africa alone records 50% of all cases.5 High levels of mental health conditions have been identified among persons living with HIV or TB or HIV-TB co-infection.6 Persons living with HIV-TB (PLHTB) co-infection have a substantial adverse impact on their mental health outcomes.7 Both standard assessments (using WHO tools) and self-assessments show that more than 70% of PLHTB have low quality of life (QoL).8 Depression is one of the most common mental health challenges confronting PLHTB, with more than half of patients experiencing various forms of mental health problems.7 Predictors of mental health disorders among co-infected patients have been found to include absence of family care, self-inflicted stigma and low socioeconomic status.9
Most deaths among PLHIV (about 20%) are as a result of co-infection with TB.10 Advanced HIV infection with a decreased immune system increases the chances of TB co-infection, which in turn increases the chances of death among patients.11 Within the sub-Saharan region, the prevalence of HIV among TB-infected persons was estimated to be around 32%.12 There is a synergistic and bidirectional association between HIV-TB co-infection and mental health.13 HIV-TB co-infection is a major risk factor for cognitive decline, and poor mental health has the propensity of increasing the risk of contracting HIV, TB or both with poor prognosis.14 The effects of mental health outcomes on the detection and management of HIV and TB can be dire, while HIV and TB can independently and jointly affect the mental well-being and mental health outcomes of patients.15
An integrated service delivery for HIV, TB and mental health is critical in improving treatment outcomes and the QoL of patients.16 17 Effective integrated service delivery is projected to avert over a million new HIV infections and 14 million TB infections and deaths by 2030.18 There is a need to involve key stakeholders including clinicians, programme implementers, policy makers and patients to understand the effects of HIV on TB treatment and vice-versa and the implications for the provision of integrated services for co-infected persons.5 This will contribute immensely to achieving the global targets of ending HIV and TB and associated morbidities.
The United Nations’ sustainable development goals (SDGs) recognise the inter-relation between physical and mental health and the tendency for either of them to derail the gains of the other.19 Target 3.3 of the SDGs seeks to end HIV/AIDS and TB by the end of 2030. This target has become difficult to attain, especially after the COVID-19 global pandemic.20 The impact of the COVID-19 pandemic further exacerbated the plights of persons living with HIV or TB, alone or co-infected. This negatively impacted the provision of services to patients and affected the achievement of the set target of 75% decline in HIV incidence.21 Similarly, the pandemic in 2020 resulted in more than 1.3 million deaths from TB alone, a larger proportion being reported among PLHTB co-infection.21 The SDG target 3.4 focuses on ensuring sound mental health for all, irrespective of one’s physical and social state.22 Only a few (<25%) of the WHO member states have been able to comprehensively integrate mental health as part of their health systems. This is woefully inadequate in addressing the global challenges posed by mental health and the capacity to achieving the goal of promoting mental health and well-being universally.23
Persons living with HIV-TB co-infection, particularly patients in SSA, experience a greater burden of mental health challenges than those with single infection.24 Patients with HIV and TB co-infection are four times more likely to experience suicidal ideation25 and poor psychological QoL compared with those with a single infection.26 27 However, the presence of psychological disorders and their effects on the well-being of PLHTB is not well examined.28 More than 10% of PLHTB co-infection were likely not to adhere to treatment, leading to extrapulmonary TB, especially among HIV patients with undisclosed status.29 Mental health support is critical to keep persons with HIV-TB co-infection healthy and in care for better outcomes.30 The rationale for conducting this systematic review is to provide stronger evidence on the impact of HIV-TB co-infection on patients’ mental health and to recommend interventions, policies and programmes to ameliorate the challenges of PLHTB in SSA.
The research team conducted searches on PROSPERO and other databases including PubMed, Google Scholar, PsycINFO, African Journals Online and HINARI for ongoing or published systematic reviews on correlates of HIV-TB co-infection and mental health outcomes. Six systematic reviews were considered potentially closely related to this systematic review.6 31–35 One systematic review covered SSA, but the focus was depression in adults living with HIV,31 and another assessed the effects of mental health disorders among HIV patients in Ethiopia,32 while the remaining four reviews focused on TB alone.6 33–35 The only systematic review that investigated PLHTB co-infection and mental health was based on patients from one country (Ethiopia)36 and not SSA.
This systematic review aims to answer the following questions: (1) What is the magnitude of mental health problems among persons co-infected with HIV and TB in countries across SSA? (2) What are the common mental health conditions reported by or identified in persons with HIV-TB co-infection? (3) What are the correlates of HIV-TB co-infection and mental health problems in these persons? The specific objectives are: (1) to assess the magnitude of mental health problems among persons co-infected with HIV and TB in countries across SSA, (2) to identify the common mental health conditions among persons with HIV-TB co-infection and (3) to assess the correlates of HIV-TB co-infection and mental health problems.
This systematic review protocol has been prepared in line with the Cochrane Handbook.37 We followed the methods and formats reported in earlier published works.38–48 We have reported the review following the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) Extension for Protocols guidelines49 (online supplemental file 1). The full review report will be guided by the PRISMA checklist and the PRISMA flow diagram to document the flow of studies during the searches and selection process50 (online supplemental file 2). The review will adopt the Patient, Intervention, Comparison, Outcome, Study (PICOS) framework to comprehensively describe the population, intervention, control/comparator, outcomes and study types. The overall quality of evidence produced from this review will be assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) (https://www.gradeworkinggroup.org/).51 The full review is planned to take, on average, 6 months to complete, starting from 1 April to 31 August 2025.
The review questions and the outcomes have been developed with patient and public involvement to make sure what matters to the patient, their families and the general community was captured and included, and any choices were informed by their priorities, experiences and preferences in line with Guidance for Reporting Involvement of Patients and the Public.52
Type of studies
This review will consider observational studies, including cohort, case-control and cross-sectional studies that focused on correlates of HIV-TB co-infection and mental health problems in the SSA region. Given that this is not an intervention effectiveness review, randomised controlled trials (RCTs) and quasiexperimental studies will not be the focus. However, if an RCT provides any mental health correlates and has a well-defined denominator, such a study will be eligible for inclusion. Reviews, opinions, commentaries, case reports and case series will not be eligible for inclusion, although these documents may serve as sources for potentially eligible primary studies. Studies reporting a country or regional-level estimate will not be eligible for inclusion as such studies do not have a well-defined source population. In instances where the results of a multicountry study have been lumped together and there is no way of disaggregating the data, such studies will be excluded from the meta-analysis. This section of the review has been reported along the formats used in earlier published works.42 45 50 52
Participants
The eligible population for inclusion in this systematic review will be adults aged 18 years and above co-infected with HIV and TB, receiving medication or not and living in the SSA region. The HIV and TB status of participants should have been determined using a validated screening/diagnostic tool. For HIV, clinical and PCR,53 and for TB, clinical54 and PCR55 diagnoses will be considered eligible for inclusion. Any special properties of the tools should have been reported. Participants should have been either registered at a recognised health facility providing antiretroviral therapy for HIV and standard services, and for TB, participants should be receiving their routine medications and standard care or not registered and not on any treatment. No restrictions will be placed on participants’ gender, ethnicity or other demographic characteristics. The participants’ inclusion section of this systematic review is in line with the reporting in earlier published works.56
Intervention/exposure
This is not an intervention effectiveness review but rather HIV-TB co-infection as an exposure and its correlates with mental health.
Persons living with HIV alone or TB alone.
Primary outcomes
The primary outcomes are:
The standardised Diagnostic and Statistical Manual of Mental Disorders 5 will be the main tool for the diagnosis of mental health conditions. However, given the paucity of data regarding which tool is most widely used across countries in SSA, the following tools will also be considered: Beck’s Depression Inventory Scale,24 Kessler Psychological Distress Scale,36 Client Diagnostic Questionnaire,57 The Montreal Cognitive Assessment,58 Depression Anxiety Stress Scale-2159 and International HIV Dementia Scale.60 The mini-mental state examination61 and Simioni Symptom Questions.62
Secondary outcomes
The following secondary outcomes will be considered:
Electronic searches
We will search all relevant electronic databases for primary studies (published and unpublished) on HIV-TB co-infection, following standards in earlier research.47 We will search PsycINFO, Google Scholar, PubMed, CINAHL, LILACS, JSTOR, Cochrane Library, SCOPUS, HINARI and African Journals Online from inception to 31 May 2025, without restriction on language. We will also search the grey literature including conference proceedings, preprint repositories, databases of dissertations, WHO and governmental databases. The key search terms to be used include ‘HIV and TB co-infection’, ‘quality of life’, ‘mental health disorder’, ‘mental health problem’, ‘HIV’, ‘tuberculosis’, ‘TB’ and ‘sub-Saharan Africa’, ‘SSA’ etc, with all applicable synonyms, alternatives, both American and British spellings, as well as plural and singular terms (see table 1 for full details). All SSA countries based on the Demographic and Health Survey Country List64 will be included as individual search terms. The PubMed search strategy has been presented in table 2. It will be adapted for the searches in the other databases.
Table 1
Search terms/concepts used for developing the search strategy
Table 2
Search strategy developed for PubMed (to be adapted for the other databases)
Other sources to be searched
We will hand search the reference lists of published systematic reviews and relevant studies on the subject. Where necessary, experts in the field will be contacted for information about completed but unpublished studies that could be included in the review.
Managing the search output and selecting studies
We will aggregate all the studies retrieved from the various sources (database and non-database sources) in Mendeley Reference Manager65 (https://www.mendeley.com) where duplicates will be removed. The deduplicated records will be exported to Rayyan QCRI66 (https://new.rayyan.ai/) for screening and selection of studies using a pretested study selection flow chart developed from the prespecified eligibility criteria (figure 1). At least two members of the review team will independently screen the titles and abstracts to exclude the most obvious irrelevant studies. Then full texts of studies found to be potentially relevant will be retrieved for further screening for inclusion in the review. In situations where there are disagreements between the independent screening authors, these will be resolved through discussion. The PRISMA flow diagram50 will be provided to show details of the study flow through the study selection process.
Figure 1
Study selection flow chart developed from the criteria for considering studies for inclusion in this review. PHLTB, persons living with HIV-TB; RCT, randomised controlled trial; SSA, sub-Saharan Africa; TB, tuberculosis.
Data extraction and management
A template will be created in Excel to extract characteristics of the studies to be included in the review such as study ID, year study was published, year study was conducted, study area/setting, the country where the study was conducted, study design, sample size, age of participants, HIV, TB and mental health status as well as information about any medication patients received or were receiving. Information on tools used for diagnosing HIV, TB and mental health conditions and scales for assessing QoL will be extracted. If two or more publications have been made from the same study, information in the articles will be extracted as one study. The article with the most information will be selected as the parent study, and the other articles will be linked with the ‘parent’ study. The data extraction will be done independently by two members of the review team and, where necessary, verified by a third reviewer for purposes of agreement, accuracy and consistency. In situations where more than two reviewers extract data, 10% of the studies extracted will be sampled and independently verified by an experienced reviewer in the team as a quality check. Where necessary, the data extraction form will be updated based on experiences learnt to incorporate any additional information. All changes made will be documented in a transparent manner. In cases of missing data, authors of the primary study will be contacted for any possibility that they could provide the needed information. Missing data will not be imputed through any computational means. Any discrepancies in data extraction will be resolved through discussion between the reviewers.
Assessment of risk of bias (quality) in the included studies
The Cochrane tool for non-randomised studies of exposure (ROBINS-E) assessment tool (V.20 June 2023) (online supplemental file 3) will be used to assess the risk of bias in the included studies based on a series of signalling questions across seven risk of bias domains: confounding, selection of study participants, measurement of exposure, postexposure intervention, missing data, measurement of outcome and selection of reported results. The signalling questions have response options ‘Yes’, ‘Probably Yes’, ‘Probably No’, 'No' and ‘No Information’. The risk of bias will be judged as ‘low’ for a domain with little or no concern about bias. Where there are some concerns about bias in a specific domain, but with no certainty of an important risk of bias, we will judge the domain as ‘Some Concerns’. For domains with some important bias concerns, we will judge as ‘High risk’. Studies with suspected serious bias will be judged as having a ‘very high risk of bias’. The results from the risk of bias assessment will be presented in a table with supporting statements from the primary studies. Disagreements between reviewers will be resolved through discussion. Where appropriate, the risk of bias tool developed by Hoy et al67 (online supplemental file 4) will also be employed to assess the quality of prevalence studies. The quality assessment tool for prevalence studies looks at the risk of bias in observational studies on four domains: selection bias, non-response bias, measurement bias and bias related to data analysis. Each domain will be judged on three levels: ‘low’, ‘high’ or ‘unclear’ risk of bias. The overall quality of evidence contributed by each study will be classified as ‘low risk of bias’ or ‘high risk of bias’. Quality assessment will be done independently by at least two of the review team members, and any disagreements will be resolved through discussion.
Plan for data analysis and heterogeneity assessment
The findings from the characteristics of the included studies will be summarised narratively with the aim to answer questions pertaining to the PICOS elements including the type and characteristics of the participants/population, type of studies, interventions and outcomes of interest. A more robust analysis will be done using Stata V.18.0 (Stata Corp LLC, Texas, USA)68 and Review Manager V.5.4.69 Dichotomous data will be presented as OR or risk ratio, and for continuous outcome data, mean difference will be used together with SD for normally distributed data. All effect estimates will be expressed with their respective 95% CIs. Data that do not follow the normal distribution will be transformed using the logarithmic transformation and analysed using the normal distribution assumption. The magnitude of heterogeneity between included studies will be assessed quantitatively using the I2 statistic. The I2 values of 25%, 50% and 75% will be considered as representing low, moderate and significant heterogeneity, respectively.70 The significance of heterogeneity will be determined by the p value of the I2 statistic, and a p value of <0.05 will be considered statistically significant. The random-effects model will be used to obtain pooled estimates of the outcome. Where a meta-analysis is not appropriate, a narrative synthesis will be conducted to summarise the evidence across studies and the results presented in tables.
Where heterogeneity is determined to be appreciable, subgroup analyses will be performed to identify the possible sources of heterogeneity. Subgroup analysis will mainly be conducted around the PICOS domains and especially around the known potential confounding variables such as age, sex, study setting (hospital or community-based) and geographical location (Central, East, Southern and West Africa).
The authors of the primary studies will be contacted in the case of missing data on pertinent variables, where possible. Where authors are unable to provide the requested information or cannot be reached, missing data will not be imputed. Instead, the potential impact of missing data on the findings of the review will be discussed in the discussion.
Where data permits, sensitivity analysis will be conducted to assess the robustness of the review findings to outlier studies with unusually high weightings. This will involve re-running the meta-analysis by excluding studies one by one to determine the impact each study has on the overall effect estimate. Additionally, we will perform sensitivity analyses by excluding studies with high risk of bias and other key methodological characteristics deemed to have inherent flaws. This process will help to identify whether any single study or group of studies will not disproportionately influence the pooled estimates, thereby the review conclusions.
The quality of evidence for the primary and secondary outcomes will be assessed using the GRADE approach.71 The GRADE system classifies the quality of evidence as high, moderate, low or very low based on the study design, risk of bias, consistency of results, directness of evidence, precision of estimates and potential publication bias. The certainty of evidence can be rated downward by one or two levels when there are serious or very serious concerns, respectively, in any of the domains.71 72 The certainty of evidence can be rated upward for high-quality observational studies, although this is usually rarely done.72 73 An overall rating of the certainty of the evidence for each outcome will be presented in a summary of findings table together with the study types, the number of studies and participants and the relative and absolute effects for each outcome.73 Each outcome will be independently evaluated by two reviewers, and any discrepancies will be resolved through discussion or consultation with a third reviewer.
This systematic review will use secondary data from published studies and no new data will be collected directly from human participants. As such, there are no direct ethical concerns related to data collection. Ethical considerations will focus on ensuring the integrity and transparency of the review process, including proper citation of sources, avoiding plagiarism, and reporting conflicts of interest. Studies with serious ethical issues will be removed from this systematic review, and the reasons for exclusion will be documented. Findings of the final review report will be disseminated through peer-reviewed journal publication and presentation at scientific conferences. It will also be shared with key stakeholders, including national and international health authorities in HIV and TB care, mental health authorities, healthcare professionals and policy makers, HIV-TB and mental health advocates and the general population.
Not applicable.
This systematic review was prepared as part of the capacity-building initiative of the Centre for Evidence Synthesis and Policy (CESP), University of Ghana in partnership with the Africa Communities of Evidence Synthesis and Translation (ACEST), Accra, Ghana, that train health professionals and scientists in Low and Middle-Income Countries (LMICs) evidence synthesis and research translation. Welbeck Amoani Twum, Mishael Yankey and Isaac Kojo Appiah are PhD students at the School of Public Health, University of Ghana specialising in Evidence Synthesis and Translation; they are mentored by Prof. Anthony Danso-Appiah (Director, Centre for Evidence Synthesis and Policy, University of Ghana, Legon, Accra).
