https://orcid.org/0009-0009-8872-732X
Figures
Abstract
Tuberculosis is a major infectious disease worldwide, but currently available diagnostics have suboptimal accuracy, particularly in patients unable to expectorate, and are often unavailable at the point-of-care in resource-limited settings. Test/treatment decision are, therefore, often made on clinical grounds. We hypothesized that contextual factors beyond disease probability may influence clinical decisions about when to test and when to treat for tuberculosis. This umbrella review aimed to identify such factors, and to develop a framework for uncertainty in tuberculosis clinical decision-making. Systematic reviews were searched in seven databases (MEDLINE, CINAHL Complete, Embase, Scopus, Cochrane, PROSPERO, Epistemonikos) using predetermined search criteria. Findings were classified as barriers and facilitators for testing or treatment decisions, and thematically analysed based on a multi-level model of uncertainty in health care. We included 27 reviews. Study designs and primary aims were heterogeneous, with seven meta-analyses and three qualitative evidence syntheses. Facilitators for decisions to test included providers’ advanced professional qualification and confidence in tests results, availability of automated diagnostics with quick turnaround times. Common barriers for requesting a diagnostic test included: poor provider tuberculosis knowledge, fear of acquiring tuberculosis through respiratory sampling, scarcity of healthcare resources, and complexity of specimen collection. Facilitators for empiric treatment included patients’ young age, severe sickness, and test inaccessibility. Main barriers to treatment included communication obstacles, providers’ high confidence in negative test results (irrespective of negative predictive value). Multiple sources of uncertainty were identified at the patient, provider, diagnostic test, and healthcare system levels. Complex determinants of uncertainty influenced decision-making. This could result in delayed or missed diagnosis and treatment opportunities. It is important to understand the variability associated with patient-provider clinical encounters and healthcare settings, clinicians’ attitudes, and experiences, as well as diagnostic test characteristics, to improve clinical practices, and allow an impactful introduction of novel diagnostics.
Citation: Basile FW, Sweeney S, Singh MP, Bijker EM, Cohen T, Menzies NA, et al. (2024) Uncertainty in tuberculosis clinical decision-making: An umbrella review with systematic methods and thematic analysis. PLOS Glob Public Health 4(7): e0003429. https://doi.org/10.1371/journal.pgph.0003429
Editor: Sok King Ong, The Chinese University of Hong Kong Faculty of Medicine, HONG KONG
Received: December 19, 2023; Accepted: June 12, 2024; Published: July 23, 2024
Copyright: © 2024 Basile 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 relevant to replicate the study have been provided in the article or uploaded as supplementary information.
Funding: Research reported in this publication was supported by the National Institute Of Allergy And Infectious Diseases of the National Institutes of Health (U01AI152084 to FWB, SS, MPS, TC, NAM, AV, PI). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Tuberculosis (TB) is a major infectious cause of morbidity and mortality globally. In 2022, 7.5 million people were diagnosed with TB, and 1.3 million people died because of the disease [1]. Missed or delayed TB diagnosis and treatment and low quality of care remain critical obstacles to disease control and improving health outcomes [2, 3].
To minimize diagnostic and treatment delays, high quality TB services should include access to rapid, affordable and accurate tests, such as the molecular WHO-recommended rapid diagnostics (mWRD) [4]. However, mWRD are seldom available at the point-of-care in resource-limited settings. Despite massive efforts to coordinate the global roll-out of GeneXpert (Cepheid, USA), recent data still show that this test is unavailable in many peripheral settings and more generally the underutilization of modern TB diagnostic technologies [5, 6].
The underutilization of diagnostics may arise due to a variety of factors, including as a consequence of providers’ know-do gap [7]. This may become particularly evident in situations where care is tailored around the patient’s perceived needs (e.g., clinicians offering a more affordable but less accurate diagnostic test), and best practices are not implemented (e.g., clinicians choosing quick symptom relief with low-cost pharmaceuticals over diagnostic certainty) [7]. Moreover, in resource-limited settings, when a patient presents with signs and symptoms suggestive of TB, clinicians may decide to start treatment based solely on clinical grounds, regardless of test availability [8, 9].
To standardize decision-making, pre- and post-test disease probabilities have been used to determine the thresholds for testing and treatment decisions [10, 11]. The provider determines a pre-test probability of disease, which varies depending on clinical signs and symptoms as well as the provider’s experience, knowledge, and health care setting. The provider then decides whether to move forward with testing or initiating treatment. Following testing, the provider determines the post-test probability of disease and decides whether to start or withhold TB therapy [11].
There have also been multiple attempts to develop scoring systems and clinical prediction models for TB screening and diagnosis [12–16]. Scoring systems can help to calculate the probability of TB disease in a reproducible way and might be particularly helpful in paediatric TB, where currently available diagnostic tests lack high sensitivity. Additionally, clinical algorithms might help determine when testing is helpful and when a negative test is insufficient to withhold treatment [17].
However, in reality, the decision to test or treat presumptive TB cases can be affected by contextual variables beyond accessibility to diagnostics, or a mere computation of disease probability [18]. Provider characteristics, including their ability to cope with complexity, risk, and uncertainty, contribute to process variability [19]. Uncertainty is an inevitable component of clinical practice and can occur throughout the decision-making process: when formulating clinical hypotheses, identifying a diagnosis, choosing a test and interpreting its result, and interpreting patient preferences [20]. Multilevel models of uncertainty emphasize the dynamic interplay between different sources and types of uncertainty at each level, and may be useful to classify the challenges of clinical decision-making [20].
Understanding uncertainty in the TB decision-making process and the reasons why a provider would initiate empiric treatment or would not utilize a microbiological test even when available, is important to develop diagnostic tools that improve TB diagnosis and care behaviours and practices, and to project the impact of the introduction of novel diagnostic aids [21]. This umbrella review of systematic reviews (SR) aimed to identify factors influencing providers’ decisions to test for TB, and initiate TB treatment in adult and paediatric patients with presumptive TB in high-TB and TB/HIV burden countries [22].
Methods
Study design rationale and methodology
An initial scoping search was conducted on MEDLINE (via OVID) for terms related to “tuberculosis” and “decision-making”, and identified several reviews relevant to our research question [23–25]. Since most records evaluated either qualitative or quantitative primary studies, and often reported complementary findings, we chose an umbrella review design to allow for the inclusion of these reviews with a broad scope of inquiry and to achieve a higher level of synthesis [26–28].
The study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [29]. The Joanna Briggs Institute (JBI) guidelines for umbrella reviews [28, 30], and the Cochrane guidance for overviews of reviews [31] were also followed to address the specific issues arising when conducting umbrella reviews. The methodology of this review was prespecified in a protocol [32].
Search strategy
Using a combination of key terms to maximize sensitivity, seven electronic databases were searched: MEDLINE (via OVID), CINAHL Complete, Embase, Scopus, Cochrane Central, the PROSPERO register, and Epistemonikos database. The search was limited from January 2007 (considering that the development Xpert MTB/RIF was completed in 2009) to the date of search, which was the 4th of July 2022. The search was rerun on the 21st of July 2023. We developed a comprehensive list of keywords and synonyms for each broad domain: 1) TB, 2) clinical decision-making. Terms were searched individually first and then combined using Boolean operators. The search was piloted in MEDLINE and repeated in all databases. Where applicable, MeSH and free text terms were combined to identify relevant studies. The search strategy was developed with the support of a librarian at the LSHTM.
Details on the search strategy are presented in S1 Appendix. Articles in English, French, Spanish, Portuguese, or Italian were considered. A search of the grey literature was not conducted.
Selection and appraisal of records
Records were selected on predefined inclusion and exclusion criteria guided by the Population, Intervention, Comparison, Outcome and Study design/setting (PICOS) framework (S1 Table) [33]. Inclusion criteria consisted of population (individuals with presumptive pulmonary TB and health care providers involved in TB diagnosis and treatment), findings/outcomes (any relevant to clinical decision-making), and setting (high TB burden countries). We considered relevant to the decision-making any intervention, action, or event that influenced the diagnosis of TB. SRs, meta-analyses, and SRs of qualitative studies (hereinafter referred to as qualitative evidence syntheses) were included. Articles exclusively on drug-resistant TB, non-review articles, and reviews that did not use systematic methods were excluded (S1 Table).
Following removal of duplicates, the title/abstract screening was carried out by a single reviewer (FWB). The full text of selected records was then examined for inclusion in the study, based on the predefined criteria (S1 Table).
Quality appraisal
Methodological quality, risk of bias and reporting quality of reviews were assessed using the JBI checklist for SRs [28, 30]. No records were excluded on grounds of quality due to lack of consensus on the most appropriate tools and approaches for managing low-quality reviews in umbrella reviews [34] (S2 Table). Where available, GRADE assessments [35, 36] were extracted and reported.
Overlap assessment
Several approaches have been proposed for overlap management in umbrella reviews [37]. We included all eligible reviews and documented the extent of overlap in primary studies using the Corrected Covered Area (CCA) index [37]. After obtaining the overall CCA, pairwise indexes were calculated (S1 Fig). For reviews with moderate to high pairwise CCA, research aims and reported outcomes were examined. If two reviews had the same aims, findings from the highest quality review were described [37].
Data extraction
Study characteristics and data of interest for included records were extracted by a single reviewer (FWB) [30, 31, 38]. These included: type of review, title, authors, publication year, number of studies and participants included in the review, aims/objectives/PICO question (or equivalent), search strategy, methodological quality/risk of bias and certainty of evidence assessment. For reviews examining global data, only findings pertinent to high-burden TB settings were extracted. Data extraction also indicated where pooled analyses included non-high TB burden countries.
Primary studies from reviews were not retrieved.
Data synthesis
Data synthesis used a systematic narrative approach for umbrella reviews [38], which involved thematic content analysis and coding of findings from each review to identify recurring themes associated with factors influencing TB clinical decision-making. Nvivo (version 1.5, 2021, QSR International Pvt. LTD, Australia) was used to iteratively code extracted key data. Themes were developed separately for quantitative and qualitative studies, then combined and presented complementarily [39, 40].
Barriers and facilitators for TB testing or treatment decisions from each review were coded first, and then grouped under common themes associated with decision-making uncertainty, based on the taxonomy developed by Eachempati et al. [20]. The taxonomy develops around macro (society and community), meso (group relationships), and micro (individual) levels of uncertainty to emphasize the dynamic interplay between different sources and types of uncertainty at each level, and may be useful to classify challenges in health care decision-making [20].
Recurring themes were further classified based on an adapted version of the WHO conceptual framework representing the TB diagnosis and care continuum [41]. The framework helped to identify four levels (patient, provider, health system, diagnostic test) of factors influencing TB clinical decision-making, including three time-points (patient-provider encounter, diagnosis, treatment initiation) for decision-making. The framework captures both the determinants (i.e., what causes decisional uncertainty) and broader sources (i.e., what contributes to the variability of decisional outcomes) of uncertainty in the decision-making process.
Definitions
Presumptive pulmonary TB was defined as clinical/pre-test suspicion or post-test suspicion despite a negative test. Diagnostic delay was defined as the time lag from first access to health system/consultation with provider to diagnosis; treatment delay was defined as the time lag from diagnosis to treatment initiation. Provider/health system delay was used to refer to any diagnostic or treatment delay attributable to provider or health systems factors (to differentiate from causes of delay attributable to patient factors).
Results
Review characteristics
Database searches yielded 8542 records. After duplicate removal, 7345 unique records were screened by title/abstract. After full text screening of 110 records, a total of 27 reviews were included (Table 1). The PRISMA flow chart detailing the phases of study selection is presented in Fig 1.
Articles were published between 2008 and 2023. Records included nine meta-analyses, three qualitative evidence syntheses, and 15 mixed-methods narrative synthesis (Table 1). Primary studies included in the reviews spanned from 1970 to 2021 and were mostly observational (Table 1).
Reviews varied in inclusion criteria, outcomes, settings, and population. Based on their primary aim, reviews were classified into four main categories: diagnostic and treatment delays (n = 8); knowledge, attitudes, and practices of TB healthcare providers and end users (n = 5); barriers and facilitators to utilization of TB diagnostic services (n = 10); diagnostic test impact on diagnosis and treatment (n = 4). Most reviews included primary studies with adult populations or did not include sub-group analysis by age. One review focused specifically on children and adolescents [25]. Key population and outcome definitions were generally consistent. Prior to the review, standardized definitions were developed allowing for direct comparison and a narrative synthesis of findings (Table 1).
Most reviews were of fair or good methodological quality. The main areas compromising methodological quality and confidence in findings were publication bias, not using consistent methods to minimize errors in data extraction, and not grading the quality of evidence (S2 Table). Based on the global CCA index, most reviews had very low to no overlap. Seven pairs had high or very high primary source overlap. The citations list from one review [42] was not available, hence not included in CCA calculations (S1 Fig).
Main findings from thematic content analysis
Through iterative thematic analysis, 15 recurring themes were identified. Applying an integrated multilevel model of uncertainty in health care [20], the themes were classified by type of uncertainty (Table 2). Findings were then classified as barriers or facilitators for testing and treatment decisions (Fig 2).
Synthesis enabled the development of a framework for uncertainty in TB decision-making, presented in Fig 3. Types of uncertainty were grouped in four macro-levels, corresponding to sources of uncertainty in TB clinical decision-making: patient-, provider-, diagnostics-, and health system-related uncertainty. The framework represents the relationship between the four sources of uncertainty and three key moments in clinical decision-making: the clinical encounter, the formulation of a diagnostic hypothesis, and the treatment initiation. Different types of uncertainty may act synergistically at given time-points (Fig 3).
Clinical uncertainty
Clinical presentation.
Half of the reviews mentioned the relationship between clinical presentation and clinicians’ suspicion of TB. Meta-analysis data from GeneXpert and urine lipoarabinomannan (LAM) diagnostic impact studies suggested a higher likelihood of being treated empirically for sicker patients requiring hospitalization [43, 44]. In the review by Getnet et al., findings from observational studies indicated that the absence of cough and the presence of atypical symptoms, fever, or good clinical conditions were associated with provider diagnostic delay [45]. Similarly, patients presenting with chronic cough and other concomitant lung disease including COVID-19 were reported to experience delays [24, 46].
Socio-demographic characteristics.
Three reviews identified several indicators of patient socio-economic status, including poor literacy, low income or unemployment, lack of health insurance, and rural residence, as factors associated with diagnostic delay [45, 47, 48]. The reviews by Yang et al. and Krishnan et al., which focused on gender-related differences in access to TB services and had moderate overlap, reported inconsistent evidence of a positive relationship between female sex and provider delay in TB diagnosis [47, 49]. Yang et al. additionally reported differences by setting. For example, providers from Thailand and Vietnam were more likely to adhere to diagnostic guidelines with male patients, whereas providers from India offered testing with similar frequency to both women and men [49]. The meta-analysis by Getnet et al. showed no evidence of a difference in the proportion of male versus female patients diagnosed with TB at the 30-day mark (pooled odds ratio (OR) = 1.08, 95% CI 0.95–1.23) [45]. Similarly, Li et al. reported no evidence of an association between female sex for patients and diagnostic delay in China (pooled OR = 1, 95%CI 0.83–1.22) [48].
TB-related risk factors.
Treatment was often delayed in patients with a previous diagnosis of TB [50] and in patients who reported antibiotic usage prior to the clinical encounter [51]. Among paediatric patients, providers were less likely to start empiric treatment in cases with unknown TB exposure [25].
Collection of diagnostic specimens.
Reviews reported that the inability of patients to produce sputum influenced decisions to initiate testing and treatment [52]. In Nathavitharana et al., the proportion of adults able to provide a sputum sample ranged between 57% and 97% in people living with HIV (PLHIV), depending on setting and severity/type of symptoms. In contrast, urine collection (for the LAM assay) was achieved in 99% of PLHIV aged 15 and above across three RCTs [43]. Challenges with specimen collection influenced decisions to withhold microbiological testing and either initiate empiric treatment or to exclude TB solely based on the clinical interview or radiological examination findings [52, 53]. Engel et al., found with high confidence in evidence that providers highly valued the possibility of using alternative samples for testing such as urine or stool, particularly for paucibacillary cases and paediatric TB [52].
Personal uncertainty
Provider attitudes, beliefs, and stigma.
Multiple reviews found that provider behaviour and discriminatory attitudes can impact TB diagnosis and treatment initiation [25, 39, 49, 52–57]. In a qualitative evidence synthesis, Barnabishvili et al. reported that providers were less rigorous when interviewing older patients or foreigners during the clinical encounter [39]. Provider discrimination towards female patients, resulting in tests underutilization and delays, emerged from narrative syntheses [39, 45, 48, 51]. Provider TB/HIV coinfection-related stigma was reported in three reviews as a factor delaying diagnosis or treatment initiation [52, 53, 55], including one review with high confidence in evidence [52].
Fear of infection.
Two reviews based on qualitative data, including one review with high confidence in evidence [52], reported that providers were generally aware of the aerosol biohazard and hesitant to test for TB because of fear of acquiring the disease [51, 52]. Fear of infection from respiratory specimen collection, particularly gastric aspiration, resulted in underutilization of diagnostic tools [52] or collection of poor quality respiratory specimens [51, 52]. In the context of the SARS-CoV-2 pandemic, some providers refused to collect respiratory specimens among presumptive TB patients presenting with COVID-19 symptoms [46].
Test characteristics and provider preference.
Diagnostic accuracy, automation, and computer-based tests were highly valued by providers based on moderate confidence in evidence [52]. Among paediatric patients, difficulties in collecting respiratory specimens (e.g., induced sputum or gastric aspirate), invasiveness of the procedure, and the lack of adequately trained staff were reported as barriers to test utilization [52, 58].
Relational and knowledge exchange uncertainty
Patient-provider communication dynamics.
Some reviews reported that provider miscommunication with patients was a potential cause of missed diagnoses [39, 42, 51–54]. The difficulty in communicating with the patient was often reported as the consequence of TB-related stigma, but it also arose from the use of metaphors for clinical explanations, resulting in patients not understanding diagnostic and therapeutic plans, and losses to follow up [53, 54]. One review reported that male providers disclosed difficulties communicating with, and understanding health concerns from, female patients during consultation [39].
Epistemic uncertainty
Provider knowledge and qualification.
Qualitative findings from twelve reviews suggested that suboptimal TB knowledge impacted providers’ ability to prescribe diagnostic tests or caused providers to delay TB diagnosis and miss treatment opportunities [24, 25, 48, 50, 52, 54, 56, 57, 59–62]. In a review on practices and knowledge of Indian providers, Satyanarayana et al. reported that the proportion of providers that suspected TB in the presence of a persistent cough of more than 2–3 weeks duration ranged from 21% to 81%, and less than 60% of patients with persistent cough were advised to undergo sputum examination [59].
A review by Teo et al. reported that poor clinical standards and low levels of knowledge of TB among providers led to delays in TB diagnosis in 12 qualitative studies, with high confidence in evidence [50]. Braham et al. reported one primary study where less than 50% of providers were aware of the principal diagnostic tools needed for TB diagnosis [61]. Poor TB knowledge and clinical skills resulted in deferral of bacteriological testing and preference for smear microscopy over mWRD, according to the narrative review by Shah et al.[57]. Additionally, the same review reported that providers’ unawareness and non-adherence to diagnostic algorithms was a reason for missed diagnoses [57].
Health care workers with particularly low levels of knowledge included informal providers [62], public providers working at the primary level, private practitioners with limited awareness of TB, and traditional healers [23, 24]. One review found that recognition of TB symptoms was associated with providers’ level of qualification and public sector employment [42]. In contrast, age, sex, years of practice, experience, and level of qualification were not associated with identification of TB symptoms [42, 54].
The meta-analysis by Amare et al. of nine intervention trials demonstrated that training interventions improve the ability of providers to diagnose TB, significantly increasing the number of bacteriologically confirmed cases [60].
Availability of policies and guidelines.
Lack of clear and updated guidelines and poor dissemination at primary healthcare levels and among private providers led to poor referral to GeneXpert testing, or inconsistency in the types of samples used [52, 57, 62]. The review by Shah et al. reported guidelines and policies variability in the private sector as one cause of missed diagnoses [57].
Test uncertainty
Utilization and impact of diagnostic tools.
Engel et al. found that in settings where low-complexity mWRDs were easily accessible, providers reported a high level of trust in the test result [52]. The meta-analysis by Lee et al reported that availability of mWRDs reduced diagnostic and treatment delays [63]. Three meta-analyses examined GeneXpert diagnostic impact [44, 58, 64], with outcomes reported only from the most recent review [44]. The use of GeneXpert (versus smear microscopy) had no effect on the proportion of participants treated for TB (risk ratio 1.10, 95% CI 0.98–1.23; GRADE: moderate confidence) [44]. This could reflect decisions to treat some patients empirically regardless of test results. The lower sensitivity of GeneXpert in paucibacillary forms of the disease, such as paediatric TB, was recognized as a limitation that would justify empiric treatment initiation [58].
Health-system uncertainty
Operational setting deficiencies.
Twelve reviews reported on challenges at the health systems level. Inadequate staff trainings, lack of diagnostic resources, lack of personal protective equipment and infection prevention control measures, and absence of private rooms for clinical assessment were mentioned as potential contributors to missed diagnosis and treatment opportunities [25, 42, 46, 51, 55–57]. Private and rural clinics not offering TB services were associated with diagnostic delays compared with public, urban facilities, where providers had better access to tests and infrastructure [39, 50, 53, 61].
Availability and timing of test results.
Sullivan et al. reported missed treatment opportunities in children due to long waiting times for culture results [25]. Reviews found that rapid test turnaround time was important to accelerate therapeutic decisions [25, 52, 56], and that offering same-day test and treat would reduce gaps in missed treatment according to providers [52].
Diagnostic test availability, accessibility, and affordability.
Reviews reported that the limited availability of resources for microbiological diagnosis (e.g., due to stock-outs, power cuts, and unreliable supply chains) was associated with GeneXpert underutilization and diagnostic delays [48, 51, 52, 57]. Engel et al. reported on providers’ perspectives regarding the impact of diagnostic accessibility and affordability on test and treatment decisions. Frequent stock-outs were reported to potentially hinder providers’ faith in the adoption of new diagnostics and hamper their reliance on prescribing diagnostic tests in the future [52]. Further, some providers disclosed a preference for initiating treatment if patients incurred excessive costs for testing, regardless of test availability [52].
Discussion
This umbrella review showed the complexity of multi-level factors that contribute to uncertainty in TB clinical decision-making, often resulting in under-utilization of diagnostic resources, misdiagnoses, empirical treatment or missed treatment opportunities, and diagnostic and treatment delays. The results of this study reinforce the concept that clinical decision-making is highly dependent on individual and interpersonal factors (provider, patient), but also closely linked to the operational context and the usability of diagnostic resources. These findings are important to inform the development of successful diagnostic aids and programs implementation strategies, and to improve TB practices in high-burden, resource-limited settings.
An important output from this study was the consolidation of a framework to present multilevel factors associated with uncertainty in TB decision-making. We found that several factors related to the local context and often beyond providers’ control were responsible for the discrepancy between TB testing and treatment decisions and scientific guidelines’ recommendations. Most of the existing literature on TB diagnostics includes diagnostic accuracy studies or randomized controlled trials that do not examine the challenges of clinical decision-making and the impact of health systems factors on diagnostic interventions. Rapid molecular diagnostics such as GeneXpert have had a great influence on TB care but there are ongoing concerns about underutilization and sustainability that need to be addressed [6]. Unfortunately, diagnostic tests, despite being cheap, fast, and accurate, are not always used as recommended–or not used at all–in high-burden settings, and it is crucial to increase our understanding of the underlying reasons [8, 65].
Reviews reported consistent evidence for patient characteristics, symptom variability and severity as primary sources of clinical uncertainty in TB decision-making [24, 47, 54, 59, 61]. When confronted with hospitalized patients, patients with advanced HIV disease, or paediatric patients, providers seemed more inclined to treat empirically, regardless of the availability of diagnostic aids, possibly also because of the complexity of obtaining clinical specimens from people in these categories [25, 43]. Additionally, history of previous TB diagnosis was associated with retreatment delays [50], potentially due to lack of confidence in diagnosis, or fear of drug side effects with injectables [66]. Further research is needed to uncover provider-related factors associated with retreatment decision-making, as rapid tests for second-line drug resistance testing and all-oral regimens become available [67, 68].
Providers’ limited knowledge of TB symptoms and approaches for clinical and diagnostic management, and insufficient familiarity with guidelines, were reported consistently as key contributors to delay in test and treatment decisions [24, 25, 48, 50, 52, 54, 56, 57, 59–62]. Epistemic uncertainty affected several aspects of the decision-making, including estimating pre-test disease probabilities, deciding to use a diagnostic test, selecting appropriate specimens based on age and disease localization, collecting good quality samples, and interpreting test results [50, 59, 61]. Conversely, the availability of highly qualified physicians, public sector facilities, and ease of access to mWRD had a positive influence on testing decisions [42, 52]. Notably, training interventions significantly improved case detection and test uptake by providers [60].
The central role of the provider in the decision-making process was also supported by extensive evidence on how interpersonal attitudes, beliefs, stigma, fear of infection, and test preferences affected test utilization and treatment decisions [25, 39, 49, 52–57]. Personal sources of uncertainty, including fear of acquiring TB through respiratory sampling, were commonly reported barriers for underutilization of diagnostics [51, 52]. As seen with other respiratory infectious diseases, fear of infection was mostly associated with poor knowledge of biohazard mitigation strategies, ambiguous guidelines, and lack of resources [69]. These findings support the importance of enhancing comprehensive national training and educational programs for providers at all levels of care, and engaging the private sector [61, 70]. Similarly, the fear of acquiring TB could be, at least partially, addressed through continuous training, and implementation of infection prevention control measures [71].
The high variability of provider-patient interactions during the clinical encounter was often reported as a source of relational uncertainty affecting the outcomes of the clinical decision-making process [39, 42, 53]. Provider personal biases could result in the inability or unwillingness to collect all necessary clinical information, diagnostic test under-utilization, misdiagnosis and diagnostic delays, especially when confronted with female patients [48, 49, 51, 61]. Although findings from meta-analyses did not confirm the association between female sex and diagnostic delays, moderate-quality qualitative sources reported the impact of gender on clinical decision-making [48, 49, 51]. Gender-related disparities in TB are well-known, especially with regards to health seeking behaviours and retention in care [51]. While TB incidence is greater in men [72], women generally face additional barriers related to care access, stigma and psychosocial consequences of the diagnosis [51]. The findings from this study confirm the importance of a gender-based approach to TB as advocated by WHO [73]. At the same time, quantitative and qualitative studies across settings and countries with different gender norms are needed to gain further insight on gaps in the TB diagnostic cascade, gender inequalities and discrimination, to inform TB interventions that have the capacity to overcome gender barriers [74].
Providers had high confidence in rapid diagnostic tests, but the confidence in mWRDs, namely GeneXpert, appeared to be generated by trust in a computer-based test, rather than from understanding of the technology and knowledge about its diagnostic accuracy [43, 52]. It should be noted that, paradoxically, a blind use of diagnostics could represent a double-edged sword, if overconfidence in results became a substitute for clinical reasoning [75]. The burden of misdiagnosis was also supported by findings from a large autopsy study, demonstrating a high prevalence of TB among children and PLHIV that were missed at clinical diagnosis [76]. Evaluating the impact of testing on clinical decisions and empiric treatment [77, 78] will be important as missing false negative patients contributes to TB morbidity and mortality, particularly among people who cannot expectorate or who have paucibacillary disease such as young children, where currently available assays have lower sensitivities [79–81]
Health system uncertainty emerged as an important driver of variability in TB decision-making. The unavailability or inaccessibility of diagnostic resources contributed to uncertainty in the decisional process and outcomes [25, 52, 55, 56].
When diagnostic tests were available, several contextual factors, such as poor infrastructure and lack of administrative resources (infection prevention control policies, insufficient trainings), represented barriers to test adoption, shifting the decisional bar towards empiric treatment initiation, particularly in children or very sick patients, or leading to missed treatment opportunities [25, 52]. The absence of locally tailored guidelines was reported to contribute to epistemic uncertainty and variability in clinical management [52, 62]. These findings confirm that resource allocation strategies, as well as trainings and guidelines, need to be more inclusive of the lower tiers of the health system [82].
This study also found that providers highly valued the possibility to use non-sputum samples for testing, such as urine or stool [52], highlighting the need for a rapid addition of sputum-free diagnostics, particularly for paucibacillary cases and paediatric TB [83].
In recent years, there has been unprecedented development of novel TB diagnostic technologies. As new products come to market, policy makers must decide which available tools to implement. Findings from this review support the idea that such decisions should not exclusively account for diagnostic assay characteristics (e.g., accuracy), but also consider acceptability and feasibility of tests within the health care infrastructure. As suggested by meta-analyses reporting inconclusive findings regarding the impact of GeneXpert on treatment initiation decisions [44, 64], it is key to understand the real-world impact of diagnostics through robust operational research at the point-of-care.
Additionally, the increasing utilization of multiple tests or different specimens in parallel, may exacerbate the challenges of results interpretation, particularly in children [84]. Understanding how clinicians manage conflicting results will be important to inform clinical algorithms.
Recently, significant progress has been made in the development and validation of clinical prediction models and algorithms to help standardize the decision-making process, particularly in contexts not yet reached by new diagnostic tools [85]. However, such tools rely on the assumption that a clinical consultation is a standardized event where relevant clinical variables or risk factors would always be disclosed and inform disease probability. Nonetheless, as suggested by the findings of this review, a clinical encounters is an event influenced by multiple uncertainties [39, 53, 55, 56]. Hence, it will be important to collect data on real-life performance of such prediction models and algorithms, and to consider setting-specific adjustments and the integration of variables beyond patient clinical and risk factors. At the same time, the complex roots of uncertainty call for integrated efforts by policy makers, researchers, and programs to combine diagnostics research and implementation with staff trainings, guidelines implementation and uptake, infrastructure development, transversal health education to combat stigma and discrimination, and investments at the most peripheral levels of health care systems globally.
Strengths and limitations
To the best of our knowledge, this is the first study to conceptualize and summarize sources and types of uncertainty in TB decision-making. The umbrella review approach allowed us to triangulate findings from varied study designs and outcomes while preserving high methodological standards. The review was conducted in a systematic manner in accordance with standardized guidance. Nonetheless, some limitations must be mentioned. First, a limitation of the umbrella review approach is our inability to conduct a detailed assessment of primary studies. Consequently, the study relied on the methods and quality of included SRs, many of which were of moderate quality. Most reviews used a narrative synthesis approach, and only a few meta-analyses and one qualitative evidence synthesis reported on the quality of the evidence. Second, it was not possible to perform a meta-analysis of quantitative review findings due to the heterogeneous inclusion criteria and outcome definitions. Third, it is possible that some relevant sources were missed, as grey literature was not included. Finally, the assessment of each record was performed by a single reviewer only, which may yield a lower sensitivity.
Conclusion
This study summarized the complex network of factors associated with decisional and outcome uncertainty in medical decision-making in TB through a synthesis and thematic analysis of the systematic review literature. Different sources of uncertainty were found to influence provider choices around testing and treatment initiation, often resulting in diagnostic and treatment delays or missed diagnoses and treatment opportunities. Further, the application of a multi-level framework to classify uncertainty revealed the extent to which findings pertaining to different sources and types of uncertainty were intertwined. Gaps in TB diagnosis and treatment suggest the need to integrate evidence from studies that consider variations in healthcare systems and end-users’ attitudes, preferences, and experiences with interventions introducing new diagnostic tools. Such considerations are important to improve TB diagnosis and treatment and quality of patient care and to allow impactful introduction of novel diagnostic aids in clinical practice worldwide.
The figure summarizes multi-level (patient, provider, health systems, diagnostic tests) factors associated with TB clinical decision-making, identified through thematic content analysis of the SRs. The factors were classified as barriers or facilitators for testing or treatment decisions, and represented using the threshold model [10]. Several facilitators positively influenced providers’ decisions to test (lower testing threshold), including the presence of typical symptoms and patient history, providers’ personal attributes and experiences, workplace (public/urban facility), and available test characteristics. Barriers to testing were the presence of confounding/atypical symptoms, inadequate TB knowledge and staff training, fear of infection, lack of resources, and challenges of respiratory specimen collection. Empiric treatment decisions (treatment threshold) were facilitated by the presence of factors generally associated with an increased risk of severe disease or negative outcomes (young age, severe symptoms), unavailability or inaccessibility (e.g., because of costs) to diagnostic tests, and lack of confidence in tests with low sensitivity. Providers were inclined to withhold treatment decisions if facing with certain elements of patient history (e.g., unknown TB exposure), waiting for test results, and in the presence of negative test results (without considering the possibility of a low negative predictive value).
During the clinical encounter, the provider assesses the patient’s clinical variables (clinical uncertainty) to determine the disease probability and evaluate therapeutic benefit-harm trade-offs. Disease probability estimates depend on the provider’s knowledge and experience (epistemic uncertainty). Provider’s ability to conduct an informative, high-quality clinical assessment is also influenced by patient-provider relation and communication strategies (relational and knowledge exchange uncertainty) as well as by provider’s attitudes and beliefs (personal uncertainty). When a decision is made to test, the probability of disease is adjusted based on diagnostic test results (post-test probability). However, a negative test result may be insufficient to withhold therapy, considering the low sensitivity of currently available diagnostic tests and the potential benefit of empiric treatment (test uncertainty). Additionally, the provider may decide not to proceed with invasive specimen collection and testing because of individual risk assessments such as fear of infection (personal uncertainty). Thus, the characteristics of diagnostic tests can impact decision-making. Clinical decisions are further limited by healthcare setting constraints such as lack of skilled staff, poor infrastructure, and scarcity of diagnostic tools (health system uncertainty), and absence of local guidelines (epistemic uncertainty).
Supporting information
S1 Appendix. Search strategy (July 2023).
The search strategy was refined and tested in the MEDLINE database, and then adapted to the other databases. To restrict the search to capture SRs, while simultaneously minimizing the capture of non-SR publications, search terms and MeSH specific to SR study designs (e.g. ‘SR’, ‘qualitative evidence synthesis’) were included. A search of grey literature was not conducted. Details on the search strategy across all databases are presented in S1 Appendix.
https://doi.org/10.1371/journal.pgph.0003429.s001
(DOCX)
S1 Table. Eligibility criteria- PICOS framework.
SRs were selected on predefined inclusion and exclusion criteria guided by the Population, Intervention, Comparison, Outcome and Study design/setting (PICOS) framework. References were pre-emptively de-duplicated in Endnote. The selected references were imported in Covidence (https://www.covidence.org/home) and re-screened for duplicates. Full texts of all potentially eligible reviews were obtained. The full text of selected papers was then examined for inclusion in the UR, based on the predefined criteria. The reason for the exclusion of each article was documented in the software for transparency and auditing purposes.
https://doi.org/10.1371/journal.pgph.0003429.s002
(DOCX)
S2 Table. Critical appraisal of methodological quality of included reviews.
JBI Critical Appraisal Checklist for Systematic Reviews and Research Syntheses.
https://doi.org/10.1371/journal.pgph.0003429.s003
(DOCX)
S1 Fig. Heatmap showing pairwise calculation of the CCA.
CCA was interpreted in banded thresholds: values below 5% indicated slight citations overlap, between 6–10% indicated moderate overlap, between 11–15% indicated high overlap and values above 15% indicated very high overlap.
https://doi.org/10.1371/journal.pgph.0003429.s004
(DOCX)
S2 Checklist. Research checklist, PRISMA 2020 checklist.
https://doi.org/10.1371/journal.pgph.0003429.s006
(DOCX)
References
- 1.
World Health Organization. WHO Global Tuberculosis Report 2023. Geneva; 2023. Available from: https://www.who.int/publications/i/item/9789240083851
- 2. Kruk ME, Gage AD, Arsenault C, Jordan K, Leslie HH, Roder-DeWan S, et al. High-quality health systems in the Sustainable Development Goals era: time for a revolution. Lancet Glob Heal. 2018 Nov 1;6(11):e1196–252. pmid:30196093
- 3. Pai M, Temesgen Z. Quality: The missing ingredient in TB care and control. J Clin Tuberc Other Mycobact Dis. 2019 Feb 1;14:12. pmid:31720411
- 4.
World Health Organization. WHO consolidated guidelines on tuberculosis. Module 3: Diagnosis ‐ Rapid diagnostics for tuberculosis detection 2021 update [Internet]. 2021 [cited 2022 Feb 1]. Available from: https://www.who.int/publications/i/item/9789240029415
- 5. Cazabon D, Pande T, Kik S, Van Gemert W, Sohn H, Denkinger C, et al. Market penetration of Xpert MTB/RIF in high tuberculosis burden countries: A trend analysis from 2014–2016. Gates Open Res. 2018;2. pmid:30234198
- 6. Albert H, Nathavitharana RR, Isaacs C, Pai M, Denkinger CM, Boehme CC. Development, roll-out and impact of Xpert MTB/RIF for tuberculosis: what lessons have we learnt and how can we do better? Eur Respir J. 2016 Aug 1;48(2):516. pmid:27418550
- 7. McDowell A, Pai M. Treatment as diagnosis and diagnosis as treatment: empirical management of presumptive tuberculosis in India. Int J Tuberc Lung Dis. 2016 Apr 1;20(4):536–43. pmid:26970165
- 8. Raizada N, Khaparde SD, Swaminathan S, Sarin S, Salhotra VS, Kalra A, et al. Catalysing progressive uptake of newer diagnostics by health care providers through outreach and education in four major cities of India. PLoS One. 2018 Mar 1;13(3). pmid:29509803
- 9. Denkinger CM, Kik S V., Cirillo DM, Casenghi M, Shinnick T, Weyer K, et al. Defining the needs for next generation assays for tuberculosis. J Infect Dis. 2015;211 Suppl 2(Suppl 2):S29–38. pmid:25765104
- 10. Djulbegovic B, van den Ende J, Hamm RM, Mayrhofer T, Hozo I, Pauker SG. When is rational to order a diagnostic test, or prescribe treatment: the threshold model as an explanation of practice variation. Eur J Clin Invest. 2015 May 1;45(5):485–93. pmid:25675907
- 11. Eisenberg JM, Hershey JC. Derived Thresholds: Determining the Diagnostic Probabilities at Which Clinicians Initiate Testing and Treatment. Med Decis Mak. 1983 Jul 2;3(2):155–68.
- 12. Jensen SB, Rudolf F, Wejse C. Utility of a clinical scoring system in prioritizing TB investigations ‐ a systematic review. Expert Rev Anti Infect Ther. 2019 Jul 3;17(7):475–88. pmid:31159621
- 13. Gunasekera KS, Vonasek B, Oliwa J, Triasih R, Lancioni C, Graham SM, et al. Diagnostic Challenges in Childhood Pulmonary Tuberculosis-Optimizing the Clinical Approach. Pathog (Basel, Switzerland). 2022 Apr 1;11(4). pmid:35456057
- 14. Van Wyk SS, Lin HH, Claassens MM. A systematic review of prediction models for prevalent pulmonary tuberculosis in adults. Int J Tuberc Lung Dis. 2017 Apr 1;21(4):405–11. pmid:28284255
- 15. Marcy O, Borand L, Ung V, Msellati P, Tejiokem M, Huu KT, et al. A Treatment-Decision Score for HIV-Infected Children With Suspected Tuberculosis. Pediatrics. 2019;144(3). pmid:31455612
- 16. Gunasekera KS, Walters E, Van Der Zalm MM, Palmer M, Warren JL, Hesseling AC, et al. Development of a Treatment-decision Algorithm for Human Immunodeficiency Virus-uninfected Children Evaluated for Pulmonary Tuberculosis. Clin Infect Dis. 2021 Aug 15;73(4):E904–12. pmid:33449999
- 17. Boyles TH, Lynen L, Seddon JA. Decision-making in the diagnosis of tuberculous meningitis. Wellcome open Res. 2020;5. pmid:32964134
- 18. Nadjib M, Dewi RK, Setiawan E, Miko TY, Putri S, Hadisoemarto PF, et al. Cost and affordability of scaling up tuberculosis diagnosis using Xpert MTB/RIF testing in West Java, Indonesia. Quinn F, editor. PLoS One. 2022 Mar 10;17(3):e0264912. pmid:35271642
- 19. Helou MA, DiazGranados D, Ryan MS, Cyrus JW. Uncertainty in Decision Making in Medicine: A Scoping Review and Thematic Analysis of Conceptual Models. Acad Med. 2020;95(1):157–65. pmid:31348062
- 20. Eachempati P, Büchter RB, Ks KK, Hanks S, Martin J, Nasser M. Developing an integrated multilevel model of uncertainty in health care: a qualitative systematic review and thematic synthesis. BMJ Glob Heal. 2022 May 1;7(5):e008113.
- 21. Lin HH, Langley I, Mwenda R, Doulla B, Egwaga S, Millington KA, et al. A modelling framework to support the selection and implementation of new tuberculosis diagnostic tools. Int J Tuberc Lung Dis. 2011 Aug;15(8):996–1004. pmid:21740663
- 22.
WHO global lists of high burden countries for tuberculosis (TB), TB/HIV and multidrug/rifampicin-resistant TB (MDR/RR-TB), 2021–2025 Background document. 2021; Available from: https://iris.who.int/handle/10665/341980
- 23. Sreeramareddy CT, Qin ZZ, Satyanarayana S, Subbaraman R, Pai M. Delays in diagnosis and treatment of pulmonary tuberculosis in India: a systematic review. Int J Tuberc Lung Dis. 2014 Mar 1;18(3):255–66. pmid:24670558
- 24. Storla DG, Yimer S, Bjune GA. A systematic review of delay in the diagnosis and treatment of tuberculosis. BMC Public Health. 2008;8. pmid:18194573
- 25. Sullivan BJ, Esmaili BE, Cunningham CK. Barriers to initiating tuberculosis treatment in sub-Saharan Africa: a systematic review focused on children and youth. Glob Health Action. 2017;10(1). pmid:28598771
- 26. Papatheodorou S. Umbrella reviews: what they are and why we need them. Eur J Epidemiol. 2019;34(6).
- 27. Biondi-Zoccai G. Umbrella reviews: Evidence synthesis with overviews of reviews and meta-epidemiologic studies. Umbrella Rev Evid Synth with Overviews Rev Meta-Epidemiologic Stud. 2016 Jan 1;1–391.
- 28. Aromataris E, Fernandez R, Godfrey CM, Holly C, Khalil H, Tungpunkom P. Summarizing systematic reviews: methodological development, conduct and reporting of an umbrella review approach. Int J Evid Based Healthc. 2015 Sep 1;13(3):132–40. pmid:26360830
- 29. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021 Mar 29;372.
- 30.
Methodology for JBI Umbrella Reviews. 2014; Available from: https://ro.uow.edu.au/cgi/viewcontent.cgi?article=4367&context=smhpapers
- 31.
Chapter V: Overviews of Reviews | Cochrane Training [Internet]. [cited 2022 Jul 7]. Available from: https://training.cochrane.org/handbook/current/chapter-v
- 32. Basile FW; Sweeney S; Bijker E; Singh M; Vassall A; Indravudh P. Medical decision-making in the diagnosis and treatment initiation of tuberculosis: protocol for an umbrella review with systematic methods [Internet]. Open Science Framework. 2023. Available from: https://doi.org/10.17605/OSF.IO/KVMHN
- 33. Methley AM, Campbell S, Chew-Graham C, McNally R, Cheraghi-Sohi S. PICO, PICOS and SPIDER: a comparison study of specificity and sensitivity in three search tools for qualitative systematic reviews. BMC Health Serv Res. 2014;14(1). pmid:25413154
- 34. Gates M, Gates A, Duarte G, Cary M, Becker M, Prediger B, et al. Quality and risk of bias appraisals of systematic reviews are inconsistent across reviewers and centers. J Clin Epidemiol. 2020 Sep 1;125:9–15. pmid:32416337
- 35. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008 Apr 26;336(7650):924–6. pmid:18436948
- 36. Lewin S, Booth A, Glenton C, Munthe-Kaas H, Rashidian A, Wainwright M, et al. Applying GRADE-CERQual to qualitative evidence synthesis findings: Introduction to the series. Implement Sci. 2018 Jan 25;13(1):1–10.
- 37. Hennessy EA, Johnson BT. Examining Overlap of Included Studies in Meta-Reviews: Guidance for using the Corrected Covered Area Index. Res Synth Methods. 2020 Jan 1;11(1):134. pmid:31823513
- 38. Hennessy EA, Johnson BT, Keenan C. Best Practice Guidelines and Essential Methodological Steps to Conduct Rigorous and Systematic Meta-Reviews. Appl Psychol Health Well Being. 2019 Nov 1;11(3):353–81. pmid:31290288
- 39. Barnabishvili M, Ulrichs T, Waldherr R. Role of acceptability barriers in delayed diagnosis of Tuberculosis: Literature review from high burden countries. Acta Trop. 2016 Sep 1;161:106–13. pmid:27311390
- 40. Sandelowski M, Voils CI, Barroso J. Defining and Designing Mixed Research Synthesis Studies. Res Sch. 2006;13(1):29. pmid:20098638
- 41. MacPherson P, Houben RMGJ, Glynn JR, Corbett EL, Kranzer K. Pre-treatment loss to follow-up in tuberculosis patients in low- and lower-middle-income countries and high-burden countries: a systematic review and meta-analysis. Bull World Health Organ. 2014 Feb;92(2):126–38. pmid:24623906
- 42. Bello S, Afolabi RF, Ajayi DT, Sharma T, Owoeye DO, Oduyoye O, et al. Empirical evidence of delays in diagnosis and treatment of pulmonary tuberculosis: systematic review and meta-regression analysis. BMC Public Health. 2019 Jun 25;19(1). pmid:31238906
- 43. Nathavitharana RR, Lederer P, Chaplin M, Bjerrum S, Steingart KR, Shah M. Impact of diagnostic strategies for tuberculosis using lateral flow urine lipoarabinomannan assay in people living with HIV. Cochrane database Syst Rev. 2021 Aug 20;8(8). pmid:34416013
- 44. Haraka F, Kakolwa M, Schumacher SG, Nathavitharana RR, Denkinger CM, Gagneux S, et al. Impact of the diagnostic test Xpert MTB/RIF on patient outcomes for tuberculosis. Cochrane database Syst Rev. 2021 May 6;5(5). pmid:34097769
- 45. Getnet F, Demissie M, Assefa N, Mengistie B, Worku A. Delay in diagnosis of pulmonary tuberculosis in low-and middle-income settings: systematic review and meta-analysis. BMC Pulm Med. 2017 Dec 13;17(1). pmid:29237451
- 46. Dlangalala T, Musekiwa A, Brits A, Maluleke K, Jaya ZN, Kgarosi K, et al. Evidence of TB Services at Primary Healthcare Level during COVID-19: A Scoping Review. Diagnostics (Basel, Switzerland). 2021 Dec 1;11(12). pmid:34943458
- 47. Cai J, Wang X, Ma A, Wang Q, Han X, Li Y. Factors associated with patient and provider delays for tuberculosis diagnosis and treatment in Asia: a systematic review and meta-analysis. PLoS One. 2015 Mar 25;10(3). pmid:25807385
- 48. Li Y, Ehiri J, Tang S, Li D, Bian Y, Lin H, et al. Factors associated with patient, and diagnostic delays in Chinese TB patients: a systematic review and meta-analysis. BMC Med. 2013 Jun 2;11(1). pmid:23819847
- 49. Yang W-T, Gounder CR, Akande T, De Neve J-W, McIntire KN, Chandrasekhar A, et al. Barriers and delays in tuberculosis diagnosis and treatment services: does gender matter? Tuberc Res Treat. 2014;2014:1–15. pmid:24876956
- 50. Teo AKJ, Singh SR, Prem K, Hsu LY, Yi S. Duration and determinants of delayed tuberculosis diagnosis and treatment in high-burden countries: a mixed-methods systematic review and meta-analysis. Respir Res. 2021 Dec 1;22(1). pmid:34556113
- 51. Krishnan L, Akande T, Shankar A V., McIntire KN, Gounder CR, Gupta A, et al. Gender-related barriers and delays in accessing tuberculosis diagnostic and treatment services: a systematic review of qualitative studies. Tuberc Res Treat. 2014;2014:1–14. pmid:24900921
- 52. Engel N, Ochodo EA, Karanja PW, Schmidt BM, Janssen R, Steingart KR, et al. Rapid molecular tests for tuberculosis and tuberculosis drug resistance: a qualitative evidence synthesis of recipient and provider views. Cochrane database Syst Rev. 2022 Apr 26;4(4). pmid:35470432
- 53. Bhatnagar H. User-experience and patient satisfaction with quality of tuberculosis care in India: A mixed-methods literature review. J Clin Tuberc Other Mycobact Dis. 2019 Dec 1;17:100127. pmid:31788569
- 54. Bell CA, Duncan G, Saini B. Knowledge, attitudes and practices of private sector providers of tuberculosis care: a scoping review. Int J Tuberc Lung Dis. 2011 Aug;15(8):1005–17. pmid:21669027
- 55. Oga-Omenka C, Wakdet L, Menzies D, Zarowsky C. A qualitative meta-synthesis of facilitators and barriers to tuberculosis diagnosis and treatment in Nigeria. BMC Public Health. 2021 Dec 1;21(1). pmid:33535990
- 56. Yasobant S, Bhavsar P, Kalpana P, Memon F, Trivedi P, Saxena D. Contributing Factors in the Tuberculosis Care Cascade in India: A Systematic Literature Review. Risk Manag Healthc Policy. 2021;14:3275–86. pmid:34408513
- 57. Shah HD, Nazli Khatib M, Syed ZQ, Gaidhane AM, Yasobant S, Narkhede K, et al. Gaps and Interventions across the Diagnostic Care Cascade of TB Patients at the Level of Patient, Community and Health System: A Qualitative Review of the Literature. Trop Med Infect Dis. 2022 Jul 1;7(7). pmid:35878147
- 58. Agizew T, Boyd R, Auld AF, Payton L, Pals SL, Lekone P, et al. Treatment outcomes, diagnostic and therapeutic impact: Xpert vs. smear. A systematic review and meta-analysis. Int J Tuberc Lung Dis. 2019 Jan 1;23(1):82–92.
- 59. Satyanarayana S, Subbaraman R, Shete P, Gore G, Das J, Cattamanchi A, et al. Quality of tuberculosis care in India: a systematic review. Int J Tuberc Lung Dis. 2015 Jul 1;19(7):751–63. pmid:26056098
- 60. Amare D, Getahun FA, Mengesha EW, Dessie G, Shiferaw MB, Dires TA, et al. Effectiveness of healthcare workers and volunteers training on improving tuberculosis case detection: A systematic review and meta-analysis. PLoS One. 2023 Mar 1;18(3). pmid:36952429
- 61. Braham CA, White PJ, Arinaminpathy N. Management of tuberculosis by healthcare practitioners in Pakistan: A systematic review. PLoS One. 2018 Jun 1;13(6). pmid:29928031
- 62. Thapa P, Jayasuriya R, Hall JJ, Beek K, Mukherjee P, Gudi N, et al. Role of informal healthcare providers in tuberculosis care in low- and middle-income countries: A systematic scoping review. PLoS One. 2021 Sep 1;16(9). pmid:34473752
- 63. Lee JH, Garg T, Lee J, McGrath S, Rosman L, Schumacher SG, et al. Impact of molecular diagnostic tests on diagnostic and treatment delays in tuberculosis: a systematic review and meta-analysis. BMC Infect Dis. 2022 Dec 1;22(1). pmid:36517736
- 64. Di Tanna GL, Raza Khaki A, Theron G, McCarthy K, Cox H, Mupfumi L, et al. Effect of Xpert MTB/RIF on clinical outcomes in routine care settings: individual patient data meta-analysis. Lancet Glob Heal. 2019 Feb 1;7(2):e191–9. pmid:30683238
- 65. Nalugwa T, Shete PB, Nantale M, Farr K, Ojok C, Ochom E, et al. Challenges with scale-up of GeneXpert MTB/RIF in Uganda: A health systems perspective. BMC Health Serv Res. 2020 Mar 4;20(1):1–7.
- 66. Paul D, Busireddy A, Nagaraja SB, Satyanarayana S, Dewan PK, Nair SA, et al. Factors Associated with Delays in Treatment Initiation after Tuberculosis Diagnosis in Two Districts of India. PLoS One. 2012 Jul 9;7(7):e39040. pmid:22792161
- 67. Penn-Nicholson A, Georghiou SB, Ciobanu N, Kazi M, Bhalla M, David A, et al. Detection of isoniazid, fluoroquinolone, ethionamide, amikacin, kanamycin, and capreomycin resistance by the Xpert MTB/XDR assay: a cross-sectional multicentre diagnostic accuracy study. Lancet Infect Dis. 2022 Feb 1;22(2):242–9. pmid:34627496
- 68. Nyang’wa B-T, Berry C, Kazounis E, Motta I, Parpieva N, Tigay Z, et al. A 24-Week, All-Oral Regimen for Rifampin-Resistant Tuberculosis. N Engl J Med. 2022 Dec 22;387(25):2331–43. pmid:36546625
- 69. Houghton C, Meskell P, Delaney H, Smalle M, Glenton C, Booth A, et al. Barriers and facilitators to healthcare workers’ adherence with infection prevention and control (IPC) guidelines for respiratory infectious diseases: a rapid qualitative evidence synthesis. Cochrane database Syst Rev. 2020;4(4):1–55. pmid:32315451
- 70. Suseela RP, Shannawaz M. Engaging the Private Health Service Delivery Sector for TB Care in India—Miles to Go! Trop Med Infect Dis. 2023 May 1;8(5). pmid:37235313
- 71. Tan C, Kallon II, Colvin CJ, Grant AD. Barriers and facilitators of tuberculosis infection prevention and control in low- and middle-income countries from the perspective of healthcare workers: A systematic review. PLoS One. 2020 Oct 1;15(10). pmid:33085717
- 72. Horton KC, MacPherson P, Houben RMGJ, White RG, Corbett EL. Sex Differences in Tuberculosis Burden and Notifications in Low- and Middle-Income Countries: A Systematic Review and Meta-analysis. PLoS Med. 2016 Sep 1;13(9). pmid:27598345
- 73.
World Health Organization. Tuberculosis in women. 2018. Availble from: https://www.who.int/publications/m/item/tuberculosis-in-women
- 74.
StopTB Partnership. Gender and TB- investment package. Available from: https://stoptb.org/assets/documents/communities/TB%20Gender%20Investment%20Package.pdf
- 75. Leeflang MMG, Allerberger F. How to: evaluate a diagnostic test. Clin Microbiol Infect. 2019 Jan 1;25(1):54–9. pmid:29906592
- 76. Garcia-Basteiro AL, Hurtado JC, Castillo P, Fernandes F, Navarro M, Lovane L, et al. Unmasking the hidden tuberculosis mortality burden in a large post mortem study in Maputo Central Hospital, Mozambique. Eur Respir J. 2019 Sep 1;54(3).
- 77. Lucian Davis J, Masae Kawamura L, Chaisson LH, Grinsdale J, Benhammou J, Ho C, et al. Impact of GeneXpert MTB/RIF on patients and tuberculosis programs in a low-burden setting. a hypothetical trial. Am J Respir Crit Care Med. 2014 Jun 15;189(12):1551–9. pmid:24869625
- 78. Ling DI, Crépeau CA, Dufresne M, Khan S, Quach C, Dendukuri N, et al. Evaluation of the impact of interferon-gamma release assays on the management of childhood tuberculosis. Pediatr Infect Dis J. 2012 Dec;31(12):1258–62. pmid:22828646
- 79. Mesman AW, Rodriguez C, Ager E, Coit J, Trevisi L, Franke MF. Diagnostic accuracy of molecular detection of Mycobacterium tuberculosis in pediatric stool samples: A systematic review and meta-analysis. Tuberculosis. 2019 Dec 1;119:101878. pmid:31670064
- 80. Kay AW, González Fernández L, Takwoingi Y, Eisenhut M, Detjen AK, Steingart KR, et al. Xpert MTB/RIF and Xpert MTB/RIF Ultra assays for active tuberculosis and rifampicin resistance in children. Cochrane Database Syst Rev. 2020 Aug 27;2020(8).
- 81. Hodille E, Maisson A, Charlet L, Bauduin C, Genestet C, Fredenucci I, et al. Evaluation of Xpert MTB/RIF Ultra performance for pulmonary tuberculosis diagnosis on smear-negative respiratory samples in a French centre. Eur J Clin Microbiol Infect Dis. 2019 Mar 4;38(3):601–5. pmid:30680567
- 82. Yellapa V, Devadasan N, Krumeich A, Pai NP, Vadnais C, Pai M, et al. How patients navigate the diagnostic ecosystem in a fragmented health system: a qualitative study from India. https://doi.org/101080/1654971620171350452. 2017;10(1).
- 83. Nathavitharana RR, Garcia-Basteiro AL, Ruhwald M, Cobelens F, Theron G. Reimagining the status quo: How close are we to rapid sputum-free tuberculosis diagnostics for all? eBioMedicine. 2022 Apr 1;78. pmid:35339423
- 84. Orikiriza P, Smith J, Ssekyanzi B, Nyehangane D, Taremwa IM, Turyashemererwa E, et al. Tuberculosis diagnostic accuracy of stool Xpert MTB/RIF and urine AlereLAM in vulnerable children. Eur Respir J. 2022 Jan 1;59(1). pmid:34140291
- 85.
World Health Organization. WHO consolidated guidelines on tuberculosis Module 5: Management of tuberculosis in children and adolescents [Internet]. 2022 [cited 2024 May 27]. Available from: https://www.who.int/publications/i/item/9789240046764