Skip to main content

Time to diagnosis and treatment of pulmonary tuberculosis in indigenous peoples: a systematic review

Abstract

Background

Time to diagnosis and treatment is a major factor in determining the likelihood of tuberculosis (TB) transmission and is an important area of intervention to reduce the reservoir of TB infection and prevent disease and mortality. Although Indigenous peoples experience an elevated incidence of TB, prior systematic reviews have not focused on this group. We summarize and report findings related to time to diagnosis and treatment of pulmonary TB (PTB) among Indigenous peoples, globally.

Methods

A Systematic review was performed using Ovid and PubMed databases. Articles or abstracts estimating time to diagnosis, or treatment of PTB among Indigenous peoples were included with no restriction on sample size with publication dates restricted up to 2019. Studies that focused on outbreaks, solely extrapulmonary TB alone in non-Indigenous populations were excluded. Literature was assessed using the Hawker checklist. Registration Protocol (PROSPERO): CRD42018102463.

Results

Twenty-four studies were selected after initial assessment of 2021 records. These included Indigenous groups from five of six geographical regions outlined by the World Health Organization (all except the European Region). The range of time to treatment (24–240 days), and patient delay (20 days–2.5 years) were highly variable across studies and, in at least 60% of the studies, longer in Indigenous compared to non-Indigenous peoples. Risk factors associated with longer patient delays included poor awareness of TB, type of health provider first seen, and self-treatment.

Conclusion

Time to diagnosis and treatment estimates for Indigenous peoples are generally within previously reported ranges from other systematic reviews focusing on the general population. However among literature examined in this systematic review that stratified by Indigenous and non-Indigenous peoples, patient delay and time to treatment were longer compared to non-Indigenous populations in over half of the studies. Studies included were sparse and highlight an overall gap in literature important to interrupting transmission and preventing new TB cases among Indigenous peoples. Although, risk factors unique to Indigenous populations were not identified, further investigation is needed as social determinants of health among studies conducted in medium and high incidence countries may be shared across both population groups.

Trial registration N/a.

Peer Review reports

Introduction

Tuberculosis (TB) is one of the leading causes of death worldwide with 1.6 million dying from the disease in 2021, an increase from 1.4 million in 2019 [1]. In many parts of the world, the COVID-19 pandemic has resulted in a substantial reduction in TB testing and case notification, with an associated increase in mortality, taking TB control back by roughly 10 years [2]. The elimination of TB among individuals requires the early detection of active disease, reduction of the reservoir of TB infection, and a high rate of treatment completion [3, 4]. Risk factors associated with progression to active disease and/or TB transmission include chronic illnesses, low socioeconomic status, inequity, poor healthcare access, and a prolonged time to diagnosis and treatment [5]. At every step, the complex interplay between biological, social, environmental, and economic issues contributes to infection and disease [5].

Delayed diagnosis and treatment has been identified as a major contributor to TB transmission [4, 6]. Early detection of TB cases acquired from recent transmission represents a critical point of intervention to further reduce the reservoir of latent infection and prevent advanced disease, further transmission, and mortality [7,8,9]. These cases include those that exhibit symptoms related to disease such as cough or subclinical cases that are diagnosed through radiological or microbiological abnormalities either by health-seeking or active case finding. The interruption of TB transmission through early diagnosis and treatment is complicated by both patient and health system level factors. Previously conducted systematic reviews have reported total delay or time from onset of symptoms to treatment of pulmonary TB averaging between 25 and 185 days for both low, middle, and high-income countries [4, 10, 11]. Risk factors for increased time to treatment included poor access to healthcare, initial visits to traditional or private versus publicly funded practitioners, poverty, substance use, extrapulmonary TB, smear-negative TB, HIV, self-treatment, stigma, and a lack of knowledge about TB [4, 7, 12,13,14,15].

Systematic reviews about time to diagnosis and treatment have not focused on Indigenous peoples, who experience rates of TB that are generally higher than non-Indigenous persons [16]. Impacts of colonization through efforts of assimilation using residential schools, loss of lands, and historical trauma related to TB have contributed in different ways to the challenges facing access to care related to TB [17,18,19,20]. These colonial impacts along with evidence of on-going transmission in Indigenous populations could indicate longer delays in diagnosis and treatment [21]. A better understanding of the factors involved in diagnosis and treatment delay among Indigenous peoples is important to identifying gaps in knowledge and practice that could reduce the rate disparity. This systematic review assesses and summarizes literature about time to diagnosis and treatment of pulmonary TB (PTB) among Indigenous peoples worldwide. PTB was the main focus of this systematic review since it is the main disease site associated with transmission events. This assessment aims to include, where possible, the comparisons of diagnosis and treatment delay estimates between Indigenous peoples and non-Indigenous persons, to identify any additional risk factors unique to Indigenous groups.

Methods

Protocol and registration

Registration of the study protocol is on PROSPERO International prospective register of systematic reviews (PROSPERO August 2018: CRD42018102463). The study protocol is accessible at https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=102463. The systematic review followed the PRISMA framework [16, 22].

Search strategy and selection criteria

The systematic review focuses on studies estimating the time, and associated risk factors (where available) for diagnosis and treatment delay among Indigenous peoples worldwide. Indigenous peoples include those who are either defined, or self-identified as belonging to this population group in each study [16, 23, 24]. Table 1 outlines terminology used to define diagnosis and treatment delays among studies assessed in the context of this systematic review.

Table 1 Glossary of definitions

Herein we report synonymous terms, diagnosis delay (DD) or time to diagnosis (TDx) as the duration between the onset of symptoms and diagnosis (see Table 1). Time to treatment (TTm) or total delay (TD) is reported as the duration between the onset of symptoms and treatment, which is commonly stratified into two main parts, patient delay (PD) and health system delay (HD). The threshold used to define the cut-off of a “delayed event” is generally reported as a median estimate within a study, though sometimes it may be an arbitrarily chosen point-estimate. As such, a “delayed event” is any that occurs at a time after the cut-off threshold (median or otherwise), and a “timely event” is any that occurs in the time before the cut-off threshold.

The review includes manuscripts in peer-reviewed journals or published conference abstracts between 1910 and 2019 (see below) since the search was conducted mid-year in 2020 and to help minimize impacts on time to diagnosis and/or treatment estimates for TB during the COVID-19 pandemic. There were no restrictions on language and sample size. Articles in non-English languages were translated to verify that the inclusion criteria were satisfied. Literature that included Indigenous and non-Indigenous populations were accepted as part of the review if time to diagnosis and/or treatment results were stratified by population group. Studies with diagnosis and/or treatment delay estimates that were based on pulmonary TB in combination with extrapulmonary TB cases were included. These studies were either annotated in result tables as including ‘pulmonary and extrapulmonary TB’ or where possible, the PTB estimate was provided. Risk factors associated with delay in diagnosis and/or treatment were assessed as a secondary focus.

Studies that reported time to diagnosis and/or treatment estimates in only extrapulmonary TB cases or included only non-Indigenous populations were excluded. Definitions that did not include start and end points as part of the definition for time to diagnosis and/or treatment terminology (see Table 1) were excluded. Outbreak studies were excluded as they report estimates for time to diagnosis for an index case, rather than providing population level estimates. The focus of this review is PTB, given its importance to transmission.

Data (literature) was collected from Ovid and PubMed in June 2020. Ovid databases accessible to the University of Alberta included Embase (1974–June 12, 2020), Global Health (1910–June 12, 2020), Ovid Medline (1946–June 12, 2020), and Journals@Ovid Full Text (up to June 12, 2020). The PubMed search was conducted June 15, 2020. The literature search included broad and narrow search terminology about TB, Indigenous peoples or groups, and time to diagnosis and treatment terminology. Table 2 describes the search terms, which are grouped into three themes: (1) TB, (2) time to diagnosis or treatment, and (3) Indigenous peoples. The terms described in Table 2 were specific to the Ovid databases. Additional file 1: Appendix A1 describes analogous MESH terms used in PubMed. The search terms for Indigenous peoples were obtained from a systematic review conducted by Tollefson et al. that quantified the burden of TB disease among Indigenous peoples [16]. The review by Tollefson et al. in turn obtained Indigenous group names from Bartlett et al. and organizations participating in the United Nations Permanent Forum on Indigenous Issues (UNPFII) [16, 24].

Table 2 Keywords used for search strategy in the systematic review of delay in diagnosis and treatment studies among Indigenous peoples for Ovid databases

M. Varughese conducted the search, literature extraction, and duplicate record removal. M. Varughese and M. Li independently screened titles and abstracts for those that warranted a full-text review. M. Varughese and C. Heffernan subsequently assessed full-texts for inclusion. Disagreements at the first and second screening stages were resolved through consultation with C. Heffernan and M. Li, respectively. Articles written in non-English language were translated and/or assessed for inclusion.

Data synthesis

Data extracted from each selected paper included the year of study, country, Indigenous population group, total sample size (and stratified sub-population size for studies that included Indigenous peoples), study design/sampling description, study’s primary objective, study setting, type of data (disaggregate/single population study), definition of the time to diagnosis and treatment terminology, estimate (e.g. mean/median), estimate type (descriptive/univariate, multivariate, and/or qualitative), and risk factors associated with longer time to diagnosis and/or treatment if available. Disaggregated studies were defined as those reporting Indigenous population group estimates independently.

For estimates, preference where possible was given to median and/or mean values. Otherwise, percentiles (rank-based estimates) were used to describe the relative position compared to the median value. For example, if a study stated that 30% of people had a delay greater than 45 days, this estimate would be analogous to a 70th percentile or “above median”. This would mean the 50th percentile/median value would be less than 45 days. If a study stated that 80% of people had a delay greater than 45 days, this estimate would be analogous to a 20th percentile or “below median”. This would indicate that the 50th percentile/median value is greater than 45 days. The use of “below median” and “above median” provide context of cut-off values and comparisons to other studies. Estimates provided as a range were included if mean, median, and/or percentile values were not available. Estimations of time to diagnosis and/or treatment across studies was preferentially reported using medians, means, 35th to 65th percentiles, and mid-point of range values.

Risk factors were grouped into four main categories: demographics, language/education/culture, TB specific, and health care access. Demographics factors included age, gender, and indigenous status. Language/education/culture factors included language barrier, education status/TB knowledge, and stigma/cultural taboos. TB specific factors included AFB smear status/extrapulmonary TB and misdiagnosis. Health care access factors included type of medical location first visited, type of medical professional first visited, self-treatment/not seeking care, distance to health facility, transportation barriers, cost of treatment, and administrative delays. Risk factors when assessed were documented by type of delay, risk factor, and those that showed significance to a delayed or timely event. Method type was stratified as multivariate, descriptive/univariate, and/or qualitative. For multivariate and descriptive/univariate methods, factors that had a p-value less than 0.05 were described as significant. The significance level would be based on how authors chose to analyze the data i.e. selecting cut-off values of time to event estimates based on central tendency, percentiles, or a priori. Themes that arose from qualitative methods were noted as having significance to a delayed or timely event.

Selected literature was qualitatively assessed using the Hawker checklist for sampling, overall methodology, and generalizability of the study. This checklist has nine criteria each rated as very poor (point = 1), poor (point = 2), fair (point = 3), and good (point = 4) totaling 36 points [25]. The nine criteria are: (1) abstract and title, (2) introduction and aims, (3) method and data, (4) sampling, (5) data analysis, (6) ethics and bias, (7) results, (8) transferability or generalizability, and (9) implications and usefulness [25]. Two reviewers, M. Varughese and C. Heffernan assessed literature independently. The assessment scores of both reviewers were averaged where possible or deferred to a lower score to provide an overall conservative assessment. Since the scoring scale is arbitrary, a score of 27 and above was considered a ‘good quality study’. This minimum value was obtained by considering an article scoring at least three (interpretation of fair) across the nine categories.

Results

Figure 1 describes the total studies obtained, screened, and selected for the systematic review. A total of 2021 records were obtained from Ovid search databases (n = 1656) and PubMed (n = 365). Removing duplicate records yielded a total of 1938 records. A total of 1825 records were excluded due to lack of information about time to diagnosis or treatment of TB among Indigenous peoples. One-hundred and thirteen records (109 articles and four conference abstracts including three requiring translation) assessed for eligibility resulted in 24 records (22 articles and two conference abstracts) included in the systematic review (see Fig. 1). A total of 89 articles (87 full-text articles and two conference abstracts) were excluded due to outbreak investigations (n = 5; 5.6%), no inclusion of pulmonary tuberculosis (n = 4; 4.5%), no clear definition of time to diagnosis/treatment related terms (n = 2; 2.2%), no quantification of time to diagnosis/treatment related terms (n = 37; 41.6%) or estimates stratified by Indigenous group (n = 41; 46.1%). One article excluded due to no estimation of delay estimates by Indigenous group could not be further assessed beyond the abstract due to full-text inaccessibility [26]. Selected studies represented five out of six World Health Organization regions: African Region (n = 6), Region of the Americas (n = 3), South-East Asia Region (n = 5), Eastern Mediterranean Region (n = 1), and Western Pacific Region (n = 9).

Fig. 1
figure 1

Search methodology results and selection of studies for the systematic review

Table 3 provides a summary of selected studies such as region, population group, sample size, primary focus, study period and design, sampling method, type of data (single/disaggregate), and delay estimates. Time to diagnosis and/or treatment estimates were reported as means, median, percentiles, and/or range values. The primary focus in six of 24 studies (25%) was not related to estimating time to diagnosis and/or treatment [27,28,29,30,31,32]. Of these studies, four provided percentile/range estimates [28, 29, 31, 32]. Across all studies, 50% (12/24) provided percentile/range estimates for time to diagnosis and/or treatment.

Table 3 Summary of delay in diagnosis by study (n = 24) among indigenous populations

The range of time to diagnosis (56–203 days; 4 of 5 studies) and treatment (24–240 days; 6 studies), and patient (20 days–2.5 years; 15 studies), health system (3–34 days; 4 studies), diagnostic (6–22 days; 4 studies), and treatment (0–9 days; 3 studies) delay estimates were variable across studies. The time to diagnosis estimate from the Van der Oest et al. study was not included above since the percentile criteria (28 days—15th percentile) was not met. In the Banerjee et al. study, time to treatment was not provided as a range, but constructed using assumptions within the paper (Range: 8 to 15 months) [32]. Since the maximum value was not explicitly stated, the minimum value (8 months) was included in the above summary range for time to treatment.

Table 4 describes the qualitative assessment of studies (not including abstracts) using the Hawker’s checklist. The scores ranged between 21 and 33 out of a total score of 36. Qualitatively 77% (17/22) of studies had a score of at least 27, which was pre-defined as a ‘good quality study’. None received a score of one (‘very poor’) in the nine criteria. Ethics/bias (8/22; 36%) and transferability/generalizability (7/22; 32%) were two sub-categories that received the most ‘poor’ ratings (score = 2) across studies. Studies that received ‘poor’ ratings for ethics/bias did not state ethics [32,33,34] and/or limited their discussions of bias related to methods and data collected. Some studies used convenience or purposive samples [3, 32, 35, 36] and others required more sampling [28, 34, 37,38,39], which impacted the bias, transferability, and generalizability of results. Remoteness and heterogeneity of Indigenous communities within and across regions is an additional limitation to generalizing results from observational studies.

Table 4 Qualitative assessment of studies (n = 22) not including abstracts (n = 2) using the Hawker checklist

Figure 2 is a forest plot that describes the distribution of estimate types (time to diagnosis/treatment, patient/health system delay, and diagnostic/treatment delay) for each study included in the systematic review. Figure 2 includes other information such as sample size, uncertainty (mean: 95% confidence interval, median: interquartile (IQR), and range limits) where available, and TB site. Studies that reported other percentile quantities (also termed “below” or “above” median value) were included in Fig. 2. In these studies, uncertainty of estimates was not generally reported. One study [40] whose estimate was less than the 35th percentile was not included in Fig. 2.

Fig. 2
figure 2

A forest plot that describes the distribution of estimate types related to time to diagnosis and/or treatment of TB among Indigenous peoples across literature included in the systematic review. a) Acronyms: pulmonary TB (PTB), smear positive pulmonary TB (S + PTB), and pulmonary and extrapulmonary TB (PTB & EPTB), b) Distribution was reconstructed to determine the interquartile range, median, and/or range values [30, 32, 35], c) Interval with no point estimate, represents a range [31, 32, 36], and d) Interquartile ranges were used to describe the spread for median values except in Basa et al. and Hussen et al. [37, 41], which was reported as a range

Uncertainty of point estimates was either reported or reconstructed from data [35] within the study (if available) in 8 out of 23 studies (35%). Three studies reported a range without a point estimate [31, 32, 36]. Two studies reported range values (instead of IQR) with a median point estimate [37, 41]. There were no differences observed for treatment (n = 3 studies) and diagnostic (n = 4 studies) delay. Point estimates (mean, median, and 45th to 55th percentiles) for total delay had the largest variation ranging between 5 and 20 weeks. Point estimates for patient delay (3 to 10 weeks) contributed more to total delay compared to health system (< 1 week to 5 weeks) delay. Among studies that reported ranges, the greatest variability was observed in patient and total delay estimates. One study, conducted in the Solomon Islands, reported patient delays of between two and three years [36]. Seeking care from traditional healers, inaccessibility of healthcare due to remoteness, and cost were some of the factors explaining significantly longer patient delays compared to other studies.

Table 5 provides a summary of risk factors associated with time to event measures in 21 of 24 studies (88%). Two studies [27, 30] did not describe associated risk factors for time to event measures. Sixty-three percent of studies (15/24) described risk factors associated with patient delay. Risk factors were stratified into four themes: (1) demographics, (2) language, education, and culture, (3) TB specific, and (4) health care access and attributes such as type of study (Indigenous (S)/disaggregated population (D)) and methodology (multivariate/descriptive (e.g. chi-square test)/univariate/qualitative) were provided. For each risk factor, total references were estimated to examine overall strength in risk factor association with time to event measures. Multivariate, descriptive, and univariate methods were mainly used to assess risk factors associated with time to treatment/total delay and health systems delay (including treatment and diagnostic delay). More variability was observed in the methods (quantitative and qualitative) used to assess risk factors associated with time to diagnosis and patient delay.

Table 5 Risk factor summary across time to event measures, direction of association (delayed/timely), and type of study (Indigenous/disaggregate population)

Age and gender had inconclusive associations across time to event measures (Table 5). In 63% (5/8) of studies, patient delays were significantly longer among Indigenous peoples compared to non-Indigenous persons. One study that described a shorter patient delay among Indigenous peoples highlighted some limitations in the interpretation of these results [40]. One study [42] that assessed patient delay in Ethiopia made comparisons within an Indigenous population group (not included in Table 5). This study described significantly longer delays among nomadic- compared to agro- pastoralists [42]. In 67% (2/3) of studies, total delay was greater among Indigenous peoples compared to non-Indigenous persons. Poor TB awareness, knowledge and/or education status was consistently associated with longer patient delay (7/10 studies; 70%). Cultural taboo, social isolation, and/or stigma was associated with longer patient (2/3 studies; 67%) and diagnosis (all 2 studies) delays. Type of health provider (i.e. traditional healer) first visited (all six studies) and self-treatment (4/5 studies; 80%) was consistently associated with longer patient delays. Distance to health facilities/remoteness (all 9 studies), transportation barriers (6/7 studies; 86%), and cost of treatment (all 6 studies) were associated with longer patient delays. Overall, risk factors assessed were generally evenly distributed between studies that solely focused on Indigenous and disaggregate populations. Type of health provider visited, transportation barriers, and cost of treatment represented risk factors that were assessed more in studies that focused on Indigenous compared to disaggregate populations.

Discussion

There is an estimated population of at least 476 million Indigenous peoples across 90 countries worldwide [43]. While the incidence of TB is generally higher for Indigenous peoples, our search strategy resulted in only 24 studies (22 journal articles and 2 abstracts) that met the inclusion criteria for this systematic review. The estimated total (24–240 days), patient (20 days–2.5 years), and health system (3–34 days) delay for Indigenous peoples were within previously reported ranges that focused on PTB in low, middle, and high-income countries in the general population [10, 11, 44, 45]. One study estimating patient delay (2–3 years) that fell outside this range was conducted in the Solomon Islands, an extremely remote community with barriers to healthcare access due to distance, cost, and preference for traditional practices [36]. Over 60% of disaggregated studies described longer patient and/or total delay among Indigenous peoples compared to non-Indigenous persons despite shared socioeconomic challenges faced in medium (15–29 cases per 100,000 population) and high (> 30 cases per 100,000 population) TB incidence countries [45, 46].

Studies that estimated time to diagnosis and treatment of TB among Indigenous peoples were sparse in terms of volume and geographical coverage. Literature among Indigenous populations in regions such as Europe and South America were not represented in this systematic review. Half of the studies that met the inclusion criteria focused solely on Indigenous peoples [27, 29,30,31,32, 35,36,37, 41, 42, 47,48,49]. The remaining studies included these estimates stratified by population group [3, 28, 33, 34, 38,39,40, 44, 50,51,52,53]. The sparseness of literature highlights a need for more research to better understand time to diagnosis and treatment of TB among Indigenous peoples since their overall TB burden is generally higher than non-Indigenous persons [16].

Patient and/or total delay (10 studies) was longer among Indigenous peoples compared to non-Indigenous persons in 60% of disaggregated studies. Patient delay was strongly tied to total delay since it contributed the most to the overall pathway from symptom onset to treatment. One study [40] that described shorter patient delays among Indigenous Australians noted caution in the interpretation of results. Excluding this study would increase the previous result to 67% or two-thirds of disaggregate studies. The three studies [34, 51, 53] that showed no differences in patient and/or total delay between Indigenous peoples and non-Indigenous persons had samples mostly comprised of Indigenous peoples (> 85% of the total sample). These comparisons are impacted by unequal sample distributions and further strengthens observed differences in patient and/or total delay between Indigenous peoples and non-Indigenous persons.

Risk factors identified in this systematic review were consistent with those in other systematic reviews [4, 10, 44, 45]. Most studies examined risk factors associated with patient delay (63%; 15/24). Risk factors significantly associated with longer patient delays included having poor knowledge/awareness of TB, type of health provider first seen (traditional healer), self-treatment/not seeking care, distance to health facility, transportation challenges, and cost of treatment. These studies were distributed evenly across those that focused on aggregated and disaggregated studies. All disaggregated studies were conducted in countries with medium and high incidence of TB and 67% (4/6) of these showed longer patient delays between Indigenous peoples and non-Indigenous persons [3, 39, 50, 54]. In these countries, population differences associated with patient delay are blurred since they face similar circumstances of increased poverty, malnutrition, overcrowding, reduced access to healthcare, and poor education [45, 46]. In these circumstances, differences in risk factors associated with patient delay may not be easily differentiated due to shared socioeconomic challenges faced by Indigenous peoples and non-Indigenous persons. Since patient delay contributes the most towards the pathway from onset of symptoms to treatment, risk factor comparisons associated with other time to event measures between Indigenous peoples and non-Indigenous persons may not be similarly impacted by having shared socioeconomic circumstances alone. This highlights a literature gap that requires further assessment.

Although increasing knowledge and training of medical centre staff and other health professionals, increasing medical facilities, and subsidizing travel costs could improve factors related to healthcare access, socioeconomic barriers such as poverty, unemployment, poor housing, food insecurity, and transportation availability can impact the feasibility and acceptance of public health interventions [55]. Understanding the relationships between these barriers and public health interventions, specifically in countries or population groups that are most impacted is important to determining ways to improve healthcare access. For some Indigenous communities, traditional medicine represents an integral part of the culture and promoting the use of non-traditional medicines can be challenging [56, 57]. Collaborative knowledge sharing between traditional healers/practitioners and health professionals could help reduce patient delays in areas where this practice is more prevalent by encouraging people with symptoms related to TB to seek additional care from non-traditional medical practitioners.

Indigenous peoples in low TB incidence countries such as Canada, United States, Australia, and New Zealand have higher rates of poverty and chronic illness, and poor housing conditions and health care access [16, 19, 21, 28, 58,59,60], that similarly impact those in medium and high incidence countries. These comparisons were challenging to make since few studies were included from low incidence countries. Only two of these studies allowed for making comparisons between Indigenous and non-Indigenous [28, 40]. In these studies, time to event measures was estimated to be comparable between Indigenous and non-Indigenous groups. In the New Zealand study, 85% of the sample had a time to diagnosis estimate of over 4 weeks [28]. This percentile-based estimate and using 4-weeks as the cut-off value to make comparisons between population groups was difficult to interpret without knowing the overall distribution of estimates (e.g. mean/median). Recall of symptom onset was another bias identified in a study impacting Indigenous populations in Australia who experience higher case fatality rates indicative of advanced disease [40]. Overall comparisons and disparities of time to event measures between Indigenous and non-Indigenous populations in low incidence countries were limited and represents a gap in literature.

This systematic review has provided a general scope of the literature about time to diagnosis and treatment of TB among Indigenous peoples worldwide. The estimates obtained in this review provide a duration of time where appreciable interventions can be made to interrupt the transmission and reduce morbidity and/or mortality of TB. Although research methodologies and time to event definitions varied across studies, efforts were made to highlight appreciable differences and align definitions with terminology described in this review.

There were limitations in making comparisons of time to event measures of TB across studies due to scarcity of studies meeting inclusion criteria, inconsistent reporting of estimate types (mean, median, range, percentile type, and a pre-determined cut-off value), inclusion of uncertainty estimates, and variability of time to event definitions. The term delay was highly interpretable and often based on the data itself (e.g. median value) or arbitrarily selected, a priori. In this review, comparisons of time to event estimates between population groups were based on how authors chose to define a “delayed event” since additional data was limited. Cut-off values are useful for multivariate analysis, however, are subject to bias on how cut-off values are chosen. Reporting the spread (e.g. IQRs, standard deviations) and central tendency (e.g. mean/median) in future studies would allow for better comparisons between and within geographical regions and/or population groups since this type of data is often skewed and ‘delayed events’ of time to event measures are not yet standardized.

Similarly observed in previously conducted systematic reviews, the operationalization of time to event terminology was not consistent across literature [4, 10, 11, 13, 44]. Among studies assessed, 13% (3/24) [31, 32, 61] did not explicitly connect time to event terms to a definition and 33% (7/21) [28, 30, 33, 36, 38, 42, 53] used definitions that did not coincide with descriptions in Table 1. One study (written in mainly Mandarin) described patient delay using the term “treatment-delay time” [53]. These inconsistencies in defining time to event terminology highlight potential misclassification of terms if they were not well defined and missing studies during the literature search.

Half of studies (4/8) that described diagnosis as the end point (time to diagnosis and/or diagnostic delay) did not clearly defined the term ‘diagnosis’ [29, 35, 39, 42]. The definition of “diagnosis” date is complex since it can be made from bacteriological, histological, and/or clinical evidence [62]. In practice and in literature, “diagnosis” or date of diagnosis in one study may not be the same in another, which makes comparisons between studies challenging. The date of diagnosis is dependent on its definition and test(s) used, e.g. culture (two to eight weeks), nucleic acid amplification test (24–48 h), and based on history of exposure, suggestive symptoms, and/or a chest radiographic abnormality consistent with TB (clinical diagnosis) [62]. As such, time to diagnosis estimates could potentially be overestimated if diagnosis was defined as a positive culture, as opposed to a positive smear. Although uncommon, if there is a high level of suspicion and/or an infection control issue, it is possible that the date of treatment could occur prior to a positive diagnostic test result. In this case, it is unknown how the date of diagnosis would be reported. The use of treatment start date as the “diagnosis” date would better estimate the potential impact of on-going transmission and the development of new TB cases and allow for improved comparisons between studies.

Future studies describing TB trends and time to event measures would benefit from well-defined start and ends points, including central tendency (e.g. mean/median) and variance for estimates especially if a subjective a priori cut-off value is used, and avoiding where possible, diagnosis date as an endpoint and using treatment date instead. A need for standardization of what constitutes a ‘delayed event’ or the use of terminology that avoids ‘delay’ where possible can help prevent possible misinterpretations since this is fraught with subjective cut-off values. Examples of terminology would include time to treatment instead of total delay, respectively.

One third of studies in this systematic review included pulmonary TB in combination with extrapulmonary TB [3, 28,29,30, 32, 35, 36, 42]. In these studies, time to event estimates could be over-estimated [4] based on the sample distribution between those who have pulmonary and extrapulmonary TB. This review explicitly highlighted these differences in the forest plot (Fig. 2) to provide additional context for comparison across studies and types of time to event measures. This systematic review was limited to peer reviewed literature and/or abstracts (not including outbreak analyses). Indigenous cultures are diverse and can use different ways to transfer and share knowledge that are not always written. The impact of not including all possible forms of knowledge that describe time to event measures for TB among Indigenous populations is unknown. The systematic review included literature published up to 2019 and excludes potential biases related to access to health services during the COVID-19 pandemic (from 2020). Time to event measures are likely to have been impacted since a large drop in TB cases with an increase in TB deaths was observed globally in 2020 and 2021 [1]. Conducting a comparison of time to event measures among Indigenous peoples and non-Indigenous persons between literature prior and post-2019 may provide additional insights to previously identified and/or new risk factors since more advanced disease was observed from 2020.

Conclusion

This is the first systematic review that aimed to describe time to diagnosis, treatment, and other time event measures of TB among Indigenous populations worldwide. While only 24 studies met the inclusion criteria, this systematic review has highlighted a need for more research especially into risk factors associated time to event measures among Indigenous peoples. Studies examining time to event estimates for TB would in general benefit from clear definitions of end points and the inclusion of variance and central tendency to allow for better comparisons between population groups and geographies. Time to event estimates among Indigenous peoples were generally within reported ranges based on previously conducted systematic review in the general population. However, among literature examined in this systematic review that stratified by Indigenous and non-Indigenous peoples, patient and total delay was longer compared to non-Indigenous persons in at least 60% of studies. Risk factors associated with patient delay were previously described in literature and shared between Indigenous peoples and non-Indigenous persons, especially in those conducted in medium and high incidence countries. In these countries, differences may be challenging to identify without more in-depth research since Indigenous and non-Indigenous population groups may face similar circumstances related to socioeconomic status and social determinants of health. The reduction of TB is multi-faceted and requires improved case finding, increased treatment completion rates, reducing risk factors associated with progression from latent infection to disease, and most importantly the prevention of new TB cases by interrupting transmission. Estimating time to diagnosis, treatment, and other time to event measures including addressing risk factors would have clear benefits to both Indigenous peoples and non-Indigenous persons in the prevention of TB.

Availability of data and materials

All data generated or analysed during this study are included in this published article.

Abbreviations

DD:

Diagnosis delay

DCD:

Health system diagnostic delay

EPTB:

Extrapulmonary tuberculosis

HD:

Health system delay

IQR:

Interquartile range

MeSH:

Medical Subject Headings

PD:

Patient delay

PTB:

Pulmonary tuberculosis

TB:

Tuberculosis

TD:

Total delay

TDx:

Time to diagnosis

TmD:

Health system treatment delay

TTm:

Time to treatment

References

  1. World Health Organization. Global Tuberculosis Report: 2022. Geneva; 2022.

  2. Drain PK, Bajema KL, Dowdy D, Dheda K, Naidoo K, Schumacher SG, et al. Incipient and subclinical tuberculosis: a clinical review of early stages and progression of infection. Clin Microbiol Rev. 2018. https://doi.org/10.1128/CMR.00021-18.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Belay M, Bjune G, Ameni G, Abebe F. Diagnostic and treatment delay among tuberculosis patients in Afar Region, Ethiopia: a cross-sectional study. BMC Public Health. 2012;12:369.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Storla DG, Yimer S, Bjune GA. A systematic review of delay in the diagnosis and treatment of tuberculosis. BMC Public Health. 2008;8:15.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Organization WH. How health systems can address inequities in priority public health conditions: the example of tuberculosis. Geneva: World Health Organization; 2010.

    Google Scholar 

  6. Uys PW, Warren RM, van Helden PD. A threshold value for the time delay to TB diagnosis. PLoS ONE. 2007;2: e757.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Lacroix C, Martin P, Turcotte S, DeRoche S, Magluilo V, Lacroix C. The delay in diagnosis of tuberculosis in the Monteregie region of Quebec, Canada. Mcgill J Med. 2008;11:124–31.

    PubMed  PubMed Central  Google Scholar 

  8. Zafran N, Heldal E, Pavlovic S, Vuckovic D, Boe J. Why do our patients die of active tuberculosis in the era of effective therapy? Tuber Lung Dis. 1994;75:329–33.

    Article  CAS  PubMed  Google Scholar 

  9. Enarson DA, Grzybowski S, Dorken E. Failure of diagnosis as a factor in tuberculosis mortality. Can Med Assoc J. 1978;118:1520–2.

    CAS  PubMed  PubMed Central  Google Scholar 

  10. Sreeramareddy CT, Panduru K, Menten J, van den Ende J. Time delays in diagnosis of pulmonary tuberculosis: a systematic review of literature. BMC Infect Dis. 2009;9:91.

    Article  PubMed  PubMed Central  Google Scholar 

  11. 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;19:820–4.

    Article  PubMed  PubMed Central  Google Scholar 

  12. 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;18:255–66.

    Article  PubMed  Google Scholar 

  13. 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;10: e0120088.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Greenaway C, Menzies D, Fanning A, Grewal R, Yuan L, FitzGerald JM, et al. Delay in diagnosis among hospitalized patients with active tuberculosis–predictors and outcomes. Am J Respir Crit Care Med. 2002;165:927–33.

    Article  PubMed  Google Scholar 

  15. Machado AC, Steffen RE, Oxlade O, Menzies D, Kritski A, Trajman A. Factors associated with delayed diagnosis of pulmonary tuberculosis in the state of Rio de Janeiro, Brazil. J Bras Pneumol. 2011;37:512–20.

    Article  PubMed  Google Scholar 

  16. Tollefson D, Bloss E, Fanning A, Redd JT, Barker K, McCray E. Burden of tuberculosis in indigenous peoples globally: a systematic review. Int J Tuberc Lung Dis. 2013;17:1139–50.

    Article  CAS  PubMed  Google Scholar 

  17. Abonyi S, Mayan M, Boffa J, Lopez C, McMullin K, Heffernan C, et al. “Finally when I started falling down”: indigenous tuberculosis patient experiences of health and illness on the Canadian Prairies. Int J Indig Health. 2017;12:3.

    Article  Google Scholar 

  18. Hick S. The enduring plague: how tuberculosis in Canadian indigenous communities is emblematic of a greater failure in healthcare equality. J Epidemiol Glob Health. 2019;9:89–92.

    PubMed  PubMed Central  Google Scholar 

  19. Cormier M, Schwartzman K, N’Diaye DS, Boone CE, dos Santos AM, Gaspar J, et al. Proximate determinants of tuberculosis in Indigenous peoples worldwide: a systematic review. The LancetGlobal health. 2019;7:e68-80.

    Google Scholar 

  20. King M, Smith A, Gracey M. Indigenous health part 2: the underlying causes of the health gap. Lancet. 2009;374:76–85.

    Article  PubMed  Google Scholar 

  21. Kunimoto D, Sutherland K, Wooldrage K, Fanning A, Chui L, Manfreda J, et al. Transmission characteristics of tuberculosis in the foreign-born and the Canadian-born populations of Alberta, Canada. Int J Tuberc Lung Dis. 2004;8:1213–20.

    CAS  PubMed  Google Scholar 

  22. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009;339: b2700.

    Article  PubMed  PubMed Central  Google Scholar 

  23. UN Permanent Forum on Indigenous Issues (UNPFII), State of the World’s Indigneous Peoples, 2nd Volume. United Nations Economic & Social Affairs; 2015.

  24. Bartlett JG, Madariaga-Vignudo L, O’Nneil JD, Kuhnlein HV. Identifying indigenous peoples for health research in a global context: a review of perspectives and challenges. Int J Circumpolar Health. 2007;66:287–307.

    Article  PubMed  Google Scholar 

  25. Hawker S, Payne S, Kerr C, Hardey M, Powell J. Appraising the evidence: reviewing disparate data systematically. Qual Health Res. 2002;12:1284–99.

    Article  PubMed  Google Scholar 

  26. Nguepy Keubo FR, Djifack Tadongfack T, Tsafack J, Djoufack G, Bianke P. Clinical profile of patients with tuberculosis in Dschang, Cameroon: Role of the therapeutic path in delaying diagnosis. Med Sante Trop. 2018;28:399–403.

    CAS  PubMed  Google Scholar 

  27. Alvarez GG, Dyk DD, Desjardlns M, Yasseen AS 3rd, Aaron SD, Cameron DW, et al. The feasibility, accuracy, and impact of Xpert MTB/RIF testing in a remote aboriginal community in Canada. Chest. 2015;148:767–73.

    Article  PubMed  Google Scholar 

  28. van der Oest AC, Kelly P, Hood D. The changing face of tuberculosis control in a rural district of New Zealand. Int J Tuberc Lung Dis. 2004;8:969–75.

    PubMed  Google Scholar 

  29. Chideya S, Harrington T, Mukasa L, Bahktawar I. Disproportionate burden of tuberculosis among Marshall Islanders living in Arkansas, 2000–2005. In: 11th Annual Conference, International Union of Tuberculsosi and Lung Disease - Norther American Region. Vancouver; 2007.

  30. Podewils LJ, Alexy E, Driver SJ, Cheek JE, Holman RC, Haberling D, et al. Understanding the burden of tuberculosis among American Indians/Alaska Natives in the U.S.: a validation study. Public Health Rep. 2014;129:351–60.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Honarvar B, Odoomi N, Rezaei A, Haghighi HB, Karimi M, Hosseini A, et al. Pulmonary tuberculosis in migratory nomadic populations: the missing link in Iran’s National Tuberculosis Programme. Int J Tuberc Lung Dis. 2014;18:272–6.

    Article  CAS  PubMed  Google Scholar 

  32. Banerjee A, Sharma BV, Ray A, Kannuri NK, Venkateswarlu TV. Acceptability of traditional healers as directly observed treatment providers in tuberculosis control in a tribal area of Andhra Pradesh, India. Int J Tuberc Lung Dis. 2004;8:1260–5.

    CAS  PubMed  Google Scholar 

  33. Lin HP, Deng CY, Chou P. Diagnosis and treatment delay among pulmonary tuberculosis patients identified using the Taiwan reporting enquiry system, 2002–2006. BMC Public Health. 2009;9:55.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Lawn SD, Afful B, Acheampong JW. Pulmonary tuberculosis: diagnostic delay in Ghanaian adults. Int J Tuberc Lung Dis. 1998;2:635–40.

    CAS  PubMed  Google Scholar 

  35. Viney KA, Johnson P, Tagaro M, Fanai S, Linh NN, Kelly P, et al. Tuberculosis patients’ knowledge and beliefs about tuberculosis: a mixed methods study from the Pacific Island nation of Vanuatu. BMC Public Health. 2014;14:467.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Massey PD, Wakageni J, Kekubata E, Maena’adi J, Laete’esafi J, Waneagea J, et al. TB questions, East Kwaio answers: community-based participatory research in a remote area of Solomon Islands. Rural Remote Health. 2012;12:2139.

    CAS  PubMed  Google Scholar 

  37. Basa S, Venkatesh S. Patient and healthcare system delays in the start of pulmonary tuberculosis treatment among tribal patients registered under DOTS, Odisha. J Clin Diagn Res. 2016;10:LC21–4.

    PubMed  PubMed Central  Google Scholar 

  38. Chang CT, Esterman A. Diagnostic delay among pulmonary tuberculosis patients in Sarawak, Malaysia: a cross-sectional study. Rural Remote Health. 2007;7:667.

    PubMed  Google Scholar 

  39. Ngamvithayapong J, Yanai H, Winkvist A, Diwan V. Health seeking behaviour and diagnosis for pulmonary tuberculosis in an HIV-epidemic mountainous area of Thailand. Int J Tuberc Lung Dis. 2001;5:1013–20.

    CAS  PubMed  Google Scholar 

  40. Ward J, Siskind V, Konstantinos A. Patient and health care system delays in Queensland tuberculosis patients, 1985–1998. Int J Tuberc Lung Dis. 2001;5:1021–7.

    CAS  PubMed  Google Scholar 

  41. Hussen A, Biadgilign S, Tessema F, Mohammed S, Deribe K, Deribew A. Treatment delay among pulmonary tuberculosis patients in pastoralist communities in Bale Zone, Southeast Ethiopia. BMC Res Notes. 2012;5:320.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Gele AA, Bjune G, Abebe F. Pastoralism and delay in diagnosis of TB in Ethiopia. BMC Public Health. 2009;9:5.

    Article  PubMed  PubMed Central  Google Scholar 

  43. United Nations. International Day of the World’s Indigenous Peoples. https://www.un.org/en/observances/indigenous-day/background. 2022.

  44. 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;17:202.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Teo AKJ, Singh SR, Prem K, Hsu LY, Yi S. Delayed diagnosis and treatment of pulmonary tuberculosis in high-burden countries: a systematic review protocol. BMJ Open. 2019;9:e029807-029807.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Figueroa-Munoz J, Ramon-Pardo P. Tuberculosis control in vulnerable groups. Bull World Health Organ. 2008;86:657.

    Article  Google Scholar 

  47. Bloss E, Holtz TH, Jereb J, Redd JT, Podewils LJ, Cheek JE, et al. Tuberculosis in Indigenous Peoples in the U.S., 2003–2008. Public Health Rep. 2011;126:677–89.

    Article  PubMed  PubMed Central  Google Scholar 

  48. Cambanis A, Yassin MA, Ramsay A, Bertel Squire S, Arbide I, Cuevas LE. Rural poverty and delayed presentation to tuberculosis services in Ethiopia. Trop Med Int Health. 2005;10:330–5.

    Article  PubMed  Google Scholar 

  49. Das D, Dwibedi B. Delay in diagnosis among pulmonary tuberculosis patients of Rayagada District, Odisha. India Int J Mycobacteriol. 2016;5(Suppl 1):S172–3.

    Article  PubMed  Google Scholar 

  50. Khan MS, Ning Y, Jinou C, Hutchison C, Yoong J, Lin X, et al. Are global tuberculosis control targets overlooking an essential indicator? Prolonged delays to diagnosis despite high case detection rates in Yunnan, China. Health Policy Plan. 2017;32(suppl_2):15–21.

    Article  Google Scholar 

  51. Rundi C, Fielding K, Godfrey-Faussett P, Rodrigues LC, Mangtani P. Delays in seeking treatment for symptomatic tuberculosis in Sabah, East Malaysia: factors for patient delay. Int J Tuberc Lung Dis. 2011;15(1231–8):i.

    PubMed  Google Scholar 

  52. Huong NT, Vree M, Duong BD, Khanh VT, Loan VT, Co NV, et al. Delays in the diagnosis and treatment of tuberculosis patients in Vietnam: a cross-sectional study. BMC Public Health. 2007;7:110.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Lingxing L, Zaiping C, Tongping Y. Analysis of health-seeking delay and relevant factors of tuberculosis diagnosis among ethnic minority areas. Chongquing Medicine. 2017;46:2535–7.

    Google Scholar 

  54. Getnet F, Demissie M, Worku A, Gobena T, Seyoum B, Tschopp R, et al. Determinants of patient delay in diagnosis of pulmonary tuberculosis in Somali pastoralist setting of Ethiopia: a matched case-control study. Int J Environ Res Public Health. 2019. https://doi.org/10.3390/ijerph16183391.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Tucker JD, Tso LS, Hall B, Ma Q, Beanland R, Best J, et al. Enhancing public health HIV interventions: a qualitative meta-synthesis and systematic review of studies to improve linkage to care, adherence, and retention. EBioMedicine. 2017;17:163–71.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Mokgobi MG. Towards integration of traditional healing and western healing: Is this a remote possibility? Afr J Phys Health Educ Recreat Dance. 2013;2013(Suppl 1):47–57.

    CAS  PubMed  PubMed Central  Google Scholar 

  57. Li R. Indigenous identity and traditional medicine: pharmacy at the crossroads. Can Pharm J (Ott). 2017;150:279–81.

    Article  PubMed  Google Scholar 

  58. The Canadian Tuberculosis Commitee. Housing conditions that serve as risk factors for tuberculosis infection and disease. Canada communicable disease report. 2007;33 ACS-9.

  59. Larcombe L, Nickerson P, Singer M, Robson R, Dantouze J, McKay L, et al. Housing conditions in 2 Canadian First Nations communities. Int J Circumpolar Health. 2011;70:141–53.

    Article  PubMed  Google Scholar 

  60. Robertus LM, Konstantinos A, Hayman NE, Paterson DL. Tuberculosis in the Australian Indigenous population: history, current situation and future challenges. Aust N Z J Public Health. 2011;35:6–9.

    Article  PubMed  Google Scholar 

  61. Alvarez S, Shell C, Berk SL. Pulmonary tuberculosis in elderly men. Am J Med. 1987;82(3):602–6.

    Article  CAS  PubMed  Google Scholar 

  62. Menzies D, Long R, Ellis E, Pai M, Wong T, Canada PHA of, et al. Canadian tuberculosis standards 7th edition. Ottawa: Public Health Agency of Canada; 2014.

Download references

Acknowledgements

Not applicable.

Funding

The research was supported by Canadian Institute of Health Research (RL: Grant #151631 & MV: Frederick Banting and Charles Best Canada Graduate Scholarship), Health Canada (http://www.cihr-irsc.gc.ca/e/193.htmlhttp://www.hc-sc.gc.ca/index-eng.php), and Alberta Innovates Graduate Studentship (MV).

Author information

Authors and Affiliations

Authors

Contributions

MV, CH, ML, and RL all contributed to the study design. MV conducted the literature search and wrote the first draft. ML and CH helped with translation of articles written in non-English languages. CH, ML, and RL provided input on figures/tables, data interpretation, writing, and editing. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Marie Varughese.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Appendix A1.

MeSH terms used by PubMed using themes outlined in Table 2.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Varughese, M., Heffernan, C., Li, M.Y. et al. Time to diagnosis and treatment of pulmonary tuberculosis in indigenous peoples: a systematic review. BMC Infect Dis 23, 131 (2023). https://doi.org/10.1186/s12879-023-08098-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12879-023-08098-y

Keywords