Mycobacterium tuberculosis Drug Resistance in Ethiopia: An Updated Systematic Review and Meta-Analysis

Background: Tuberculosis (TB) remains a significant global public health issue, despite advances in diagnostic technologies, substantial global efforts, and the availability of effective chemotherapies. Mycobacterium tuberculosis, a species of pathogenic bacteria resistant to currently available anti-TB drugs, is on the rise, threatening national and international TB-control efforts. This systematic review and meta-analysis aims to estimate the pooled prevalence of drug-resistant TB (DR-TB) in Ethiopia. Materials and Methods: A systematic literature search was undertaken using PubMed/MEDLINE, HINARI, the Web of Science, ScienceDirect electronic databases, and Google Scholar (1 January 2011 to 30 November 2020). After cleaning and sorting the records, the data were analyzed using STATA 11. The study outcomes revealed the weighted pooled prevalence of any anti-tuberculosis drug resistance, any isoniazid (INH) and rifampicin (RIF) resistance, monoresistance to INH and RIF, and multidrug-resistant TB (MDR-TB) in newly diagnosed and previously treated patients with TB. Results: A total of 24 studies with 18,908 patients with TB were included in the final analysis. The weighted pooled prevalence of any anti-TB drug resistance was 14.25% (95% confidence interval (CI): 7.05–21.44%)), whereas the pooled prevalence of any INH and RIF resistance was found in 15.62% (95%CI: 6.77–24.47%) and 9.75% (95%CI: 4.69–14.82%) of patients with TB, respectively. The pooled prevalence for INH and RIF-monoresistance was 6.23% (95%CI: 4.44–8.02%) and 2.33% (95%CI: 1.00–3.66%), respectively. MDR-TB was detected in 2.64% (95%CI: 1.46–3.82%) of newly diagnosed cases and 11.54% (95%CI: 2.12–20.96%) of retreated patients with TB, while the overall pooled prevalence of MDR-TB was 10.78% (95%CI: 4.74–16.83%). Conclusions: In Ethiopia, anti-tuberculosis drug resistance is widespread. The estimated pooled prevalence of INH and RIF-monoresistance rates were significantly higher in this review than in previous reports. Moreover, MDR-TB in newly diagnosed cases remained strong. Thus, early detection of TB cases, drug-resistance testing, proper and timely treatment, and diligent follow-up of TB patients all contribute to the improvement of DR-TB management and prevention. Besides this, we urge that a robust, routine laboratory-based drug-resistance surveillance system be implemented in the country.


Introduction
Tuberculosis (TB) has been known as a human pathogen for decades and is still a significant global health problem [1][2][3]. Despite advances in diagnostic tools, the availability of effective anti-TB therapy, and substantial global efforts, about 10.0 million people contracted TB in 2019, resulting in over 1.4 million deaths [1]. The burden of national TB epidemics varies significantly between countries, with Asia and Africa being the most affected continents [1]. Due to a multitude of causes, TB has the highest mortality rate of all infectious diseases in "low-and middle-income countries" (LMICs). Poor access to healthcare facilities, weak TB preventive and control systems, overcrowded living conditions, occupational hazards, individuals' poor nutritional status, a high prevalence of human immunodeficiency virus (HIV), and other comorbidities and drug addiction all lead to the high incidence of TB in LMICs [4,5]. Ethiopia, like other LMICs, is experiencing an increase in TB incidence since 2001, posing major challenges for the public healthcare system and national TB-control efforts [3].
Drug-resistant tuberculosis (DR-TB) strains, particularly multidrug-resistant TB (MDR-TB), continue to pose a serious threat to public healthcare systems, mainly in resourceconstrained nations such as Ethiopia, where innovative molecular diagnostic technologies and well-equipped laboratory settings are lacking [3,6,7]. DR-TB usually occurs due to the patient's delay in early diagnosis and treatment, previous anti-TB drug exposure [8], inappropriate drug regimens [9], the patient's poor adherence to anti-tuberculosis drug regimens [10], and primary infection with DR-TB strains [11][12][13]. Antibiotic resistance develops in Mycobacterium tuberculosis strains due to spontaneous gene alterations that decrease the susceptibility of the bacterium to the most widely used anti-tuberculosis drugs. These genes can encode drug targets or mechanisms of drug metabolism, affecting the efficacy of anti-TB therapy [1, 14,15]. Failure to diagnose drug-resistance and subsequent improper therapy of TB-patient increase the risk of developing drug resistance and the direct transmission of DR-TB strains to other individuals [1,14]. Treatment of DR-TB, particularly MDR-TB, is time-consuming and costly, requiring the use of second-line anti-TB medicines that are more toxic and ineffective [1,3].
In resource-constrained settings such as Ethiopia, DR-TB, particularly MDR-TB detection rate, is under-reported due to a lack of reliable, quick, and affordable diagnostic testing. Nearly half a million people worldwide contracted RIF-resistant TB (RR-TB) in 2019, with 78.0% of those suffering from MDR-TB [1]. According to the recent World Health Organization (WHO) 2019 report, MDR/RR-TB was found in 3.4% of newly tested and 18.0% of patients previously treated with TB worldwide [3]. Ethiopia is among the top 30 countries with the highest rates of TB, DR-TB, and TB/HIV co-infection [3]. According to a recent national report, Ethiopia's TB incidence was reported to be 157 per 100,000 people, with 23,800 people dying from the disease [1]. In Ethiopia, MDR-TB was found in 0.71% of newly diagnosed cases and 12.0% of patients previously treated with TB in 2019 [1].
To effectively treat and prevent the spread of DR-TB strains, all patients with TB must undergo anti-TB drug susceptibility testing (DST), either by standard phenotypic or by using effective and rapid molecular diagnostic methods [3,16]. However, mycobacterial culture on a liquid or solid medium and conventional DST is time-consuming and requires well-equipped laboratory settings and extensive biosafety resources to obtain results that inform the initiation of proper anti-TB drug treatment. This is impractical in several resource-constrained countries, including Ethiopia [1, 15,17]. Moreover, phenotypic DST methods usually lack accuracy and reproducibility [17]. For these reasons, the use of reliable and rapid molecular diagnostic methods is recommended and widely practiced globally. Molecular testing assays, such as the "GeneXpert ® MTB/RIF and Ultra assays" ("Cepheid, Sunnyvale, CA, USA") and the line probe assays "GenoType ® MTBDRplus" and "GenoType ® MTBDRsl" ("Hain Life-science GmbH, Nehren, Germany"), have been shown to significantly reduce the time required to initiate therapy in patients with TB and DR-TB and for the establishment of the appropriate treatment regimen [1]. GeneXpert is the best available quick diagnostic test since it simultaneously detects M. tuberculosis and the RIF resistance gene. The negative predictive value and accuracy of GeneXpert were found to be superior to "acid-fast bacilli" smear microscopy [18]. The "Xpert MTB/XDR", the other recently developed molecular diagnostic assay, provides high accuracy for the diagnosis of INH and FLQ resistance and can support the choice of the appropriate treatment regimen for TB patients [19]. Similarly, " a broad-range PCR coupled with electrospray ionization mass spectrometry" offers an alternative to existing diagnostic techniques for the rapid detection of genetic markers in INH and RIF resistance in M. tuberculosis strains [20].
Following the recent WHO recommendation to use GeneXpert and line probe assays, which are effective and rapid molecular testing technologies, as part of an initial TB diagnostic test [3], there has been a steady increase in the reporting of high numbers of detected TB and DR-TB cases in developing countries, including Ethiopia [3,6,13,21]. Ethiopia's Federal Ministry of Health has rolled out the GeneXpert ® MTB/RIF ("Cepheid, Sunnyvale, CA, USA") molecular diagnostic assay into the country's current national TB diagnostic system, with line probe assay performed in each regional referral laboratory center, resulting in increased TB case detection, early identification of DR-TB isolates, and the provision of effective and appropriate therapy of TB patients [3,22]. Thus, the rate of TB and DR-TB case detection at the national level is improving [1,23]. Several previous systematic reviews and meta-analysis studies undertaken in Ethiopia have estimated the pooled prevalence of DR-TB using phenotypic DST data or data derived from both phenotypic and molecular DST findings [24][25][26][27]. However, both phenotypic and molecular DST methods have different degrees of sensitivity and specificity, and many factors can affect the accuracy of DST performance. Given the paucity of data on the weighted pooled prevalence of anti-tuberculosis drug resistance using molecular diagnostic laboratory results, the main aim of this meta-analysis is to provide an up-to-date data analysis on the weighted pooled prevalence of any anti-TB drug-resistance, any isoniazid (INH) and rifampicin (RIF) resistance, monoresistance to INH and RIF, and MDR-TB, in newly diagnosed TB cases and retreated patients with TB in Ethiopia. This systematic review and meta-analysis provides relevant data to better understand the magnitude of DR-TB and helps the country to pursue evidence-based measures for DR-TB control and to establish robust and routine laboratory-based DR-TB surveillance using affordable and rapid molecular diagnostic methods.

Study Protocol
The search strategy for potential articles, screening them by title and abstract, and evaluating their eligibility to be included in the final analysis, was conducted using the "Preferred Reporting Items for Systematic Reviews and Meta-Analysis" Protocol [28], (Supplementary Table S1). This review protocol has been submitted to and registered in the "International Prospective Register of Systematic Reviews" (ID#: CRD42020176713) (https: //www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=176713) (accessed on 12 March 2022).

Selection Criteria and Data Extraction
The records found through database searching were merged and the duplicates were removed using EndNote X7 ("Thomson Reuters, New York, NY, USA"). Three reviewers (screeners 1, 2, and 3) screened all titles and abstracts independently and excluded irrelevant data, then independently assessed the remaining articles for inclusion. The information extracted from each study included (a) author's name, (b) publication year, (c) study period, (d) study area, (e) types of patients with TB (pulmonary TB and extra-pulmonary TB), (f) study design, (g) molecular diagnostic methods (LPA, sequencing, and other PCRbased), (h) type of TB cases (newly diagnosed and retreated), (i) the number of patients, (j) total TB positive cases, (k) total patients/isolates with available DST results, (l) any drug-resistance, (m) monoresistance, and (n) MDR-TB cases in newly diagnosed TB cases and retreated patients with TB ( Table 1).
Studies that addressed any of the following criteria were included: (a) studies that used WHO-approved molecular DST methods; (b) studies that reported data on the prevalence of drug-resistance in retreated patients with TB or newly diagnosed TB cases [among both pulmonary TB (PTB) and extra-pulmonary TB (EPTB) patients]; (c) studies reporting the prevalence of any anti-tuberculosis drug-resistance, monoresistance, or MDR/or extensively drug-resistant-TB (XDR-TB); (d) studies conducted in Ethiopia and published in the English language. Whereas studies were excluded if they met the following conditions: (a) studies with non-tuberculous mycobacteria data; (b) studies that did not perform DST of first-and second-line anti-TB drugs; (c) studies that only used phenotypic DST methods to detect DR-TB. This study also excluded the following: editorial papers, narrative reviews, meta-analysis and/or systematic reviews, conference abstracts, and case reports. Qualitative studies and citations without full text were also excluded.

Quality Assessment
Two writers (MAR and BAT) independently evaluated the quality of the included studies using an updated version of the tool proposed by the "Joanna Briggs Institute (JBI)" [29]. A third author (BBA) acted as an arbiter and adjudicated in any cases where there was disagreement. Studies (case-control, cross-sectional, and cohort) with an average score of four or higher were considered high quality and included, whereas studies with an average score of three or lower were considered low quality and excluded (Supplementary  Table S3).

Definitions
The following standard definitions of anti-TB drug resistance were used [2,30].

Statistical Analysis
Important data from included studies were recorded/extracted using a standard data extraction format prepared in a Microsoft Excel spreadsheet and then exported to Stata/SE software (version 11, "StataCorp, College Station, TX, USA") for final analysis. A randomeffects model was used in the meta-analysis due to the heterogeneity of the studies. The heterogeneity of included studies was evaluated using Cochran's Q test and the I 2 index. The presence of studies' publication bias was assessed utilizing Begg's and Egger's tests, and funnel plots of the standard error of the Logit event rate were provided to show the presence of publication bias. All statistical results and interpretations were reported with a 95% confidence interval (CI) basis. A statistical test with a p < 0.05 level of significance was considered statistically significant.

Study Outcomes
The main outcomes of the review were the weighted pooled prevalence of any antituberculosis drug resistance, any resistance to INH and RIF, monoresistance to INH and RIF, and MDR-TB among newly diagnosed cases, retreated patients with TB, and overall TB cases. Figure 1 shows a total of 1524 relevant studies recorded from searched electronic databases. Of the total, 1328 studies were non-duplicated and subjected to further evaluation; 1227 were assessed and excluded based on their title, abstract, and for other reasons (review, non-English papers, outdated, etc.), while 101 papers were retained for full-text review. After full-text evaluation, the final analysis (meta-analysis) included 24 articles that reported on drug-resistant M. tuberculosis isolates.

Statistical Analysis
Important data from included studies were recorded/extracted using a standard data extraction format prepared in a Microsoft Excel spreadsheet and then exported to Stata/SE software (version 11, "StataCorp, College Station, TX, USA") for final analysis. A randomeffects model was used in the meta-analysis due to the heterogeneity of the studies. The heterogeneity of included studies was evaluated using Cochran's Q test and the I 2 index. The presence of studies' publication bias was assessed utilizing Begg's and Egger's tests, and funnel plots of the standard error of the Logit event rate were provided to show the presence of publication bias. All statistical results and interpretations were reported with a 95% confidence interval (CI) basis. A statistical test with a p < 0.05 level of significance was considered statistically significant.

Study Outcomes
The main outcomes of the review were the weighted pooled prevalence of any antituberculosis drug resistance, any resistance to INH and RIF, monoresistance to INH and RIF, and MDR-TB among newly diagnosed cases, retreated patients with TB, and overall TB cases. Figure 1 shows a total of 1524 relevant studies recorded from searched electronic databases. Of the total, 1328 studies were non-duplicated and subjected to further evaluation; 1227 were assessed and excluded based on their title, abstract, and for other reasons (review, non-English papers, outdated, etc.), while 101 papers were retained for full-text review. After full-text evaluation, the final analysis (meta-analysis) included 24 articles that reported on drug-resistant M. tuberculosis isolates.

Characteristics of Included Studies
Twenty-four potential studies with a total of 18,908 patients with TB (PTB, n = 16,223; EPTB, n = 958; and PTB + EPTB, n = 1727) were included in the final analysis . The types of patients with TB (new and retreated cases) were 3844 and 3401, respectively, while few studies did not mention the types of patients with TB. However, the rate of drug resistance was evaluated in 4101 M. tuberculosis isolates with available DST results in the included studies. In total, 911 M. tuberculosis isolates resistant to any anti-TB drug were identified in those included studies. The rate of MDR-TB was estimated using data from 16 studies comprising a total of 2818 M. tuberculosis isolates with complete DST profiles [31][32][33][34][35]37,40,44,[46][47][48][49][51][52][53][54]. On the other hand, a total of 2818 M. tuberculosis isolates with complete DST results reported in 16 studies were examined for the occurrence of any INH resistance [31][32][33][34][35]37,40,44,[46][47][48][49][51][52][53][54]. Most of the studies included in this review evaluated the occurrence of any RIF resistance, and we estimated the weighted pooled prevalence of any RIF resistance among a total of 4086 M. tuberculosis isolates [31,.

Prevalence of Any Anti-TB Drug Resistance
Any anti-tuberculosis drug resistance was reported in 24 studies, with a weighted pooled prevalence of 14.25% (95% CI: 7.05-21.44%) among 4101 M. tuberculosis isolates with available DST results; a substantial heterogeneity was (I 2 = 98.5%) (Figure 2). The funnel plot, as well as Egger's test result, which indicates a publication bias, and the sensitivity analysis of the included papers, are all presented in a supplementary file (Supplementary Figures S1A-E). Seven of the included studies reported the highest proportion of any anti-TB drug resistance, ranging from 15.80% to 71.10% [34,35,37,39,48,53,54], and three of them reported from the Amhara region [34,39,53]. The highest prevalence of any anti-TB drug resistance, with 71.10%, was reported by Abate et al. in central Ethiopia, Addis Ababa [54], while the lowest proportion, with 2.70%, was reported by Fanosie et al. in the Amhara region [44]. The high rate of any DR-TB observed by Abate et al. [54] could be due to the study's inclusion of exclusively retreated patients with TB. Hence, the occurrence of DR-TB is anticipated to be prominent among retreated patients with TB (Figure 2).

Prevalence of Any INH Resistance
Any INH resistance was reported in 16 studies, with a weighted pooled prevalence of 15.62% (95%CI: 6.77-24.47%; I 2 = 98.8%) among a total of 2818 M. tuberculosis isolates (Figure 3). The funnel plot, as well as Egger's test result, which indicates a publication bias, and the sensitivity analysis of the included papers, are all presented in the supplemental file ( Supplementary Figures S2A-D). Geographically, the highest proportion of any INH resistance was recorded in seven studies, with rates ranging from 13.33% to 65.35% [34,35,37,47,48,53,54]; three of the studies were conducted in central Ethiopia, Addis Ababa [37,47,54], while two studies were conducted in the Amhara region [34,53]. The highest prevalence of any-INH resistance, with 65.35%, was reported in Addis Ababa [54]. This may be because all patients with TB in the study done by Abate et al. (2014) had been previously treated, which increases the likelihood of developing any INH resistance.  Figures S2A-S2D). Geographically, the highest proportion of any INH resistance was recorded in seven studies, with rates ranging from 13.33% to 65.35% [34,35,37,47,48,53,54]; three of the studies were conducted in central Ethiopia, Addis Ababa [37,47,54], while two studies were conducted in the Amhara region [34,53]. The highest prevalence of any-INH resistance, with 65.35%, was reported in Addis Ababa [54]. This may be because all patients with TB in the study done by Abate

Prevalence of Any RIF Resistance
In this meta-analysis, almost all included studies [31,33-54] reported any-RIF-resistant M. tuberculosis isolates ( Table 1). The weighted pooled prevalence of any RIF resistance was 9.75% (95% CI: 4.69-14.82%; I 2 = 98.2%) among a total of 4086 M. tuberculosis isolates with available DST results (Figure 4). The funnel plot, as well as Egger's test result, which indicates a publication bias, and the sensitivity analysis of the included papers, are all presented in a supplementary file (Supplementary Figures S3A-S3D). Geographically,

Prevalence of Any RIF Resistance
In this meta-analysis, almost all included studies [31, reported any-RIF-resistant M. tuberculosis isolates (Table 1). The weighted pooled prevalence of any RIF resistance was 9.75% (95% CI: 4.69-14.82%; I 2 = 98.2%) among a total of 4086 M. tuberculosis isolates with available DST results (Figure 4). The funnel plot, as well as Egger's test result, which indicates a publication bias, and the sensitivity analysis of the included papers, are all presented in a supplementary file (Supplementary Figures S3A-D). Geographically, the prevalence of any RIF resistance varied. Two studies [35,54] reported the highest prevalence of any RIF resistance. The highest incidence of any RIF resistance, with 65.35%, was reported by Abate et al. [54] in central Ethiopia, Addis Ababa. This might be because the study only included retreated patients with TB, and it is anticipated that the incidence of DR-TB is high among patients previously treated with TB [54]. The lowest rate of any-RIF resistance, with 1.09%, was reported by Haile et al. [40] in the Aris zone, Oromia Region (Figure 4).

Discussion
The increasing incidence of DR-TB strain is one of the most serious threats to global TB control efforts [1,15], particularly in resource-constrained countries, where innovative molecular diagnostic technologies and well-equipped laboratory settings are lacking [3,6,7]. In particular, the emergency and spread of MDR/XDR-TB strains in both developing and developed countries need substantial global efforts, resources, and breakthrough diagnostic technology to halt transmission and control the disease [1,55]. Ethiopia, like other LMICs, is experiencing an increase in DR-TB, posing challenges for its public healthcare system and national TB-control efforts [3]. Ethiopia is one of the top 30 countries with the highest rates of TB, DR-TB, and TB/HIV co-infection [3], and MDR-TB was found in 0.71% of newly diagnosed cases and 12.0% of patients previously treated with TB in 2019 [1]. To identify the drug-resistance pattern of M. tuberculosis isolates and to ensure effective therapy for patients with TB, all TB-positive individuals must undergo DST for first-and second-line anti-TB drugs [1]. Monitoring drug resistance is critical for developing TB control policies and strategies, as treating drug resistance requires the use of more complicated, expensive, toxic, and less effective second-line antibiotics [1].
In this review, we estimated the weighted pooled prevalence of any anti-TB drug resistance, any INH and RIF resistance, monoresistance to INH and RIF, and MDR-TB in newly diagnosed cases and retreated patients with TB in Ethiopia using molecular DST results from 24 studies included. The pooled prevalence of any anti-TB drug resistance (14.25%) found in our review was lower than Benin's national survey report, which found

Discussion
The increasing incidence of DR-TB strain is one of the most serious threats to global TB control efforts [1,15], particularly in resource-constrained countries, where innovative molecular diagnostic technologies and well-equipped laboratory settings are lacking [3,6,7]. In particular, the emergency and spread of MDR/XDR-TB strains in both developing and developed countries need substantial global efforts, resources, and breakthrough diagnostic technology to halt transmission and control the disease [1,55]. Ethiopia, like other LMICs, is experiencing an increase in DR-TB, posing challenges for its public healthcare system and national TB-control efforts [3]. Ethiopia is one of the top 30 countries with the highest rates of TB, DR-TB, and TB/HIV co-infection [3], and MDR-TB was found in 0.71% of newly diagnosed cases and 12.0% of patients previously treated with TB in 2019 [1]. To identify the drug-resistance pattern of M. tuberculosis isolates and to ensure effective therapy for patients with TB, all TB-positive individuals must undergo DST for first-and second-line anti-TB drugs [1]. Monitoring drug resistance is critical for developing TB control policies and strategies, as treating drug resistance requires the use of more complicated, expensive, toxic, and less effective second-line antibiotics [1].
In this review, we estimated the weighted pooled prevalence of any anti-TB drug resistance, any INH and RIF resistance, monoresistance to INH and RIF, and MDR-TB in newly diagnosed cases and retreated patients with TB in Ethiopia using molecular DST results from 24 studies included. The pooled prevalence of any anti-TB drug resistance (14.25%) found in our review was lower than Benin's national survey report, which found a rate of 54% [56], as well as the pooled estimated rates in China (20.1%) [30], and Burundi (16.1%) [57]. However, it was higher than the national survey results in Mozambique (11.4%) [58], Rwanda (12.0%) [59], and the weighted pooled prevalence estimate in Nigeria (12.0%) [6]. This discrepancy may be due to the types of studies used. Our review included only studies that used molecular DST methods, whereas Benin and Rwanda's national surveys included phenotypic DST results. The other possible reason could be due to study conditions, especially laboratory setups, the sensitivity of drug-resistance diagnostic methods, and the skill of laboratory staff.
INH is a selective prodrug used to treat active TB in the first stage of the disease, and it is commonly given together with RIF, pyrazinamide, and either streptomycin or ethambutol [54]. INH resistance must be assessed regularly because it lowers the chances of TB-treatment success, raises the risk of developing MDR-TB, and decreases the efficacy of INH preventive therapy [60]. The weighted pooled prevalence of any INH resistance found in our review (15.62%) was lower than the 27.9% reported in a national survey in Benin [56], but higher than the 7.9% reported in a national survey in Mozambique [58], Burundi (6.3%) [57], and the pooled prevalence estimated in China (12.0%) [30]. In agreement with our result, one previous study conducted using WHO data (1994WHO data ( -2009) estimated the INH resistance burden and found that INH resistance was considerably higher in the Eastern European region than in any other WHO region, at 44.9%; of those, 33.5% were among new cases and 61.4% were among patients retreated with TB [60]. According to the same report, INH resistance in Africa was found to be 6.3% among newly diagnosed cases and 20% among retreated patients with TB [60].
Despite its prevalence, INH monoresistance has gained less attention than RIF resistance until recently, since INH resistance is more difficult to diagnose using molecular testing and the clinical consequences of INH resistance are unknown [61]. A DST for INH-monoresistance could be conducted on patients with TB, as INH-monoresistance increases the chance of a poor treatment outcome and progression to MDR-TB if not addressed properly [60]. INH-monoresistance is the most common form of DR-TB globally, with estimates rising to 7.0% among newly diagnosed TB cases and 8.0% to 11.0% among previously treated TB cases [62]. From 1994 to 2009, monoresistance to INH was estimated to occur in between 6.4% and 33.5% of new TB cases [60]. The pooled prevalence of INH monoresistance in our review (6.23%) was lower than the 9.3% reported in Benin [56], and Nigeria (11.0%) [6]. However, it was higher than the 1.1% reported in Rwanda [59], and 2.3% in Mozambique [58]. Notably, our review demonstrated that INH-monoresistant TB is prevalent in Ethiopia. The high pooled prevalence of INH monoresistance observed in this review may be due to the use of INH as empiric treatment for HIV-positive and TB-exposed individuals in Ethiopia. Also, further study is required to better comprehend the prevalence and phenotypic and molecular resistance features of INH-monoresistant M.tuberculosis strains in Ethiopia.
RIF is a highly effective sterilizing agent against resistant M. tuberculosis isolates. Resistance to RIF is most commonly conferred by mutations in the rpoB gene, which codes for the RNA polymerase ß-subunit [63]. Besides direct transmission, RIF resistance is acquired by the selection of mutant strains during anti-TB treatment, which is usually the result of improper medication, stock-outs, poor adherence, and/or drug penetration [63]. RIF resistance is the basis of both MDR and XDR-TB, and the key indicator for MDR-TB, and makes patient treatment more difficult [64,65]. In our review, we found that the weighted pooled prevalence of any RIF resistance was 9.75%, which was lower than the reported rate in Benin [56]. Similarly, our result was slightly lower than the prevalence of any RIF-resistant M. tuberculosis reported in a meta-analysis study (11.0%) [66]. According to the same study, the African region had a 3.0% prevalence of RIF resistance, while the Western Pacific, European, South-East Asian, and American regions had 23.0%, 10.0%, 6.0%, and 1.0%, respectively [66]. In contrast, our finding was higher than the report in Rwanda, where the rate was 4.0%, with 3.4% in new cases and 19.6% in previously treated patients with TB [59].
Currently, RIF-resistant TB is incorrectly classified as MDR-TB; however, this method could lead to incredibly long and toxic anti-TB drug treatment regimens for patients with TB with RIF-monoresistant TB strains [67]. The availability of diagnostic tests that can detect RIF resistance quickly has increased awareness of the presence of patients with RIF-monoresistant TB, which was previously thought to be uncommon [67]. In our study, we found that the pooled prevalence of RIF monoresistance was 2.33%, which was comparable to the rate of RIF monoresistance reported in Benin (2.2%) [56]. According to Weldegebreal and colleagues, Ethiopia has a prevalence of RIF monoresistance ranging from 0.0% to 2.2% [27]. In contrast, our review result was higher than those reported in other African countries, e.g., Rwanda (0.1%) [59] and Mozambique (0.4%) [58]. Interestingly, our study highlighted that RIF-monoresistance is widespread in Ethiopia, resulting in the misclassification of many patients with RIF-monoresistant TB as MDR-TB patients and their exposure to more toxic, ineffective second-line treatment. Hence, it is crucial to accurately detect RIF-monoresistant TB strains using proper diagnostic methods in order to appropriately treat patients with TB and achieve a good treatment outcome.
A more worrisome feature of TB is the spread of MDR-TB strains. MDR-TB is a form of TB that is difficult and expensive to treat since it is resistant to two important first-line antibiotics, RIF and INH [1]. MDR-TB is an emerging threat that has always played a significant role in the prevention and control of infectious diseases [1]. In our study, we found that MDR-TB had a weighted pooled prevalence of 2.64% among newly diagnosed cases. Our result was consistent with recent reports in Ethiopia by Girum et al. [24] and Eshetie et al. [25], who reported 2.2% and 2.0%, respectively. Similarly, according to the national report, MDR-TB was detected in 2.7% of newly diagnosed patients in Ethiopia [68]. In agreement with this, the first Ethiopian national TB survey found that MDR-TB prevalence was 2.7% [69]. However, our result was marginally higher than a meta-analysis finding in Sub-Saharan African countries (1.5%) [26], and the national survey reports of Ethiopia [3] and Benin [56]. Nevertheless, our result was lower than the reports from elsewhere: Mozambique [58], Rwanda [59], and Nigeria [6].
On the other hand, the weighted pooled prevalence of MDR-TB among previously treated patients with TB (11.54%) in our review was lower than the reports by Eshetie et al.  [24], as well as the WHO (2019) national survey report estimated for Ethiopia (16%) [3]. Similarly, it was lower than a Nigerian meta-analysis study, which found a pooled prevalence of 19.0% [6]. It was, however, similar with the survey results from Benin (11.1%) [56], Mozambique (11.2%) [58], and Rwanda (9.4%) [59]. While previously treated patients with TB are more likely to contract a new infection and develop resistance by gene mutation while on medication, their risk of developing MDR-TB is higher than that of newly diagnosed patients with TB [1, 24,25]. On the other hand, several previous studies have demonstrated that MDR is associated with specific M. tuberculosis genotypes that have effectively adapted to a particular geographical area and are capable of spreading to the entire population, like the Beijing genotype [66]. Thus, it is necessary to define the genetic diversity of M. tuberculosis in a specific geographical area and to analyze its drug-resistance profile to provide effective TB-control measures.
The results of this review may be hampered by the fact that the study was undertaken in TB treatment facilities, which means that its generalizability for the general population is limited. Because the review was confined to studies published in English and within the given data sources, as well as within a few geographical areas, generalization for the entire country may be challenging. Owing to the facility-based nature of the primary study, presumed TB cases may be included, thereby increasing the prevalent estimate.
In conclusion, DR-TB continues to be a major public health problem in Ethiopia. MDR-TB was found to be marginally lower among retreated cases and in overall TB patients compared to the WHO national report (2020); however, MDR-TB among newly diagnosed TB cases remained high. Besides, this review found that monoresistance to INH and RIF is higher than in earlier studies in Ethiopia. Early TB-case detection, proper treatment of both drug-susceptible and DR-TB, and a strict TB-patient follow-up strategy are all important.
Rapid and effective molecular diagnostic methods for both drug-susceptible and DR-TB are also critical for the early detection and proper treatment of patients with TB.  Table S1: PRISMA-2020-checklist; Table S2: Search Strategy Medline/PubMed; Table S3: Quality assessments.

Data Availability Statement:
The datasets analyzed during this review can be accessed from the corresponding author upon reasonable request.