Prevalence of Latent Tuberculosis Infection in the Middle East and North Africa: A Systematic Review

Objective Data on the prevalence of latent tuberculosis infection (LTBI) in Middle Eastern and North African countries are scarce. We aimed to review all relevant published data in countries belonging to this region to determine the overall prevalence of LTBI in the Middle East and North Africa (MENA) region. Methods In this systematic review PubMed and Google Scholar databases were searched for observational, prospective, retrospective, cross-sectional, and cohort studies providing prevalence data of LTBI in any MENA country. Studies fulfilling the search criteria were incorporated in the review. Overall prevalence of LTBI with 95% confidence intervals (CI) was calculated using the random-effects model; heterogeneity was assessed using I2 statistics. Gender and age group-based subgroup analyses were performed to evaluate the basis of heterogeneity. Results The total number of overall LTBI studies identified was 956, of which 31 studies from ten countries within the MENA region were included that represented 12,439 subjects. The overall prevalence was 41.78% (95% CI 31.18% to 52.78%, I2 = 99.31%). By gender-based subgroup analysis, the prevalence of LTBI was 33.12% (95% CI 18.97% to 49.04%, I2 = 99.25%) and 32.65% (95% CI 19.79% to 47%, I2 = 98.89%) in males and females, respectively, while in the age-based subgroup analysis, the prevalence of LTBI was 0.44% (95% CI -0.05% to 0.9%), 3.37% (95% CI 2.23% to 4.74%, I2 = 0%), and 43.81% (95% CI 33.09% to 54.82%, I2 = 99.18%) for children, adolescents, and adults, respectively. Conclusion This systematic review reveals a high prevalence of LTBI in the MENA region; enhanced LTBI surveillance and prompt infection prevention steps are urgently needed to prevent active tuberculosis, this would help achieve the World Health Organization End TB Strategy 2035, and the United Nations Sustainable Development Goals 2030 target in the MENA region.


Introduction
Tuberculosis (TB) is a major health problem, with an estimated 10 million people (range 9 to 11.1 million) developing TB disease in 2018, of which 5.8 million, 3.2 million, and 1 million were men, women, and children, respectively. Twothirds of cases were from eight countries, India (27%), China (9%), Indonesia (8%), Philippines (6%), Pakistan (5%), Nigeria (4%), Bangladesh (4%), and South Africa (3%) [1]. Latent tuberculosis infection (LTBI) does not induce infectious expression of the disease, although it causes continuous immune response generated towards TB antigens. LTBI has a 10% probability of progressing into active TB disease, 5% during the first two years of acquiring the infection, and 5% during the rest of the individual's lifetime. The detection of LTBI and prevention before it becomes infectious is a crucial component of the WHO-End TB strategy. It has been reported from mathematical models that approximately 30% of the population worldwide are LTBI carriers [2]. Previous studies have documented the rates of LTBI to be 31.2% in Ethiopia [3], 49% in Uganda [4], 55.2% in South Africa [5], 11.2% in Spain [6], 50% in India [7], 51% in Korea [8], and 7.6% in England [9];however, very few studies have been undertaken to estimate the prevalence of LTBI in the Middle East and North Africa (MENA) region.
In previous studies, it has been observed that patients belonging to lower socioeconomic groups, refugees, and migrants [10], patients with abnormal immune responses (post-organ transplant, hemodialysis patients, people living with HIV, etc.), and chronic inflammatory conditions have an increased risk of acquiring TB and its progression to active disease [11][12][13];further, LTBI in people living with HIV has a 10% probability of progressing into active TB, when left untreated, annually; furthermore, it has been shown that a significant geographical variation in TB infection rates persists across the world, implying that health care workers (HCW) in various countries encounter different risks of acquiring TB [14]. In 2018, 87% of new TB cases occurred in the top thirty high TB burden countries, of which eight countries accounted for two-thirds of all new TB cases, they include India, China, Indonesia, Philippines, Pakistan, Nigeria, Bangladesh, and South Africa, while the occurrence was extremely low in the MENA regions [1], it has also been reported that HCW are at particular risk of LTBI, and hence, annual screening is performed in most standardized health care facilities. In addition, the prevalence of LTBI in HCW has been reported to be higher than that of other community groups around the world [15,16].
Currently, the direct diagnosis of LTBI is not fully possible [17]. The diagnosis of memory T-cell response against LTBI is performed by either the tuberculin skin test (TST) or interferon-gamma release assays (IGRA) [18]. At present, no gold standard test has been developed to measure LTBI; however, there are increasing advancements in this field looking into tumor necrosis factor, chemokines, interleukin growth factors, and other factors that could enhance LTBI diagnosis [19]. With TST, TB-purified protein derivative (PPD) stimulates a type IV hypersensitivity-delayed type reaction [20][21][22], its advantage is that it is inexpensive and generally accepted especially in low economic countries including Africa [3], but has several disadvantages, as it has demonstrated poor response in individuals with reduced immunity and those with active TB, requires two-step verification, is operative dependent, and exhibits low specificity in determining reactivation of TB in Bacillus Calmette-Guérin (BCG) vaccinated individuals, it can also cause falsepositive results in patients sensitized to naturally existing nontuberculous mycobacteria [18,23].
On the other hand, IGRA has greater specificity compared to TST [17], it involves only one blood test after incubation with Mycobacteria tuberculosis-specific antigens, following which T-cell mediated immune response and interferon-(IFN-) gamma release are measured. The Quanti-FERON®-TB-Gold-in-Tube (QFT-GIT) and T-SPOT.TB assay tests are the two commercially available IGRA, in which the former is based on ELISA (enzyme-linked immunosorbent assay) and comprises of peptides from the ESAT-6, CFP-10, and TB7.7 antigens of TB. T-SPOT.TB assay is preferred in immunocompromised patients [24][25][26]. IGRA provides more conclusive results that would help in decision-making, with only a single visit required for the test, it also eliminates false-positive results in people vaccinated with BCG or sensitized with nontuberculous mycobacteria.
Several previous studies have documented the prevalence of LTBI in many countries of the Middle East and North Africa, in a wide range of population, including HCW, household contacts, people living with HIV, prisoners, refugees, and in patients with varied health problems; however, to our knowledge, there are no published studies that have assessed the overall prevalence within the whole MENA region; hence, we performed a systematic review to evaluate the prevalence of LTBI in the MENA region in different population groups belonging to various age groups.  The use of medical subject heading (MeSH) terms for LTBI was employed in the database search combined with the following search terms: (latent tuberculosis OR TB OR LTBI OR Mycobacterium tuberculosis) AND (Prevalence OR Epidemiology OR "Country name"). The Middle East countries included were Iran, Iraq, Saudi Arabia, Yemen, Syria, Jordan, United Arab Emirates, Israel, Lebanon, Oman, Kuwait, Qatar, Bahrain, Palestine, Cyprus, and Turkey. North African countries included were Egypt, Libya, Algeria, Morocco, Tunisia, Sudan, Western Sahara, and Mauritania. A broad search strategy was used to ensure that all relevant studies were identified, with no filters included in the searches. Following this, the author independently analyzed the title of the study and its abstract and keywords outlining the record, based on which studies were either included or excluded. No minimal sample size was required to be included in the analysis; however, a sample size of ≥200 was considered as adequate, and a sample size of <200 was considered as inadequate.

Study Selection and Data Extraction.
A detailed search of PubMed and Google Scholar databases by employing various search terms was performed. The duplicate citations were removed, and the studies for inclusion in the review were selected. The initial screening was based on the citation titles and abstracts, following which, the articles were selected and picked up and their complete text obtained, reviewed, and assessed for their eligibility for inclusion. The bibliographic information of the included studies was also screened to identify additional relevant articles for inclusion; furthermore, the data from relevant studies were abstracted using 2 Pulmonary Medicine a data extraction form, and the applicable items for the review were reported in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist.
The following key information has been presented in the data extraction template: first author, period of study and year of publication, country where the research was conducted, study design, number of participants, age at assessment, tools used for assessment, and key findings.

Quality (Risk of Bias)
Assessment. The Mirza and Jenkins [27] checklist were referred to for investigating the quality of included studies. The assessment was based on the following nine criteria: clear study aims, adequate sample size, representative sample, inclusion and exclusion criteria, adequate assessment of outcome, response rate reported, adequate description of data, appropriate statistical analysis, and appropriate informed consent obtained. A final total score was calculated for each of the criteria, scored 0 if absent and 1 if present. Thus, the minimum and maximum obtainable scores would be 0 and 9, respectively.

Statistical Analysis.
Analysis was performed using STATA software. The effect sizes were reported as proportions with 95% confidence intervals. The heterogeneity of effects was assessed and quantified by the I 2 . The I 2 values greater than 50% were considered to represent substantial heterogeneity. The random-effects model was subjected in cases exhibiting substantial heterogeneity. Subgroup analysis based on sex (male and female), by age strata, and by quality score of the studies (<5 and ≥5) was also performed. A p value less than 0.05 was considered statistically significant for all the analyses undertaken.

Search Results and Study
Selection. The database search resulted in a total of 956 citations, of which 384 citations were eliminated due to their duplication, and the rest of the 572 citations were examined. After screening, examination of titles and abstracts resulted in the elimination of 362 citations from the study. Following this, 210 full-text citations were retrieved, and after subjecting them to inclusion and exclusion criteria, a total of 31 studies were identified ( Figure 1).

Study Characteristics.
Thirty-one studies representing 12,439 subjects from ten countries within the MENA region were included: thirteen from Turkey, five from both Iran and Saudi Arabia, two from Egypt, and one each from Syria, Israel, Oman, Qatar, Tunisia, and United Arab Emirates. These studies were conducted between 2005 till 2018. The sample size ranged from 34 to 2,650 (Table 1).

Publication Bias.
From the 31 studies, the minimal checklist score was 5 in two studies, while the highest was 9. Details of all included studies clarity, adequacy of sample size, and other details are outlined in Table 2.        The subgroup analyses revealed the existence of heterogeneity. In the gender-based subgroup analysis, some of the studies failed to mention the gender-based prevalence of LTBI, and hence 14 and 15 studies were excluded from the subgroup analysis of males and females, respectively; hence, the subgroup analysis of males was performed with 17 studies, and that of females with 16 studies. The analysis revealed that the proportion of LTBI ranged from 0.32% to 86.04% and from 0.54% to 90.90% in males and females, respectively. The overall prevalence was estimated to be 33.12% (95% CI 18.97% to 49.04%, I 2 = 99:25%) and 32.65% (95% CI 19.79% to 47%, I 2 = 98:89%) in males and females, respectively.
For the evaluation of age-based prevalence, the WHO classification for age groups was utilized, and the age range for children, adolescents, and adults was taken as <10 years, between 10 and 19 years, and >19 years, respectively; further, three studies, Shitrit et al. [28], Yilmaz et al. [29], and Jam et al. [30], were excluded from this subgroup analysis as the age of subjects in those studies overlapped the age range for children, adolescents, and adults, i.e., 12 years and above, 13 to 67 years, and 1 month to above 60 years, respectively. Moreover, there was no differentiation in the age range for the prevalence of LTBI in these studies; hence, the subgroup analysis of children, adolescents, and adults was performed with 1, 2, and 27 studies, respectively. The prevalence of LTBI in children was observed to be 0.44% (95% CI -0.05% to 0.9%); the prevalence of LTBI in adolescents and adults ranged from 2.46% to 3.55% and 6.93% to 88.15%, respectively. The overall prevalence was observed to be 3.37% (95% CI 2.23% to 4.74%, I 2 = 0%) and 43.81% (95% CI 33.09% to 54.82%, I 2 = 99:18%) for adolescents and adults, respectively.

Discussion
After screening 956 studies, a total of 31 scientific papers from ten countries within the MENA region were included in this systematic review . The subjects included in these studies were healthcare workers, laboratory staff, medical school students, people living with HIV, and patients with chronic inflammatory diseases. The detection of LTBI in these studies was performed by TST or IGRA or both; furthermore, the studies covered the incidence of LTBI among populations belonging to varying age groups, including children, adolescents, and adults.
In the present study, LTBI prevalence was evaluated by employing the random effects model since high heterogeneity was encountered among studies. The existence of high heterogeneity may have possibly been due to variations in study settings, subjects or participants, methodologies involved, exposure to TB patients, and the control measures taken across the studies.
The overall prevalence of LTBI in the MENA region was found to be 41.78%. In the gender-based subgroup analyses, the prevalence of LTBI was found to be 33.12% and 32.65% in males and females, respectively. As for the age-based prevalence, it was assessed to be 0.44%, 3.37%, and 43.81% in children, adolescents, and adults, respectively; therefore, this systematic review implies a high prevalence of LTBI in the MENA region irrespective of gender, and in order to achieve the WHO End TB 2035 objective, there is an immediate need to scale up measures to stop TB disease and enhance LTBI detection within the MENA region.
There are some strengths and limitations within this study that needs to be highlighted; first, as per our findings, this is the first systematic review on the epidemiology and prevalence of LTBI in the MENA region. As for limitations, studies published in English alone have been included, therefore, other reports from countries with high TB incidence that are published in native or other languages other than English, in national or local journals, have not been included; additionally, studies published in journals indexed in PubMed and Google Scholar were included, while other studies may exist that were published in other indexing databases.
To conclude, this review indicates a high prevalence of LTBI in the MENA region despite the high heterogeneity observed. Future studies should aim towards more rigorous assessment of LTBI prevalence within the MENA region to reach exact estimates as the first important step to hamper TB disease diffusion in these countries.

Data Availability
All data are included in the manuscript.

Conflicts of Interest
The authors declare that they have no conflicts of interest.