Laboratory diagnostic systems used in the diagnosis of tuberculosis in Ethiopia: A systematic review

s reviewed


INTRODUCTION
Tuberculosis (TB) is one of the world's most important causes of morbidity and mortality among adults.It is caused by the bacterium Mycobacterium tuberculosis most commonly affects the lungs (pulmonary TB), but may affect any organ or tissue outside of the lungs (extra pulmonary TB).When TB is detected and effectively treated, the disease is largely curable.One of the largest challenges in preventing morbidity and mortality from TB is the difficulty in making a timely diagnosis.Diagnostic approaches relying on symptoms, chest radiographs, tuberculin skin tests, microscopy or cultures, all have particular challenges (World Health Organization (WHO), 2011, 2008).
Early, accurate and rapid diagnosis of TB is critical for reducing TB transmission and incidence.Even though World Health Organization (WHO) has endorsed a novel, rapid and automated diagnostic systems that are simple enough to be run in basic laboratories and clinics outside of a reference laboratory setting in 2010, smear microscopy is the sole method used for TB diagnosis in most laboratories in developing countries, where over 95% of TB-related deaths occur.In patients with active pulmonary TB, only an estimated 45% of infections are detected by sputum microscopy.In most cases, mycobacterial culture is generally considered to be the best available reference standard diagnostic test for TB diagnosis and is the most important method first step in detecting drug resistance (WHO, 2011).
TB remains the leading cause of morbidity and mortality in developing countries including Ethiopia, despite the availability of short-course therapy that can be both inexpensive and effective.Early and proper diagnoses are essential for effective tuberculosis control programs; to improve treatment, to reduce transmission and to control development of drug resistance.Laboratories and laboratory networks are fundamental components of tuberculosis control, providing testing for diagnosis, surveillance and treatment monitoring at every level of the health-care system (Ridderhof et al., 2007;Steingart et al., 2007) Care of patients with tuberculosis starts with a quality assured diagnosis.Arguably, the weakest component of health systems is laboratory services, which have been grossly neglected, understaffed and underfunded over time.Low-income and middle-income countries, which bear most of the global burden of tuberculosis, rely heavily on outdated tuberculosis diagnostic tests; including sputum smear microscopy, solid culture, and chest radiography (WHO, 2011).
In high-incidence countries, TB control relies on passive case finding among individuals self-presenting to health care facilities, followed by either diagnosis based on clinical symptoms or laboratory diagnosis using sputum smear microscopy.Serial sputum specimens are required (one taken on the spot and the second brought in the following morning), which means that the people are asked to make repeated visits to the health care centers for specimen delivery and collection of results.For many patients, the costs of repeated visits to health care facilities are prohibitive, and patient dropout is a significant problem (Steingart et al., 2007).
Many smear microscopy laboratories are single room and understaffed with poorly maintained microscopes, and some of these laboratories lack consistent sources of electricity and clean water.There are few opportunities for the training of staff and little staff capacity to handle high-volume workloads.Quality assurance programs including quality control and external quality assessments are often lacking.Thus, there is a critical need for new, sensitive, easy, and rapid point-of-care diagnostics and also for investments in laboratory infrastructure, quality assurance programs and well-trained staff (Parsons et al., 2011).
Ethiopia is ranked 7th among the 22 countries with a high-burden of TB and third in Africa (WHO, 2009) but almost all TB laboratories in Ethiopia have only been equipped with the acid-fast staining and lack resources for culture, identification and drug susceptibility testing of mycobacteria, which present a huge hindrance for tuberculosis control in the country.Culture and drug susceptibility testing for Mycobacterium tuberculosis are not performed routinely in clinical microbiology laboratories.Smear microscopy contributes little to the diagnosis of pediatric TB and does not, by definition, identify smearnegative TB which may account for 24 to 61% of all pulmonary cases in people living with HIV (Getahun et al., 2007;Tessema et al., 2012).
Sputum smear microscopy remains the cornerstone of TB diagnosis in developing countries.The method depends upon the quality and bacterial load of the sputum specimen and the training and motivation of laboratory technicians.Although highly specific in most countries, smear microscopy is insensitive -it detects roughly 50% of all the active cases of TB.Sensitivity can be as low as 20% in children and HIV infected people.Furthermore in addition, smear microscopy cannot detect bacterial resistance to anti-microbial drugs (World Health Organization, 2008).
Sputum smear examination for acid-fast bacilli (AFB) can diagnose up to 50 to 680% of cases of pulmonary tuberculosis in well-equipped laboratories (Aber et al., 1980).In low-income countries, poor access to highquality microscopy services contributes to even lower rates of AFB detection.Furthermore, in countries with high prevalence of both pulmonary tuberculosis and HIV infection, the detection rate is even lower owing to the paucibacillary nature of pulmonary tuberculosis in patients with HIV infection.In the absence of positive sputum smears for AFB, at primary care level, most cases of pulmonary tuberculosis are diagnosed on the basis of clinical and radiological indicators (Getahun et al., 2007).Smear microscopy also places a burden on the patient.A person with suspected TB typically has to visit the clinic at least twice before a diagnosis can be made, and then has to return again for the results.This process can be expensive in terms of transportation costs and lost wages from time off work.Some people never complete the testing process, or do not return for their results.Culture is a more sensitive method for TB diagnosis than smear microscopy and it permits testing for drug resistance but it has limitations and requires biosafety facilities that are expensive to build and maintain and specially trained laboratory technicians to perform the procedure (WHO, 2008).
Some national TB programmes in developing countries have no functioning TB culture facility at all.In others, TB culture is performed only at national reference laboratories or in hospital laboratories in large cities.Even where capacity exists, diagnosing TB with culture can take weeks because of the slow growth rate of TB bacilli.In most countries TB culture is reserved for retreatment cases.Specimens are often sent to distant laboratories.This can delay processing of specimens and lead to inaccurate results.Furthermore, test results must travel long distances back to reach the clinic and the patient (WHO, 2008).

METHODOLOGY
A systematic search of studies addressing the diagnosis of tuberculosis in Ethiopia was performed.All returned titles were reviewed and articles that obviously did not involve diagnosis or isolation of tuberculosis were excluded.The authors then reviewed abstracts of remaining articles to determine which studies examined diagnostic systems used in the diagnosis of tuberculosis.Bibliographies of relevant articles were also reviewed for potential articles.The two investigators independently reviewed the remainning articles, independently deciding on inclusion in the review using a predetermined eligibility criterion.Disagreements were resolved by consensus.For inclusion, the articles needed to describe a study involving the use of diagnostic system to diagnose tuberculosis.Only English language articles were included.Each article was analyzed to determine the study setting, study design, sample characteristics, type of diagnostic system used, reference or gold standard used for comparison, and findings of the diagnostic system.Duplicate publications of the same findings were excluded.Studies analyzing the diagnosis of tuberculosis without actually performing the diagnosis or isolating tuberculosis from study subjects were also excluded from the review.

Search strategy
We searched MEDLINE and EMBASE database for reports published in English up to the end of July, 2013.The first search was done on 20th April, 2013 and repeated on 7th July, 2013.The searches yielded 267 citations of which 38 were duplicate papers, 229 were subject to title and abstract review, 76 were subject to full text review and 33 articles were included in the final review.An over inclusive search strategy was used to ensure that no papers were missed.The key search words used were: "Tuberculosis or TB, "Mycobacterium tuberculosis or MTB", "diagnosis or screening or isolation or tests or assays", "and Ethiopia".We also reviewed references of the selected papers to ensure that no papers were missed.We have also searched the website of the STOP TB Partnership's New Diagnostic Working Group.We reviewed studies cited by articles identified by this search strategy and selected those we identified as relevant.The researcher reviewed titles and abstract in duplicate to exclude ineligible articles.Papers meting the inclusion criteria were subject to full-text review (Figure 1).

Inclusion and exclusion criteria
To be eligible, articles needed were: peer-reviewed publications, published in English, conducted in Ethiopia, conducted after 2000, focused on diagnosis or isolation of tuberculosis.Use of clear screening and diagnostic methods and algorithms in the diagnosis of TB was adopted and a survey methodology for data collection applied.

Data extraction and classification
The following data was extracted and summarized in evidence tables: Citation, year of publication, settings, study design, study population, type of diagnosis and outcomes.

RESULTS
The online search of MEDLINE and EMBASE yielded 267 articles.Additional potential studies were identified through searches of bibliographies.After articles that did not address the diagnosis of tuberculosis in Ethiopia were excluded, 229 articles remained.Further, 153 articles were excluded upon closer review because they did not include a diagnosis system or focused on screening of drug resistant tuberculosis.We retrieved full text articles for 76 citations, of which 43 studies were excluded based on a review of title and abstract.Articles that briefly mentioned a diagnosis system but did not give details or include how it was used in the study were also excluded.Thirty-two published peer-reviewed articles and one country wide survey from Ethiopia met the general study criteria and included in the synthesis review (Table 1).We have also performed an updated search on 7 July, 2013 that yielded 62 titles, all of which had been identified during the previous search or were ineligible based on title or abstract.Repeated searches using the same search strategies were employed until the end of July, 2013 to ensure that no new publications were missed.Thus, we included 33 relevant studies in this review (Figure 1).Of the total 33 studies, one was population based country study [14] carried out at three (Urban, Rural and Pastoralists) study setting.

Description of included studies
Of the 267 citations identified by literature searches, a total of 33 English-language publications (32 articles, and 1 population based survey) met the eligibility criteria (Figure 1).Of the 33 studies, ten (30.3%) studies utilized culture (Aderaye et al., 2003(Aderaye et al., , 2007;;Alebachew et al., 2011;Beyene et al., 2008;Shah et al., 2009;Bruchfeld et al., 2002;Abebe et al., 2010Abebe et al., , 2012;;Legesse et al., 2011) and one study employed polymerase chain reaction (PCR)  (PCR) as the reference standard (Beyene et al., 2009).Eighteen (54.5%) studies used a cross sectional study design, three (9.1%)used comparative, two (6.1%) used a retrospective design, two (6.1%) randomized trial and one (3%) did not describe the study design employed method.Missing data were treated as "not reported" (indicated in the table as "not reported) (Table 1).Three studies presented the data based on two separate microscopy methods (Ziehl-Neelsen [ZN] and Light Emitting Diode Fluorescence Microscopy [LED-FM]) (Alebachew et al., 2011;Yassin et al., 2013;Moges et al., 2012).Twenty five (785.8%)studies were published after 2008 and few studies included information about patients infected with HIV.Descriptive information on population characteristics, such as sex and age category, was often reported.Most studies have used more than one diagnostic system.Twenty-four (72.73 %) of the 33 studies utilized ZN stained microscopy and three 3 (9.4%)studies used light emitting diodes fluorescent microscope (LED-FM).Four (12.15%) reported the use of some processing methods (Aderaye et al., 2007;Ali et al., 2012), three (9.14%)studies used tuberculin skin tests (TST) and QuantiFERON-TB gold In-Tube (QFTGIT) together and two (6.13%) studies employed cytological techniques.Chest X-ray was employed in the majority of studies especially for smear negative results with symptomatic patients suggestive of tuberculosis.For the majority of studies, information about timing, manner of sputum collection, mycobacterial culture methods and quality assurance for microscopy was reported while sputum characteristics (macroscopic appearance of sputum samples) and appropriate instructions how to produce quality sputum was not described except one study (Federal Ministry of Health, 2010).Eighteen (54.5 %) of the 33 studies reported about PTB alone, seven (21.29%) studies reported extra-pulmonary tuberculosis (EPTB) alone, four (12.15%) studies reported both pulmonary tuberculosis (PTB) and EPTB, three (9.14%)studies reported latent tuberculosis infection (LTBI) alone and one study reported both PTB and LTBI (Table 1).

DISCUSSION
Our systematic review of literature of 33 studies on diagnosis of tuberculosis showed that there is a considerable similarity between diagnosis systems used, study design and the types of TB isolated.However, comparisons of these variables were not significantly different.The diagnostic systems included the inclusion or exclusion of laboratory testing, and even their diagnostic focus (that is, pulmonary TB alone, pulmonary and extra pulmonary TB or latent TB infections) were varied.Because publication dates of the articles range over the last thirteen years, some systems were developed and evaluated after the HIV epidemic and focused specifically on co-infected patients.
Smear-positive TB case detection rate is 36% in Ethiopia (Federal Ministry of Health, 2010).Lack of accurate and rapid diagnostics remains a major obstacle to progress in this regard.Health care facilities still heavily rely  heavily rely on sputum smear microscopy for the diagnosis of TB.This technique has low sensitivity and specificity (Aderaye et al., 2007), however, efforts were made to ensure quality AFB diagnosis through appropriate instruction of symptomatic individuals on how to produce quality sputum sample from their lung.In the laboratory, the macroscopic appearance of a sputum sample was checked and poor specimens were replaced with an immediate spot collection (Tadesse et al., 2011).The gold standards chosen to evaluate the validity of these diagnostic systems also varied widely.laboratory, the macroscopic appearance of a sputum sample was checked and poor specimens were replaced with an immediate spot collection (Tadesse et al., 2011).
The gold standards chosen to evaluate the validity of these diagnostic systems also varied widely.Ten studies used culture as the gold standard these diagnostic systems also varied widely.Ten studies used culture as the gold standard (Aderaye et al., 2003;Aderaye et al., 2007;Alebachew et al., 2011;Beyene et al., 2008;Shah et al., 2009;Bruchfeld et al., 2002;Abebe et al., 2010Abebe et al., , 2012;;Legesse et al., 2011;Deribew et al., 2012), others used TST and QFTGIT (Legesse et al., 2010;Dagnew et al., 2012;Wassie et al., 2013), or histology, cytology and histopathology (Eshete et al., 2011;Muluye et al., 2013;Federal Ministry of Health, 2010).Unfortunately, laboratory diagnosis is likely to depend strongly upon the experience and knowledge base of the laboratory technician; it may be less reliable in settings where technicians have less training.To allow for comparison of diagnostic tests across different studies and settings, future studies need to employ a more consistent and rapid gold standard systems.Ideally, culture would be gold standard, as it is a standard for validation that could be reliably reproduced across settings.
However, because cultures are difficult to obtain in resource limited settings and can lead to a delay in treatment, performing studies with culture as the gold standard may be difficult.In addition to using a variety of gold standards, the various studies often included very different sample populations.Some studies did not clearly describe the characteristics of the patient population or how they were selected.Some were retrospective, often utilizing record review (Alene et al., 2013;Muluye et al., 2013).
Ideally, prospective studies of diagnostic systems would evaluate a clearly defined sample of participants with a spectrum of disease representative of the study subjects to which the diagnostic systems would be applied in the study settings.It is essential that researchers should clearly describe the sampling process and inclusion criteria in such studies to allow for more accurate comparison of diagnosis systems across different populations or settings and to promote the utility of these systems.
QFTGIT and TST were also used as part of their diagnostic gold standard to differentiate latent TB from no TB from active TB disease (Legesse et al., 2010;Dagnew et al., 2012;Wassie et al., 2013).This makes it difficult to interpret the accuracy of a diagnostic system and its ability to predict a diagnosis of TB in a particular patient or patient population.This overlap also causes difficulty in determining the relative importance of particular signs or symptoms within the diagnostic system.The largest shift in the newer diagnostic systems as compared to smear microscopy is the focus on both pulmonary tuberculosis and extra pulmonary tuberculosis including latent tuberculosis.Diagnostic systems focusing simply on pulmonary TB, such as (Cambanis et al., 2006;Aderaye et al., 2007) have demonstrated higher sensitivities than those developed to diagnose both extra pulmonary and pulmonary TB.Because children have a higher incidence of extra pulmonary TB, using diagnostic systems targeted at pulmonary TB only addresses part of the diagnostic challenge.
The findings of this systematic review are limited by the study settings design and quality of the studies included.The lack of consistent and sometimes clearly defined inclusion criteria among the studies makes it difficult to compare sensitivity and specificity across the different diagnostic systems.Most of the various diagnostic systems have only been evaluated in specific geographic locations or single populations; few studies evaluate a particular diagnostic system in multiple geographic regions or patient populations.Fewer studies have compared the diagnostic tests with different specimens (Bronchoalveolar lavage, pre and post-bronchoscopy sputum for pulmonary tuberculosis) in the same population (Aderaye et al., 2007).There may be probably more TB cases in the surveyed communities or settings than reported in these studies.Furthermore, sputum samples may not be collected and examined in symptomatic individuals that may further aggravate the underestimation of the true TB prevalence.
With a better screening method and diagnostic facility, the prevalence could even be more than what was reported.However, one thing is very certain; the passive case detection approach currently implemented in Ethiopia leaves many (two-thirds of smear-positive TB cases) undetected and untreated TB cases in the community; favoring continuous spread of the disease through maintaining active TB transmission (Tadesse et al., 2011).

Conclusion
The results of this review suggest that there is a need for revising the diagnostic systems.Most of the laboratory tests employed was based on direct smear microscopy, which is insensitive and can only detect 36% of tuberculosis cases in Ethiopia.This may result in misdiagnosis of the disease and further transmission of the disease especially in children and populations with high HIV co-infection.Therefore, there is an urgent need for laboratories to find a rapid and efficient method for TB diagnosis as a complement to the smear microscopy.The low case detection rates observed nationally can be improved by introducing enhanced case detection mechanisms and promoting favorable health seeking behaviors

Figure 1 .
Figure 1.Search strategy and paper selection flowchart.

Table 1 .
Diagnostic systems used in the diagnosis of tuberculosis and characteristics of selected studies in Ethiopia.