Objective To evaluate the application value of Xpert Mtb/RIF detection assay in the diagnoses of tuberculosis (TB) and Rifampin resistance. Methods Clinicians continuously recruited 2142 TB suspects in four counties (districts) level TB dispensary (Xiangtan, Yueyang, Beilin, Lanxi). Three sputum specimens each patient provided were performed smear, solid culture and Xpert Mtb/RIF test. The TB patients with definite diagnoses were followed up for two months. The staff in provincial laboratory completed rifampin susceptibility test in 555 culture-positive strains using proportional method. The staff in national reference laboratory sequenced the rifampin-resistant determinant region of rpoB genes in the strains with the inconsistent results between Xpert Mtb/RIF and traditional drug susceptibility test. Using clinical diagnosis as gold standard, the performance of smear, solid culture and Mtb/RIF test were evaluated. The performance of Xpert Mtb/RIF for rifampin resistance detection was compared with traditional drug susceptibility test.   Results Using clinical diagnosis result as the gold standard, the sensitivities of smear, solid culture and Xpert Mtb/RIF were 25.68%（284/1106）, 51.44%（555/1079） and 58.82%（650/1105）, respectively. The sensitivity of Xpert Mtb/RIF was higher than smear (χ²=360.10, P<0.05) and the solid culture(χ²=50.13, P<0.05). Compared with traditional drug susceptibility test, the sensitivity and the specificity of Xpert Mtb/RIF for rifampin resistance detection were 87.10% (27/31) and 97.95% (477/487), respectively. Conclusion Xpert Mtb/RIF test is simple, the sensitivity of Xpert Mtb/RIF is higher than smear microscopy and solid culture，it can also diagnose patients of rifampin resistance, so it has very good application prospects in our county (district) level laboratories.

Objective To evaluate the performance of GeneXpert Mtb/RIF assay for the diagnosis of Mycobacterium tuberculosis (Mtb) and rifampin-resistant tuberculosis (TB) in clinical samples. Methods The sputum samples were collected from all TB suspects who visited the outpatient departments of the 309th Hospital of Chinese People’s Liberation Army and Guangzhou Chest Hospital, as well as the TB clinic of He’nan Provincial Center for Disease Control and Prevention (He’nan Provincial CDC) from January to May 2011. Smear microscopy, culture (golden-standard method), drug susceptibility testing (DST, conventional proportion method) and GeneXpert Mtb/RIF assay were respectively performed on each clinical sample. SPSS software version 11.5 was used for statistical analysis to compare the diagnostic sensitivity and specificity of GeneXpert Mtb/RIF assay and other examination methods, P<0.05 was regarded as a significant difference. Results Sputum samples were collected from 1971 TB suspects. The samples from 3 suspected cases were contaminated and were excluded from the analysis. Thus 1968 suspected cases were involved in the analysis. Compared with the results of smear microscopy and culture, the sensitivity and specificity of GeneXpert Mtb/RIF assay for diagnosis of TB was 93.27%(748/802), 91.93%(797/867) and 91.60%(1068/1166), 95.55%(1052/1101) respectively; compared with the results of clinical diagnosis, the sensitivity and specificity of GeneXpert Mtb/RIF assay for diagnosis of TB was 78.66%(822/1045) and 98.67%(888/900) respectively; compared with the results of conventional DST, the sensitivity and specificity of GeneXpert Mtb/RIF assay for diagnosis of rifampin-resistant TB was 97.92%(188/192) and 100.00%(595/595) respectively. Conclusion GeneXpert Mtb/RIF assay is a rapid, simple and hand-free molecular diagnostic tool with high sensitivity and specificity for the diagnosis of TB and rifampin-resistant TB. It can be used for clinical diagnosis of TB and rifampin-resistant TB.

To evaluate the application value of GeneXpert MTB/RIF system in the diagnosis of pulmonary tuberculosis (TB) among the TB suspects who sought health care for the first time. A total of 361 TB suspects, who visited Jiaxing 1st Hospital to seek health care for the first time from Jan 2015 to Dec 2016, were consecutively enrolled in the study. The sputum specimens were collected from those patients and smear microscopy, liquid culture, GeneXpert MTB/RIF test and solid drug susceptibility testing (DST) by using proportion method were performed. Finally, 184 cases (50.97%) were confirmed to have pulmonary TB while 162 cases (44.88%) were diagnosed to be pulmonary infections and 15 cases (4.15%) were diagnosed to be pulmonary infections caused by atypical mycobacteria. Among 361 TB suspects, the positive rate of GeneXpert MTB/RIF was 32.41% (117/361), which was significantly higher than the smear-positive rate (22.71%, 82/361) (χ ²=8.49, P<0.05); but the difference of the positive rate between GeneXpert and liquid culture (26.32%, 95/361) was not statistically significant (χ ²=3.23, P=0.072); the smear-positive rate was lower than that of liquid culture (26.32%, 95/361), but the difference was not statistically significant (χ ²=1.26, P=0.261). Among 184 cases with pulmonary TB, the positive rate of GeneXpert MTB/RIF was 63.59% (117/184) which was significantly higher than the smear-positive rate (36.41% (67/184)) and the liquid culture positive rate (48.91% (90/184)) respectively (χ ²=27.17, P<0.01; χ ²=8.05, P<0.05). Among 117 smear-negative pulmonary TB patients, the positive rate of GeneXpert MTB/RIF was 42.74% (50/117), which was significantly higher than that of liquid culture (24.79%, 29/117) (χ ²=8.43, P=0.004). If the clinical diagnosis result was regarded as a standard, the sensitivity of GeneXpert in TB suspects, who sought health care for the first time, was 63.59% (117/184), its specificity was 100.00% (177/177) and its correct indices was 0.64; if the result of solid DST with proportion method was regarded as a standard, the sensitivity of GeneXpert in TB suspects, who sought health care for the first time, was 4/5, its specificity was 97.70% (85/87) and its correct indices was 0.97. GeneXpert MTB/RIF technique has demonstrated a high capacity on detection of tuberculosis and rifampicin resistant TB in TB suspects, especially in smear-negative pulmonary TB patients.

Objective To evaluate the value and significance of Xpert MTB/RIF in the diagnosis of tuberculosis in general hospitals.Methods From June 2016 to August 2017, 376 specimens of patients with suspected tuberculosis in Zhuzhou Central Hospital were collected, including 227 sputum specimens (60.37%), 109 alveolar lavage fluid samples (28.99%), 21 cerebrospinal fluid samples (5.59%), 15 pleural effusion samples (3.99%), and 4 other specimens (1.06%, including drainage fluid, pus, secretions, and urine, 1 case each). GeneXpert MTB/RIF test (referred to as “Xpert method”), sandwich cup liquid-based bacteria acid-fast staining smear microscopy (referred to as “sandwich cup method”) and traditional solid Roche culture method (referred to as “solid culture method”) were conducted for each specimen, and the results were compared. Statistical analysis was performed using SPSS 19.0 software. The count data were compared by χ ² test, and the difference was statistically significant at P<0.05. Results Among the 376 suspected cases, 349 were clinically diagnosed as tuberculosis and 27 were non-tuberculosis. The positive detection rate of Xpert method (37.50%, 141/376) was significantly higher than that of sandwich cup method (22.87%, 86/376) and solid culture method (17.82%, 67/376) (χ ² values were 19.09 and 36.39, respectively; P values <0.05). Taking clinical diagnosis as gold standard, the sensitivity of Xpert method (40.11%, 140/349) was significantly higher than that of sandwich cup method (23.50%, 82/349) and solid culture method (19.20%, 67/349) (χ ² values were 19.86 and 35.98, respectively; P values were <0.05). The diagnostic consistency of Xpert method (K=0.080) was higher than that of sandwich cup method (K=0.016) and solid culture method (K=0.033). Among the 140 positive specimens detected by Xpert method, proportion of resistance to rifampicin was 15.71% (22/140). Among the 63 positive specimens detected by solid culture method, the proportion of resistance to rifampicin was 14.29% (9/63). The difference in proportion of resistance to rifampicin between Xpert and solid culture method was not statistically significant (χ ²=0.07, P>0.05). For the 9 rifampicin-resistant samples detected by solid culture method, consistent results were achieved by Xpert method. Conclusion The Xpert detection method is simple and rapid and shows high sensitive. It has high clinical value for early diagnosis and treatment of tuberculosis in general hospitals.

Objective To investigate the value of GeneXpert MTB/RIF (GeneXpert) and line probe assay for rapid diagnosis of osteoarticular tuberculosis and detection of rifampicin resistance. Methods A total of 172 suspected osteoarticular tuberculosis patients who received focal cleaning or paracentesis between March to December 2018 were selected. GeneXpert, line probe assay and BACTEC MGIT 960 mycobacterial culture were used to test the samples. Using clinical comprehensive diagnosis as the reference standard, the efficacy of GeneXpert, linear probe assay and combination of the two methods in diagnosis of osteoarticular tuberculosis were evaluated. Using mycobacterial culture drug sensitivity test as the reference standard, the efficacy of GeneXpert, linear probe assay and combination of the two methods in detecting rifampicin resistance were evaluated. Results According to the clinical comprehensive diagnosis of the 172 suspected osteoarticular tuberculosis patients, 112 were diagnosed with osteoarticular tuberculosis, whereas 60 were non-tuberculosis. Using clinical comprehensive diagnosis as the reference standard, the sensitivity and specificity of GeneXpert were 82.14% (92/112) and 96.67% (58/60). The results of GeneXpert was consistent with that of clinical comprehensive diagnosis (Kappa=0.74). The sensitivity and specificity of line probe assay were 69.64% (78/112) and 90.00% (54/60). The consistency with clinical comprehensive diagnosis was not good (Kappa=0.54). The sensitivity and specificity of combination of the two methods were 82.14% (92/112) and 90.00% (54/60). The combination method had good consistency with clinical comprehensive diagnosis (Kappa=0.68).The area under the receiver operating characteristic (ROC) curve of GeneXpert, line probe assay and combination of the two methods were 0.89, 0.80 and 0.86, respectively, indicating good diagnosis performance. Using mycobacterial culture drug sensitivity test as the reference standard, the sensitivity and specificity of GeneXpert in inspection to rifampicin resistance were 100.00% (18/18) and 92.31% (24/26), the results was highly consistent (Kappa=0.91), and the area under the ROC curve was 0.96. The sensitivity and specificity of line probe assay in inspection to rifampicin resistance were 100.00% (18/18) and 88.46% (23/26), the results was highly consistent (Kappa=0.86), and the area under the ROC curve was 0.94. The results of combination of the two methods was consistent with that of the linear probe assay. Conclusion GeneXpert and line probe assay can rapidly and accurately diagnose osteoarticular tuberculosis and detect rifampicin resistance, having good clinical application value.

To evaluate the auxiliary diagnostic value of GeneXpert MTB/RIF (GeneXpert) for detecting extrapulmonary tuberculosis based on fine needle biopsy tissue eluent, a total of 103 patients with extrapulmonary tuberculosis in Wuhan Pulmonary Hospital between January 2016 and December 2017 were enrolled. All of these patients were operated through fine needle puncture and the biopsy tissue eluent were tested by tuberculosis-DNA PCR assay, GeneXpert, tuberculosis solid culture and smear method, respectively. The results showed that the sensitivity of GeneXpert (78.6% (81/103)) was significantly higher than that of tuberculosis solid culture (26.2% (27/103)) and smear method (16.5% (17/103)), and the differences were statistically significant (χ ² values were 56.76 and 79.72, respectively; all P values <0.05). Hence, it is clear that GeneXpert showed a higher diagnostic sensitivity for detecting extrapulmonary tuberculosis based on fine needle biopsy tissue eluent.

Objective To compare the cost of solid culture, drug susceptibility test and Xpert Mtb/RIF test in the primary laboratories in China. Methods We collected the cost data of solid culture for 3 times and Xpert Mtb/RIF test from 4 county (district) level laboratories, and strain identification and rifampin (RFP) susceptibility test for 3 times from 2 province reference laboratories. The element method was used to calculate the unit cost of each method, the detection cost of each patient. Of the population with different TB prevalence rate, the solid culture was used as the gold standard to compare the cost of Xpert Mtb/RIF test to find 1 TB patient under different detection efficiency. Of the population with different rifampicin-resistant prevalence rate, the traditional drug susceptibility test was used as the gold standard to compare the cost of Xpert Mtb/RIF test to find 1 RFP-resistant patient under different detection efficiency. After price adjustment of Xpert Mtb/RIF detection equipment and reagents, the unit cost of Xpert Mtb/RIF test was analyzed. Results The unit costs of solid culture, strain identification, RFP susceptibility test and Xpert Mtb/RIF detection were 47.87, 46.73, 82.86 and 118.62 Yuan, respectively. The cost per TB patient by traditional methods (2 solid cultures, 1 sub-culture, 1 PNB strain identification) was 172.57 Yuan. The average cost of each patient with drug-resistant TB by traditional methods (2 solid cultures, 1 sub-culture, 1 PNB strain identification and 1 drug susceptibility test) was 208.84 Yuan. When the specificity of Xpert Mtb/RIF test for the detection of Mtb and RFP was 85%, its sensitivity is greater than 70%, its cost of detecting one TB case or RFP-resistant case among the population with 1% to 70% prevalence rate of TB or RFP resistance was lower than traditional methods. The influence of reagent price changes on the testing cost was far greater than the change of equipment price. Conclusion The Xpert Mtb/RIF test is a rapid and more economic method than traditional tests, and suitable for the use in the primary laboratories.

Tuberculosis (TB) is one of the leading infectious diseases worldwide. As a rapid nucleic acid amplification test, Xpert MTB/RIF (Xpert) assay can detect mycobacteria bacilli and rifampicin resistance simulta-neously, and provide reliable information for TB diagnosis and treatment due to its high sensitivity and specificity. However, Xpert also demonstrated some imperfections during the clinical practice. Xpert MTB/RIF Ultra (Xpert Ultra) assay is a next-generation assay of Xpert, WHO recommended it in March, 2007. In contrast to Xpert, Ultra has acquired increased sensitivity with paucibacillary specimens collected from patients with extra-pulmonary TB, TB/HIV co-infection or pediatric TB, which is of great significance for regimen establishment. In this review, the limitation of Xpert, the similarity and difference between Xpert and Xpert Ultra are discussed, which will throw light on the value of Xpert Ultra’s clinical application.

Objective To compare the results of Mycobacterium tuberculosis (MTB) resistance to rifampicin (RFP) by Micropore-plate method (Microdrug sensitivity, MicroDST) and GeneXpert MTB/RIF test. Methods Among the 6893 suspected MTB patients who admitted to or visited Outpatient Department in Shaanxi Provincial Tuberculosis Control and Prevention Hospital, 6086 patients were diagnosed as MTB cases according to the clinical diagnostic criteria. A total of 425 clinical specimens were tested by BACTEC MGIT 960 liquid culture method and GeneXpert method, including 217 sputum specimens, 170 alveolar lavage specimens, and 38 extrapulmonary fluid specimens (such as pleural and peritoneal effusion, pus and cerebrospinal fluid). After excluding 3 specimens that reported errors or could not be interpreted and 2 specimens that were contaminated when cultured, there were 420 valid specimens. Three hundred and seventy-two clinical isolates with positive results by GeneXpert and MGIT 960 method were tested for drug resistance by MicroDST, GeneXpert MTB/RIF and MGIT 960 method, including 191 sputum specimens, 158 alveolar lavage specimens, and 23 extrapulmonary fluid specimens (such as pleural and peritoneal effusion, pus and cerebrospinal fluid). Taking MGIT 960 test as the gold standard, the performance of MicroDST and GeneXpert MTB/RIF in detecting MTB resistance to RFP were compared. Results The overall positive rates of RFP resistance by GeneXpert MTB/RIF, MicroDST and MGIT 960 method were 18.8% (70/372), 16.7% (62/372), and 16.7% (62/372), respectively. Taking MGIT 960 test as the gold standard, the sensitivity, specificity, coincidence rate and Kappa value of RFP resistance test were 91.9% (57/62), 98.4% (305/310), 97.3% (362/372) and 0.8 by MicroDST assay and 96.8% (60/62), 96.8% (300/310), 96.8% (360/372) and 0.8 by GeneXpert MTB/RIF assay. The sensitivity, specificity, coincidence rate and Kappa value using sputum, alveolar lavage and extrapulmonary specimens were 96.7% (29/30), 98.1% (158/161), 97.9% (187/191), 0.9; 93.3% (28/30), 96.3% (155/161), 95.8% (183/191), 0.9; and 91.7% (22/24), 98.5% (132/134), 97.5% (154/158), 0.9 by MicroDST assay and 100.0% (24/24), 97.0% (130/134), 97.5% (154/158), 0.9; 75.0% (6/8), 100.0% (15/15), 91.3% (21/23), 0.8; and 100.0% (8/8), 100.0% (15/15), 100.0% (23/23), 1.0 by GeneXpert MTB/RIF assay, respectively. Conclusion Taking MGIT 960 test as the gold standard, MicroDST and GeneXpert MTB/RIF assay have a good consistency in overall positive detection rate of RFP resistance, as well as in different samples.

Objective To analyze the role of phenotypic drug sensitivity test(DST) and GeneXpert MTB/RIF in the diagnosis of drug-resistant spinal tuberculosis. Methods Data of 514 patients diagnosed as spinal tuberculosis in Beijing Chest Hospital, Capital Medical University from 2015-2018 were retrospectively investigated. Pus, granulation tissue and caseous necrotic tissue in the focus were obtained by puncture and surgery in all the patients, then tubercle bacillus among them were cultured by modified L?wenstein-Jensen medium without potato starch and phenotypic drug sensitive tests by concentration methods were carried out, and GeneXpert MTB/RIF test was performed at the same time. Results In 514 patients, 260 were males (50.6%),254 were females (49.4%), ranging from 1 to 86 years old with an average age of 46.2±18.7 years. There were 200 (38.9%) cases of lumbar tuberculosis, 186 (36.2%) cases of thoracic tuberculosis, 60 (11.7%) cases of lumbosacral tuberculosis,46 (8.9%) cases of thoracolumbar tuberculosis,14 (2.7%) cases of cervical tuberculosis and 8 (1.6%) cases of cervical and thoracic tuberculosis. The results of drug sensitivity test were obtained in 109 cases, the positive ratio was 21.2%, 25 bacterial strains were resistance to one and more drugs, the positive ratio was 4.9%, 9 of the positive culture strains were multidrug-resistance tuberculosis and extensively-resistant tuberculosis (XDR-TB) (8.3%). Three hundred and eighty-three cases (74.5%) were positive for GeneXpert MTB/RIF, of which 47 cases (9.1%) were tested rpoB mutation including that of 9 MDR-TB patients confirmed by DST. Conclusion The detection rate of drug-resistant Mycobacterium tuberculosis in spinal tuberculosis focus is low through DST, and drug-resistant cases can not be detected early. GeneXpert MTB/RIF method has a high value for the early diagnosis of drug-resistant spinal tuberculosis patients.

Objective To assess the diagnostic value of GeneXpert MTB/RIF assay using the lapping suspension of thoracoscopic biopsy tissue in detecting tuberculous pleurisy.Methods From January 1, 2017 to December 31, 2017 in the pleurisy ward of Shenyang Chest Hospital, 51 cases who showed unexplained pleural effusion under medical thoracoscopy and were not treated by any anti-tuberculous therapy were selected as the subjects of this study. A routine pathological examination of the pleural biopsy specimen was conducted for each patient. Also, BACTEC MGIT 960 liquid culture system and GeneXpert MTB/RIF detections were conducted using the suspension of the other part of specimen after grinding. The diagnosis accuracy of GeneXpert MTB/RIF assay using biopsy tissue suspension in detecting tuberculous pleurisy was evaluated, taking BACTEC MGIT 960 positive results (stain identification as Mycobacterium tuberculosis) and pleural tissue biopsy results (granulomatous inflammation and anti-acid staining positive) as diagnostic criteria.Results Of the 51 patients, 34 were diagnosed with tuberculous pleurisy. MGIT 960 positive culture and strain identification confirmed 20 cases (39.2%, 20/51), and 17 cases (33.3%, 17/51) were pleural biopsy pathological positive. Among them, 3 cases were confirmed by the two detection me-thods. Seventeen (33.3%, 17/51) patients were diagnosed with non-tuberculous pleural effusion. Taking the finial diagnosis as the gold standard, the sensitivity and specificity of GeneXpert MTB/RIF, MGIT 960 culture system and pathological diagnosis were 64.7% (22/34) and 100.0% (17/17), 58.8% (20/34) and 100.0% (17/17), and 50.0% (17/34) and 100.0% (17/17), respectively. The sensitivity of GeneXpert MTB/RIF was superior to the MGIT 960 culture system and pathological diagnosis; however, the difference was not statistically significant (χ ²=1.53, P=0.466).Conclusion The sensitivity and specificity of GeneXpert MTB/RIF assay using the lapping suspension of thoracoscopic biopsy tissue taken under thoracoscopy is high. It is of great significance for the diagnosis of tuberculous pleurisy.

Drug-resistant tuberculosis (DR-TB) especially multidrug-resistant tuberculosis (MDR-TB) and extensively drug-resistant tuberculosis (XDR-TB) are one of the most difficult problems to be solved in clinical practice. Anti-tuberculous drug susceptibility tests (DST) provide experimental basis for the diagnosis of DR-TB. This expert consensus briefly introduces the common M.tuberculosis (MTB) DST methods (including phenotypic DST and molecular DST) used in clinic, proposes the strategies of MTB drug resistance detection and correct interpretation of the results, points out the problems of DST detection technology and clinical application, and prospects the development of DST in the future.

Objective To assess the performance of GeneXpert MTB/RIF (GeneXpert) assay in the detection of Mycobacterium tuberculosis (MTB) and its rifampicin resistance, and further analyze the relationship between the false-positive rifampicin-resistant results of GeneXpert assay and the low bacterial load in the clinical samples. Methods From April 2016 to September 2017, a total of 2707 clinical samples (1945 sputum specimens and 762 alveolar lavage fluid specimens) of suspected pulmonary tuberculosis cases from Xi’an Chest Hospital were collected for GeneXpert detection, BACTEC MGIT 960 liquid culture and drug susceptibility test. Among them, GeneXpert detected rifampicin-resistant samples were further tested by GenoType MTBDRplus assay. Results The positive rate of GeneXpert assay in detecting MTB was 39.8% (1077/2707). Seven cases had “indeterminate” rifampicin resistance results and “Very low” bacterial load level. Considering MGIT 960 drug susceptibility test as the gold standard, the sensitivity and specificity of GeneXpert test in detecting rifampicin resistance was 89.4% (135/151) and 94.7% (721/761), respectively. The inconsistent results of two methods were found in 56 cases. Among them, 40 cases were rifampicin-resistant detected by GeneXpert but were rifampicin-sensitive detected by MGIT 960. Of the 40 patients, 10, 6, 14 and 10 showed a MTB load level of “Very low”, “Low”, “Medium” and “High”, respectively. Taking MGIT 960 test result as gold standard, the false-positive rates of rifampicin-resistance by GeneXpert in the four MTB load levels were 58.8% (10/17), 16.7% (6/36), 19.2% (14/73) and 20.4% (10/49), respectively; the difference was statistically significant (χ ²=13.981, P<0.05). Among the 40 cases, 30 cases with Low, Medium and High level of MTB load were rifampicin-resistant when detected by the MTBDRplus test, which was consistent with the GeneXpert results. As to the 10 cases with Very low level of MTB load, 2 cases were negative and 8 were positive by MTBDRplus test, and in the cases with positive results, 5 cases were rifampicin-sensitivity and 3 rifampicin-resistance. Conclusion The detection of MTB rifampicin-resistance with GeneXpert assay has a higher sensitivity and specificity. However, when the tested load of MTB is Very low, false drug resistant results may occur.

Objective  To evaluate the performance of Xpert Mtb/RIF test on detecting tuberculosis and rifampin resistance for patients.   Methods  Recruited 588 initial diagnosis patients, selected sputum samples to do smear, solid culture, traditional proportion method drug susceptibility test and Xpert Mtb/RIF test. Analyze the sensitivity and specificity of Xpert Mtb/RIF method in detecting tuberculosis and refampin resistance.  Results  According to the solid culture result, the sensitivity and specificity of Xpert Mtb/RIF test in detecting tuberculosis are 95.0% (345/363) and 89.6% (147/164). According to the conventional drug susceptibility test result, the sensitivity and specificity in detecting rifampin resistance are 89.1% (41/46) and 96.3% (289/300) respectively.  Conclusion  The Xpert Mtb/RIF test has very high sensitivity and specificity. It has very good application prospects in rapidly detection tuberculosis and rifampin resistance.

Objective To evaluate the diagnostic performance of GeneXpert MTB/RIF (“Xpert”) for tuberculous pleurisy.Methods Pleural effusion samples collected with ultrasound guidance from 180 patients with suspected tuberculous pleurisy from Anshan Tuberculosis Hospital between January 2017 and December 2017 were tested with smear acid-fast bacilli testing (“smear”), liquid culture by BACTEC MGIT 960(“MGIT 960”) and GeneXpert. With clinical diagnosis results as the reference standard, sensitivity, specificity, positive prediction value, negative prediction value and consistent rate were evaluated for performances of smear test, MGIT 960 and GeneXpert test on detecting Mycobacterium tuberculosis (MTB) in those pleural effusion samples.Results Among those 180 suspected patients, 162 cases (90.00%) were confirmed as tuberculous pleurisy (TBP), 18 cases (10.00%) were excluded of TBP. Among the 162 cases of TBP, 61 cases (37.65%) were MGIT 960 culture-positive, 101 cases (62.35%) were diagnosed as TBP basing on standard clinical diagnosis criteria, including clinical symptoms, biochemical indicators of pleural effusion samples, results of chest imaging examination, and diagnostic anti-TB treatment results. Eighteen cases excluded of TBP had lung cancer pleural metastasis in 9 cases, sarcoidosis in 1 case, systematic lupus pleurisy in 1 case, pleural mesothelioma in 2 cases, nephropathy in 2 cases, cardiac insufficiency in 2 cases and hypoproteinemia in 1 case. With clinical diagnosis as the reference standard, the sensitivity of smear test, MGIT 960 and GeneXpert test were 0.00% (0/162), 37.65% (61/162) and 54.94% (89/162) separately; The specificity of three tests were all 100.00% (18/18); Positive prediction value of three tests were 0.00% (0/0),100.00% (61/61) and 100.00% (89/89); Negative prediction value of three tests were 10.00% (18/180),15.13% (18/119) and 19.78% (18/91);Consistent rate of three tests with clinical diagnosis results were 10.00% (18/180), 43.89% (79/180) and 59.44% (107/180) separately. Sensitivity of GeneXpert test was significantly higher than that of the smear and MGIT 960 culture (χ ²=119.82,P=0.000). Conclusion GeneXpert has high diagnostive value in detecting tuberculous pleurisy.