Diagnostic accuracy of nucleic acid amplification tests (NAATs) in urine for genitourinary tuberculosis: a systematic review and meta-analysis

Abstract

Background: Genitourinary tuberculosis is the third most common form of extrapulmonary tuberculosis. Diagnosis is difficult because of unspecific clinical manifestations and low accuracy of conventional tests. Unfortunately, the delayed diagnosis impacts the urinary tract severely. Nucleic acid amplification tests yield fast results, and among these, new technologies can also detect drug resistance. There is lack of consensus regarding the use of these tests in genitourinary tuberculosis; we therefore aimed to assess the accuracy of nucleic acid amplification tests in the diagnosis of genitourinary tuberculosis and to evaluate the heterogeneity between studies.

Methods: We did a systematic review and meta-analysis of research articles comparing the accuracy of a reference standard and a nucleic acid amplification test for diagnosis of urinary tract tuberculosis. We searched Medline, EMBASE, Web of Science, LILACS, Cochrane Library, and Scopus for articles published between Jan 1, 1990, and Apr 14, 2016. Two investigators identified eligible articles and extracted data for individual study sites. We analyzed data in groups with the same index test. Then, we generated pooled summary estimates (95% CIs) for sensitivity and specificity by use of random-effects meta-analysis when studies were not heterogeneous.

Results: We identified eleven relevant studies from ten articles, giving information on PCR, LCR and Xpert MTB/RIF tests. All PCR studies were "in-house" tests, with different gene targets and had several quality concerns therefore we did not proceed with a pooled analysis. Only one study used LCR. Xpert studies were of good quality and not heterogeneous, pooled sensitivity was 0·87 (0·66-0·96) and specificity was 0·91 (0·84-0·95).

Conclusion: PCR studies were highly heterogeneous. Among Xpert MTB/RIF studies, specificity was favorable with an acceptable confidence interval, however new studies can update meta-analysis and get more precise estimates. Further high-quality studies are urgently needed to improve diagnosis of genitourinary tuberculosis.

Protocol registration: PROSPERO CRD42016039020.

Keywords: Genitourinary tuberculosis; Nucleic acid amplification test; Systematic review.

Fig. 1

Selection of studies reporting on the use of a NAAT for GUTB diagnosis in urine

Fig. 2

Summary of Quality Assessment tool for Diagnostic Accuracy tests (QUADAS-2). The point estimates of sensitivity and specificity from each study are shown as solid circles. Error bars are 95% confidence intervals. A statistical continuity correction of 5% has been applied

Fig. 3

Forest plots of the diagnostic accuracy of PCR, LCR and Xpert MTB/RIF for the diagnosis of GUTB. Each triangle represents a study in the meta-analysis and the circle the summary estimate. The light line is the confidence interval

Fig. 4

Summary Receiver-Operating Characteristic (SROC) curves for Xpert MTB/RIF assays