Performance Evaluation, Comparison and Degree of Effectiveness of Five Different Techniques for Detection and Isolation of Mycobacterium Tuberculosis
Authors/Creators
Description
This study was aimed to evaluate and compare the clinical performance and
effectiveness of five different techniques for detection and isolation of
Mycobacterium tuberculosis to determining the best technique for the
detection and isolation of MTB at resource limited province of Pakistan,
Balochistan. A cross-sectional retrospective cohort study on 307 samples
collected from 2016 to 2018 was conducted at Fatima Jinnah General and
Chest Hospital Quetta (FJHQ). All the samples were cultured according to
the WHO protocol on LJ media and BACTEC 960 system. The fluorescence
microscopy, ICT (MPT64 Kit) and ZN microscopy was used as confirmatory
techniques. IBM-SPSS version 23 was used to analyze the recorded data and
compute all the statistics. Multi Variate Binary Logistic Regression
(MVBLR) analysis was used to determine the technique of choice best for
the detection and isolation of MTB in clinical isolates. The frequency of
positive samples processed by BACTEC 960, ICT (MPT64 Kit), ZN
Microscopy, fluorescence Microscopy and LJ media culture were 39.7%
(122/307), 38.1% (117/307), 39.4% (121/307), 39.1% (120/307) and 45.3%
(139/307) respectively. The ICT (MPT64 Kit) and LJ media culture was
found with lowest and highest recovery rate, sensitivity, specificity, PPV and
NPV respectively. The overall comparative recovery rate was as follow ICT
(MPT64 Kit) < fluorescence microscopy < ZN microscopy < BACTEC 960
< LJ media. The LJ media has highest recovery rate and reversal time with
minimum contamination rate of 14/351 (3.98 %) as compared to BACTEC
960 system having the highest contamination rate of 46/351 (13.10 %) and
exception of analyzing blood and urine samples. ICT (MPT64 Kit),
fluorescence microscopy and ZN microscopy are found an essential
confirmatory technique.
Files
1-OPS+Journal_27_170225_June3.pdf
Files
(518.3 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:2b012d6b1e7f9a56514ae5abd42e10d3
|
518.3 kB | Preview Download |