Should the 'bleach microscopy method' be recommended for improved case detection of tuberculosis? Literature review and key person analysis.

SETTING: It has been proposed that the sensitivity of direct sputum smear microscopy can be improved if sputum is liquefied with sodium hypochlorite (NaOCl or household bleach), and concentrated by centrifugation before acid-fast staining. OBJECTIVE: To summarise the results of the studies of the bleach method for improved sensitivity of sputum microscopy and to describe the opinions and knowledge of key persons in National Tuberculosis Control Programmes (NTPs) about this method. DESIGN: We searched Medline, EMBASE and Web of Science for studies comparing the bleach method to direct sputum smear microscopy in low- or middle-income countries. Each study was assessed regarding methodology and field applicability. We also sent out questionnaires concerning the bleach method to key persons in NTPs in 85 countries. RESULTS: In 15 of the 19 studies identified there was a statistically significant improvement in the proportion of positive tests or sensitivity ranging from 7-253%. The majority (73%) of the key persons had heard of the bleach method. Forty-four per cent thought it could improve case detection in their countries, while 49% did not know; 93% of them would promote the bleach method; the most common reasons for doing so would be recommendations from the WHO or the IUATLD, or favourable studies performed in their own country. The bleach method was used routinely in only three countries. CONCLUSION: There is enough evidence to recommend the evaluation and introduction of the bleach method in most settings where mycobacterial culture is not performed routinely.

Evaluation of the BacT/ALERT 3D system for recovery and drug susceptibility testing of Mycobacterium tuberculosis.

We evaluated the BacT/ALERT 3D system for recovery and drug susceptibility testing (DST) of Mycobacterium tuberculosis (MTB). Of 2659 clinical specimens, MTB was detected in 92 using BacT/ALERT, compared to 94 using Löwenstein-Jensen culture. Detection time was 25% shorter with BacT/ALERT. Sensitivities were 92%, 96%, 78% and 100% for resistance to rifampicin, isoniazid, streptomycin and ethambutol, respectively, while specificity was 100% for all antibiotics, when BacT/ALERT was compared with the BACTEC 460 method on 50 MTB isolates. The BacT/ALERT system is fully automated and creates no radioactive waste. It seems to be a valid alternative for primary isolation, but further evaluation is needed regarding DST.