|Year : 2016 | Volume
| Issue : 2 | Page : 148-154
Use of RODAC plates to measure containment of Mycobacterium tuberculosis in a Class IIB biosafety cabinet during routine operations
Géraldine Daneau1, Elie Nduwamahoro1, Kristina Fissette1, Patrick Rüdelsheim2, Dick van Soolingen3, Bouke C de Jong4, Leen Rigouts5
1 Mycobacteriology Unit, Department of Biomedical Sciences, Institute of Tropical Medicine, Antwerp, Belgium
2 Perseus BVBA, Technologiepark-Zwijnaarde, Zwijnaarde, Belgium
3 Mycobacteria Reference Laboratory, National Institute for Public Health and the Environment, Bilthoven, The Netherlands
4 Mycobacteriology Unit, Department of Biomedical Sciences, Institute of Tropical Medicine, Antwerp, Belgium; Division of Infectious Diseases, New York University, New York, NY, USA; Vaccination Theme, Medical Research Council Unit, Fajara, The Gambia
5 Mycobacteriology Unit, Department of Biomedical Sciences, Institute of Tropical Medicine; Department of Biomedical Sciences, University of Antwerp, Antwerp, Belgium
|Date of Web Publication||9-Feb-2017|
Institute of Tropical Medicine, Nationalestraat 155, 2000 Antwerpen
Source of Support: None, Conflict of Interest: None
Objective/background: Guidelines for the manipulation of Mycobacterium tuberculosis (MTB) cultures require a Biosafety Level 3 (BSL-3) infrastructure and accompanying code of conduct. In this study, we aimed to validate and apply detection methods for viable mycobacteria from surfaces in a BSL-3 MTB laboratory.
Methods: We evaluated phenotypic (Replicate Organism Detection and Counting [RODAC] plates) and molecular (propidium monoazide [PMA]-based polymerase chain reaction [PCR]) approaches for the detection of viable mycobacteria, as well as the effect of 70% ethanol applied for 5min for disinfection against mycobacteria. For validation of the method, recovery of serial dilutions of Mycobacterium bovis bacillus Calmette–Guérin from glass slides was measured. Subsequently, we stamped surfaces in and around the biosafety cabinet (BSC) after different technicians had manipulated high bacterial load suspensions for routine drug-susceptibility testing in a Class II BSC.
Results: RODAC stamping could detect as few as three bacteria on slides stamped either 5min or 60min after inoculation. PMA-based PCR, tested in parallel, did not pass validation. Mycobacteria were still detected after 5-min disinfection with ethanol 70%. In the BSL-3, from 201 RODAC-stamped surfaces, MTB was detected in four: three inside a BSC—on a tube cap and on an operator's gloves—and one outside, on an operator's gown.
Conclusion: RODAC plates detect mycobacteria at low numbers of microorganisms. In addition, this method allowed us to show that 70% ethanol does not reliably kill mycobacteria when applied for 5min to a dried surface, and that MTB bacilli may arrive outside a Class II BSC during routine practice, although the route could not be documented.
Keywords: Biosafety, Environmental sampling, Ethanol, Propidium monoazide, Replicate Organism Detection and Counting, Tuberculosis
|How to cite this article:|
Daneau G, Nduwamahoro E, Fissette K, Rüdelsheim P, van Soolingen D, de Jong BC, Rigouts L. Use of RODAC plates to measure containment of Mycobacterium tuberculosis in a Class IIB biosafety cabinet during routine operations. Int J Mycobacteriol 2016;5:148-54
|How to cite this URL:|
Daneau G, Nduwamahoro E, Fissette K, Rüdelsheim P, van Soolingen D, de Jong BC, Rigouts L. Use of RODAC plates to measure containment of Mycobacterium tuberculosis in a Class IIB biosafety cabinet during routine operations. Int J Mycobacteriol [serial online] 2016 [cited 2020 Jun 1];5:148-54. Available from: http://www.ijmyco.org/text.asp?2016/5/2/148/199923
| Introduction|| |
Manipulation of Mycobacterium tuberculosis (MTB) cultures requires Biosafety Level 3 (BSL-3) infrastructure, equipment, and practices . Use of liquid culture media and inoculating-loop manipulation increase the risk of creating aerosols, while only a few mycobacteria are sufficient to infect a laboratory worker ,. The emergence of multidrug-resistant and extremely resistant strains requires particular attention to containment, both by the importance an infection may imply and by the increased volume of drug-susceptibility testing (DST) performed ,. Despite the general recognition that BSL-3 measures effectively contain airborne, high-risk pathogens, there is little information on the concrete value of individual measures commonly incorporated in BSL-3 practices ,.
Culture-based tools are commonly used for detection of microorganisms in the environment. The presence of viable airborne particulates can be estimated using settling plates , whereas monitoring for surface-bound particulates can be done using Replicate Organism Detection and Counting (RODAC) plates, without the need for a decontamination step that could decrease the sensitivity of detection ,,,,,. To allow for detection of slow-growing mycobacteria, a modified medium selective for preventing overgrowth by nonmycobacterial contaminants and enriched for supporting the growth of these fastidious microorganisms should be used . Although this method has to date not been validated for mycobacteria, RODAC recovery rates described for other bacteria are highly variable; for example, for Listeria monocytogenes, the recovery rate is only 4.15%, and for Clostridium difficile, it is dependent on the medium used, and ranges from 0.3% to 45.8% ,.
Molecular methods are powerful and fast tools to detect traces of genetic material . In particular, propidium monoazide (PMA)–polymerase chain reaction (PCR) is useful in this regard as it selectively amplifies DNA from viable bacteria . This method has already been successfully applied to diagnose patients with drug-resistant MTB infection ,.
Surface decontamination of materials leaving the biosafety cabinet (BSC) is recommended, despite lack of evidence that MTB can be reaerosolized from surfaces, and subsequently establish infection ,. Where carried out, such decontamination should be done with a fast-acting disinfectant that does not affect bacilli inside culture tubes. Ethanol (70%) is used in some settings, however, data on its effectiveness are contradictory, with viable bacilli detected 1min, 5min, or 10min after decontamination in two seminal studies ,.
Although BSL-3 standards call for a Class II cabinet (open, with laminar flow), this is based primarily on data from flow measurements rather than experimental evidence with bacterial detection, and these Class II cabinets are used in many laboratories handling MTB worldwide, in both high- and low-resource settings, where most laboratories do not meet all BSL-3 standards .
In this study, we evaluated both a culture-based and a molecular method for detection of mycobacteria from surfaces. We subsequently used the validated culture-based method to test the effect of ethanol decontamination of surfaces and to sample different surfaces in the environment after routine manipulation of MTB in a BSC Class II B1 .
| Material and methods|| |
Validation experiments were performed using the nonpathogenic MTB complex (MTBC) Mycobacterium bovis bacillus Calmette–Guérin (M. bovis BCG; ITM-110295), which was cultured on home-made Stonebrink medium (Löwenstein–Jensen with pyruvate) for 3–4weeks.
For quality control of media, we used the MTB reference strain (H37Rv; ITM-083715), M. bovis BCG (ITM-110295), and Mycobacterium terrae (ITM-110258) from the BCCM mycobacteria culture collection (Belgian Coordinated Collections of Microorganisms, Institute of Tropical Medicine, Antwerp, Belgium).
Because the sampling was part of a risk analysis in our TB laboratory, in accordance with our biosafety consultant, ethical approval was neither required for the BCCM samples nor for anonymous sampling in the BSL-3 laboratory, including those stamped on technicians' laboratory coats.
RODAC plates were prepared using Middlebrook 7H11 agar, consisting in DIFCO Mycobacteria 7H11 agar (Becton Dickinson, Erembodegem, Belgium) supplemented with 2.1%, Middlebrook oleic acid–albumin–Catalase–dextrose enrichment 10% [1.5% alpha-d(+)-glucose (Acros Organics, Geel, Belgium), 0.06% oleic acid (Sigma–Aldrich, Diegem, Belgium), 5% albumin bovine fraction V with pH modified at 6.3 (Acros Organics)], glycerol 0.5% (Merck, Overijse, Belgium), ticarcillin 50mg/L, trimethoprim 10mg/L, amphotericin B 10mg/L, and colistin 200mg/L (all antibiotics from Sigma–Aldrich). Each plate was prepared with 16mL of medium, shaped as a parabolic meniscus, and stored at 4–8°C for a maximum of 6weeks. Each lot was validated by direct inoculation of one plate with MTB H37Rv or M. bovis BCG, and another one with M. terrae. Both plates should yield growth after 7–14-day incubation at 34–38°C with 7.5% CO2.
To validate the recovery of mycobacterial bacilli from surfaces, a suspension (1mg/mL) of M. bovis BCG was prepared [about 106–107 colony-forming units (CFU)/mL according to positive control plates] with serial dilution until achieving 1CFU on 3 different days (each starting from a different BCG subculture). Nonsterile microscopy slides were inoculated with 50μL of those dilutions and placed in the BSC for subsequent stamping by the RODAC plates at different time points: within 10s, about 5min, or 60min after pipetting. Each slide was stamped once, and the stamping covered about 1 cm2. At this point, slides had not yet completely dried except for the 60-min time point. The contact time between slide and RODAC plate was 1s in the two first adjustment experiments and subsequently 1min. Where mentioned, slides were moistened with 50μL of sterile water and added after drying time. All BCG stamping validation experiments were performed in triplicate. Direct inoculation of the dilutions on a RODAC plate served as the positive control to calculate the capture rate with stamping. In the first experiment, an additional control was performed on Petri dishes with plain 7H11 medium. Colonies were counted after 28days' incubation and considered negative at 10weeks.
The capture rate was calculated as the mean number of colonies on the stamped plates, divided by the mean number of colonies on the directly inoculated RODAC plate of the respective bacterial suspension. When the number of colonies was >50 /plate, it was categorized as 1+ (50–100 colonies), 2+ (100–200 colonies), 3+ (>200 colonies), or ∞ (confluent).
In three different experiments, half of the slides were inactivated by adding 50μL of 70% ethanol (with 2.1% isopropyl alcohol; from Yvsolab, Turnhout, Belgium) after 5min or 60min, and subsequently stamped after 5-min ethanol contact.
A 1-mg/mL M. bovis BCG suspension was serially diluted in molecular-grade water and 10μL of each dilution (about 103 to <1CFU) was incubated with PMA (Biotium, Hayward, CA, USA), for various combinations of incubation time (3–15min) and PMA concentration (10–100μM). The suspensions were subsequently centrifuged for 20min at 4000g, washed, and resuspended in Tris–EDTA buffer. All tubes were exposed to 400+230W halogen light oriented at 45° at 20-cm distance for different durations (2–5min). Negative controls were obtained by boiling part of the serial dilutions for 5min before adding PMA, to kill all mycobacteria (internal validation). Mixtures of dead and live bacteria were obtained before PMA treatment by mixing boiled and nonboiled suspensions at ratios of 1:1, 5:1, and 25:1 at different microorganism concentrations.
All light-treated suspensions were lysed by overnight incubation with proteinase K, followed by boiling for 10min. DNA was extracted following the modified Mangiapan method , which involved fishing DNA with a bead-bound probe specific for the 16S ribosomal RNA gene of MTBC (biotin–triethylene glycol [TEG]-CGTGGGTGATCTGCCCTGCAC and biotin–TEG-CGTCAATCCGAGAGAACCCGGA) and capturing bound DNA with streptavidin magnetic beads. PCR was performed with SYBR Green MTBC-specific primers for the 16S gene (GCGCTTTAGCGGTGTGGGATGA and TCGTCGCCTTGGTAGGCCGT). A DNA-positive control for the PCR and for the extraction was based on purified MTB H37Rv DNA at known concentration, and was used to validate the extraction and PCR method (without PMA). Sterile water was used as a negative control at each step.
Sampling in the BSL-3 TB laboratory
Sampling was performed in the BSL-3 laboratory at the ITM Mycobacteriology Unit in Antwerp, Belgium. After completion of manipulations for DST, multiple places on the operator and in and around the BSC Class IIB were stamped by RODAC plates, with a 1-min contact time. Those DST manipulations, considered at risk, were performed on routine or research samples growing MTB, using serial 102× and 104× dilutions from a 1-mg/mL bacterial suspension (approximately 106–107 CFU/mL) for first-line DST, and 10× and 103× dilutions for second-line DST. As for the RODAC validation step, colonies were counted after 28days' incubation, and considered negative at 10weeks. Strains were identified as MTBC with the MPT 64 antigen test (SD Bioline, Yongin, Korea).
| Results|| |
Validation of the RODAC method
Growth of the positive control, the direct inoculum, was similar between the antibiotic-containing RODAC plate and Petri dish with plain 7H11 medium ([Table 1]). The CFUs obtained on control plates were in line with CFUs from routine positive control plates.
With a 1-s contact time between RODAC plates and the slides, the capture rate by stamping was not linear. Compared with direct inoculation, RODAC stamping with a 1-s contact time detected bacteria directly or after 5min, but none were detected when stamping after 60min, even after moistening (data not shown). When the contact time was increased to 1min, the capture rate increased, especially at 60min. Upon analyzing plates with countable numbers of colonies, the capture rate was found to be 60% (standard deviation [SD] 35%) after 5min, and 32% (SD 38%) after 60min (confidence interval 0–100% for both stamping times). Individual measurements are shown in [Figure 1] A. In addition, a suspension with 2+ growth (on control) showed a similar growth when stamped after 5min, and a 1+ growth when stamped after 60min. No contamination was observed.
|Figure 1: Validation experiments. (A) Stamping efficiency and (B) ethanol effect. Graph showing the number of colonies growing on each of the three RODAC plates on 2 days at the most discriminatory dilution (103× in 1 experiment and 103× and 104× on another day). Individual measures are represented by symbols: black circles (Series 1), light grey squares (Series 2), dark grey triangles (Series 3), whereas mean and standard deviation from triplicates are represented by horizontal lines: solid (Series 1), long-dashed (Series 2), or short-dashed (Series 3). (A) Control represents the direct inoculation on a RODAC plate; Sampling 5′ and Sampling 60′ are from RODAC plates stamped after 5min and 60 min, respectively, with a 1-min contact time. (B) Control 5′ represents RODAC plates stamped after 5 min, without adding ethanol. Ethanol 70% was added 5 min after bacterial inoculation on a microscopy slide, and stamped 5 min after ethanol addition. RODAC = Replicate Organism Detection and Counting.|
Click here to view
Inactivation by ethanol
On average, slides inoculated with diluted M. bovis BCG grew seven colonies; addition of ethanol 5min after inoculation reduced the colony growth to six (87%, SD 25%). Individual measurements are shown in [Figure 1]B. When ethanol was added 60min after inoculation, the number of colonies decreased from 1+ to 0–3 colonies (range) or, for low dilutions (3 and 14 colonies on average), to zero.
Optimization of the PMA–PCR method
Linearity of the DNA extraction and quantitative PCR without PMA was established from 10 to 104 DNA copies (1 DNA copy/bacterium). Detection of <10 bacteria was possible, but not consistently: five of the ten experiments had a signal for the theoretic point of one bacterium. Although PMA has a theoretical ability to improve discrimination between viable and nonviable bacteria, we observed that PCR signals from dead bacteria were not completely inhibited after reaction with PMA, whereas PCR signals from live bacteria were inhibited in proportions similar to dead bacteria after reaction with PMA.
Sampling in BSL-3 TB laboratory with RODAC plates
On 30 different days, after one of eight different technicians had just completed manipulations for DST of MTB, we sampled from a number of different locations, including the technician's double-gloved little finger, both (gowned) forearms and thighs, two positions on the arm rest of the BSC, and the caps of inoculated MTB culture tubes (MGIT, LJ, or 7H11). In total, 201 RODAC plates were stamped. The exact locations are shown in [Figure 2]. [Figure 2] also shows how the little finger is used during DST manipulations. The schedule of sampling is summarized in [Table 2].
|Figure 2: Locations of RODAC plates. (A) Stamping on the forearm, lap, and arm rest. (B) Stamping on tube caps. (C) The manner in which the little finger is used during the manipulation to open tube caps. (D) Stamping of the little finger. RODAC = Replicate Organism Detection and Counting.|
Click here to view
From those 201 stampings, five RODAC plates were positive on different days: two colonies and 1+ growth, respectively, on two plates stamped on little fingers, one colony from a tube cap (on the same day as the finger plate which grew 2 colonies), one colony from a forearm, and one colony from a thigh. All plates grew MTBC, except the positive RODAC plate from the forearm, with one colony that failed to grow on subculture and could not be further identified. No contamination by non-mycobacteria was observed.
| Discussion|| |
The assessment of TB biosafety by monitoring the TB laboratory environment for the presence of viable mycobacteria is hampered by the absence of an established methodology and scanty data ,. We developed an approach to assess the presence of mycobacteria on surfaces in the TB laboratory, and present data on the efficacy of decontamination with 70% ethanol.
Stamping with RODAC plates using a 1-min contact time showed a high sensitivity on glass, down to 3CFU, yet with a wide confidence interval for capture rate; some replicates could not be cultured, suggesting that stamping at least three plates is necessary to evaluate the absence of viable mycobacteria. Sensitivity is presumably increased by extending sampling to multiple surfaces.
Our capture rate with MTBC appears to be in line with results obtained with other bacteria, detecting bacteria at a variable rate up to 50% ,. Unlike C. difficile, our results on positive controls suggest that antibiotics used for selectivity did not have a detectable inhibitory effect on mycobacterial growth . This finding suggests that a specific liquid growth medium could be developed for inoculation from non-decontaminated swabs, potentially shortening growth time ,. As a different application, this medium might be tested in settings where TB cultures present a high contamination rate, not sufficiently managed when following the World Health Organization recommendations .
Our RODAC approach was considered sufficiently sensitive and selective to detect mycobacteria on surfaces in the BSL-3 TB laboratory, and therefore could be used for biosafety assessment in TB laboratories worldwide. In our case, we applied the RODAC method to operators (technicians) and the environment following DST, in which high concentrations of bacteria (1mg/mL, about 106–107 CFU/mL) are manipulated in liquid medium. This experiment took place in a routine environment that respects all current recommendations for TB manipulation: technicians had been trained on biosafety and specific techniques according to the good laboratory practices in place in our ISO 15189-accredited (BSL-3) laboratory, and the BSC is checked and maintained every 6months according to recommendations. It is not clear whether the MTB colony on the technician's gown arrived directly by aerosol from the (open front) Class II BSC, or by contact with a contaminated finger, which might occur if unnoticed contaminated gloves are not removed inside the BSC prior to taking one's hand out to retrieve additional material not stored in the BSC. Even when an (contaminated) outer glove is removed inside the BSC, inner gloves may also be contaminated by colonies on the tube caps when specimens are removed from the BSC. Laminar flow might also have been disturbed by movements around the BSC, like incubator or laboratory door opening . In our case, stamping on the arm rest, at risk of contamination in case of an airborne leak, was always negative, making airborne contamination, which would constitute a higher biosafety risk, less likely. Contamination on the laboratory gown represents an additional risk of transporting mycobacteria away from the BSC, with potential reaerosolization when removing the gown.
The present results prompted us to retrain our staff on always removing the outer gloves upon taking hands out of the BSC, even if only briefly. In addition, a customized closed-front Class II BSC has been installed in our BSL-3 laboratory. A closed-front BSC would better contain mycobacteria at the source, dramatically reducing the potential risk of any direct escape of aerosols . Moreover, using wrist cuffs and double gloves reinforces removal of the outer gloves before exiting the arm sleeves. Such an approach may provide an alternative to a negative pressure laboratory in settings with limited resources, with priorities focused on a well-functioning BSC, its maintenance and calibration, training, and its periodic reinforcement .
Detection of bacteria on laboratory coat, gloves, and tubes also showed that RODAC stamping with our adapted medium works on surfaces other than glass, even if capture rate may be different.
PMA had been shown to be a promising approach to identify DNA from live bacteria, including mycobacteria ,,. Previous results were however from experiments with large numbers of bacteria, and PMA was used to distinguish between dead and living bacteria, as an alternative to DST. Our results suggest that detecting scarce live mycobacteria among a dead mycobacterial population is much more difficult, and PMA–PCR does not appear to be suitable for environmental monitoring.
In line with the results of Best et al. , 70% ethanol does not appear to be effective at killing mycobacteria, although our sample size is too small to draw firm conclusions.
| Conclusions|| |
In conclusion, we validated an approach—mycobacterium-specific RODAC stamping—that can be readily applied in TB laboratories to monitor biosafety. Our results further call into question whether open Class II BSC and current recommendations are sufficient to contain mycobacteria at the source in a routine environment, particularly when handling high-risk specimens. While ethanol did not completely kill mycobacteria, further experiments are required to compare disinfectants suitable for decontaminating culture tubes without affecting their content's viability.
| Conflicts of interest|| |
| Acknowledgments|| |
This work was partially supported by European Research Infrastructure on Highly Pathogenic Agents of European Commission (Grant No 262042); a European Research Council Starting Grant to B.C.dJ. (Grant No 311725), and the Fonds Wetenschappelijk Onderzoek by Flemish Government from Belgium. We thank Professor Bruno Gryseels for helpful suggestions and discussions regarding approaches to evidence-based biosafety improvements, and Dr. Kevin Peterson for his useful feed-back on the manuscript.
G.D., P.R., B.C.dJ., and L.R. designed the study; D.v.S. developed the RODAC medium; G.D., E.N., and K.F. performed the laboratory experiments; G.D., B.C.dJ., and L.R. wrote the article; all authors approved the final version.
| References|| |
World Health Organization (WHO), Tuberculosis Laboratory Biosafety Manual, WHO, Geneva, Switzerland, 2012.
R.L. Riley, C.C. Mills, W. Nyka, et al, Aerial dissemination of pulmonary tuberculosis. A two-year study of contagion in a tuberculosis ward. 1959, Am. J. Epidemiol. 142 (1995) 3–14.
D.L. Sewell, Laboratory-associated infections and biosafety, Clin. Microbiol. Rev. 8 (1995) 389–405.
World Health Organization (WHO), Global Tuberculosis Report, WHO, Geneva, Switzerland, 2013.
G.B. Migliori, A. Matteelli, D. Cirillo, et al, Diagnosis of multidrug-resistant tuberculosis and extensively drugresistant tuberculosis: current standards and challenges, Can. J. Infect. Dis. Med. Microbiol. 19 (2008) 169–172.
T.M. Shinnick, C. Gilpin, A risk assessment-based approach to defining minimum biosafety precautions for tuberculosis laboratories in resource-limited settings, Appl. Biosaf. 17 (2012) 6–10.
T.G. Kimman, E. Smit, M.R. Klein, Evidence-based biosafety: a review of the principles and effectiveness of microbiological containment measures, Clin. Microbiol. Rev. 21 (2008) 403– 425.
A. Lu, L.J. Alto, M.G. Prusia, Reliability of the settling-plate method in monitoring laminar-airflow benches, Am. J. Hosp. Pharm. 40 (1983) 271–273.
D.M. Hacek, W.E. Trick, S.M. Collins, et al, Comparison of the RODAC imprint method to selective enrichment broth for recovery of vancomycin-resistant enterococci and drugresistant Enterobacteriaceae from environmental surfaces, J. Clin. Microbiol. 38 (2000) 4646–4648.
L. Poletti, C. Pasquarella, M. Pitzurra, et al, Comparative efficiency of nitrocellulose membranes versus RODAC plates in microbial sampling on surfaces, J. Hosp. Infect. 41 (1999) 195–201.
J.A. Brewer, A.G. Turner, Replicating RODAC plates for identifying and enumerating bacterial contamination, Health Lab. Sci. 10 (1973) 195–202.
E.B. Smith, I.J. Raphael, M.G. Maltenfort, et al, The effect of laminar air flow and door openings on operating room contamination, J. Arthroplasty 28 (2013) 1482–1485.
D. D'Alessandro, F. Cerquetani, M.G. Deriu, et al, Evaluation of fungal contamination in operating rooms using a dusting cloth pad: comparison among different sampling methods, Am. J. Infect. Control 41 (2013) 658–660.
A.F. Guiteras, L.H. Flett, R.L. Shapiro, A quantitative method for determining the bacterial contamination of dishes, Appl. Microbiol. 2 (1954) 100–101.
R.J. Dubos, G. Middlebrook, Media for tubercle bacilli, Am. Rev. Tuberc. 56 (1947) 334–345.
B.P. Buggy, K.H.Wilson, R. Fekety, Comparison of methods for recovery of Clostridium difficile from an environmental surface, J. Clin. Microbiol. 18 (1983) 348–352.
D. Gomez, A. Arino, J.J. Carraminana, et al, Comparison of sampling procedures for recovery of Listeria monocytogenes from stainless steel food contact surfaces, J. Food Prot. 75 (2012) 1077–1082.
K.B. Mullis, Target amplification for DNA analysis by the polymerase chain reaction, Ann. Biol. Clin. (Paris) 48 (1990) 579–582.
A. Nocker, C.Y. Cheung, A.K. Camper, Comparison of propidium monoazide with ethidium monoazide for differentiation of live vs. dead bacteria by selective removal of DNA from dead cells, J. Microbiol. Methods 67 (2006) 310– 320.
S. Pholwat, S. Heysell, S. Stroup, et al, Rapid first- and secondline drug susceptibility assay for Mycobacterium tuberculosis
isolates by use of quantitative PCR, J. Clin. Microbiol. 49 (2011) 69–75.
P. Miotto, S. Bigoni, G.B. Migliori, et al, Early tuberculosis treatment monitoring by Xpert MTB/RIF, Eur. Respir. J. 39 (2012) 1269–1271.
Y. Qian, K. Willeke, S.A. Grinshpun, et al, Performance of N95 respirators: reaerosolization of bacteria and solid particles, Am. Ind. Hyg. Assoc. J. 58 (1997) 876–880.
M. Best, S.A. Sattar, V.S. Springthorpe, et al, Efficacies of selected disinfectants against Mycobacterium tuberculosis
, J. Clin. Microbiol. 28 (1990) 2234–2239.
P. Chedore, C. Th'ng, D.H. Nolan, et al, Method for inactivating and fixing unstained smear preparations of Mycobacterium tuberculosis
for improved laboratory safety, J. Clin. Microbiol. 40 (2002) 4077–4080.
T. Hanscheid, M.P. Grobusch, Biosafety and tuberculosis laboratories in Africa, Lancet Infect. Dis. 10 (2010) 582–583.
NSF International Standard/American National Standard, Biosafety Cabinetry: Design, Construction, Performance, and Field Certification, Annex E (NSF/ANSI 49-2011), NSF International, Ann Arbor, MI, 2011.
G. Mangiapan, M. Vokurka, L. Schouls, et al, Sequence capture-PCR improves detection of mycobacterial DNA in clinical specimens, J. Clin. Microbiol. 34 (1996) 1209– 1215.
D.J. Malik, K.V. Patel, M.R. Clokie, et al, On the difficulties of isolating Clostridium difficile from hospital environments, J. Hosp. Infect. 84 (2013) 181–183.
G. Middlebrook, Z. Reggiardo, W.D. Tigertt, Automatable radiometric detection of growth of Mycobacterium tuberculosis
in selective media, Am. Rev. Respir. Dis. 115 (1977) 1066–1069.
G.E. Pfyffer, H.M. Welscher, P. Kissling, et al, Comparison of the mycobacteria growth indicator tube (MGIT) with radiometric and solid culture for recovery of acid-fast bacilli, J. Clin. Microbiol. 35 (1997) 364–368.
S. Kumari, R.L. Ichhpujani, Guidelines on Standard Operating Procedures for Microbiology, World Health Organization, New Delhi, India, 2000 (pp. 113–118).
D. van Soolingen, H.J. Wisselink, R. Lumb, et al, Practical biosafety in the tuberculosis laboratory: containment at the source is what truly counts, Int. J. Tuberc. Lung. Dis. 18 (2014) 885–889.
[Figure 1], [Figure 2]
[Table 1], [Table 2]