Decontamination of a BSL3 laboratory by hydrogen peroxide fumigation using three different surrogates for Bacillus anthracis spores

Authors

  • O. Kaspari,

    Corresponding author
    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    • Correspondence

      Oliver Kaspari, Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Nordufer 20, 13353 Berlin, Germany.

      E-mail: kaspario@rki.de

    Search for more papers by this author
  • K. Lemmer,

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author
  • S. Becker,

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author
  • P. Lochau,

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author
  • S. Howaldt,

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author
  • H. Nattermann,

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author
  • R. Grunow

    1. Division Highly Pathogenic Microorganisms, Centre for Biological Threats and Special Pathogens, Robert Koch-Institute, Berlin, Germany
    Search for more papers by this author

Abstract

Aims

Two independent trials investigated the decontamination of a BSL3 laboratory using vaporous hydrogen peroxide and compared the effect on spores of Bacillus cereus, Bacillus subtilis and Bacillus thuringiensis as surrogates for Bacillus anthracis spores, while spores of Geobacillus stearothermophilus served as control.

Methods and Results

Carriers containing 1·0 × 106 spores were placed at various locations within the laboratory before fumigation with hydrogen peroxide following a previously validated protocol. Afterwards, carriers were monitored by plating out samples on agar and observing enrichment in nutrient medium for up to 14 days. Three months later, the experiment was repeated and results were compared. On 98 of 102 carriers, no viable spores could be detected after decontamination, while the remaining four carriers exhibited growth of CFU only after enrichment for several days. Reduction factors between 4·0 and 6·0 log levels could be reached.

Conclusions

A validated decontamination of a laboratory with hydrogen peroxide represents an effective alternative to fumigation with formaldehyde. Spores of B. cereus seem to be more resistant than those of G. stearothermophilus.

Significance and Impact of the Study

The results of this study provide important results in the field of hydrogen peroxide decontamination when analysing the effect on spores other than those of G. stearothermophilus.

Introduction

Biosafety level 3 and 4 laboratories are required by law (EC Directive 2009/41/EC) to use a reliable and validated method of disinfection for whole room treatment, following accidental spillage or routine maintenance. Further, since the anthrax attacks in 2001, decontamination of rooms or facilities by fumigation became a major issue. In this regard, the fumigation procedures had to be effective against highly resistant Bacillus spores. Using paraformaldehyde, chlorine dioxide or hydrogen peroxide as fumigants, each shows advantages and disadvantages from the practical point of view (Canter 2005; Canter et al. 2005). While efficacy is of paramount importance, it is also essential that such a method does not affect technical equipment which is always present in laboratories, and that surfaces and materials do not corrode or otherwise deteriorate.

Room decontamination with formaldehyde is a widely accepted procedure specified as being effective against bacteria and viruses in the list of tested and approved disinfectants by the Robert Koch-Institute (Anon 2013). However, it is known that formaldehyde fumigation is not a secure method to kill spores and conditions of 30–65°C and 60–99% humidity or long dwell times are applied to reach sufficient sporicidal efficiency (Munro et al. 1999; Rogers et al. 2007; Schaal et al. 2009). These parameters are difficult to maintain when conducting decontamination of a whole laboratory; however, these are not necessary when using the vaporous hydrogen peroxide method (Dietz et al. 1980; Jahnke and Lauth 1996). In case of hydrogen peroxide fumigation, the necessary parameters are generated automatically by the commercially available devices. Even for hydrogen peroxide, a high humidity is obligate for an effective fumigation. In case of the technique used in our study, humidity in the sterilization phase is about 85–95%. Currently, using hydrogen peroxide is more and more favoured as a technique for decontamination of biosafety cabinets, larger possibly contaminated spaces or when dealing with biological agents including bacterial spores. For evaluating the method, predominantly spores are used as bioindicators (Fey et al. 2010; Meyer et al. 2013). Additionally, hydrogen peroxide decomposes completely into water and oxygen although some materials may absorb vapours and release them for some time after fumigation (Beswick et al. 2011). In contrast, formaldehyde vapours leave precipitations, especially when neutralized by ammonia, that have to be and cleaned by hand which overall might be more harmful to surfaces and technical equipment (Dietz et al. 1980; Cheney and Collins 1995).

Decontamination of rooms and equipment with gaseous hydrogen peroxide is used increasingly, especially in health care, pharmaceutical facilities and laboratories (Klapes and Vesley 1990; French et al. 2004; Kahnert et al. 2005; Andersen et al. 2006; Krishnan et al. 2006; Hall et al. 2007). However, most studies in this field have focused on clinically relevant agents such as Clostridium difficile (Boyce et al. 2008; Otter and French 2009; Davies et al. 2011), MRSA (French et al. 2004; Pottage et al. 2012) and Mycobacterium tuberculosis (Kahnert et al. 2005; Andersen et al. 2006; Hall et al. 2007; Grare et al. 2008).

The biocidal effect of vaporous hydrogen peroxide is based on its oxidizing properties. Oxygen radicals are generated which severely damage membrane proteins, nucleic acids or other essential bacterial or viral proteins (McDonnell and Russell 1999). The effect on bacterial spores is not yet fully understood. However, it has been shown that treated spores prove unable to take up water, thus preventing germination (Melly et al. 2002).

In this article, we show almost complete decontamination of a BSL3 laboratory after distribution of carriers coated with test organisms serving as surrogates for anthrax spores and making it possible to translate results into a scenario in which a contamination with Bacillus anthracis occurs. To analyse the efficacy of vaporous hydrogen peroxide, spores of Bacillus thuringiensis, Bacillus subtilis and Bacillus cereus were used as surrogates, while spores of Geobacillus stearothermophilus (international standard for decontamination by heat and hydrogen peroxide) served as a control agent.

Materials and methods

Preparation of carriers

Spores of B. thuringiensis DSM 350, B. subtilis ATTC 6633 and B. cereus ATTC 10987 were prepared according to the European norm (EN) 14347:2005 which was slightly modified in that cell culture flasks with filter caps were used to ensure constant ventilation, while centrifugation was performed in 50-ml tubes at 4°C. Spore suspensions obtained were adjusted to 108 spores per millilitre by counting of colony forming units (CFU) on tryptone soy agar (TSA; Oxoid, Hampshire, UK).

A 10-μl drop of the respective spore solution, which contained 1·0 × 106 spores, was spread onto a glass slide and dried overnight inside a microbiological safety cabinet. Actual amounts of spores on the carriers were determined by immersing the glass slides along with glass pearls (diameter:3 mm) in five ml of Tryptone Soy Broth (TSB; Oxoid) and putting them on a shaker for 5 min at 400 rounds per min. One hundred microlitres of the supernatant was subsequently plated out on TSA, and spore numbers were determined by counting CFU after incubation at 37°C overnight. These numbers, representing spores that could be detached mechanically from the glass slides by shaking with glass pearls, served as the basis for calculating reduction factors after decontamination and at the same time as positive controls to ensure that carriers contained viable spores. Additionally, the same experiment was performed in neutralization medium instead of pure TSB. This medium consisted of TSB modified with 9% Tween 80, 0·9% lecithin and 3% histidine per litre and was used later to neutralize potential residues of hydrogen peroxide, while no difference concerning obtained CFU was detectable when comparing the neutralization medium to pure TSB. This should rule out a negative impact on bacteria growth through residual oxygen radicals. Prepared carriers were put into petri dishes and placed in various locations for hydrogen peroxide fumigation.

The G. stearothermophilus ATTC 12980 carriers used are commercially available (Apex Laboratories Inc.; Sanford, NC) and consist of small metal discs coated with 1 × 106 spores and enveloped in Tyvek®.

Hydrogen peroxide vaporization according to validated programme

For our experiment, we used a Geschko MLT 07 hydrogen peroxide gas generator (PEA; Koblenz, Germany) (Website: http://www.pea-gmbh.de/en_prod_geschko_MLT_07.htm), which nebulizes a 35% solution of hydrogen peroxide (AppliChem, Darmstadt, Germany). The procedure used was validated thrice, before the tests described here were performed, using 30 carriers coated with 1·0 × 106 spores of G. stearothermophilus as bioindicator at various locations and featured the following parameters: First, air humidity was withdrawn by the generator until it dropped below 32% relative humidity (rH), which took 20 min on average. Next, hydrogen peroxide vapour was delivered at a rate of 5·0 g min−1 for 60 min, followed by 2·5 g min−1 for 420 min with a targeted hydrogen peroxide concentration of at least 300 ppm. Afterwards, hydrogen peroxide vapours were neutralized by leading the gas stream over a catalysator for 720 min. An even distribution of hydrogen peroxide fumes was achieved by rotating fans (see Fig. 1) and tested by chemical indicators specific for hydrogen peroxide (Steris, Mentor, OH), although these indicators only show the presence of vapours and do not necessarily show that the target concentration has been reached at that position. The air stream of these fans was directed against the gas flow to spread vapours into corners and force the gas to take a complex route through the laboratory. During the decontamination procedure, air humidity reached levels of 85–95% in the sterilization phase and a rise of room temperature from 23 to 26°C was recorded. Decontamination was confirmed by incubating gassed carriers in nutrient medium for 7 days. This validation was done thrice, and no bacteria growth could be detected on any of the tested carriers after fumigation.

Figure 1.

Schematic of BSL3 laboratory rooms with preparations for hydrogen peroxide decontamination. The gas generator was placed into Laboratory 1, while hoses were laid out as above. Arrows show direction of the gas flow, indicating that hydrogen peroxide vapours were exhausted into Laboratory 2 and evenly distributed by three rotating fans placed in the open doorways. Room air is fed back into the generator via the hose in the anteroom. The short piece of hose ending in front of the emergency exit was plugged, as it was not used in this validation.

Preparation of the laboratory

Prior to hydrogen peroxide decontamination, freezers and the emergency door were hermetically sealed with tape; computers and microscopes were disinfected by hand and wrapped in plastic bags; and safety cabinets were turned on with the front shield open. Further, doors of incubators and centrifuges were left ajar; power plug lids, drawers, cabinets as well as inner laboratory doors were opened; and negative pressure and ventilation inside the BSL3 laboratory were switched off.

Gas hoses were laid out inside the laboratory rooms according to the validated protocol, and an even distribution of the hydrogen peroxide vapour was achieved by placing rotating fans in the open doorways (Fig. 1). Hoses touched the ground in as few places as possible to avoid hydrogen peroxide condensation on the floor, which is several degrees cooler than the room air.

Placement of carriers

Prepared carriers were placed inside open petri dishes according to the following table (Table 1 and Fig. 2). Two carriers of the same type were placed in each location to allow incubation in TSB as well as neutralization medium after room decontamination. These locations included spots that were easy as well as difficult to access and, in some cases, carriers were placed inside petri dishes with closed lids. In addition, commercially available carriers with G. stearothermophilus were placed in nine additional locations to serve as a control.

Table 1. Placement of carriers
Carrier No.LocationAgentHeight (m)
  1. Columns 1–3 indicate the corresponding carrier number, their location in the rooms and the test agent. Column 4 shows the distance of the carriers from the floor in metres. Carriers at positions 22–30 served as a control, as the validation of the fumigation procedure had been performed with G. stearothermophilus.

1Top of air duct Bacillus cereus 2·15
2Top of air duct Bacillus subtilis 2·15
3Laboratory table (closed lid) Bacillus thuringiensis 0·76
4Box on laboratory table B. cereus 1·00
5Top of material cabinet B. subtilis 2·10
6On incubator B. thuringiensis 1·75
7Laboratory table B. cereus 0·76
8Floor below safety cabinet (closed lid) B. subtilis 0
9Top of air duct B. thuringiensis 2·15
10Top of air duct B. cereus 2·15
11Laboratory chair B. subtilis 0·45
12On freezer B. thuringiensis 1·97
13On H2O2 generator (closed lid) B. cereus 1·40
14On incubator (closed lid) B. subtilis 1·83
15On refrigerator B. thuringiensis 1·80
16Top of air duct (closed lid) B. cereus 2·15
17On fuse box B. subtilis 1·77
18Inside fuse box B. thuringiensis 1·34
19Top of storage shelf B. cereus 1·65
20Lowest board of storage shelf B. subtilis 0·40
21On incubator B. thuringiensis 1·20
22Top of storage shelf Geobacillus stearothermophilus 1·65
23On fuse box G. stearothermophilus 1·42
24Inside air duct G. stearothermophilus 2·00
25On incubator G. stearothermophilus 1·00
26Inside air duct G. stearothermophilus 2·00
27On the front of H2O2 generator G. stearothermophilus 1·12
28On the front of H2O2 generator G. stearothermophilus 0·78
29On the back of safety cabinet G. stearothermophilus 1·40
30Inside incubator G. stearothermophilus 0·23
Figure 2.

Placement of carriers inside the rooms of the BSL3 laboratory. Numbers given in the room schematic show placement of the prepared carriers according to Table 1. Per location two carriers of the same type were placed to allow incubation in different media after fumigation with hydrogen peroxide.

Treatment and evaluation of carriers after hydrogen peroxide fumigation

When measurement of hydrogen peroxide concentrations in the air showed that a level had been reached that made entering safe, carriers were collected and analysed. Control carriers coated with 106 spores of G. stearothermophilus were removed from the Tyvek® envelopes, put into five ml of TSB and incubated at 37°C for 7 days according to the recommendations of the provider PEA. After this time period, a sample was plated out on TSA plates as an additional control. All other types of carriers were immersed in five ml of TSB or neutralization medium. Glass pearls (3 mm in diameter) were added, and samples were put on a shaker for 5 min at 400 rounds per min to remove adherent spores from the carriers and resuspend them. Afterwards, a sample of 100 μl was plated out on TSA, and CFU were counted after incubation at 37°C overnight. Reduction factors were calculated by subtracting the log value of the CFU ml−1 obtained by plate counting after decontamination from the log value of the untreated controls. Then, enrichment of the samples was monitored by incubating the media containing the detached spores and the carriers at 37°C for additional 14 days after which the medium was investigated for cloudiness. After that time another 100 μl were plated out on TSA plates, incubated at 37°C overnight and analysed for CFU.

Results

First testing

Before starting the first test run, detachable numbers of spores on the prepared carriers were analysed. Median percentages recovered were 42% (±3·0%) for B. cereus, 12% (±0·5%) for B. subtilis and 46% (±1·0%) for B. thuringiensis, respectively (numbers in parentheses indicate standard deviation from three countings) (Table 2).

Table 2. Results of first hydrogen peroxide decontamination
Carrier No.Direct analysisEnrichment (14 days)Observed log reduction
TSBNeutralization mediumTSBNeutralization medium
  1. Carrier numbers (column 1) refer to the position within the rooms as indicated in Table 1. Columns 2 and 3 refer to growth analysis of the carriers and indicate whether CFU were observed after resuspending spores either in TSB or neutralization medium and plating out a sample on TSA directly after fumigation. Columns 4 and 5 show whether clouding of the medium occurred due to bacterial growth after incubating detached spore samples for 14 days at 37°C, while column 6 gives the calculated reduction factor (RF), taking into account both plate counting and enrichment.

1+4·6
2≥5·1
3≥5·6
4+4·6
5≥5·1
6≥5·6
7≥5·6
8≥5·1
9≥5·6
10≥5·6
11≥5·1
12≥5·6
13≥5·6
14≥5·1
15≥5·6
16≥5·6
17≥5·1
18≥5·6
19≥5·6
20≥5·1
21≥5·6
22  ≥6·0
23  ≥6·0
24  ≥6·0
25  ≥6·0
26  ≥6·0
27  ≥6·0
28  ≥6·0
29  ≥6·0
30  ≥6·0

After the fumigation of the BSL3 laboratory, carriers were analysed as described previously and no CFU could be obtained when plating out samples of 100 μl on agar plates directly after fumigation, independent of organism or location. The only indication of bacterial growth was given by a clouding of the TSB medium containing carriers at positions 1 and 4. However, as this growth did not occur in neutralization medium and only showed after 3 days of enrichment, it can be considered as survival of very few, probably damaged spores of B. cereus. In the case of the neutralization media and the enrichment of all other carriers, no bacterial growth was detectable within the 14 days of observation. Regarding the control carriers coated with G. stearothermophilus, all samples were negative, and a reduction of more than 6·0 log10 levels could be shown, while for the other carriers, reduction was between 4·6 and 6·0 log10 levels.

Second testing

After this first test, another one was conducted 3 months later, following exactly the same conditions as applied during the first one. Again, detachable numbers of spores on the prepared carriers were investigated before starting the experiment. Median percentages recovered were 11% (±6·0%) for B. cereus, 46% (±35·8%) for B. subtilis and 37% (±3·0%) for B. thuringiensis, respectively (Table 3).

Table 3. Results of second hydrogen peroxide decontamination
Carrier No.Direct analysisEnrichment (14 days)Observed log reduction
TSBNeutralization mediumTSBNeutralization medium
1≥5·0
2≥5·6
3≥5·6
4≥5·0
5≥5·6
6≥5·6
7≥5·0
8≥5·6
9≥5·6
10≥5·0
11≥5·6
12≥5·6
13≥5·0
14≥4·6
15≥5·6
16+4·0
17≥5·6
18+4·6
19≥5·0
20≥5·6
21≥5·6
22  ≥6·0
23  ≥6·0
24  ≥6·0
25  ≥6·0
26  ≥6·0
27  ≥6·0
28  ≥6·0
29  ≥6·0
30  ≥6·0

Again, the efficacy of the decontamination proved to be very high as 93% of the carriers showed no bacterial growth, even when incubated for 14 days. Only in the case of two test carriers, medium cloudiness could be observed: one at position 16 after 4 days of incubation and the other at position 18 but as early as after 24 h. However, this can be considered as a very low survival rate, as no CFU could be detected in samples plated out directly after fumigation.

Taken together, this suggests again a very high decontamination efficacy, with reduction factors between 4·0 and more than 6·0 log levels.

Discussion

When conducting decontamination of a BSL3 or BSL4 laboratory by fumigation, it is of topmost importance to achieve the highest possible reduction of biological agents. However, other issues also have to be considered, such as the time during which the rooms cannot be used or whether technical equipment and surfaces will suffer from deterioration caused by the disinfecting agent. While usage of formaldehyde fumigation is an internationally validated procedure which is very effective against bacteria and viruses (Anon 2013), its application in laboratory decontamination holds several disadvantages. First, while the amount of time needed is comparable to that needed for the hydrogen peroxide procedure, formaldehyde leaves a white layer of residues after neutralization with ammonia gas on all surfaces and on equipment which may be damaged and have to be laboriously cleaned (Dietz et al. 1980; Cheney and Collins 1995). In that regard, hydrogen peroxide represents the better alternative, because it completely decomposes into water and oxygen over time without requiring the application of a neutralization agent, although this process is normally sped up using a catalyst. Secondly, in rooms that are contaminated with bacterial spores, hydrogen peroxide proves to be more practical, because for a formaldehyde fumigation to be effective against spores, it is necessary to maintain air humidity of 60–99% and temperatures between 30 and 65°C over several hours (Munro et al. 1999; Rogers et al. 2007; Schaal et al. 2009). H2O2 fumigation can be conducted at room temperature although humidity reaches similar levels (Jahnke and Lauth 1996; Rogers et al. 2005). This is easier to achieve under practical conditions and more gentle to equipment in the room. Hydrogen peroxide vapour proved to be very effective in killing bacterial spores (Melly et al. 2002), Myco. tuberculosis (Kahnert et al. 2005; Andersen et al. 2006; Hall et al. 2007; Grare et al. 2008) and even prions (Fichet et al. 2004). It has been shown that H2O2 fumigation can also be used under clinical conditions to efficiently sterilize rooms contaminated with Clostridium spores (Boyce et al. 2008; Otter and French 2009; Davies et al. 2011) or MRSA (French et al. 2004; Pottage et al. 2012).

In addition, most surfaces (except, for example copper and alloys thereof) are not negatively affected by hydrogen peroxide vapour (Jahnke and Lauth 1996; McDonnell and Russell 1999; Rogers et al. 2005, 2008; Grare et al. 2008). Neither is electronic equipment damaged by dry hydrogen peroxide vapours (Dietz et al. 1980; Heckert et al. 1997; Krause et al. 2001; Bates and Pearse 2005) and can even be left running during the decontamination procedure, a fact which is becoming increasingly important in modern laboratories.

However, a disadvantage of using hydrogen peroxide for room decontamination is that a validation has to be done and confirmed by three independent test runs every time a new facility is to be gassed. Some BSL3 laboratories avoid using hydrogen peroxide in case there are windows in the room which represent surfaces that are much cooler than the room air, especially in the colder seasons. This could possibly lead to condensation of the vapours and thus to a loss of efficacy. However, several experiments with cold surfaces that contained 106 spores of B. thuringiensis were performed at our institution and we could not confirm problems with condensation (data not shown). Another point is that condensed hydrogen peroxide vapour can damage the equipment upon repeated fumigations (Beswick et al. 2011). Therefore such spots need to be dried by hand after the fumigation procedure.

The IARC/WHO categorizes formaldehyde as belonging to the most cancerogenic group 1 (Cancer 2006), while H2O2 is classified as a member of group 3 (non-classifiable) (Cancer 1999). However, although it is not cancerogenic, hydrogen peroxide is a highly potent irritant due to its potent oxidizing nature and is likely to cause severe damage to mucous membranes and the lining of the lungs if airborne exposure is high. While formaldehyde is still produced and used for decontamination in most European countries, France prohibited use of formaldehyde for room decontamination altogether in 2006 and production was stopped in 2007 (Grare et al. 2008). Taken together, these statements suggest that formaldehyde represents the more harmful disinfectant when compared with hydrogen peroxide.

As test agents, spores from three different bacteria of the Bacillus species were chosen along with G. stearothermophilus spores, with the latter representing the international standard for decontamination experiments concerning heat and hydrogen peroxide. Bacillus cereus and B. thuringiensis both belong to the B. cereus group, which also includes B. anthracis. Therefore, they can serve as surrogates from which an inference can be made on anthrax spores. Also previous experiments performed by our group showed that on average B. thuringiensis spores show a higher resistance to peracetic acid than those of B.anthracis (Lemmer et al. 2012). B. subtilis was chosen as an additional member of the Bacillus genus because it exhibits similar resistance to liquid disinfectants (Majcher et al. 2008). Decontamination of spores represents a great challenge for disinfectants as spores have been shown to be highly stable under diverse environmental conditions and highly resistant to various decontamination procedures (McDonnell and Russell 1999; Nicholson et al. 2000; Melly et al. 2002; Rogers et al. 2005; Majcher et al. 2008; Otter and French 2009). If a decontamination procedure is effective against spores, it is likely that other biological agents (except for prions which would have to be tested separately) will be inactivated as well.

Our results clearly demonstrate that decontamination of the indicated BSL3 laboratory with hydrogen peroxide was successful. We obtained reductions between 4·0 and 6·0 log10 levels, which is consistent with other studies in the field of decontamination of bacterial spores (Rogers et al. 2005; Majcher et al. 2008; Barbut et al. 2009; Otter and French 2009; Davies et al. 2011). Spores were inactivated regardless of organism and position in the room and in most cases even inside closed petri dishes which demonstrates the efficacy of the procedure as well as the even distribution of the H2O2 vapours inside the laboratory during fumigation. In the very few cases in which bacteria growth occurred after decontamination, survival of spores can be regarded as marginal, as CFU appeared only after enrichment and never following direct plating out of samples. This means that the amount of surviving spores recovered was below the detection limits of the agar plates which are 10 spores per millilitre due to the fact that 100 μl were plated out. Carriers at positions 16 and 18, which were positive after enrichment, had been placed inside closed petri dishes or inside a fuse box, which might explain why some viable spores were obtained there after fumigation. This shows that for successful hydrogen peroxide decontamination, it is important to make sure that every location in the laboratory is accessible to the vapours and all doors and drawers are opened. When this is not easily possible, decontamination has to be done by hand. Obtained reduction factors are sufficient to achieve a deactivation of most biological agents. However, regarding organisms with a very low infection dose, such as Yersinia pestis, Francisella tularensis or Brucella sp., it might be advisable to repeat the decontamination procedure. In such cases, reduction values should be higher, as vegetative bacteria do not exhibit resistances comparable to spores.

In addition, three of four test carriers that showed growth after sample enrichment had been coated with B. cereus, which might suggest a slightly higher resistance to hydrogen peroxide. Other experiments performed at our facility showed that clearly that a certain percentage of spores of B. cereus ATCC 12826 exhibit a higher resistance against peracetic acid than spores of B. subtilis or an investigated B. anthracis strains (data not shown). Thus, results obtained from this study raise the question; why is G. stearothermophilus used as a standard decontamination challenge when there are other bacterial spores available that exhibit a higher resistance to hydrogen peroxide? Finally, no damage to surfaces of any kind could be detected in the gassed rooms, despite the presence of vaporous hydrogen peroxide for 8 h during each test.

Taken together, our results suggest that a validated decontamination of a laboratory with hydrogen peroxide represents an effective alternative to fumigation with formaldehyde.

Acknowledgements

The authors would like to thank Michael Bischoff (PEA, Koblenz) for performing the set-up of the hydrogen peroxide generator and validation of the fumigation protocol.

Conflict of Interest

No conflict of interest declared.

Ancillary