Comparative performance of impactor air samplers for quantification of fungal contamination

Journal of Hospital Infection (2001) 47: 149–155 doi:10.1053/jhin.2000.0883, available online at http://www.idealibrary.com on

Comparative performance of impactor air samplers for quantification of fungal contamination Didier Nesa, Jacques Lortholary*, Adel Bouakline†, Muriel Bordes‡, Jacques Chandenier§, Francis Derouin† and Jean-Pierre Gangneux† Laboratoire de Parasitologie-Mycologie, Hôpital Saint-Antoine; *Service central d’Hygiène de l’Assistance Publique, †Laboratoire de Parasitologie-Mycologie, Hôpital Saint-Louis, Paris; ‡Laboratoire central, Hôpital Albert Chenevier, Créteil; and §Laboratoire de Parasitologie-Mycologie, Hôpital Sud, Amiens, France

Summary: The aim of this study was to assess the performance of different impactor air samplers for fungal spore collection in the hospital environment. Four recent impactor air samplers were selected: Samplair (AES, Combourg, France); Air Test Omega (LCB, France); Air Samplair Mas-100 (Merck, France); and BioImpactor 100–08 (AES). They were compared with one another at three different hospital sites with varying levels of contaminated air. No significant difference in the efficiency of spore recovery was found between Air Test Omega, Mas-100 and BioImpactor, whereas Samplair was significantly less efficient. BioImpactor was then selected to represent the three superior impactors and was compared with the single-stage Andersen disposable sampler and the Collectron MD8 air sampler (Sartorius, France) and the High Flow Air Sample (BioTest, France), which are based on filtration and centrifugation methods, respectively. No significant difference was observed in terms of spore recovery. On the basis of their performance, unit sampling cost, autonomy and simplicity of use, we conclude that Air Test Omega, Air Samplair Mas-100 and BioImpactor 100-08 are suitable for routine indoor evaluation of fungal contamination of air in hospitals. © 2001 The Hospital Infection Society

Keywords: Impactor air samplers; bioimpactors; air sampler; fungal contamination; hospital air; Aspergillus; invasive aspergillosis.

Introduction Invasive aspergillosis (IA) is a major opportunistic fungal infection in severely immunocompromised patients, especially those receiving cytotoxic therapy for haematological malignancies or undergoing organ or bone marrow transplantation.1–5 The hospital-acquired origin of aspergillosis has been Received 5 July 2000; revised manuscript accepted 18 October 2000. Author for correspondence: Professor F. Derouin, Laboratoire de Parasitologie-Mycologie, Faculté de Médecine Lariboisière-Saint-Louis, 15 rue de l’Ecole de Médecine, 75270 Paris Cedex 06, France. Fax: 33 1 43 29 51 92; E-mail: [email protected]

0195-6701/01/020149;07 $35.00

demonstrated in epidemic situations, with airborne spore transmission via unfiltered air, and massive environmental contamination during construction and renovation work both inside and outside the hospital.6–8 Elimination of Aspergillus spores from the environment by the use of high-efficiency particulate air (HEPA) filtration procedures, with or without Laminar Air Flow (LAF) systems, reduces the incidence of hospital-acquired aspergillosis.9–11 Monitoring of environmental fungal contamination is strongly recommended to detect increases in conidia density and to assess air filtration efficiency.12–13 There is no consensus, however, on the density at which the risk of invasive aspergillosis is increased. Rhame et al. reported a higher risk of IA when the © 2001 The Hospital Infection Society

D. Nesa et al.

150

average density of A. fumigatus was above 0.9 colony-forming unit/m3.14 Arnow et al. noted a marked reduction in the incidence of IA when Aspergillus density fell from 1–2 to less than 0.2 cfu/m3.15 Sheretz et al. reported no cases of aspergillosis when conidia density was reduced to 0.009/m3.9 The ‘safe’ spore density threshold is difficult to establish, as the sampling conditions depend upon the type of sampler, sampling time and the culture conditions.16 At least 10 air samplers are currently marketed, and are based on different physical principles (mainly impaction, centrifugal acceleration or filtration), with different conditions of use for Aspergillus spore collection. Impactors are the most widely used type of sampler, of which the Andersen Cascade Sampler or the Casella Slit Sampler (Casella Ltd, UK) have been used for some time. These samplers are relatively impractical, however, and new generations of impactors offer advantages in term of cost, performance and simplicity of use. In this study, four ‘new generation’ impactors were compared experimentally for their capacity to collect fungal spores from indoor and outdoor air. The performance of a representative impactor was then compared with that of the Andersen sampler and two other samplers, one based on filtration and the other on centrifugation. Methods Selection of air samplers The choice of samplers for this study was based on the following criteria: commercial availability; collection principle; cost of each air sampling; presence or absence of a battery giving at least 1 h of autonomy; flow rate of 100 L/min; weight; simplicity of use under various conditions; and cost of the sampler. The characteristics and performance description of 17 samplers were examined. Most samplers were impactors, whereby sampling technology is based upon impacting particles from an airstream onto an agar surface. The advantage of direct impaction on to agar is that agar plates used for collection, preferentially standard 90 mm diameter Petri dishes, can be incubated without further treatment. For this study, we selected four impactors: Samplair (AES, France); Air Test Omega (LCB, France); Air Samplair Mas-100 (Merck, France); and BioImpactor 100–08 (AES). These samplers

were considered most appropriate as they offer at least 1 h of autonomy, use standard Petri dishes for collection (thus reducing the cost of each sampling procedure), sample at a flow rate of 100 L/min, and are considered simple to handle and use by the technicians (Table I). The Andersen Sampler and two other samplers with different sampling methods were also used for comparison (Table I). The Collectron MD8 sampler (Sartorius, France) is based on filtration of air onto a gelatine membrane which is then cultured on an agar Petri dish. The other sampler, High Flow Air Sample (BioTest Diagnostics, France), is based on centrifugal acceleration and impaction. Airborne particles are drawn in by an impeller and thereafter impacted onto an agar coated plastic strip lining the internal periphery of the impeller housing. Agar strips are subsequently removed for incubation. At the time of this study the remaining samplers either failed to fulfill the above criteria or were obsolete, namely SAS standard and SAS 90 (Bioblock, France), Airtest and Isobio (LCB), M-AIR T (Millipore, France), Ochlovar 92, Ochlovar 95 (AES), RCS and RCS plus (BioTest Diagnostics, France) and the Joubert device (Joubert, France). In the first set of experiments, the four selected impactor air samplers were compared. In the second set of experiments, the BioImpactor was chosen to represent these impactors, and was then compared with the widely used single-stage Andersen impactor, the Collectron MD8 device and the High Flow Air Sample. Sampling sites and volumes This study was performed in three Paris hospitals. Several sites expected to have different levels of fungal contamination were selected from each hospital. The design of the comparative assessment is shown in Table II. Samplers were used at a flow rate of 100 L/min, except for the Andersen sampler, which was used at the maximum flow rate of 28 L/min. Sampling times were determined according to the sampling volumes and flow rates. During each experiment and at each site of sampling, the air samplers were used simultaneously. After each sampling, samplers were successively permuted. Thus, measures were made in triplicate at each site, yielding 6 or 12 repeated measures when two or four samplers were compared, respectively.

Impactor air samplers for fungal quantification

151

Table I Characteristics of the studied samplers Name

Manufacturer

Principle

Andersen

Andersen

Impaction

Samplair

AES

Air Test Omega

Programmable time/volume

Spore collection

Rechargeable batteries (autonomy)

Weight (kg)

28

No

Petri dish 90 mm Ø

No

6.4

Impaction

100

Yes

Petri dish 90 mm Ø

Yes (14 h)

1.8

LCB

Impaction

100

Yes

Petri dish Yes (4 h) 90 or 65 mm Ø

1.2

Air Samplair Mas-100

Merck

Impaction

100

Yes

Petri dish 90 mm Ø

Yes (7 h)

2.2

BioImpactor 100-08

AES

Impaction

100

Yes

Petri dish 90 mm Ø

Yes (1 h)

2

Collectron MD8 High Flow

Sartorius

Filtration

Yes

No

6.5

Biotest

Centrifugal acceleration and impaction

Filter; 90 mm Ø Petri dish Strips

Yes (2 h)

1.5

Table II

Flow rate (L/min)

42–133 100

Yes

Study design

Air samplers

Air sampling site

Sampling volume (L)

Number of air samples/air sampler

Samplair vs. Air Test Omega vs. Mas-100 vs. BioImpactor

Room with laminar air flow* (Saint-Louis Hospital) Conventional room (Chenevier Hospital) Archive room (SaintAntoine Hospital) Entrance hall (SaintLouis Hospital) Outdoors (SaintAntoine Hospital)

1000

12

100, 250

12, 12

100, 250

12, 12

100, 250

12, 12

100, 250

12, 12

Andersen vs. BioImpactor

Corridor (Saint-Louis Hospital)

100, 250

6, 6

Collectron MD8 vs. BioImpactor

Corridor (Saint-Louis Hospital)

250, 500, 1000

6, 6, 6

High Flow vs. BioImpactor

Archive room (SaintAntoine Hospital)

250, 500, 1000

6, 6, 6

* Provided with HEPA filtered air (99.99% efficiency).

Culture and counting Petri dishes containing 20 mL malt agar (BioRad, France) were used. They were positioned upon the impactors according to the manufacturers’ recommendations. With the Collec-tron MD8 device, cellulose nitrate filters kindly provided by the manu- facturer were placed upon Petri dishes. With the High Flow Air Sample, strips of Sabouraud

agar, kindly provided by the manufacturer, were used as recommended. Petri dishes and strips were incubated for three days at 32°C. Colony counts were expressed as colony-forming units (cfu) per unit volume of air. Means
D. Nesa et al.

152

Statistical analysis The Kruskal–Wallis test was used to compare mean spore counts recovered with the four impactors in the first experiment. If a significant difference was observed, data for each sampler were then compared with each other, using the Mann–Whitney test. In the second experiment, data obtained with the BioImpactor vs. Andersen, vs. Collectron MD8 or vs. High Flow were compared using the Mann– Whitney test. A P value :0.05 was considered statistically significant.

to 30 cfu/sample, with no significant difference between the two air sampling volumes tested, i.e., 100 and 250 L Air Test Omega, Mas-100 and BioImpactor had similar efficiencies (Figs 2, 3), whereas Samplair was sig-nificantly less efficient than these three devices (P:0.001). Similar results were obtained from a heavily contaminated archive 30

Results

20

cfu

Comparative evaluation of four impactor air samplers The four impactors were first evaluated in a patient room equipped with LAF. Twelve measures with a sampling volume of 1000 L were made with each sampler. No spores were collected with any of the four impactors. In the entrance hall, mean counts ranged from approximately 1 to 5 cfu/sample. A sampling volume of 250 litres collected more spores than 100 L with BioImpactor and Mas100 (P:0.05), although spore counts were not directly proportional to the sample volume (Fig. 1). In a conventional patient room, and outdoors, mean counts ranged from approximately 10

*

0

BioImpactor

Samplair

Mas-100

Omega

Figure 2 Number of fungal colony-forming units collected in a conventional room (mean
15

30

10

20

cfu

cfu

*

10

5

* 0

*

10

BioImpactor

*

*

Samplair

Mas-100

Omega

Figure 1 Number of fungal colony-forming units collected in the entrance hall (mean
0

BioImpactor

Samplair

Mas-100

Omega

Figure 3 Number of fungal colony-forming units collected in outdoor air (mean
Impactor air samplers for fungal quantification

153

room (50–100 cfu/ sample) (Fig. 4). At this site the Petri dishes were saturated by a sampling volume of 250 L. No significant difference was observed in terms of the fungal species collected (mainly moulds and rarely Aspergillus), whatever the sampler, sampling site or sampling volume (data not shown). Comparison of the BioImpactor with the Andersen device, and of the impaction method vs. filtration and centrifugation methods The first set of experiments showed that BioImpactor was a good representative of the newgeneration impactors, and it was thus selected for 120

the comparisons with the Andersen, Collectron MD8 and High Flow samplers. Laminar air flow rooms were not sampled, as they had produced totally negative results. Samples of 250, 500 and 1000 L were collected with the BioImpactor, Collectron MD8 and High Flow Air Sample from two other sites. Fungal contamination ranged between two and 100 spores/sample, with no significant difference between the efficiency of the three devices (Table III). The BioImpactor and Andersen samplers were compared at one site with moderate fungal contamination. Only volumes of 100 and 250 L were sampled, as sampling 1000 L with the Andersen device would have taken 35 min, resulting in partial drying out of the agar. No difference in the efficiency of the two samplers was observed at the two volumes tested (Table III). Discussion

cfu

90

60

30 *

0

BioImpactor

*

Samplair

Mas-100

Omega

Figure 4 Number of fungal colony-forming units collected in an archive room (mean
This study shows that several ‘new-generation’ air samplers are equally efficient at collecting fungal spores from indoor and outdoor air with varying degrees of fungal contamination. The assessment was not performed under controlled conditions (i.e., in an experimental room with defined levels of airborne spores) as our intention was to reflect the practical conditions in which these collectors are used for monitoring fungal contamination in the hospital environment. For optimal comparison of the samplers, experiments were performed at several hospital sites and each sampling procedure was repeated at each site to assess reproducibility. Besides the intrinsic performance of each sampler, we also observed ease of use, particularly portability, the need for an external power supply and the cost of each sampling procedure.

Table III Spore recovery by the BioImpactor compared to the Andersen, Collectron MD8 and High Flow Air Samplers from various hospital sites. For each sampling volume, mean spore counts
BioImpactor vs. Andersen (entrance hall)

100 L

250 L

500 L

1000 L

0.83<0.41 1<0.89

3.83<1.17 2.50<1.38

– –

– –

2.50<2.74 2<1.79

5<5.55 2.67<2.25

72.83<26.96 43.67<22.71

105.5<18.09 75.67<20.77

BioImpactor vs. Collectron MD8 (corridor)

– –

BioImpactor vs. High Flow (archive room)

– –

3<3.79 1<1.55 29<11.37 57.50<37.8

D. Nesa et al.

154

The impactors fulfilled most of the criteria for optimal use in routine practice and low cost. Samplair, Air Test Omega, Air Samplair Mas-100 and BioImpactor 100-08 are truly portable devices, as they are light and powered by rechargeable batteries (with an autonomy of at least 1 h); in addition, they are easy to clean and maintain and use classical Petri dishes. Results showed good reproducibility among 12 measurements made with each sampler, except in the case of high-level contamination. Sampling of large volumes (1000 L in a room equipped with HEPA-filtered air yielded systematically negative results. This was probably due to the absence of fungal contamination in such environments, but may also indicate the limits of biocollectors in areas with a very low level of spores. In such cases, particulate counting is probably more appropriate to assess the efficacy of the filtration system. In sites with low-level fungal contamination, sampling of 100 or 250 L was equally efficient in terms of spore recovery, whereas in heavily contaminated atmospheres, sampling of 250 L saturated the Petri dishes. Air Test Omega, Air Samplair Mas-100, and BioImpactor 100-08 were equally efficient for collecting fungal spores from different sites. By contrast, Samplair gave significantly lower spore counts than the other three devices (P:0.001), whatever the site, sampling volume or user. This sampler can be equipped with either a standard grid (394 holes), or a high-speed sampling grid (106 holes), which allow a flow rate-impaction on agar of 3 or 16 m/s, respectively, when a sampling flow rate of 100 L/min is selected (data provided by the manufacturer). We used the standard grid, as initially recommended by the manufacturer, but in the view of these results and other unpublished data, the manufacturer now recommends the use of the high speed sampling grid for optimal performance. The results obtained with these new-generation samplers were compared with those of the Andersen sampler. In addition to its flow rate of only 28 L/min, which is a disadvantage for high sampling volumes, the Andersen sampler requires a mains electricity supply and is heavy and noisy. It is still used in numerous hospitals, however, and was thus considered as the ‘reference impactor’ for this study.17,18 The BioImpactor, chosen to represent the new-generation impactors, was as efficient as the Andersen sampler (Table III). Other sampler devices based on different physical spore collection principles, were also compared to the BioImpactor. The filtration method, as used

in the Collectron MD8, and the centrifugal acceleration and impaction method used in the High Flow Air Sample, are considered highly efficient, but require the use of a specific culture support (gelatine filters or agar strips).19–25 The cost of each sampling procedure is therefore four to 10 times as high as impactors using classical Petri dishes. In terms of performance, the impaction method was as efficient as the filtration and centrifugation techniques (Table III). In conclusion, the impaction method is at least as efficient as other sampling methods for fungal spore collection. The Air Test Omega (LCB), Air Samplair Mas-100 (Merck) and BioImpactor 100-08 (AES) are both efficient and convenient for routine evaluation of hospital indoor air quality. In routine practice, sampling of 100 L appears to be sufficient in conventional rooms or outdoors as it is reproducible and reliable for the detection of Aspergillus spores in air. In rooms equipped with filtered air, volumes of up to 500–1000 L must be collected to show the absence of fungal contamination. The selected samplers can sample such volumes simply, in a matter of minutes, and at low cost. Acknowledgments We are grateful to the manufacturers AES, Biotest, LCB, Merck and Sartorius for providing the Samplair and BioImpactor, High Flow Air Sample, Air Test Omega, Air Samplair Mas-100 and Collectron MD8 air samplers, respectively, and to Dr A. Datry (La Pitié-Salpêtrière Hospital, Paris, France) for providing the Andersen air sampler. We thank David D. Young for reviewing the manuscript and Jean-Louis Poirot for fruitful discussion. References 1. Abbasi S, Shenep JL, Hughes WT, Flynn PM. Aspergillosis in children with cancer: a 34 year experience. Clin Infect Dis 1999; 29: 1210–1219. 2. Denning DW. Invasive aspergillosis. Clin Infect Dis 1998; 26: 781–805. 3. Paterson D, Singh N. Invasive aspergillosis in transplant patients. Medicine 1999; 78: 123–138. 4. Ribaud P, Chastang C, Latgé JP et al. Survival and prognostic factors of invasive aspergillosis after allogenic bone marrow transplantation. Clin Infect Dis 1999; 28: 322–330. 5. Wald A, Leisenring W, van Burik JA, Bowden RA. Epidemiology of Aspergillus infections in a large

Impactor air samplers for fungal quantification

06. cohort of patients undergoing bone marrow transplantation. J Infect Dis 1997; 175: 1459–1466. 06. Carter CD, Barr BA. Infection control issues in construction and renovation. Infect Control Hosp Epidemiol 1997; 18: 587–596. 07. Vandenbergh MFQ, Verweij PE, Voss A. Epidemiology of nosocomial fungal infections: invasive aspergillosis and the environment. Diagn Microbiol Infect Dis 1999; 34: 221–227. 08. Walsh TJ, Dixon DM. Nosocomial aspergillosis: environmental microbiology, hospital epidemiology, diagnosis and treatment. Eur J Epidemiol 1989; 5: 131–142. 09. Sheretz RJ, Belani A, Kramer BS et al. Impact of air filtration on nosocomial Aspergillus infections: unique risk of bone marrow transplant recipients. Am J Med 1998; 83: 709–718. 10. Barnes RA, Rogers TR. Control of an outbreak of nosocomial aspergillosis by laminar air-flow isolation. J Hospital Infect 1989; 14: 89–94. 11. Cornet M, Levy V, Fleury L et al. Efficacy of prevention by high-efficiency particulate air filtration or laminar airflow against Aspergillus airborne contamination during hospital renovation. Infect Control Hosp Epidemiol 1999; 20: 508–513. 12. Leenders ACAP, van Belkum A, Behrendt M, Luijendijk A, Verbrugh HA. Density and molecular epidemiology of Aspergillus in air and relationship to outbreaks of Aspergillus infection. J Clin Microbiol 1999; 37: 1752–1757. 13. Iwen PC, Davis JC, Reed EC, Winfield BA, Hinrichs SH. Airborne fungal spore monitoring in a protective environment during hospital construction, and correlation with an outbreak of invasive aspergillosis. Infect Control Hosp Epidemiol 1994; 15: 303–306. 14. Rhame FS. Prevention of nosocomial aspergillosis. J Hosp Infection 1991; 18: 466–472.

155

15. Arnow PM, Sadigh M, Costas C, Weil D, Chudy R. Endemic and epidemic aspergillosis associated with in-hospital replication of Aspergillus organisms. J Infect Dis 1991; 164: 998–1002. 16. Mishra SK, Ajello L, Ahearn DG et al. Environmental mycology and its importance in public health. J Med Mycol Vet Mycol 1992; 30: 287–305. 17. Andersen AA. New samplers for the collection, sizing and enumeration of viable airborne particles. J Bacteriol 1958; 76: 471–484. 18. Peto S, Powell EO. The assessment of aerosol concentrations by means of the Andersen sampler. J App Bacteriol 1970; 33: 582–598. 19. Parks SR, Bennett AM, Speight SE, Benbough JE. An assessment of the Sartorius MD8 microbiological air sampler. J Appl Bacteriol 1996; 80: 529–534. 20. Clark S, Lach V, Lidwell OM. The performance of the Biotest RCS centrifugal air sampler. J Hosp Infect 1981; 2: 181–186. 21. Nakhla LS, Cummings RF. A comparative evaluation of a new centrifugal air sampler (RCS) with a slit air sampler (SAS) in a hospital environment. J Hosp Infect 1981; 2: 231–266. 22. Macher JM, First MW. Reuter centrifugal air sampler: measurement of effective airflow rate and collection efficiency. Appl Environ Microbiol 1983; 45: 1960–1962. 23. Smid T, Schokkin E, Boleij JS, Heederik D. Enumeration of viable fungi in occupational environments: a comparison of samplers and media. Am Ind Hyg Assoc J 1989; 50: 235–239. 24. Benbough JE, Bennett AM, Parks SR. Determination of the collection efficiency of a microbial air sampler. J Appl Bacteriol 1993; 74: 170–173. 25. Ljungqvist B, Reinmuller B. The Biotest RCS air samplers in unidirectional flow. J Pharm Sci Technol 1994; 48: 41–44.