Evaluation of the effectiveness of commercially available contact ...

Letters in Applied Microbiology ISSN 0266-8254

NOTE TO THE EDITOR

Evaluation of the effectiveness of commercially available contact plates for monitoring microbial environments F. Pinto1, S. Hiom2, S. Girdlestone2 and J.-Y. Maillard1 1 Welsh School of Pharmacy, Cardiff University, Cardiff, UK 2 Cardiff & Vale NHS Trust, SMPU, Quadrant Centre, Cardiff Business Park, Llanishen, Cardiff, UK

Keywords contact plates, manufacturing, recovery, surface contamination. Correspondence Jean-Yves Maillard, Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff, CF10 3NB, UK. E-mail: [email protected]

2008 ⁄ 1223: received 16 July 2008, revised and accepted 2 September 2008 doi:10.1111/j.1472-765X.2008.02534.x

Abstract Aim: The aim of this study was to measure the efficiency of contact plates to recover microbial contaminants from stainless steel surface. Materials and Methods: Three commercially available contact plates were used to recover two biological indicators from stainless steel sheets. The method used was standardised and validated to provide robust results. Parameters such as wetness, fertility and loss of water were also investigated for possible correlation with recovery efficiency. Results: The percentage of recovery from the contact plates was low and differences in recovery efficiency between brands depended upon the test organism. The poor recovery was probably due to the inability of the dried micro-organism to transfer to the plate, rather than the inability of the plate to grow the micro-organism. Wetness might help in improving recovery. Conclusions: The use of a validated protocol allowed robust investigations into the recovery efficiency of contact plates. Significance and Impact of the Study: The poor and variable recovery rates are of concern as they cast doubt on the comparability and reliability of environmental monitoring results where different commercial contact plates have been used.

The preparation and administration of parenteral products can be at risk from contamination and result in infections. Surface bioburden is considered one of the five major sources of final product contamination (Beaney 2006). These factors together with other risk elements (e.g. the need for complex calculations) have led to the recommendations that, where possible, parenteral products are prepared in dedicated hospital pharmacy aseptic units (Anonymous 2001). These units prepare parenteral products under microbiologically controlled conditions, as set out in National Standards and Guidelines (Anonymous 2002; Beaney 2006). The nature of aseptic processing means that products will often be released and administered before standard microbiological testing can be carried out. Therefore all assurance of quality must be provided by adequately validated systems of work. Part of this quality assurance programme is to monitor the microbial environment, which includes the use of agar contact plates to assess surface bioburden. The National Standards and Guidelines (Anonymous 2002) set limits for surface microbes within described

classes of clean rooms. Units manufacturing outside these microbial limits risk closure from the associated Regulatory Authority (Medicines and Healthcare Products Regulatory Agency). There are several commercially available contact plates which are acceptable for use in environmental monitoring within the pharmaceutical industry. However, content and preparation of these plates are variable between manufacturers and anecdotal evidence suggests there is also variability with microbial recovery rates. Overall, there is little documented evidence to support the efficiency of these and other contact plates (Niskanen and Pohja 1977; Poletti et al. 1999; Lemmen et al. 2001). The aim of this study was to compare and contrast the fertility and recovery rates of three commercially available contact plates, using standardized micro-organisms and validated techniques. Staphylococcus epidermis (NCIMB8853) and Staphylococcus aureus (NCIMB9518) were used as biological indicators following recent concerns with these microorganisms. They were grown on tryptone soya agar (TSA; Oxoid Ltd, Basingstoke, UK) slopes for 24 h at 37C. New

ª 2009 The Authors Journal compilation ª 2009 The Society for Applied Microbiology, Letters in Applied Microbiology 48 (2009) 379–382

379

Evaluation of the effectiveness of commercially available contact plates

slopes were resuspended daily with 5 ml of buffer (tryptone sodium chloride) to prepare a fresh bacterial suspension. Contact plates were obtained from BioMerieux (CountTact Irradiated Agar; BioMerieux, Marcy l’Etoile, France), Cherwell (T ⁄ V Contact Plate; Cherwell, Bicester, UK) and Oxoid (Tryptone soya agar contact plate). Plates were stored in accordance to the manufacturers’ recommendation. Validation of bacterial recovery from surface was performed using stainless steel discs (AISI 304; finish 2B; 20 mm diameter, thickness of 1Æ5 mm thickness; Goodfellow Cambridge Limited, Huntingdon; UK). One 10 ll drop of a diluted bacterial suspension (10)6 dilution) was inoculated at the centre of a disc, which was placed in a six well-plate (Corning Inc., NY, USA), and then dried for 1 h under a laminar air flow cabinet. The surface of the disc was then flushed 12 times with 10 ml of buffer, and 10 drops (10 ll) of this suspension were plated on the surface of a TSA plate. After 24 h incubation at 37C, colonies per drop were counted and the percentage of recovery calculated. To measure the recovery rates of contact plates, the surface of three steel sheets (100 · 350 mm; AISI 304; finish 2B; 2 mm thickness; Goodfellow Cambridge Limited) was divided into four box-grids. In each box, four drops (10 ll) of a bacterial suspension (approx. 160 CFU) were inoculated. Twelve contact plates from different manufacturers were then randomly assigned to each of the 12 boxes. An applicator (Count-Tact; Biomerieux, France) was used to standardise surface testing by applying a uniform pressure of 500 g on the surface of the contact plate for 10 s. After 24 h incubation at 37C, colonies per drop were counted. In order to assess differences between contact plates, other parameters were measured such as fertility, loss of water, wetness of the plates, but also the effect of neutralizer and expiring dates on recovery. Fertility was measured by inoculating 10 ll of the diluted bacterial suspension directly onto the contact plate. The fertility rate was measured as the ratio between the viable count on TSA and the number of CFU ml)1 recovered from contact plates. Loss of water was measured by weighing

F. Pinto et al.

the plates prior to, and after, 24 h incubation at 37C. The wetness of the contact plate surface was measured semi-quantitatively with the following scale: high: presence of droplets on the surface; medium: presence of some moisture on the surface; low: dry surface. The effect of neutralizer on recovery was only analysed with the Biomerieux plates. anova and t-test (Minitab Ltd, Coventry, UK) were used to analyse differences in recovery between contact plates and the effects of the different parameters on recovery. The validation of the counting technique and the validation of the recovery method were central to this study design. It was important to use a dilution that will allow the recovery of a countable number per 10 ll drop. A dilution of 10)6 allowed the recovery of 22Æ9 ± 5Æ5 CFU per drop and there was no significant difference (P = 0Æ465; n = 82) in the number of colonies recovered per drop. Using the stainless steel disks to measure the efficiency of bacterial recovery from surfaces, there was no significant differences (P = 0Æ073; n = 88) in the number of colonies recovered between disks when a set bacterial concentration was inoculated and dried on the surface. However, the recovery from the stainless steel surface for Staph. epidermidis was 72Æ6% and for Staph. aureus 58Æ1%. These percentages of recovery from the steel discs were used to adjust the original control inoculum concentrations used for the calculation of percentage recovery by contact plates. The three contact plate brands investigated had overall a low percentage of recovery (Table 1). The number of Staph. epidermidis colonies recovered from each brand was significantly different (P = 0Æ000; n = 24). Cherwell contact plates showed the highest recovery rate with the two biological indicators, 38% and 56% respectively (Table 1). However, for Staph. aureus the percentage recovery values were not significantly different between the three brands (P = 0Æ08; n = 16), although a higher variability in results was observed. In addition, the recovery of Staph. aureus as compared with Staph. epidermidis was better for all three contact plate brands (P = 0Æ02; n = 16). The ability of the different plates to support bacterial growth was not a

Table 1 Percentage of recovery from contact plates Staphylococcus epidermidis

Staphylococcus aureus

Manufacturer

CFU per drop expected (Log10)

CFU per drop measured (Log10)

% recovery*

CFU per drop expected (Log10)

CFU per drop measured (Log10)

% recovery*

Biomerieux Oxoid Cherwell

2Æ13 ± 0Æ18 2Æ13 ± 0Æ18 2Æ13 ± 0Æ18

1Æ54 ± 0Æ19 1Æ46 ± 0Æ21 1Æ69 ± 0Æ18

26 ± 5 23 ± 7 38 ± 14

1Æ76 ± 0Æ41 1Æ76 ± 0Æ41 1Æ76 ± 0Æ41

1Æ63 ± 0Æ22 1Æ55 ± 0Æ22 1Æ74 ± 0Æ16

46 ± 23 38 ± 18 56 ± 18

n = 24 for Staph. epidermidis; n = 16 for Staph. aureus. *Calculated from individual number; the percentage of recovery from the controls (i.e. recovery from stainless steel discs) was taken into account.

380


F. Pinto et al.


Table 2 Differences in fertility and contact plate characteristics between brands (n = 24) Manufacturers

Fertility

Wetness

Loss of water

Biomerieux Oxoid Cherwell

0Æ88 ± 0Æ09 0Æ98 ± 0Æ17 0Æ92 ± 0Æ11

Medium Low High

0Æ62 ± 0Æ09 0Æ77 ± 0Æ08 0Æ64 ± 0Æ08

Fertility: ratio between viable count and colonies recovered from direct inoculation on contact plates. Fertility was only performed with Staphylococcus epidermidis. Loss of water: difference in weight before and after 24 h incubation at 37C. Wetness: high: presence of droplets on the surface; medium: presence of some moisture on the surface; low: dry surface.

factor affecting recovery since the fertility was very high and not significantly different (P = 0Æ24; n = 29) between the three brands (Table 2). Other parameters may affect the effectiveness of bacterial recovery from surface. The different contact plates showed significantly different wetness and water evaporation following incubation (P = 0Æ00; n = 71) (Table 2). Although there was no apparent correlation between water loss and percentage of recovery, one might argue that an increase initial wetness might be favourable (Table 2). The Oxoid and Biomerieux brands which were stored at 4C and then allowed to reach room temperature before use, appeared drier than the Cherwell contact plates. It was observed that the Cherwell plates adhered better to the surface, possibly allowing a better transfer of dried bacteria to the surface of the plate. Finally, the presence or not of neutralizer in the contact plate and the expiry dates did not affect the efficiency of the recovery of bacterial inoculum dried on a stainless steel surface (data not shown). Contact plates are critical in quality assurance programmes used for the aseptic manufacture of parenteral products. This study demonstrated that contact plates only recover a low percentage of micro-organisms from stainless steel surfaces. There were some differences in the recovery depending upon the biological indicator studied. At present, it is unclear why the plates performed better when recovering Staph. aureus. Since there was no difference in the ability of the contact plate to support bacterial growth, such low recovery percentage might be caused partly by the inability of dried bacteria to transfer to the surface of the contact plate. An increased wetness might have allowed an enhanced recovery of Staph. epidermidis and this observation is in agreement with the study by Scott et al. (1984) who observed a correlation between the percentage of recovery from contact plates and the wetness of surfaces. The overall low percentage of recovery might also be affected by the contact time between the contact plates on the surface. Here a 10 s contact time was used at a set pressure. Longer contact time would

have enabled an enhanced recovery as demonstrated by Foschino et al. (2003), although longer contact time would not represent usage in practice. One of the major drawbacks of using contact plates is the inability to differentiate bacterial clumps and single organisms (Favero et al. 1968; Scott et al. 1984), which might lead to an underestimation of microbial contamination. On surfaces, micro-organisms are more likely to be found as clumps, which would produce only one colony on the contact plate. One can argue that the presence of moisture would facilitate the dispersion of bacterial clumps and allow for a higher count. Bacterial adherence to stainless steel due to surface charge, topography and hydrophobicity can not be underestimated (Flint et al. 2000), although the role of such events in the lack of effectiveness of contact plates to recover micro-organisms is difficult to ascertain in the present study. Contact plates have been deemed to be more efficient in recovery low number of Gram-positive micro-organisms (Niskanen and Pohja 1977; Scott et al. 1984). Here, we observed that contact plates were not necessarily efficient in recovering a low bacterial inoculum. The incubation time and temperature for environmental samples recovered by contact plate has been described to be important and optimal at 32C for 43 h (Vesley et al. 1966). In our study, an incubation time of 24 h at 37C was deemed to be sufficient and longer incubation time might have resulted in difficulties in counting single distinct colonies in a small area (i.e. 10 ll drop were placed on the surface). The efficiency of swabs as a comparison to the contact plate was not investigated in this study since swabs are not recommended for licensed facility (Beaney 2006). It has to be noted that the use of contact plates imposes a costly delay between the end of sanitation and the start of production. The use of rapid microbiology method such as chemoluminescence would circumvent this problem, although their usage must be validated for in hospital pharmacy aseptic suits. Underestimating the microbial contamination of surfaces might have serious consequences on the quality assurance of aseptically prepared hospital pharmaceutical products. Conversely, pharmacists must have confidence that they are able to recover any potential microbial contaminants from ‘clean’ areas, although an increase in recovery efficiency might lead to some issue in quality assurance and the potential of licensed units falling outside the limits stated in standard documents (Anonymous 2002; Beaney 2006). The use of a validated protocol allowed robust investigations into the recovery efficiency of the different contact plates to recover dried micro-organisms from surfaces. The poor and variable recovery rates are of concern as they cast doubt on the comparability and reliability of environmental monitoring results where different commercial contact plates have been used.


381


Acknowledgement This study was supported by the Society for Applied Microbiology under the ‘Student into Work Grant Scheme’. References Anonymous. (2001) A Spoonful of Sugar. Audit Commission Report. London: Audit Commission. Anonymous. (2002) Rules and Guidance for Pharmaceutical Manufactures and Distributors. Medicines Control Agency. London: HMSO. Beaney, A. (2006) NHS Quality Control Committee. In The quality Assurance of Aseptic Preparation Services, 4th edn. ed. Beaney, A.M. London: Pharmaceutical Press. Favero, M.S., McDade, J.J., Robertsen, J.A., Hoffman, R.K. and Edwards, R.W. (1968) Microbiological sampling of surfaces. J Appl Bacteriol 31, 336–343. Flint, S.H., Brooksa, J.D. and Bremer, P.J. (2000) Properties of the stainless steel substrate, influencing the adhesion of thermo-resistant streptococci. J Food Eng 43, 235–242. Foschino, R., Picozzi, C., Civardi, A., Bandini, M. and Faroldi, P. (2003) Comparison of surface sampling methods and

382

F. Pinto et al.

cleanability assessment of stainless steel surfaces subjected or not to shot peening. J Food Eng 60, 375–381. Lemmen, S.W., Ha¨fner, H., Zolldann, D., Amedick, G. and Lu¨tticken, R. (2001) Comparison of two sampling methods for the detection of Gram-positive and Gram-negative bacteria in the environment: moistened swabs versus Rodac plates. Int J Hyg Environ Health 203, 245–248. Niskanen, A. and Pohja, M.S. (1977) Comparative studies on the sampling and investigation of microbial contamination of surfaces by the contact plate and swab methods. J Appl Bacteriol 42, 53–63. Poletti, L., Pasquarella, C., Pitzurra, M. and Savino, A. (1999) Comparative efficiency of nitrocellulose membranes versus RODAC plates in microbial sampling on surfaces. J Hosp Infect 41, 195–201. Scott, E., Bloomfield, S.F. and Barlow, C.G. (1984) A comparison of contact plate and calcium alginate swab techniques for quantitative assessment of bacteriological contamination of environmental surfaces. J Appl Bacteriol 56, 317– 320. Vesley, D., Keenan, K.M. and Halbert, M.M. (1966) Effect of time and temperature in assessing microbial contamination on flat surfaces. Appl Microbiol 14, 203–205.
