Letters to the Editor

References 1. Van Looveren M, Goossens H, The ARPAC Steering Group. Antimicrobial resistance of Acinetobacter spp. in Europe. Clin Microbiol Infect 2004;10:684—704. 2. Borgmann S, Wolz C, Grobner S, et al. Metallo-betalactamase expressing multi-resistant Acinetobacter baumannii transmitted in the operation area. J Hosp Infect 2004;57: 308—315. 3. Wagenvoort JH, De Brauwer EI, Toenbreker HM, et al. Epidemic Acinetobacter baumannii strain with MRSA-like behaviour carried by healthcare staff. Eur J Clin Microbiol Infect Dis 2002;21:326—327. 4. Walsh TR, Bolmstrom A, Qwarnstrom A, et al. Evaluation of a new E-test for detecting metallo-b-lactamases in routine clinical testing. J Clin Microbiol 2002; 40:2755—2759. 5. Zeana C, Larson E, Sahni J, et al. The epidemiology of multidrug-resistant Acinetobacter baumannii: does the community represent a reservoir? Infect Control Hosp Epidemiol 2003;24:275—279.

B. Schultea, C. Goerkea, P. Weyrichb, S. Gro ¨bnera, C. Bahrsc, C. Wolza, I.B. Autenrietha, S. Borgmanna,* a Institute of Medical Microbiology and Hygiene, ¨bingen, Tu ¨bingen, Germany University of Tu b Department of Internal Medicine IV, University of ¨bingen, Tu ¨bingen, Germany Tu c ¨bingen, BG Trauma Centre, University of Tu ¨bingen, Germany Tu E-mail address: stefan.borgmann@med. uni-tuebingen.de Available online 6 October 2005

*Corresponding author. Tel.: C49 7071 29 81528; fax: C49 7071 29 3435. Q 2005 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jhin.2005.05.009

A new screening method for carbapenem-resistant Acinetobacter baumanii and Pseudomonas aeruginosa Sir, With the rapid spread of Acinetobacter baumanii strains resistant to nearly all antibiotics1 and bearing

357 potentially transmissible carbapenemase determinants (TCDs), there is a need to develop and employ appropriate screening methods to prevent such strains from becoming endemic in further hospitals. We have developed a highly selective screening method that could be adopted by hospitals that do not yet have endemic TCD-bearing A. baumanii. When devising a medium for detecting TCDbearing A. baumanii and Pseudomonas aeruginosa, there are two key advantages. One is that both species will grow on media, for example that of Berlau et al.,2 that have insufficient nutrients to support the growth of most other pathogens. The other advantage is that TCD-bearing strains will grow with concentrations of antibiotics that would inhibit nearly all other bacteria. We added antifungal (amphotericin B 2 mg/L) and antibiotic (meropenem 5 mg/L and vancomycin 5 mg/L) to the medium of Berlau et al. which consists of inorganic salts, sodium acetate, glucose and agar. As ours is a liquid medium, we omit the agar. We dissolve each salt separately then autoclave a mixture of the solutions to which the antibiotics and filter-sterilized glucose and acetate solutions are added. The meropenem in the medium is unstable so 10-mL bottles of the liquid medium are stored frozen before use. This acetate, vancomycin, meropenem and amphotericin (AVMA) medium supports the growth of TCD-bearing A. baumanii and P. aeruginosa. Bottles containing 10 mL AVMA are inoculated with swab material and incubated overnight in air at 37 8C. The resulting suspension is subcultured on to Oxoid chromogenic isosensitest agar (under development) to which a 10-mg meropenem disc is applied. Oxoid chromogenic UTI agar (CM0949) could also be used but the kaolin present (white) makes A. baumanii colonies less conspicuous and interferes with antibiotic action, so we use chromogenic isosensitest agar which is more likely to give appropriate meropenem zones. A. baumanii grows as opaque, white colonies that test oxidase negative; those that are growing close to the meropenem disc are tested for sensitivity to meropenem and are identified by API 20E. A. baumanii strains testing resistant to meropenem by disc test are further tested by meropenem E-test and E-test MBL. To detect P. aeruginosa, chromogenic plates are examined under long wavelength ultraviolet light. Only fluorescent oxidase-positive colonies growing close to the meropenem disc are considered to be possible TCD-bearing P. aeruginosa and are identified by API 20NE. Antibiotic sensitivity is tested using the same method as for A. baumanii. We have devised AVMA medium for the screening of patients who are unlikely to carry

358 TCD-bearing bacteria, so it is important that growth of most irrelevant bacteria is not supported. Stenotrophomonas maltophilia is meropenem resistant and is a species that we do not wish to isolate since its resistance is chromosomally coded and therefore not readily transmitted. Unfortunately, AVMA does sometimes support limited growth of S. maltophilia. When 14 pools, each of five nose and throat swabs taken at Great Ormond Street Hospital as methicillin-resistant Staphylococcus aureus screens, were cultured in AVMA medium, only one pool yielded colonies growing within 11 mm of the centre of the meropenem disc. The colonies tested oxidase negative and were translucent; they were subsequently identified as S. maltophilia. The colonial appearance did not fit with A. baumanii and the oxidase reaction did not fit with P. aeruginosa. In a blind testing of four strains each of A. baumanii, P. aeruginosa and S. maltophilia cultured on chromogenic isosensitest agar, we were able to assign the correct species for all strains on the basis of appearance and oxidase reaction. S. maltophilia is recognized because colonies are clear, sometimes faintly turqoise (and therefore not A. baumanii), and oxidase negative (and therefore not P. aeruginosa). As TCD-bearing A. baumanii has only been isolated once at our hospital, we supplied the media for evaluation purposes to a nearby Trust where there has been a recent outbreak. Swabs from 38 patients were tested and meropenem-resistant A. baumanii was isolated from eight samples with the new method and seven samples with the Trusts standard method. We hope the approach described will be widely adopted and help to slow the spread of TCD-bearing bacteria in and between hospitals.

Acknowledgements We thank the following people for their help: Neil Woodford, Tyrone Pitt, Annette Jepson, and Sarah Cheeseman and Peter Stephens of Oxoid’s Research and Development Department.

References 1. Coelho J, Woodford N, Turton J, et al. Multiresistant Acinetobacter in the UK: how big a threat? J Hosp Infect 2004;58:167—169. 2. Berlau J, Aucken H, Malnick H, et al. Distribution of Acinetobacter species on skin of healthy humans. Eur J Clin Microbiol Infect Dis 1999;18:179—183.

Letters to the Editor P. Parthasarathy, J. Soothill* Great Ormond Street Hospital, London, UK E-mail address: [email protected]

*Corresponding author. Tel.: C44 207 405 9200 ext. 5237; fax: C44 207 813 8268. Q 2005 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jhin.2005.06.011

Why investigate nosocomial diarrhoea?

Sir, The articles by Bruins et al. and Rao et al. were very informative.1,2 It was interesting to read both negative and positive aspects of applying the ‘three-day rule’, which could result in failure to detect nosocomial outbreaks caused by conventional enteric pathogens but save on unnecessary laboratory expenses. Such studies are of great importance for tropical countries like India where bacterial pathogens are frequent, and documented rates of nosocomial diarrhoea are very high.3,4 I would like to share my view on this subject in light of a study on nosocomial diarrhoea conducted in the paediatric wards of LN Hospital, New Delhi, India.5 During the study period, 60 cases of sporadic nosocomial diarrhoea were detected among 6000 patient admissions (1% incidence). Enteropathogenic Escherichia coli was the most common bacterial pathogen (12%). Salmonella and Shigella were not isolated from the patients, thus supporting the ‘three-day rule’. However, surveillance is important as this can be a yardstick for assessment of the quality of hospital hygiene, as the pathogens involved in the causation of nosocomial diarrhea are acquired exogenously. As such, increasing rates of nosocomial diarrhoea reflect hygiene malpractice. In her article on ‘How to assess hospital cleaning’, Dancer suggested the presence of ‘indicator organisms’ and high total aerobic colony counts as ways to assess surface hygiene in hospitals, and also addressed the issue of whether sampling should be carried out routinely or in response to an infection incident.6 As an increase in the incidence of nosocomial diarrhoea is a pointer towards hygiene malpractice, its surveillance is important and can serve as an indicator.