4. PL/CMO 2003/3: Hospital pharmacy initiative for promoting prudent use of antibiotics in hospitals. 2003. P. Seetulsingh a, * S. Collier b a Department of Medical Microbiology, Hemel Hempstead General Hospital, Hemel Hempstead, UK b Department of Microbiology, University College Hospital, London, UK E-mail address: seetul@aol.com Available online 10 July 2006 * Corresponding author. Address: Department of Medical Microbiology, Hemel Hempstead General Hospital, Hemel Hempstead HP2 4AD, UK. Tel.: þ44 1442 287834; fax: þ44 1442 287089. ª 2006 The Hospital Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jhin.2006.05.007 Neonatal astrovirus gastroenteritis during an inborn nursery outbreak Madam, Astroviruses are a common cause of childhood diarrhoea but neonatal cases are rare. 1 Astro- viruses were first described in 1975 following electron microscopy studies of an outbreak of diarrhoea in the nursery of a maternity ward, but there have been no such reports since. 2 We de- scribe an outbreak that occurred between 7 and 21 July 2004 in healthy neonates admitted to an 18-bed nursery of a maternity ward of Ramathibodi Hospital, Bangkok. After close postnatal observation for approxi- mately 6 h, a neonate born in the hospital is usually transferred to this maternity ward and is ‘roomed’ with its mother in an individual room. A neonate born by Caesarean delivery is generally kept in the nursery for the first 24e48 h of life, while a neonate born by vaginal delivery is ‘roomed’ with its mother straight away. The epidemic started when a neonate, born on 1 July 2004, developed diarrhoea on 7 July 2004 (three days after being discharged). Within two weeks, 16 neonates, admitted to the same nursery, developed diarrhoea either whilst on the ward (nine cases) or two to four days after they went home (seven cases). An epidemiological investigation and a programme to control the diarrhoeal outbreak were implemented fully on 14 July 2004. 3,4 All neo- nates with diarrhoea were isolated with enteric pre- cautions. Neonates without diarrhoea in the nursery were placed in one cohort, and newly born neonates were accommodated in a well-cleaned temporary nursery. The importance of assiduous handwashing was emphasized. The nursery was closed for three days. Environmental decontamination was per- formed, and after thorough cleaning, all nursery materials and surfaces were disinfected with hypo- chorite. The parents of all discharged neonates at risk were telephoned to check for any observable ill- ness. An active surveillance programme for early detection of new cases was established. None of the neonates admitted to the nursery after 14 July 2004 developed gastroenteritis during a three- month surveillance period. On 16 and 19 July 2004, between zero and nine days after the onset of diarrhoea, stool samples from the 13 neonates and one member of nursery staff who had diarrhoea were collected and cul- tured for enteric bacterial pathogens, and tested for enteric viruses. Rotavirus was detected by RNA- polyacrylamide gel electrophoresis and the silver stain method, 5 and calicivirus was detected by polymerase chain reaction (PCR). 6 Astrovirus and enteric adenovirus antigens were tested by en- zyme-linked immunosorbent assays (ELISA) using commercial test kits (Amplified IDEIAä Astrovirus and IDEIAä Adenovirus, DakoCytomation, Carpin- teria, CA, USA). Although astrovirus was only detected in the stool samples of four of the 13 neonates and the staff member who had diarrhoea, all 16 neonates and the staff member were thought to be infected with astrovirus. Enteric bacterial pathogens, rota- virus, enteric adenovirus and calicivirus were not found in any of the stool samples tested. All 16 neonates who had been hospitalized on the same postnatal ward had similar clinical manifestations with onset within a two-week period, and nursery- associated diarrhoea had previously been very unusual in this setting. The reasons for negative astrovirus detection in the other cases may have been because the stool samples from each case were only tested once on various days after the onset of diarrhoea, and ELISA is not the most sensitive method. Stool collection during the first two to three days of illness and repeated collec- tion thereafter has been shown to facilitate the detection of viral pathogens. 3 Reverse transcrip- tase PCR was found to double the detection rate 196 Letters to the Editor