Contribution of bacteriology and virology in sudden unexpected death in infancy.

OBJECTIVE: To appraise the inter-agency protocol used in sudden unexpected death in infancy (SUDI) cases in South Yorkshire, UK. DESIGN: A retrospective audit of 121 postmortems carried out over a 3-year period was completed to assess adherence to local guidelines introduced in 2005 specifying the required microbiological specimen set to be collected at postmortem in cases of SUDI. Data on organisms isolated was also collated and assessed for significance. SETTING: Sheffield Children's Hospital Histopathology Department is the South Yorkshire referral centre for SUDI. Post-mortem samples were processed by Sheffield Teaching Hospital's microbiology and virology departments. PATIENTS: All postmortems of SUDI in children less than 2 years of age performed between January 2004 and December 2007. RESULTS: 116/121 cases had samples sent for microbiological and/or virological investigation: 90% of cases had a blood culture and 68% had a cerebrospinal fluid sample taken. Of the 116 cases, 49% had a potentially pathogenic organism isolated, 73% had post-mortem flora and 10% had no organisms isolated (32% had both post-mortem flora and a potential pathogen). 27% of cases were found to have middle ear exudate requiring sampling, from 48% of which a potentially pathogenic organism was isolated. CONCLUSIONS: Our finding of a potential pathogen in 57/116 (49%) of our cases, although not necessarily the cause of death, confirms the relevance of performing multisite and virology investigations in all cases of SUDI. Standardised protocols with agreed definitions are necessary for a consistent approach.

Hospital-based epidemiology: a strategy for 'dealing with Clostridium difficile'.

Clostridium difficile-associated diarrhoea (CDAD) remains a major infection control problem. Uncertainty remains over methods of diagnosis and definitions for ascertaining provenance of cases. We undertook a prospective epidemiological study to better ascertain local epidemiology of 275 new cases (general practitioner and hospital) diagnosed in a large teaching hospital in the UK. The highest incidence of cases was found in haematology and critical care and a surprisingly high proportion, 29%, of hospital cases occurred in those aged <65 years. Fifty-five cases were diagnosed within 48h of admission. Of these, those defined as 'community-acquired' varied between 9 and 25 according to various proposed definitions relating to acquisition and onset of diarrhoea. Of 48 community-onset cases, this number varied between 19 and 25, the variability making comparisons between National Health Service (NHS) trusts potentially inaccurate. Cases were followed for 90 days after diagnosis and all cause mortality data obtained. Of 227 cases diagnosed in hospital, 56 (25%) died within 30 days, 29% of whom were aged <65 years. Death certification data were available in 86% of these cases. C. difficile was recorded on 15 (31%) certificates and as a primary cause (1a or 1b) in 8 (17%) cases. Our study shows the value of local epidemiology for planning infection prevention and control strategies within an NHS trust and for contributing to the evidence base for national targets and policies.