Send letter to journal:
Re: MRSA bacteraemia is different on a neonatal unit compared to a paediatric unit.

Dear Editor

Methicillin resistant Staphylococcus aureus (MRSA) bacteraemia among children in a district general hospital: Neonatal unit and Paediatric Unit data are different.

We were very interested to read Khairulddin et al's report. MRSA now causes 40% of all Staphylococcus aureus bacteraemia (SAB) in England and Wales, and is increasing in frequency among the under 15-year-old population. Khairulddin et al. comment on the data collected nationally through the mandatory reporting of SAB. They conclude that the increasing proportion of MRSA among children is a cause for concern in view of the associated higher mortality and longer hospital stay. However the nationally collected data misses important differences between infants on neonatal units and those on paediatric wards.

We performed a retrospective study of all SAB in children and neonates at Birmingham Heartlands Hospital (a district general hospital) over a period of 10 years from 1993 to 2003. We focus particularly on the cases of MRSA bacteraemia.

In a neonatal intensive care unit serving 6000 deliveries per year there were 33 cases of SAB. Three were discounted as contaminants. Of the remaining 30, 8 cases (27%) were due to MRSA. The babies with MRSA were of lesser gestation (median 28 weeks vs 30) and similar post-delivery age (12 days) when compared to those with methacillin sensitive SAB. However they were more likely to be known to be colonised with MRSA (37% vs 14%), and they were more likely to have central venous access (87% vs 45%). Mortality was 12.5% in the MRSA group, and 9% in the methacillin sensitive group. The first case of MRSA bacteraemia was identified in 1997 but we found no clear trend of increasing incidence.

The paediatric unit serves a population of approximately 100,000 children (up to 16 years). We identified 64 cases of SAB over 10 years and discounted 13 cases as contaminated. Two of the contaminants were MRSA. Only 3 (6%) of the remaining 51 cases were due to MRSA. Of these 3, one was community acquired without any identifiable risk factors. Two were hospital acquired after admissions of 21 and 11 days. The first had bronchiolitis and central venous access, the second had Harlequin ichthyosis and was known to be colonised with MRSA. Only one paediatric patient died, with methacillin sensitive SAB.

The overall proportion of SAB caused by MRSA was 13.5% in our study – comparable to the nationally reported rates. However our data shows large differences in incidence and outcome between the neonatal unit and paediatric ward. Our data suggests that central venous access (or long lines) is a risk factor for MRSA bacteraemia on neonatal units and supports the urgent need for improving infection control measures.

Future data on SAB in children needs to be separated into neonatal and paediatric populations.

Reference