Send letter to journal:
Re: PICU - a lesson for the next phase of development

Dear Editor

Henderson et al produce data on critically sick children which show a clear, and unambiguous, improvement in outcome for the most sick children card for in a regional PICU. Their data however reveals little (if any) impact of these expensive services on outcome for those children with PRISM scores which indicate less severe physiological derangement[1].

The underlying message therefore from this study is similar to almost all of the previous studies[2,3,4] - that the most sick children do well and others appear to get better and may not need a complicated potentially hazardous, and expensive interhospital transfer.

The ramifications of the continuing centralisation of PICU services has meant that many District General Hospitals have stopped caring for chidren at the lowest tiers of intensive care for any significant duration. A number of authors have pointed to concerns with regards to loss of skills in managing acutely sick children and in some circumstances a loss of confidence amongst staff. This aspect of paediatric emergency care and the effct on outcome for all children has not been studied.

Many working in tertiary PICU centres express the (almost certainly correct) view that most parents would not mind travelling to ensure that outcome for their sick child is optimised. paediatricians should accept that the (at present) the evidence very strongly suggests that the most sick children should be transported to large regional PICU's. These should concentrate on the most sick children and larger District General Hospitals should be supported to meet standards to provide Level 2 intensive care - working within a networked provision.

Tilford reveals a clear relationship between volume and outcome in a PICU setting[5] such that units with less that 1000 children annually must surely have a questionable future - unless governance arrangements, with large neighbouring providers, exist to maintain the skills of the medical and nursing staff. Another marker of intensity might could be number of "ventilator days".

The real message evolving from these studies is that development of such services should be along the lines of managed clinical networks[6] - with transfer according to predetermined clinical criteria and with support for those hospitals which accept children through the casualty door in the provision of emergency, HDU and, in some cases, level 2 provision. This surely is the message to optimise outcome for our most sick children.

References

(1) Henderson AJ, Garland L, Warne S et al Risk Adjusted mortality of critical illness in a defined geographical region. Arch Dis Child 2002 86 194-199

(2) Pollack MM, Alexander SR, Clarke N et al Improved outcome from tertiary centre paediatric intensive care. Crit Care med 1991 19 150-159

(3) Gemke RJ, Bonsel GJ. The paediatric intensive care assesmentof outcome study group (PICASSO) Crit Care med 1995 23 238-245

(4) Pearson G, Shann F, Barry P et al Should paediatric intensive care be centralised. Lancet 1997 349 1213-1217

(5) Tilford JM, Simpson PM, Green J et al Volume outcome relationships in paediatric intensive care units. Pediatrics 2000 106 289 -94

(6) Clinical networks a discussion paper. The NHS confederation 2001 Publ: NHS Confedration.