KEY

A	Cambridge University Hospitals NHS Foundation Trust
B	Brighton & Sussex University Hospitals NHS Trust
C	Cardiff & Vale NHS Trust
D	Central Manchester & Manchester Children's University Hospitals NHS Trust
E	Great Ormond Street Hospital for Children NHS Trust
F	Guy's & St. Thomas' NHS Foundation Trust
G	Hull & East Yorkshire Hospitals NHS Trust
H	King's College Hospital NHS Trust
I	Leeds Teaching Hospitals NHS Trust
J	The Lewisham Hospital NHS Trust
K	Newcastle upon Tyne Hospitals NHS Foundation Trust
K1	Newcastle General Hospital
K2	Newcastle Freeman Hospital
K3	Newcastle Royal Victoria Infirmary
L	University Hospital of North Staffordshire NHS Trust
M	Nottingham University Hospitals NHS Trust
N	Oxford Radcliffe Hospitals NHS Trust
O	Royal Brompton & Harefield NHS Trust
P	Royal Liverpool Children's NHS Trust
Q	Sheffield Children's NHS Foundation Trust
Q1	Sheffield Children's Hospital (NICU)
Q2	Sheffield Children's Hospital (PICU)
R	Southampton University Hospitals NHS Trust
S	South Tees Hospitals NHS Trust
T	St. George's Healthcare NHS Trust
U	St. Mary's NHS Trust
V	Birmingham Children's Hospital NHS Trust
W	United Bristol Healthcare NHS Trust
X	University Hospitals of Leicester NHS Trust
X1	Leicester Glenfield Hospital
X2	Leicester Royal Infirmary
Y	NHS Lothian - University Hospitals Division
Z	Barts and the London NHS Trust
ZA	NHS Greater Glasgow and Clyde - Women and Children's Division
ZB	The Royal Group of Hospitals and Dental Hospital HSS Trust

Published in the UK by the Paediatric Intensive Care Audit Network (PICANet). This work is copyright. Apart from any use as permitted under the Copyright, Designs and Patents Act 1988, no part may be reproduced by any process without permission from PICANet. Requests and enquiries concerning reproduction rights should be directed to PICANet at:

PICANet
Paediatric Epidemiology Group
Centre for Epidemiology and Biostatistics
The Leeds Institute of Genetics, Health and Therapeutics
Room 8.49, Worsley Building
University of Leeds
Clarendon Way
Leeds
LS2 9JT
0113 343 8125
picanet@leeds.ac.uk

In all cases PICANet must be acknowledged as the source when reproducing or quoting any part of this publication. Please use the following format when citing this report:

National Report of the Paediatric Intensive Care Audit Network January 2005 - December 2007 (published June 2008): Universities of Leeds and Leicester. ISBN 978 0 85316 275 9.

1 CONTENTS

2 ACKNOWLEDGEMENTS

We are acutely aware that the success of this national clinical audit is highly dependent on the hard work and commitment of a large number of individuals working within the paediatric intensive care community. We are very grateful to all the audit clerks, secretaries, nurses and doctors who support and contribute to the Paediatric Intensive Care Audit Network (PICANet) from their own paediatric intensive care units (PICUs).

PICANet was established in collaboration with the Paediatric Intensive Care Society (PICS) and their active support continues to be a key component of our successful progress. The PICANet Steering Group (SG) has patient, academic, clinical, government and NHS members all of whom are thanked for their continuing assistance and advice. Members of our Clinical Advisory Group (CAG) are PICANet's formal interface with clinical care teams and their valuable support and contribution is gratefully acknowledged.

PICANet is funded by the Healthcare Quality Improvement Partnership (HQIP), Health Commission Wales Specialised Services, NHS Lothian / National Service Division NHS Scotland, the Royal Belfast Hospital for Sick Children and the Pan Thames PICU Commissioning Consortium.

The organisation and functioning of PICANet is dependent on IT programming and development from Martin Perkins (University of Leicester), who we thank for his essential contributions.

3 FOREWORD

I am delighted to write a foreword for this PICANet Annual Report. The importance of timely and accurate information to audit health services has recently been re-emphasised, with a focus on clinical outcomes and patient reported experience - the latter a challenge in the children’s world which I am sure we can rise to in due course.

PICANet is well established and highly valued. The information is increasingly used with a wider audience.

I congratulate all those who work hard to collect and analyse these data - clinicians, managers, researchers and their teams. Your efforts to improve the care of the sickest children are to be admired.

I encourage and applaud you both in your current endeavours and in considering future developments to build on this excellent work in timely and cost effective ways, ensuring benefits are achieved from a local, regional and national perspective.

Sheila Shribman

Dr Sheila Shribman
National Clinical Director for Children, Young People and Maternity
Services

4 EXECUTIVE SUMMARY

PICANet is a clinical audit of paediatric intensive care (PIC) activity in England and Wales aiming to improve patient outcomes through providing information on delivery of care to critically ill children and an evidence base for clinical governance. PICANet was established in 2001 and functions in close collaboration with members of the PIC clinical community.
The specific objectives of PICANet are to identify best practice, monitor supply and demand, monitor and review outcomes of treatment episodes, facilitate strategic health care planning, quantify resource requirements and study the epidemiology of critical illness in children.
The national PICANet dataset continuously records details of admission, discharge, diagnoses (coded using Clinical Terms 3 (The Read Codes)), medical history, physiology, interventions and outcome. The outcome information is adjusted by 'case mix' to provide reliable evidence on patients' outcomes for clinicians, managers, patients. From 2006 the casemix adjustment tool is the updated Paediatric Index of Mortality 2.
Rigorous data quality procedures, incorporating iterative feedback loops between PICANet and the units, continue to ensure the dataset is of high quality.
PICANet have developed and expanded the core dataset in response to changes in the infrastructure and funding streams of the NHS. PICANet has provided software for units to record the Paediatric Critical Care Minimum Dataset (PCCMDS) to support the Paediatric Critical Care Healthcare Resource Groups (HRGs) and Payment by Results (PbR). The flexibility for the collection of unit specific additional items will remain, whilst additional modules, such as that on retrievals, are under construction.
Data are presented on 43,841 paediatric intensive care admissions to 25 NHS trusts in England and Wales and the Royal Hospitals for Sick Children in Edinburgh and Glasgow over the three year period January 2005 to December 2007. Detailed tables present information nationally, by Strategic Health Authority/Health Board (SHA / HB), Primary Care Organisation (PCO) and named individual NHS trust. Data are again, as last year, available for downloading from the Web in spreadsheet format.
Children under 1 year comprise 47% of all admissions with an overall excess of boys (58%) compared to girls (41%). The majority of admissions (54%) are unplanned. Retrievals of 76% of children are by specialist paediatric intensive care teams.
Invasive ventilation procedures are recorded for 67% of admissions. This varies by trust between 6% and 94% over the three years.
A total of 253,554 bed days were delivered between 2005 and 2007. Length of stay has been calculated to the minute and presented as numbers of admissions by length of stay category ranging from less than an hour (0.6%) to 7 days or longer (16%). A 'bed census' has been calculated for children actually occupying a bed at 10 minutes past midnight on each day to provide a more accurate assessment of daily occupancy in the PIC service.
It is extremely rare for a child to die in paediatric intensive care and 95% are discharged alive. Risk-adjusted performance of all trusts fell within acceptable limits in each individual year and aggregated across the three year period.
The re-organisation of the NHS into Primary Care Organisations in 2006 is reflected in this report. Maps by SHA and PCO illustrate considerable variation in the geographical distribution of the volume of patients receiving paediatric intensive care and the percentage of children invasively ventilated.
PICANet acknowledge that data on status 30-day post discharge is incomplete for 55% of children discharged alive.
Eleven recommendations arising from this report are outlined in the next section.

5 RECOMMENDATIONS

PICANet recommend that

high quality clinical audit data on children receiving intensive care in the UK continue to be collected to optimise the delivery of care, to facilitate future planning, permit ongoing audit and describe the epidemiology of critically ill children. <\li>
PICANet develop links with relevant patient / parent groups to facilitate the development of patient orientated audits to optimise quality of patient care.<\li>
links with the clinical community and professional organisations continue to be strengthened and expanded via collaborative audit using the PICANet dataset.<\li>
links with PIC commissioners are enhanced to facilitate the planning of PIC services.<\li>
the PICANet dataset should be used to recalibrate the mortality risk-adjustment algorithms in paediatric intensive care and publish these on an annual basis for the UK.<\li>
Trusts provide support for the collection of child status at 30 days following discharge from PIC especially in those trusts with little or no follow-up data.<\li>
Trusts share their experiences of the collection of NHS numbers to improve this data collection to a level in excess of 95%.<\li>
continued efforts to capture complete national data on children admitted to adult intensive care units.<\li>
PICANet data collection is integrated with detailed data concerning retrievals to facilitate national audit of retrievals data.<\li>
international collaborations should be established to enable the development of large-scale audit comparisons between countries that will inform clinical practice.<\li>
all PICUs should be encouraged to supply the components of the PCCMDS to PICANet to enable more detailed analysis of activity and level of care at a national level.

6 BACKGROUND

PICANet was established in 2002, following a tender in 2000 by the DOH for a national paediatric intensive care database that would allow core data to be collected in a standardised way throughout all PICUs in the country.

Since November 2002, all NHS PICUs within England and Wales outside the Pan Thames region have been collecting data on consecutive admissions to their units. The Pan Thames units began data collection in March 2003, whilst the PICU at the Royal Hospital for Sick Children, Edinburgh began in December 2004. The Royal Hospital for Sick Children, Glasgow began in March 2007. The Royal Belfast Hospital for Sick Children began in April 2008. A full list of participating PICUs can be found in Appendix A.

PICANet receives support and advice from a Clinical Advisory Group consisting of doctors and nurses working within the speciality. A Steering Group (SG), comprising professionals from Health Services Research, the Royal Colleges of Paediatrics & Child Health, Nursing and Anaesthetics, and user groups such as Action for Sick Children, monitors PICANet and offers additional support and advice. Appendices B and C provide a full list of CAG and SG members. Additional support from the clinical community is provided through the Paediatric Intensive Care Society.

7 INTRODUCTION AND AIMS

This is the fifth national report produced by PICANet on data submitted by participating PICUs in the UK.

The report has been published in two formats:

As a .pdf document, downloadable from http://www.picanet.org.uk/.
As a web document with tables and figures available for download in Microsoft Excel format, again, available from http://www.picanet.org.uk/.

We have decided not to produce a print document for environmental and cost reasons. The downloadable format means that individuals can select specific sections of the report to print if necessary and the tables and figures can be manipulated and used in presentations and reports. Please ensure that PICANet is acknowledged as the source of this information using the format given on the inside cover.

In collaboration with participating units, PICANet remains committed to achieving the following objectives:

Identifying best practice.
Monitoring supply and demand.
Monitoring and reviewing outcomes of treatment episodes.
Facilitating strategic health care planning and quantifying resource requirements.
Studying the epidemiology of critical illness in children.

Since data collection commenced in 2002, one of the main aims of PICANet has been to provide a national database of paediatric intensive care activity of a consistently high quality, in order to help achieve the above objectives. With the addition of the Royal Belfast Hospital for Sick Children, PICANet covers all UK PICU admissions from April 2008. The expansion of the dataset to include the Paediatric Critical Care Minimum Dataset (PCCMDS) will mean that PICU activity can be assessed by level of care in the future. We hope that all units will be able to supply this data to PICANet in future for national comparisons.

The data collected allows comparisons of activity at a local level with nationwide benchmarks. PICANet therefore provides an important evidence base on paediatric intensive care outcomes, processes and structures, permitting planning for future practice, research and interventions. PICANet is a resource available to clinicians and service providers, amongst others, and is being used for research, audit and commissioning (Appendix D). The provision of comprehensive, routinely available information to such parties is extremely important and is a powerful tool for supporting clinical governance. PICANet is also used to provide data to provide baseline information for clinical trials.

This year we have included a short piece written by the parents of a child admitted to PICU in the course of her illness. We hope that we will continue to receive comments and opinions from parents of children admitted to PIC (or the children themselves) in order that a more complete picture of the care delivered is provided in this report.

8 PICU: ONE PARENT’S IMPRESSIONS

The background

Our daughter was admitted to PICU 36 hours after we discovered that she had cancer. The enormous explosion of cancer cells into her blood caused tumour lysis to set in as the cells became starved of oxygen, died and collapsed releasing toxins into her body that threatened to stop her heart and cause her kidneys to fail. She was four and a half and had been healthy and happy until two days before. To save her she was admitted to PICU and put on a continuous dialysis machine to reduce the toxins threatening to swamp her body. So, for us, PICU was part of the chaotic fear of those first few days. Our lives had turned upside down and we had not yet evolved the coping mechanisms we needed.

Our experience

On the positive side PICU saved our daughter’s life and we are eternally grateful for that. The nursing care she received was superb and it was helpful at times to have nurses who just took over and did everything she needed. The best PICU nurses in circumstances like ours were quietly assertive and organised us all with no-nonsense warmth and compassion. They inspired trust and trust is crucial to coping. People were very kind to us, and things like the quilt she received and visits from nurses who had looked after her on the one day she had been on the cancer ward made us feel other people felt she was special too. It was a lifeline being allowed visitors to come and sit with me especially for those interminable evenings as either I or my husband was always at home with our other daughter.

On the negative side her terror and distress, and our own, meant the whole period was a hormonal roller coaster. I hated having to leave her to sleep so I hardly slept – I dreaded her waking and not seeing a familiar face. I hated having to leave her to eat so I hardly ate. I wasn’t even allowed to drink tea near her. I hated not having stuff to do to look after her, just having to sit there – that was a killer. I found the way the PICU consultant swept in and out trailing people behind him intimidating – I was always close to tears and too many new interactions was too much. I needed to trust him but felt no connection between him and my daughter so it was difficult. Also we had to relate to a PICU consultant, a renal consultant and an oncology consultant. It was horrendous not being able to hold her and rock her – she was all tubes and wires, panic and pain. I felt completely cut off from fresh air, the natural world and the rest of my life. It would not have been much more dislocating to have woken up as a cockroach. I was not in control of her treatment and was forced to trust people I had never met before with her life. She did not even have a full diagnosis yet, and we had had no chance to learn the names of the medicines and their roles – we were always told, but my brain was too full of stress hormones to fully compute and I could not ask questions without the risk of crying.

Overall it was worth anything to save her life, but these are our impressions and are obviously a product of our specific circumstances.

9 A CLINICIAN'S COMMENTARY
Dr Michael Marsh

It is just over 10 years since the publication of A Framework for the Future and A Bridge to the Future documents^1,2 that followed Nicholas Geldard’s death in 1995. Paediatric Intensive Care in the United Kingdom has made significant strides in developing a high-class modern service. The issue of capacity has largely been addressed when compared to the situation the UK was in at the time of the 2nd BPA report on Paediatric Intensive Care³ when only approximately 50% of children requiring intensive care were cared for in a designated children’s facility. However over the last 3 winters there has still been a large number of children transferred out of region in order to receive intensive care in a PICU. With the publication of this fifth annual PICANet report we have the opportunity to turn our attention to a number of issues and areas that have not been specifically focused on as well as reconsidering capacity.

Capacity – Do we have adequate capacity for PIC in the UK? Winter pressure on beds still appears to be an issue despite the developments of the last 10 years. If we were to look back at the letter to the Lancet written by Frank Shann⁴ in 1993 the answer would probably be yes – but it is still in the wrong place! The concept of large general PICUs admitting over 1,000 children a year versus a larger number of small units remains unresolved with professional opinion still divided. Lord Darzi in his report High Quality Care For All⁵ published on the 60th anniversary of the NHS may not help – as the concept of treating centrally where necessary and local where possible can be interpreted to support either model.

Quality and Outcome – It should be the aim of everyone working in paediatric intensive care to deliver the highest quality care with the objective of obtaining the best outcomes possible. Lord Darzi’s report places a new emphasis on these issues and as the NHS focuses more on quality as well as outcomes. We need to concentrate on both patient experience (including parental experience) and staff experience. Hence we should heed the words of the parent who so eloquently recalls their child’s admission to a UK PICU carefully.

One Parent’s Impression may make uncomfortable reading for some within the community. Our nurses however can feel proud of the work they do and the care they provide. Over the past 15 years my experience is that their work, dedication, empathy, compassion and individualised care is widely appreciated by our patient’s families. We should note that the stress families experience whilst their loved one is under our care is considerable and the way in which we, the medical profession, conduct ourselves has far reaching consequences. Patient and family experiences needs our constant attention and are a concern for the whole team not just the nurses, and we can learn from each other but only if we share information and encourage research in this area.

Safety – There is increased emphasis within the NHS on safety and this is something all PICUs have been focusing on with the use of critical incident reporting, root cause analysis and rigorous governance systems. The concept of creating an environment of continual improvement is key to delivering a world-class service. It is exciting that the PICANet data now covers the whole of the United Kingdom with the recent addition of data from Glasgow so the picture is more complete. The challenge for the PIC community and the PICANet Team is to consider what metrics should be developed to audit and study the areas of experience, safety and outcomes beyond simple mortality.

Dr Michael Marsh
Divisional Clinical Director, Women and Children’s Services
Consultant Paediatric Intensivist, Southampton General Hospital
Honorary Secretary, Paediatric Intensive Care Society

9.1 References

A Framework for the Future. Report from National Coordinating Group on Paediatric Intensive Care to The Chief Executive of the NHS Executive.
A Bridge to the Future. Nursing Standards, Education and Workforce Planning in Paediatric Intensive Care Report to the Chief Nursing Officer’s Taskforce.
British Paediatric Association. The care of critically ill children. Report of a multidisciplinary working party on intensive care. 1993, London: BPA.
Shann F. Australian view of paediatric intensive care in Britain. The Lancet 1993; 342 (8863):68.
High Quality Care For All. NHS Next Stage Review Final Report. Lord Darzi, Gateway reference: 10106.

10 THE PICANet DATASET

10.1 Development and description of the current dataset

The PICANet dataset was established in consultation with members of the PICANet CAG, representing the paediatric intensive care community, and the Department of Health. The overriding criteria for inclusion of specific variables were that they provided key information on activity, case mix, demographics and outcome at a national and local level, that they were feasible to collect and that the wider paediatric intensive care community supported their inclusion in the national database.

The current PICANet dataset consists of 137 variables (including five address elements, the option for a second family name and 6 optional variables). These variables and their definitions are given in the PICANet Data Definitions Manual, obtainable from http://www.picanet.org.uk/. The data collection form is included in Appendix E. The dataset was expanded in summer 2007 when PICANet software was enabled to collect the Paediatric Critical Care Minimum Dataset.

10.2 The Paediatric Critical Care Minimum Dataset

The Paediatric Critical Care Minimum Dataset (PCCMDS) has been developed by the Information Centre for health and social care (IC) under the guidance of the Paediatric Critical Care Expert Working Group (PCCEWG) and was issued as an NHS dataset change notice (DSCN) in January 2007. The PCCMDS has been developed to support the new Paediatric Critical Care Healthcare Resource Groups (HRGs) and Payment by Results (PbR). This dataset has many common elements with the PICANet dataset but collects information on interventions and treatment on a daily basis as opposed to an episode summary. This dataset has been mandated from October 2007.

With the support of the CAG, PICANet has agreed to enable collection of the PCCMDS using its software. The current intervention fields will be populated using the new data items. This will ensure comparability with historical PICANet data and will reduce duplication of data collection effort. In the future, PICANet will also have more detailed information on daily activity which will provide better information for clinical audit and commissioning. The software will also enable PICUs to export the PCCMDS for processing by their trust to enable accurate returns for PbR. The additional burden of data collection is estimated at 1 minute 45 seconds per patient per day based on the pilots carried out to develop the PCCMDS. PICANet will not be responsible for completing data returns for PbR from the central database.

10.3 Retrievals dataset

PICANet has not collected detailed information on retrievals of critically ill children in the past, concentrating on their experience in PICU. With the support of PICANet, the Clinical Advisory Group and the Paediatric Intensive Care Society, Dr Allan Wardhaugh has developed detailed proposals for a dataset that will capture information on this important sub-population of children during the retrieval process. These proposals are outlined in a separate chapter in this report.

10.4 Data collection and validation

PICANet has developed a paper data collection form and bespoke data entry software to enable a consistent national dataset to be assembled. Those units who use their own or commercial data collection software have been provided with an export file specification to enable data to be imported by the PICANet software. Training sessions were organised over two days to familiarise clinical and data entry staff with data definitions, data collection issues and software. Since the original training sessions, ad hoc training has been provided by the PICANet team for new staff concerned with data collection and entry.

The PICANet software performs internal logical consistency and range checks as data are entered and provides an on-screen summary of outstanding validation checks on the completion of a record. Units importing data from their own databases are provided with an import log, detailing which records have been imported and any outstanding validation issues. Central validation and data quality issues are dealt with in the section on data quality.

10.5 Clinical coding

Clinical diagnoses and procedures are coded using Clinical Terms 3 (The Read Codes) referred to as CT3. CT3 encompasses a huge range of diagnostic, procedural and context-dependent clinical codes designed to reflect all aspects of clinical care in the population in general. The long-term strategy of the NHS is to use SNOMED CT^® for clinical coding of diagnostic information (see http://www.connectingforhealth.nhs.uk/ for further details). PICANet will migrate to SNOMED CT^® when the appropriate support architecture is in place but will continue to use CT3 in the meantime. There are plans to develop a SNOMED subset for PICU, an initiative supported by Connecting for Health. This issue is being taken forward by representatives of the Paediatric Intensive Care Society Study Group Health Informatics Group, with the support of PICANet.

10.6 Confidentiality

PICANet collects patient identifiable information including name, address, date of birth and NHS number. With this information, PICANet can identify multiple admissions for the same individual, making the dataset person and episode-based. Personally identifiable information held by PICANet has been linked with death registration details, obtained from the Office for National Statistics (ONS), to assess long-term mortality in children admitted to paediatric intensive care. National census and other geographical data have been linked with validated postcodes of individual children to enable PICANet to assess the association between social class, population density and other geo-demographic and environmental information and paediatric intensive care admissions.

To comply with the provisions of the Data Protection Act, 1998, PICANet has implemented stringent confidentiality and data protection arrangements. The Patient Information Advisory Group (PIAG) has granted PICANet exemption from gaining signed parental consent under Section 60 of the Health and Social Care Act, 2001. This class support enables PICANet to collect and process patient identifiable information for the purpose of auditing, monitoring and analysing patient treatments, to ensure that adequate and appropriate paediatric intensive care services are available for all children admitted for paediatric intensive care. Exemption was given under specified conditions in December 2002 and is due for review in June 2008.

Posters providing information about PICANet are displayed in PICUs, and information leaflets for parents / guardians and children are available (see Appendix F for a copy of the information leaflet).

10.7 Data transmission

The PICANet data entry software includes the facility to transmit data electronically via the NHS intranet if local IT infrastructure can be configured appropriately. The data are first encrypted using public key encryption and then placed on the server. The uploaded data is regularly moved to a secure holding area, decrypted and uploaded onto the central server database.

Where local IT departments have been unable or unwilling to configure their systems and firewalls to allow electronic transfer, the data is encrypted and placed in a local folder and then sent as an email attachment.

11 DATASET DEFINITIONS FOR THIS REPORT

This report covers the three year period January 2005 - December 2007. During this time, there were 44,836 admissions to participating PICUs.
There are 27 participating NHS trusts (located in England, Wales and Scotland), 24 of whom collected data for the entire reporting period. Barts and the London,Edinburgh and Glasgow did not contribute over the whole period.
Trusts are identified in this report, with agreement from all participating trusts' Chief Executives.
A key enabling identification of each trust can be found on the inside cover.
The main focus of this report are admissions aged 0 - 15 years of which there were a total of 43,841 over the three year period. In addition there were 995 admissions aged 16 years and above.
Unless stated otherwise, the proportions in tables throughout the report are row percentages, except in the total column where they are column percentages.
'Unknown' includes cases where the unit have specifically recorded 'not known' and also cases where a required value has been left blank.

12 DESCRIPTION OF TABLES AND FIGURES

A brief description of the data contained in the tables and figures is given below, together with hyperlinks to the beginning of each section. In the .pdf version of this report, the hyperlink will bring you to the first page of the section. In the web document, the hyperlink will take you to an Excel spreadsheet that contains links to all the tables and figures in the section. These are all downloadable for use by individuals and organisations but please acknowledge the source of this data as indicated on the inside of the front cover. In some cases, individual figures are not described separately, as they clearly relate to the data in the tables on the same worksheet.

The PICANet dataset is dynamic and updated regularly. This means that overall admission figures have changed for 2005 and 2006 since the publication of the fourth national report. The data in this report are those supplied to PICANet up to June 2nd, 2008, when the dataset was frozen.

13 ADMISSION DATA

13.1 Admission numbers by age, sex, month and year of admission, NHS trust and diagnostic group

Tables 1 - 9 give numbers of admissions by age, sex, month of admission, NHS trust and diagnostic group. The primary diagnosis for the whole admission has been categorised into 13 diagnostic groups to enable a simple comparison between NHS trusts. The classification is based on CT3 (The Read Codes). Within these mutually exclusive thirteen groups:

Infection excludes any respiratory or gastrointestinal infection but includes meningitis
Neurological disorders include neurovascular complications
Oncology includes neuro-oncology (brain tumours)
Other includes those diagnoses not covered by the other 12 groups.

Read codes are five characters in length and can be made up of numbers, letters, or periods. The ordering of the individual characters does not indicate the hierarchy (e.g. patent ductus arteriosus (P70..) is a subset of congenital abnormality of ductus arteriosus (Xa6aC)). Table 8 and figure 8 focus on admissions for respiratory conditions by year and month.

13.2 Admissions by Strategic Health Authority (SHA) / Health Board (HB)

Table 10 gives numbers of admissions by SHA / HB, prior to and following the July 2006 NHS reorganisation. These were obtained by linking the validated home address of children admitted to PICU to SHA / HB via the National Statistics Postcode Directory (NSPD) (http://www.statistics.gov.uk/geography/nspd.asp). These tables present column percentages. Of the total number of admissions 96.8% had addresses which were validated. The remaining 3.2% included foreign addresses (2%) and missing addresses (1.2%). Figure 10 shows the SHA / HB boundaries overlaid by the primary care structure.

13.3 Admissions by mortality risk category

Table 11 gives numbers of admissions by mortality risk group by NHS trust. The expected probability of mortality was estimated using the paediatric index of mortality (PIM)¹, using recalibrated coefficients supplied by UK PICOS². The categorization into <1%, 1-<5%, 5%-<15%, 15-<30% and 30% plus expected probability of mortality reflects those used by the Australian and New Zealand Intensive Care Society (ANZPICS)³ for comparability.

13.4 Admissions by admission type

Tables 12 - 15 present numbers by admission type overall and by trust and year and a breakdown of the source of admission and care area admitted from by trust and year for emergency admissions (see below).

We have used the following definitions for type of admission:

An admission that is 'planned - following surgery' is one that the unit is aware of before the surgery begins and one that could have been delayed for 24 hours without risk (e.g. spinal surgery).
An admission that is 'unplanned - following surgery' is one that the unit was not aware of before surgery began and one that could not have been delayed without risk (e.g. bleeding tonsillectomy).
A 'planned - other' admission is any other planned admission that is not an emergency (e.g. liver biopsy).
An 'unplanned - other' admission is one that the unit was not expecting and is therefore an emergency admission (e.g. status epilepticus).

NB: Surgery is defined as undergoing all or part of a procedure or anaesthesia for a procedure in an operating theatre or anaesthetic room. Patients admitted from the operating theatre where surgery is not the main reason for admission (e.g. a patient with a head injury who is admitted from theatre after insertion of an ICP monitor) are not included here. In such patients the main reason for admission is head injury and thus the admission type would be 'unplanned - other'.

13.5 Admissions by primary diagnostic group

Tables 16 - 22 present a breakdown of admissions by diagnostic group, overall, by trust and year and further by trust and year for each of the admission types listed above.

Tables 23 - 25 present the twenty most common Read Codes returned to PICANet for primary reason for admissions overall (these represent 15,984 (37%) of all admissions) and for unplanned admissions (after surgery and 'other') by sex without any attempt to group them further.

PICANet has not imposed an arbitrary grouping of codes but present the raw data for the top 20 codes. The level of precision in the coding method makes interpretation of these data difficult without some form of aggregation. However, PICANet has allowed the flexibility to code very specifically to enable prospective audit to focus on particular conditions; for example, respiratory syncytial virus (RSV) positive bronchiolitis. Some units have chosen to code diagnoses in more detail to allow them to use this information locally, others have coded a single diagnosis at a general level. For most reporting purposes, the broad diagnostic groups used in this report are sufficient. Further disaggregation is not always possible due to the variation in coding practice between individual units.

13.6 References

Shann F, Pearson G, Slater A, Wilkinson K, Paediatric index of mortality (PIM): a mortality prediction model for children in intensive care. Intensive Care Med 1997; 23:201-207.
Brady AR, Harrison D, Black S, Jones S, Rowan K, Pearson G, Ratcliffe J, Parry GJ, on behalf of the UK PICOS Study Group. Assessment and Optimization of Mortality Prediction Tools for Admissions to Pediatric Intensive Care in the United Kingdom. Pediatrics 2006; 117: 733-742.
Australian and New Zealand Intensive Care Society. Report of the Australian and New Zealand Paediatric Intensive Care Registry 2005. ISBN: 1876980184 [Online] [Accessed 23/02/2007] Available from the World Wide Web at http://www.anzics.com.au/uploads/2005ANZPICRReport.pdf.

14 RETRIEVAL DATA

Tables 26 - 28 present retrieval data by team type and age, by diagnostic group for nonspecialist team retrievals (see below) and by team type and trust.

Data are collected on whether or not a child was retrieved / transferred into the PICU. We have used the following definitions:

'Own team' identifies that your own team collected the child from the referring hospital.
'Other specialist team (PICU)' identifies that another PICU retrieval team transferred the child to your unit.
'Other specialist team (non PICU)' identifies that another transport team, not a PICU team (e.g. Accident and Emergency Department (A&E), theatre teams or neonatal teams), transferred the child to your unit.
'Non-specialist team' identifies that a non-PICU, non-specialist team transported the child to your unit (e.g. ward staff).

In the majority of PICUs, doctors and nurses who work on the unit undertake retrieval of critically ill children. Within London, there are two specific transport teams, the Children's Acute Transfer Service (CATS) and the South Thames retrieval team. CATS is based at Great Ormond Street Hospital (GOSH), and is staffed separately from the intensive care units at GOSH. For PICANet, any child retrieved by CATS into a PICU at GOSH is recorded as 'other specialist team (PICU)'. The South Thames retrieval team is based at Evelina Children's Hospital and is staffed by doctors and nurses from within the PICU. For PICANet, any child retrieved by the South Thames team into the PICU at Evelina Children's Hospital is classed as 'own team'.

The Central Manchester and Manchester Children's University Hospitals NHS Trust has two sister hospitals (Booth Hall and the Royal Manchester Children's Hospital). For local reporting reasons, hospital transfers between the two hospitals are classed as internal admissions (admissions from the 'same hospital') but as the hospitals are 6 miles apart, any transfer requires a 'retrieval' by ambulance and crew.

15 INTERVENTION DATA

Tables 29 - 31 present summary data relating to interventions carried out on PICU. Most of the interventions described are available in all PICUs, although a few specialist interventions (such as extra corporeal membrane oxygenation (ECMO) or left ventricular assist device to support cardiac function (LVAD)) are only available in a PICU where invasive cardiac procedures are routinely performed. Note that table 30 contains aggregated data for 2005 - 2007. This, however, does not include Birmingham Children's Hospital as no intervention data was returned for 2005.

Length of ventilation was calculated in whole days. Any ventilation during the period

00:00 to 23:59 was counted as one complete day of ventilation (e.g. a child intubated and ventilated at 23:45 on 7 March, and extubated at 02:30 on 8 March, would count as two days of ventilation). Intubation and extubation times are not recorded in the PICANet dataset.

Figures 31a - 31b map the percentage of children receiving invasive ventilation by SHA / HB and by PCO for 2006 and 2007. Data for 2005 are not mapped as, no intervention data were returned by Birmingham Children's hospital in 2005. The proportion of children invasively ventilated has been used as a very rough proxy for level of care.

16 BED ACTIVITY AND LENGTH OF STAY

Tables 32 - 33 present data on total bed days delivered by age and sex overall and by age by trust. The total number of bed days delivered is calculated as the sum of children receiving intensive care in a PICU each day. Tables 34 - 35 and their associated figures present summary data by year and month and by trust and year on a 'bed census': the number of children present in a PICU bed at 10 minutes past midnight. Tables 36 - 37 present data we describe as 'bed activity' by month and by trust, where a bed is counted as occupied if a child was present on a unit for any part of a the day. This inevitably results in higher figures than the bed census data as a bed may have more than one child occupying it in any one day. Tables 38 - 39 present summary data on length of stay by trust and age group and trust and diagnostic group. Table 40 groups the number of admissions by length of stay by trust, calculated to the minute in categories ranging from less than 1 hour to over 1 week. Children admitted prior to the report period, but discharged during it, are counted from 00:00 on 1 January 2005 until their discharge (or until 24:00 on 31 December 2007 if not discharged). Children admitted during the report period but discharged in 2008 (or who are still on the PICU) are counted from their admission date until 24:00 on 31 December 2007.

The number of bed days, bed census, bed activity and length of stay data are summarised by median and interquartile range (IQR). Median daily bed census figures and daily bed activity are plotted using a box and whisker graph by month and year, and by NHS trust. This type of graph indicates the median by a line within the coloured box, the ends of which give the IQR. The 'whiskers' indicate values beyond the IQRs, although extreme outside values are not plotted.

17 OUTCOME DATA

PICU mortality data are described in terms of unit discharge status by age and sex for England, Wales and Edinburgh combined, and by trust in tables 41 - 45 and also using unadjusted and risk-adjusted standardized mortality ratios (SMRs). Table 46 describes the discharge destination of children discharged alive from PICU. Unadjusted SMRs are calculated by dividing the expected number of deaths, based on the national data by the observed number of deaths in each trust. In addition, risk-adjusted SMRs are calculated by dividing the expected number of deaths predicted by PIM¹ by the observed number of deaths in each trust. We have used the original version of PIM with revised coefficients supplied by UK PICOS² that give a better calibration as these coefficients are based on a more recent dataset. We have also produced SMRs using PIM 2³ for 2006 and 2007.

Unadjusted and risk-adjusted SMRs are presented by trust and year for 2005, 2006, 2007 and combined years in tables 47 - 49. PICU mortality funnel plots for the same periods are presented in figures 47a - 50b to provide a visual means of comparing unadjusted and adjusted SMRs between trusts, without imposing the ranking observed in league tables. Figures 48c and 49c presents risk-adjusted mortality using PIM 2.

The SMRs are plotted on the y-axis against the number of admissions to the trust on the x-axis. Higher mortality rates are represented by points plotted above the line of unity, with those appearing outside the upper control limit indicating an unusual excess mortality. Lower mortality rates are represented by points plotted below the line of unity and those falling below the lower control limit indicate unusually low mortality. In order to satisfy the condition, that if the overall distribution of the mortality ratios is random, there exists an approximately 5% chance of a unit falling outside the control limits, then the upper and lower control limits constructed at an individual unit level must represent not 95% confidence intervals, but 99.9% confidence intervals around a mortality ratio of one by number of admissions.⁴ This is analogous to increasing the confidence interval (or significance level) when correcting for multiple comparisons in data containing numerous groups. This means that the funnel plots are drawn in such a way that there is an approximately 5% chance of a unit falling outside the control limits if the distribution of SMRs is random.

In Figures 50, risk-adjusted SMRs by SHA / HB have been produced by allocating children to the SHA in which they were living based on their address at admission. These ratios have then been expressed as a percentage and mapped to illustrate the range of variability in SMRs between SHAs. It should be noted that these ratios have not been subject to any spatial smoothing and confidence intervals are relatively wide in areas of low population. Tables 51 - 55 present 30-day follow-up data by age, sex and trust.

17.1 References

Shann F, Pearson G, Slater A, Wilkinson K, Paediatric index of mortality (PIM): a mortality prediction model for children in intensive care. Intensive Care Med 1997; 23:201-207.
Brady AR, Harrison D, Black S, Jones S, Rowan K, Pearson G, Ratcliffe J, Parry GJ, on behalf of the UK PICOS Study Group. Assessment and Optimization of Mortality Prediction Tools for Admissions to Pediatric Intensive Care in the United Kingdom. Pediatrics 2006; 117: 733-742.
Shann F, Slater A, Pearson G. PIM 2: a revised version of the Paediatric Index of mortality. Intensive Care Med 2003; 29:278-285
Spiegelhalter D. Funnel plots for institutional comparison. Quality and Safety in Health Care 2002;11(4):390-391.

18 DATA ON INDIVIDUAL CHILDREN

In all other chapters of this report, PICU activity is presented for episodes of care or admissions. This chapter describes activity related to 32,490 individual patients representing the 43,841 admissions (0 - 15 years) during 2005 - 2007.

Firstly, Table 56 summarises admissions by the source of their previous admission (same or other trust or single admission only). Table 57 reports the number of children having repeat admissions by trust and Table 58 the number of children admitted by diagnostic group. Table 59 summarises the number of children admitted by diagnostic group either once to a single trust, more than once to the same trust or more than once to more than 1 trust.

19 PREVALENCE FOR ADMISSION

Age and sex specific prevalence for admission to PICUs in England and Wales has been calculated with 95% Poisson confidence intervals using population counts from the 2001 Census¹ (table 60). Age-sex standardised prevalence for the childhood population (less than 16 years) by SHA / HB has been calculated (table 61). This is mapped in figure 61a.

Children were allocated to an SHA / HB using their residential address at admission. Addresses were validated using AFD Postcode Plus² address validation software to obtain a correct postcode. Using the National Statistics Postcode Directory (http://www.statistics.gov.uk/geography/nspd.asp), postcodes were then linked to SHA / HB.

We have also presented age-sex standardised prevalence by PCO in figure 61b.

Prevalence for Scotland is not presented as PICANet only has data from the PICUs in Edinburgh and Glasgow for part of the reporting period.

Office for National Statistics. 2001 Census : Census Area Statistics (England and Wales) [computer file]. ESRC/JISC Census Programme, Census Dissemination Unit, MIMAS (University of Manchester).
AFD Refiner Q.2/08. AFD Software Ltd, Lough House, Approach Road, Ramsey, ISLE OF MAN, IM8 1RG, UK, 2008.

20 CHILDREN RECEIVING CARE IN ADULT INTENSIVE CARE UNITS

Data on children (under 16 years) treated in adult intensive care units (AICUs), including age in months, sex, date of admission and discharge, outcome and discharge location and admission diagnosis, were provided by the Intensive Care National Audit & Research Centre (ICNARC) and the South West Audit of Critically Ill Children (SWACIC). These data are summarised in tables 62 - 67. Analysis is restricted to 2005 and 2006. ICNARC receives data from 74% of AICUs in England.

Signed consent was obtained from the unit director of each AICU. ICNARC was able to release data from more AICUs in 2006 than in 2005. One AICU providing data to SWACIC did not give explicit permission for PICANet to receive their data.

21 DATA QUALITY

Data quality continues to be of paramount importance to PICANet as we expand our datacollection to include the Paediatric Critical Care Minimum Dataset (PCCMDS). PICANet is now in its seventh year of data collection and continues to assess and feedback on data quality at regular intervals. This is essential if we are to maintain the high standards now expected from us by the paediatric intensive care community.

Considerable effort is made by both PICU staff and the PICANet team to ensure that the data is of the highest quality. As units have acclimatised to the data collection process the overall quality of the data has improved. During previous years, the PICANet team visited individual units to review a sample of records to cross check that the data submitted to PICANet corresponded to that data held in the unit’s paper records and clinical information systems. This form of cross checking, although neglected due tostaffing shortage in previous years, has now been reinstated. Units are being visited by members of the PICANet team with a structured method for data quality validation.

This chapter details improvements in data quality during last year and highlights areasneeding attention. The results are presented by unit as well as by NHS Trust to acknowledge the importance of unit level data management.

21.1 Data quality assurance processes

Internal logical, consistency and range checks are carried out at input by the PICANet software with an on-screen summary of outstanding validation checks on completion of a record. Units importing data from their own databases or commercial software are provided with an import log detailing which records have been imported and outstanding validation issues.
Data transmitted to the PICANet central server in Leeds are subject to a series of additional validation checks (including address and postcode validation and clinical coding verification). Data validation reports (DVRs) are returned via email (Appendix G).
Units are provided with monthly admission reports (Appendix H) and asked to cross check these with local patient registers (e.g. unit admission book).
Units are provided with data status reports (Appendix I) which highlight particular dimensions of data quality that require attention, these include the number of missing values returned.

Full details of the PICANet data quality control and assurance processes are provided in the PICANet National Report 2003 - 2004.

The completeness for all data items collected by PICANet are given in Table DQ1, showing a 95.4% completeness level of the data items. Table DQ2 details the completeness of the data by month by year for the last 3 years, while table DQ3 provides a breakdown by individual unit for the combined 3 years. The PICANet dataset contains 4.2% of exception values (i.e. data collected as 'not recorded' or 'not known') and with 0.5% left blank. Figure DQ1 highlights thirteen data items found to have the largest number of invalid, exception or blank values. Completeness, overall, has increased slightly from last year.

Table DQ1 Data completeness

FIELD	Eligible	Complete						Incomplete
		Valid		Exceptions		Total		Invalid		Blank		Total
		n	%	n	%	n	%	n	%	n	%	n	%
ADDATE	44836	44836	(100.0)	0	(0.0)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
ADDRESS1	44834	44406	(99.0)	0	(0.0)	44406	(99.0)	0	(0.0)	428	(1.0)	428	(1.0)
ADNO	44836	44836	(100.0)	0	(0.0)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
ADTIME	44836	44835	(100.0)	0	(0.0)	44835	(100.0)	0	(0.0)	1	(0.0)	1	(0.0)
ADTYPE	44836	44805	(99.9)	29	(0.1)	44834	(100.0)	0	(0.0)	2	(0.0)	2	(0.0)
APDIAG	44836	44836	(100.0)	0	(0.0)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
BASEEXCESS	30846	28281	(91.7)	2556	(8.3)	30837	(100.0)	0	(0.0)	9	(0.0)	9	(0.0)
BGFIRSTHR	40020	38654	(96.6)	1348	(3.4)	40002	(100.0)	0	(0.0)	18	(0.0)	18	(0.0)
BPSYS	44836	38577	(86.0)	6127	(13.7)	44704	(99.7)	0	(0.0)	132	(0.3)	132	(0.3)
CAREAREAAD	44295	43053	(97.2)	1239	(2.8)	44292	(100.0)	0	(0.0)	3	(0.0)	3	(0.0)
CASENO	44836	44836	(100.0)	0	(0.0)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
DELORDER	1427	1205	(84.4)	221	(15.5)	1426	(99.9)	0	(0.0)	1	(0.1)	1	(0.1)
DISPALCARE	42578	42049	(98.8)	526	(1.2)	42575	(100.0)	0	(0.0)	3	(0.0)	3	(0.0)
DOB	44424	44424	(100.0)	0	(0.0)	44424	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
DOBEST	44836	44425	(99.1)	0	(0.0)	44425	(99.1)	1	(0.0)	410	(0.9)	411	(0.9)
DOD	2739	2690	(98.2)	0	(0.0)	2690	(98.2)	0	(0.0)	49	(1.8)	49	(1.8)
ECMO	44836	43735	(97.5)	1081	(2.4)	44816	(100.0)	0	(0.0)	20	(0.0)	20	(0.0)
ETHNIC	44836	44420	(99.1)	0	(0.0)	44420	(99.1)	0	(0.0)	416	(0.9)	416	(0.9)
FAMILYNAME	44836	44425	(99.1)	0	(0.0)	44425	(99.1)	0	(0.0)	411	(0.9)	411	(0.9)
FIO2	30643	25601	(83.5)	4976	(16.2)	30577	(99.8)	0	(0.0)	66	(0.2)	66	(0.2)
FIRSTNAME	44836	44424	(99.1)	0	(0.0)	44424	(99.1)	0	(0.0)	412	(0.9)	412	(0.9)
FU30DISSTATUS	41237	21076	(51.1)	20113	(48.8)	41189	(99.9)	0	(0.0)	48	(0.1)	48	(0.1)
FU30LOCATION	20704	17948	(86.7)	2752	(13.3)	20700	(100.0)	0	(0.0)	4	(0.0)	4	(0.0)
FU30LOCHOSP	3536	3429	(97.0)	102	(2.9)	3531	(99.9)	0	(0.0)	5	(0.1)	5	(0.1)
GEST	25794	16834	(65.3)	8955	(34.7)	25789	(100.0)	0	(0.0)	5	(0.0)	5	(0.0)
HEADBOX	30643	29362	(95.8)	1251	(4.1)	30613	(99.9)	0	(0.0)	30	(0.1)	30	(0.1)
ICPDEVICE	40020	39196	(97.9)	804	(2.0)	40000	(100.0)	0	(0.0)	20	(0.0)	20	(0.0)
INTTRACHEOSTOMY	44836	43587	(97.2)	1229	(2.7)	44816	(100.0)	0	(0.0)	20	(0.0)	20	(0.0)
INTUBATION	30643	30095	(98.2)	520	(1.7)	30615	(99.9)	0	(0.0)	28	(0.1)	28	(0.1)
INTUBEVER	44836	44836	(100.0)	0	(0.0)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
INVVENT	44826	43639	(97.4)	1183	(2.6)	44822	(100.0)	0	(0.0)	4	(0.0)	4	(0.0)
INVVENTDAY	29586	29506	(99.7)	59	(0.2)	29565	(99.9)	0	(0.0)	21	(0.1)	21	(0.1)
LVAD	44836	43724	(97.5)	1092	(2.4)	44816	(100.0)	0	(0.0)	20	(0.0)	20	(0.0)
MECHVENT	44836	44387	(99.0)	432	(1.0)	44819	(100.0)	0	(0.0)	17	(0.0)	17	(0.0)
MEDHISTEVID	44836	44430	(99.1)	391	(0.9)	44821	(100.0)	0	(0.0)	15	(0.0)	15	(0.0)
MULT	44836	33997	(75.8)	10832	(24.2)	44829	(100.0)	0	(0.0)	7	(0.0)	7	(0.0)
NHSNO	44836	35738	(79.7)	1913	(4.3)	37651	(84.0)	0	(0.0)	7185	(16.0)	7185	(16.0)
NONINVVENT	44836	43508	(97.0)	1310	(2.9)	44818	(100.0)	0	(0.0)	18	(0.0)	18	(0.0)
NONINVVENTDAY	5705	5660	(99.2)	43	(0.8)	5703	(100.0)	0	(0.0)	2	(0.0)	2	(0.0)
PAO2	30846	23833	(77.3)	6983	(22.6)	30816	(99.9)	0	(0.0)	30	(0.1)	30	(0.1)
POSTCODE	44836	44805	(99.9)	0	(0.0)	44805	(99.9)	0	(0.0)	31	(0.1)	31	(0.1)
PREVICUAD	44836	44150	(98.5)	676	(1.5)	44826	(100.0)	0	(0.0)	10	(0.0)	10	(0.0)
PRIMDIAG	44836	44626	(99.5)	0	(0.0)	44626	(99.5)	69	(0.2)	141	(0.3)	210	(0.5)
PRIMREASON	40020	39379	(98.4)	624	(1.6)	40003	(100.0)	0	(0.0)	17	(0.0)	17	(0.0)
PUPREACT	44836	41029	(91.5)	3790	(8.5)	44819	(100.0)	0	(0.0)	17	(0.0)	17	(0.0)
RENALSUPPORT	40020	39235	(98.0)	762	(1.9)	39997	(99.9)	0	(0.0)	23	(0.1)	23	(0.1)
RETRIEVAL	44836	44700	(99.7)	126	(0.3)	44826	(100.0)	0	(0.0)	10	(0.0)	10	(0.0)
RETRIEVALBY	15270	14971	(98.0)	299	(2.0)	15270	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
SEX	44836	44813	(99.9)	23	(0.1)	44836	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
SOURCEAD	44836	44735	(99.8)	100	(0.2)	44835	(100.0)	0	(0.0)	1	(0.0)	1	(0.0)
TIMEDTH	2246	2244	(99.9)	0	(0.0)	2244	(99.9)	0	(0.0)	2	(0.1)	2	(0.1)
UNITDISDATE	44826	44825	(100.0)	0	(0.0)	44825	(100.0)	0	(0.0)	1	(0.0)	1	(0.0)
UNITDISDEST	42578	42358	(99.5)	218	(0.5)	42576	(100.0)	0	(0.0)	2	(0.0)	2	(0.0)
UNITDISDESTHOSP	41330	37644	(91.1)	3686	(8.9)	41330	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
UNITDISSTATUS	44836	44824	(100.0)	2	(0.0)	44826	(100.0)	0	(0.0)	10	(0.0)	10	(0.0)
UNITDISTIME	44826	44816	(100.0)	0	(0.0)	44816	(100.0)	0	(0.0)	10	(0.0)	10	(0.0)
VASOACTIVE	44836	43569	(97.2)	1245	(2.8)	44814	(100.0)	0	(0.0)	22	(0.0)	22	(0.0)
Total	2156706	2056861	(95.4)	89618	(4.2)	2146479	(99.5)	70	(0.0)	10157	(0.5)	10227	(0.5)

Figure DQ1 Percentage of exception or blank values in the PICANet dataset

Note: Full description of variables are provided in the PICANet Data Definitions Manual

Some of these data items are used in the calculation of the Paediatric Index of Mortality (PIM) 2. PICANet is investigating the impact of missing data on this risk adjustment index. Thirty-day follow-up status is a standard patient care outcome measure used across the NHS. Within PICANet, 30 day follow-up data is 99.9% complete, however 49% of this data is recorded as 'not known'. The distribution of 30 day follow-up data collection across PICANet units is detailed in figure DQ2.

Figure DQ2 Data completeness for 30-day follow-up information

The NHS Number is a unique patient identifier that provides a common link between patient records across the NHS. The number can be used by Trust Patient Administration Systems/Patient Information Systems to easily and reliably link to the PICANet database.

Table DQ2 Data completeness by year (all variables)

Year	Month	Eligible	Completion
			Complete						Incomplete
			Valid		Exceptions		Total		Invalid		Blank		Total
			n	%	n	%	n	%	n	%	n	%	n	%

2005	1	55688	51951	(93.3)	3331	(6.0)	55282	(99.3)	1	(0.0)	405	(0.7)	406	(0.7)
	2	52516	49005	(93.3)	3183	(6.1)	52188	(99.4)	1	(0.0)	327	(0.6)	328	(0.6)
	3	56359	52455	(93.1)	3593	(6.4)	56048	(99.4)	0	(0.0)	311	(0.6)	311	(0.6)
	4	52028	48649	(93.5)	3076	(5.9)	51725	(99.4)	0	(0.0)	303	(0.6)	303	(0.6)
	5	55160	51780	(93.9)	3109	(5.6)	54889	(99.5)	4	(0.0)	267	(0.5)	271	(0.5)
	6	58107	54654	(94.1)	3157	(5.4)	57811	(99.5)	2	(0.0)	294	(0.5)	296	(0.5)
	7	57770	54303	(94.0)	3133	(5.4)	57436	(99.4)	4	(0.0)	330	(0.6)	334	(0.6)
	8	53870	50709	(94.1)	2872	(5.3)	53581	(99.5)	0	(0.0)	289	(0.5)	289	(0.5)
	9	56578	53163	(94.0)	3115	(5.5)	56278	(99.5)	6	(0.0)	294	(0.5)	300	(0.5)
	10	55289	52097	(94.2)	2882	(5.2)	54979	(99.4)	1	(0.0)	309	(0.6)	310	(0.6)
	11	61521	57947	(94.2)	3283	(5.3)	61230	(99.5)	2	(0.0)	289	(0.5)	291	(0.5)
	12	61092	57301	(93.8)	3500	(5.7)	60801	(99.5)	5	(0.0)	286	(0.5)	291	(0.5)
2005 Total		675978	634014	(93.8)	38234	(5.7)	672248	(99.4)	26	(0.0)	3704	(0.5)	3730	(0.6)

2006	1	65204	62498	(95.8)	2540	(3.9)	65038	(99.7)	0	(0.0)	166	(0.3)	166	(0.3)
	2	59205	56756	(95.9)	2301	(3.9)	59057	(99.8)	1	(0.0)	147	(0.2)	148	(0.2)
	3	63177	60680	(96.0)	2341	(3.7)	63021	(99.8)	4	(0.0)	152	(0.2)	156	(0.2)
	4	57528	55103	(95.8)	2260	(3.9)	57363	(99.7)	2	(0.0)	163	(0.3)	165	(0.3)
	5	60250	57920	(96.1)	2167	(3.6)	60087	(99.7)	2	(0.0)	161	(0.3)	163	(0.3)
	6	57991	55790	(96.2)	2025	(3.5)	57815	(99.7)	0	(0.0)	176	(0.3)	176	(0.3)
	7	56686	54565	(96.3)	1951	(3.4)	56516	(99.7)	0	(0.0)	170	(0.3)	170	(0.3)
	8	56141	53956	(96.1)	2019	(3.6)	55975	(99.7)	0	(0.0)	166	(0.3)	166	(0.3)
	9	55050	52959	(96.2)	1923	(3.5)	54882	(99.7)	2	(0.0)	166	(0.3)	168	(0.3)
	10	59883	57629	(96.2)	2092	(3.5)	59721	(99.7)	3	(0.0)	159	(0.3)	162	(0.3)
	11	61984	59714	(96.3)	2068	(3.3)	61782	(99.7)	0	(0.0)	202	(0.3)	202	(0.3)
	12	61041	58616	(96.0)	2201	(3.6)	60817	(99.6)	0	(0.0)	224	(0.4)	224	(0.4)
2006 Total		714140	686186	(96.1)	25888	(3.6)	712074	(99.7)	14	(0.0)	2052	(0.3)	2066	(0.3)

2007	1	63313	61021	(96.4)	2083	(3.3)	63104	(99.7)	0	(0.0)	209	(0.3)	209	(0.3)
	2	57293	55144	(96.2)	1942	(3.4)	57086	(99.6)	3	(0.0)	204	(0.4)	207	(0.4)
	3	62474	60167	(96.3)	2145	(3.4)	62312	(99.7)	3	(0.0)	159	(0.3)	162	(0.3)
	4	62296	59643	(95.7)	2099	(3.4)	61742	(99.1)	3	(0.0)	551	(0.9)	554	(0.9)
	5	67149	64340	(95.8)	2238	(3.3)	66578	(99.1)	5	(0.0)	566	(0.8)	571	(0.9)
	6	63070	60425	(95.8)	2094	(3.3)	62519	(99.1)	1	(0.0)	550	(0.9)	551	(0.9)
	7	65044	62442	(96.0)	2085	(3.2)	64527	(99.2)	2	(0.0)	515	(0.8)	517	(0.8)
	8	61158	58595	(95.8)	1968	(3.2)	60563	(99.0)	4	(0.0)	591	(1.0)	595	(1.0)
	9	58763	56375	(95.9)	2021	(3.4)	58396	(99.4)	1	(0.0)	366	(0.6)	367	(0.6)
	10	68784	66426	(96.6)	2154	(3.1)	68580	(99.7)	1	(0.0)	203	(0.3)	204	(0.3)
	11	71046	68473	(96.4)	2331	(3.3)	70804	(99.7)	5	(0.0)	237	(0.3)	242	(0.3)
	12	66198	63610	(96.1)	2336	(3.5)	65946	(99.6)	2	(0.0)	250	(0.4)	252	(0.4)
2007 Total		766588	736661	(96.1)	25496	(3.3)	762157	(99.4)	30	(0.0)	4401	(0.6)	4431	(0.6)

Total		2156706	2056861	(95.4)	89618	(4.2)	2146479	(99.5)	70	(0.0)	10157	(0.5)	10227	(0.5)

The distribution of NHS number recording in PICANet units is detailed in table DQ4 and in figure DQ3 below. 20% (improving from 23% to 12% over the three years) of patients in this report do not have NHS numbers, an improvement upon the figure of 25% in the last report.

Table DQ3 Data completeness by PICU

PICU	Eligible	Complete						Incomplete
		Valid		Exceptions		Total		Invalid		Blank		Total
		n	%	n	%	n	%	n	%	n	%	n	%
A	64777	59049	(91.2)	5590	(8.6)	64639	(99.8)	0	(0.0)	138	(0.2)	138	(0.2)
B	29357	28307	(96.4)	980	(3.3)	29287	(99.8)	0	(0.0)	70	(0.2)	70	(0.2)
C	44843	44111	(98.4)	727	(1.6)	44838	(100.0)	0	(0.0)	5	(0.0)	5	(0.0)
D	89211	87601	(98.2)	1576	(1.8)	89177	(100.0)	0	(0.0)	34	(0.0)	34	(0.0)
E	221383	213428	(96.4)	7008	(3.2)	220436	(99.6)	0	(0.0)	947	(0.4)	947	(0.4)
F	165671	156901	(94.7)	7660	(4.6)	164561	(99.3)	45	(0.0)	1065	(0.6)	1110	(0.7)
G	6373	6263	(98.3)	107	(1.7)	6370	(100.0)	0	(0.0)	3	(0.0)	3	(0.0)
H	46287	42719	(92.3)	2971	(6.4)	45690	(98.7)	0	(0.0)	597	(1.3)	597	(1.3)
I	131447	129863	(98.8)	1513	(1.2)	131376	(99.9)	0	(0.0)	71	(0.1)	71	(0.1)
J	13709	12929	(94.3)	589	(4.3)	13518	(98.6)	0	(0.0)	191	(1.4)	191	(1.4)
K1	43432	41925	(96.5)	1380	(3.2)	43305	(99.7)	0	(0.0)	127	(0.3)	127	(0.3)
K2	50672	48930	(96.6)	1734	(3.4)	50664	(100.0)	0	(0.0)	8	(0.0)	8	(0.0)
K3	40755	39183	(96.1)	1571	(3.9)	40754	(100.0)	0	(0.0)	1	(0.0)	1	(0.0)
L	46433	45600	(98.2)	766	(1.6)	46366	(99.9)	0	(0.0)	67	(0.1)	67	(0.1)
M	55133	53412	(96.9)	1554	(2.8)	54966	(99.7)	0	(0.0)	167	(0.3)	167	(0.3)
N	43740	42348	(96.8)	1203	(2.8)	43551	(99.6)	0	(0.0)	189	(0.4)	189	(0.4)
O	94258	88901	(94.3)	4757	(5.0)	93658	(99.4)	0	(0.0)	600	(0.6)	600	(0.6)
P	158973	155009	(97.5)	3889	(2.4)	158898	(100.0)	0	(0.0)	75	(0.0)	75	(0.0)
Q1	12270	11884	(96.9)	386	(3.1)	12270	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
Q2	69423	67255	(96.9)	2133	(3.1)	69388	(99.9)	0	(0.0)	35	(0.1)	35	(0.1)
R	105426	104194	(98.8)	1013	(1.0)	105207	(99.8)	0	(0.0)	219	(0.2)	219	(0.2)
S	26835	25890	(96.5)	913	(3.4)	26803	(99.9)	0	(0.0)	32	(0.1)	32	(0.1)
T	59201	55461	(93.7)	3362	(5.7)	58823	(99.4)	0	(0.0)	378	(0.6)	378	(0.6)
U	56698	53505	(94.4)	2776	(4.9)	56281	(99.3)	0	(0.0)	417	(0.7)	417	(0.7)
V	150731	133205	(88.4)	15970	(10.6)	149175	(99.0)	24	(0.0)	1532	(1.0)	1556	(1.0)
W	100622	93200	(92.6)	6609	(6.6)	99809	(99.2)	0	(0.0)	813	(0.8)	813	(0.8)
X1	66740	63603	(95.3)	2949	(4.4)	66552	(99.7)	1	(0.0)	187	(0.3)	188	(0.3)
X2	52228	47901	(91.7)	4257	(8.2)	52158	(99.9)	0	(0.0)	70	(0.1)	70	(0.1)
Y	63547	61573	(96.9)	1974	(3.1)	63547	(100.0)	0	(0.0)	0	(0.0)	0	(0.0)
Z	16984	16057	(94.5)	904	(5.3)	16961	(99.9)	0	(0.0)	23	(0.1)	23	(0.1)
ZA	29547	26654	(90.2)	797	(2.7)	27451	(92.9)	0	(0.0)	2096	(7.1)	2096	(7.1)
Grand Total	2156706	2056861	(95.4)	89618	(4.2)	2146479	(99.5)	70	(0.0)	10157	(0.5)	10227	(0.5)

Table DQ4 Data completeness for NHS number by NHS trust

NHS trust	Eligible	Valid		Exceptions		Invalid		Blank
NHS trust	Eligible	n	%	n	%	n	%	n	%
A	1403	513	(36.6)	753	(53.7)	0	(0.0)	137	(9.8)
B	644	576	(89.4)	0	(0.0)	0	(0.0)	68	(10.6)
C	908	903	(99.4)	0	(0.0)	0	(0.0)	5	(0.6)
D	1834	1696	(92.5)	138	(7.5)	0	(0.0)	0	(0.0)
E	4580	3633	(79.3)	0	(0.0)	0	(0.0)	947	(20.7)
F	3440	2418	(70.3)	0	(0.0)	0	(0.0)	1022	(29.7)
G	131	130	(99.2)	0	(0.0)	0	(0.0)	1	(0.8)
H	966	373	(38.6)	0	(0.0)	0	(0.0)	593	(61.4)
I	2718	2648	(97.4)	0	(0.0)	0	(0.0)	70	(2.6)
J	291	119	(40.9)	2	(0.7)	0	(0.0)	170	(58.4)
K	2806	2611	(93.1)	61	(2.2)	0	(0.0)	134	(4.8)
L	986	919	(93.2)	0	(0.0)	0	(0.0)	67	(6.8)
M	1138	1121	(98.5)	0	(0.0)	0	(0.0)	17	(1.5)
N	887	713	(80.4)	0	(0.0)	0	(0.0)	174	(19.6)
O	1914	1318	(68.9)	0	(0.0)	0	(0.0)	596	(31.1)
P	3239	3165	(97.7)	0	(0.0)	0	(0.0)	74	(2.3)
Q	1754	1743	(99.4)	5	(0.3)	0	(0.0)	6	(0.3)
R	2137	1944	(91.0)	0	(0.0)	0	(0.0)	193	(9.0)
S	569	561	(98.6)	0	(0.0)	0	(0.0)	8	(1.4)
T	1270	892	(70.2)	0	(0.0)	0	(0.0)	378	(29.8)
U	1149	768	(66.8)	0	(0.0)	0	(0.0)	381	(33.2)
V	3143	1869	(59.5)	0	(0.0)	0	(0.0)	1274	(40.5)
W	2072	1259	(60.8)	0	(0.0)	0	(0.0)	813	(39.2)
X	2544	2506	(98.5)	0	(0.0)	0	(0.0)	38	(1.5)
Y	1319	365	(27.7)	954	(72.3)	0	(0.0)	0	(0.0)
Z	364	358	(98.4)	0	(0.0)	0	(0.0)	6	(1.6)
ZA	630	617	(97.9)	0	(0.0)	0	(0.0)	13	(2.1)
Total	44836	35738	(79.7)	1913	(4.3)	0	(0.0)	7185	(16.0)

In the absence of the NHS Number it is difficult to definitively link patients with additional data repositories. PICANet is establishing a linked data set with Hospital Episode Statistics data. The NHS number is a crucial item of data which will enable long term follow-up and outcomes study of PICU patients, as well as effective aggregation of the PCCMDS in Trusts.

Figure DQ3 Data completeness for NHS number

With the implementation of the Paediatric Critical Care Minimum Data set, greater demands willbe placed on the data collection and quality assurance processes within units. A collaborativeapproach to data quality control and assurance, with regular and timely feedback to units, will ensure that the PICANet dataset remains of the highest standard. We hope that all units willwork with PICANet to meet this goal.

22 USES AND DISSEMINATION OF PICANet DATA

PICANet was established in collaboration with clinical colleagues from all participating NHS trusts, with a view to providing timely and accurate national and local information on PICU activity for those who deliver the service and those who plan the delivery of care. In common with all datasets the use of the data inevitably improves its quality. No data are ever provided or presented which allows an individual to be identified. In this, we act in accordance with the guidelines provided by ONS.

Information on PICANet is available to clinical care teams and parents through posters that are displayed in units and leaflets that are produced in 'parent packs'. The PICANet website address is given in this material and provides a further source of general information and copies of the national reports. Newsletters on progress are distributed regularly to lead nurses and consultants in each unit.

PICANet is pleased to report an increasing number of requests for data and information (Appendix D). Some requests have only requested aggregated, anonymised data from the entire dataset. For other requests, for example those that identify individual PICUs, PICANet always ensures that lead clinicians are informed and seeks permission for their data to be used.

Requests have been received from individual clinicians, groups of researchers and NHS commissioners. Some of the reports produced have required complex data processing and analyses and this has incurred additional costs which have been charged accordingly.

Dissemination of information from PICANet has been of prime importance to the team and Appendix K details specific talks given at various venues, a list of abstracts that have been presented at conferences and papers published by members of the PICANet team on PICANet and related topics. We welcome the opportunity to present data in these forums: please contact one of the team if you would like us to speak at local or national meetings.

23 TABLES AND FIGURES

Tables and Figures

APPENDIX A PARTICIPATING NHS TRUSTS AND HOSPITAL CHARACTERISTICS

NHS Trust	Participating Hospital	Unit / Ward	Number of ITU beds	Number of HDU beds	Type of unit
Barts and the London NHS Trust	Barts and the London Children's Hospital	PCCU	2 ventilated beds	4	General
Birmingham Children's Hospital NHS Trust	Birmingham Children's Hospital	PICU	19	0	General & Cardiac
Brighton & Sussex University Hospitals NHS Trust	The Royal Alexandra Children's Hospital	L8 PICU	1.5¹	8²	General
Cambridge University Hospitals NHS Foundation Trust	Addenbrooke's Hospital	PICU	6	2	General
Cardiff & Vale NHS Trust	University Hospital of Wales	PICU	7	0	General
Central Manchester & Manchester Children's University Hospitals NHS Trust	Royal Manchester Children's Hospital	PICU	15	0	General
Great Ormond Street Hospital for Children NHS Trust	Great Ormond Street Hospital for Children	CCCU	14-16³	0	Cardiac
Great Ormond Street Hospital for Children NHS Trust	Great Ormond Street Hospital for Children	PICU & NICU	21	0	General & Neonatal Unit
Guy's & St. Thomas' NHS Foundation Trust	Evelina Children's Hospital	PICU	15	0	General & Cardiac
Hull & East Yorkshire Hospitals NHS Trust	Hull Royal Infirmary	PICU beds on AITU	0	4⁴	Adult ICU providing General PICU
King's College Hospital NHS Trust	King's College Hospital	PICU	6⁵	0	General & Hepatic & Neurosurgical
Leeds Teaching Hospitals NHS Trust	Leeds General Infirmary	Wards 2 & 4	17⁶	0	General & Cardiac
Leeds Teaching Hospitals NHS Trust	St. James's University Hospital	PICU	17⁶	0	General
Newcastle Upon Tyne Hospitals NHS Foundation Trust	Newcastle General Hospital	PICU	10⁷	6⁷	General
	Royal Victoria Infirmary	Ward 3	10⁷	6⁷	Surgical ICU
	Freeman Hospital	PICU Freeman	7⁸	0	Cardiothoracic surgery & ECMO
NHS Lothian - University Hospitals Division	Royal Hospital for Sick Children, Edinburgh	PICU	7⁹	6⁹	General
NHS Greater Glasgow and Clyde - Women and Children's Division	Royal Hospital for Sick Children, Yorkhill	PICU	16¹⁰	6¹⁰	General
Oxford Radcliffe Hospitals NHS Trust	The John Radcliffe Hospital	PICU	7	2¹¹	General & Cardiac
Nottingham University Hospitals NHS Trust	Queen's Medical Centre	PICU	6	4	General (plus regional neurosurgical, spinal and cleft lip & palate services)
Royal Brompton & Harefield NHS Trust	Royal Brompton Hospital	PICU	10	4	Cardiac & Respiratory
Royal Liverpool Children's NHS Trust	Royal Liverpool Children's Hospital	PICU	21	0	General & Cardiac
Sheffield Children's NHS Foundation Trust	Sheffield Children's Hospital	PICU	9	2	General
Sheffield Children's NHS Foundation Trust	Sheffield Children's Hospital	Neonatal Surgical Unit	2	0	Neonatal Surgical Unit
Southampton University Hospitals NHS Trust	Southampton General Hospital	PICU	10¹²	0	General & Cardiac
South Tees Hospitals NHS Trust	James Cook University Hospital	PICU	4	0	General
St. George's Healthcare NHS Trust	St. George's Hospital	PICU	5	0	General & Neurosurgical
St. Mary's NHS Trust	St. Mary's Hospital	PICU	8	2	General
The Lewisham Hospital NHS Trust	University Hospital, Lewisham	PICU	1	2¹³	General & Surgery
The Royal Group of Hospitals and Dental Hospital HSS Trust	Royal Belfast Hospital for Sick Children	PICU	7¹⁴	0	General
United Bristol Healthcare NHS Trust	Bristol Royal Hospital for Children	PICU	14¹⁵	0	General & Cardiac
University Hospitals of Leicester NHS Trust	Leicester Royal Infirmary	CICU	6	2	General
University Hospitals of Leicester NHS Trust	Glenfield Hospital	PICU	5	0	Cardiac, General & ECMO
University Hospital of North Staffordshire NHS Trust	University Hospital of North Staffordshire	PICU	6	1	General

1	Capacity for 3 beds, 1.5 funded.
2	HDU is a separate unit.
3	The actual figure depends on the number of ECMO patients and HDU patients.
4	With capacity to ventilate two patients on the Adult ICU.
5	A new unit opens in April 2008, which will ultimately house 8 PICU beds and 8 HDU beds
5	(the latter to be phased in gradually).
6	Nurses / beds used flexibly across the sites.
7	Total bed numbers split between two hospital sites.
8	From March 2008, this unit will have 8 PICU beds (possibly increasing to 10 by the end of the year).

9	This change in bed complement (ie from 6 to 7 beds) was effective as of 1st November 2007.
9	ITU/HDU beds used flexibly (e.g. ITU 7 + HDU 6; 9 ITU + 4 HDU; 11 ITU + 2 HDU).
10	Beds used flexibly between areas.
10	HDU will increase to 4 beds in February 2008 in a separate unit.
12	The 10th bed opened Jan/Feb 2008 and a further bed will be opening in summer 2008.
13	Flexed by a further 2 beds to support winter pressures.
14	The unit is anaesthetist-led and only admits patients under 15 years.
14	A total of 7 ITU/HDU beds are used flexibly depending on demand.
15	This change in bed complement (ie from 13 to 14 beds) was effective as of 1st April 2007.

APPENDIX B CLINICAL ADVISORY GROUP MEMBERSHIP

Name	Position	NHS Trust / Hospital	Period served
Dr Paul Baines	Consultant in Paediatric Intensive Care	Royal Liverpool Children's NHS Trust	2002 - present
Dr Paul Baines	Consultant in Paediatric Intensive Care	Alder Hey Hospital	2002 - present
Ms Corenna Bowers	Sister	Cardiff & Vale NHS Trust	2002 - 2004
Ms Corenna Bowers	Sister	University Hospital of Wales	2002 - 2004
Dr Anthony Chisakuta	Lead Clinician	The Royal Group of Hospitals and Dental Hospital HSS Trust	2008 - present
Dr Anthony Chisakuta	Lead Clinician	Royal Belfast Hospital for Sick Children	2008 - present
Dr Peter Davis	Consultant in Paediatric Intensive Care	United Bristol Healthcare NHS Trust	2006 - present
Dr Peter Davis	Consultant in Paediatric Intensive Care	Bristol Royal Hospital for Children	2006 - present
Dr Andrew Durward	Consultant in Paediatric Intensive Care	Guy's & St Thomas' NHS Foundation Trust	2002 - present
Dr Andrew Durward	Consultant in Paediatric Intensive Care	Evelina Children's Hospital	2002 - present
Ms Georgina Gymer	Research Nurse	Nottingham University Hospitals NHS Trust	2005 - 2006
Ms Georgina Gymer	Research Nurse	Queen's Medical Centre	2005 - 2006
Dr James Fraser	Consultant in Paediatric Intensive Care	United Bristol Healthcare NHS Trust	2002 - 2006
Dr James Fraser	Consultant in Paediatric Intensive Care	Bristol Royal Hospital for Children	2002 - 2006
Dr Hilary Klonin	Consultant in Paediatric Intensive Care	Hull & East Yorkshire Hospitals NHS Trust	2002 - present
Dr Hilary Klonin	Consultant in Paediatric Intensive Care	Hull Royal Infirmary	2002 - present
Ms Christine Mackerness	Sister	Newcastle Upon Tyne Hospitals NHS Foundation Trust	2002 - present
Ms Christine Mackerness	Sister	Newcastle General Hospital	2002 - present
Ms Tina McClelland	Audit Sister	Royal Liverpool Children's NHS Trust	2006 - present
Ms Tina McClelland	Audit Sister	Alder Hey Hospital	2006 - present
Dr Jillian McFadzean	Consultant in Paediatric Intensive Care	NHS Lothian - University Hospitals Division	2005 - present
Dr Jillian McFadzean	Consultant in Paediatric Intensive Care	Edinburgh Royal Hospital for Sick Children	2005 - present
Ms Victoria McLaughlin	Audit Nurse	Central Manchester & Manchester Children's University Hospitals NHS Trust	2002 - 2007
Ms Victoria McLaughlin	Audit Nurse	Royal Manchester Children's Hospital	2002 - 2007
Dr Roddy O'Donnell	Consultant in Paediatric Intensive Care	Cambridge University Hospitals NHS Foundation Trust	2002 - present
Dr Roddy O'Donnell	Consultant in Paediatric Intensive Care	Addenbrooke's Hospital	2002 - present
Ms Geralyn Oldham	Information Support Manager	Great Ormond Street Hospital for Children NHS Trust	2002 - present
Ms Geralyn Oldham	Information Support Manager	Great Ormond Street Hospital for Sick Children	2002 - present
Dr Gale Pearson (Chair)	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital NHS Trust	2002 - present
Dr Gale Pearson (Chair)	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital	2002 - present
Dr Damian Pryor	Consultant in Paediatric Intensive Care	Cardiff & Vale NHS Trust	2002 - 2004
Dr Damian Pryor	Consultant in Paediatric Intensive Care	University Hospital of Wales	2002 - 2004
Ms Chloe Rishton	CHiP Nurse	Central Manchester & Manchester Children's University Hospitals NHS Trust	2008 - present
Ms Chloe Rishton	CHiP Nurse	Royal Manchester Children's Hospital	2008 - present
Dr Allan Wardhaugh	Consultant in Paediatric Intensive Care	Cardiff & Vale NHS Trust	2004 - present
Dr Allan Wardhaugh	Consultant in Paediatric Intensive Care	University Hospital of Wales	2004 - present
Ms Debbie White	Sister	Cambridge University Hospitals NHS Foundation Trust	2002 - present
Ms Debbie White	Sister	Addenbrooke's Hospital	2002 - present

APPENDIX C STEERING GROUP MEMBERSHIP

Name	Position	Organisation	Representation	Period Served
Mrs Pamela Barnes	Chair of Action for Sick Children	Action for Sick Children	Lay Member	2002 - present
Professor Nick Black (Chair)	Head of Health Services Research Unit	London School of Hygiene and Tropical Medicine	Health Services Research / Public Health	2002 - 2007
Mr William Booth	Clinical Nurse Manager	United Bristol Healthcare NHS Trust	Royal College of Nursing	2002 - present
Mr William Booth	Clinical Nurse Manager	Bristol Royal Hospital for Children PICU	Royal College of Nursing	2002 - present
Ms Bev Botting	Child Health and Pregnancy Statistics	Office for National Statistics	Office for National Statistics (data protection)	2002 - 2003
Dr Jean Chapple	Consultant in Perinatal Epidemiology / Public Health	Westminster Primary Care Trust	PICNET founder	2002 - 2006
Dr Bill Chaudhry	Consultant Paediatrician	Newcastle Upon Tyne Hospitals NHS Trust	Clinical IT	2002 - 2003
Dr Bill Chaudhry	Consultant Paediatrician	Newcastle General Hospital PICU	Clinical IT	2002 - 2003
Dr Mark Darowski	Consultant Paediatric Anaesthetist	Leeds Teaching Hospitals NHS Trust	Royal College of Anaesthetists	2002 - present
Dr Mark Darowski	Consultant Paediatric Anaesthetist	Leeds General Infirmary PICU	Royal College of Anaesthetists	2002 - present
Mr Noel Durkin	Department of Health	Child Health Services Directorate	Department of Health	2002 - present
Dr Ian Jenkins	Consultant in Paediatric Intensive Care	United Bristol Healthcare NHS Trust	Paediatric Intensive Care Society	2006 - present
Dr Ian Jenkins	Consultant in Paediatric Intensive Care	Bristol Royal Hospital for Children PICU	Paediatric Intensive Care Society	2006 - present
Dr Steve Kerr	Consultant in Paediatric Intensive Care	Royal Liverpool Children's NHS Trust	Chair of PICS	2003 - present
Dr Steve Kerr	Consultant in Paediatric Intensive Care	Alder Hey Hospital PICU	Chair of PICS	2003 - present
Ms Helen Laing	Clinical Audit	Healthcare Commission	Healthcare Commission	2004 - 2006
Mr Ian Langfield	Audit Co-ordinator	National Assembly of Wales	National Assembly of Wales	2002 - 2003
Dr Michael Marsh	Consultant in Paediatric Intensive Care	Southampton University Hospitals NHS Trust	Royal College of Paediatrics and Child Health	2002 - present
Dr Michael Marsh	Consultant in Paediatric Intensive Care	Southampton General Hospital PICU	Royal College of Paediatrics and Child Health	2002 - present
Dr Jillian McFadzean / Ms Laura Reekie	Consultant in Anaesthesia & Intensive Care / PA	NHS Lothian - University Hospitals Division	Edinburgh Royal Hospital for Sick Children	2005 - present
Dr Jillian McFadzean / Ms Laura Reekie	Consultant in Anaesthesia & Intensive Care / PA	Edinburgh Royal Hospital for Sick Children	Edinburgh Royal Hospital for Sick Children	2005 - present
Dr Roddy McFaul	Medical Advisor	Child Health Services Directorate	Department of Health	2002 - 2003
Dr Kevin Morris	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital NHS Trust	Clinical Lead for the West Midlands Medicines for Children Local Research Network	2006 - present
Dr Kevin Morris	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital PICU		2006 - present
Professor Jon Nicholl	Director of Medical Care Research Unit	School of Health and Related Research	Health Services Research / Statistics	2002 - 2006
Professor Jon Nicholl	Director of Medical Care Research Unit	University of Sheffield	Health Services Research / Statistics	2002 - 2006
Dr Gale Pearson	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital NHS Trust	Chair of PICANet CAG	2002 - present
Dr Gale Pearson	Consultant in Paediatric Intensive Care	Birmingham Children's Hospital PICU	Chair of PICANet CAG	2002 - present
Ms Tanya Ralph	Nursing Research Lead	Sheffield Children's NHS Foundation Trust	PICS	2002 - 2006
Ms Tanya Ralph	Nursing Research Lead	Sheffield Children's Hospital PICU	PICS	2002 - 2006
Dr Kathy Rowan (on sabbatical 2004 - present, represented by Lucy Scott)	Director	ICNARC	Intensive Care National Audit & Research Centre	2002 - present
Mr Stuart Rowe	PCT Commissioner	Commissioning Department	PCT Commissioner (Pan-Thames)	2003 - present
Mr Stuart Rowe	PCT Commissioner	Hammersmith & Fulham PCT	PCT Commissioner (Pan-Thames)	2003 - present
Ms Dominique Sammut	Audit Co-ordinator	Health Commission Wales	Health Commission Wales	2003 - present
Dr Jennifer Smith	Medical Advisor	Office Project Team	Commission for Health Improvement	2002 - 2004
Dr Charles Stack	Consultant in Paediatric Intensive Care	Sheffield Children's NHS Foundation Trust	PICS	2002 - 2006
Dr Charles Stack	Consultant in Paediatric Intensive Care	Sheffield Children's Hospital PICU	PICS	2002 - 2006
Professor Stuart Tanner	Medical Advisor in Paediatrics and Child Health	Child Health Services Directorate	Department of Health	2003 - 2006
Professor Stuart Tanner	Medical Advisor in Paediatrics and Child Health	Department of Health	Department of Health	2003 - 2006
Dr Robert Tasker	Lecturer in Paediatrics	Department of Paediatrics	PICS SG	2004 - present
Dr Robert Tasker	Lecturer in Paediatrics	University of Cambridge Clinical School	PICS SG	2004 - present
Dr Edward Wozniak	Medical Advisor in Paediatrics and Child Health	Child Health Services Directorate	Department of Health	2006 - present
Dr Edward Wozniak	Medical Advisor in Paediatrics and Child Health	Department of Health	Department of Health	2006 - present

APPENDIX D DATA/INFORMATION REQUESTS RECEIVED TO DATE

Data and Information Requests

APPENDIX E DATA COLLECTION FORM

PICANet Data Collection Form

APPENDIX F INFORMATION LEAFLET

PICANet Information Leaflet

APPENDIX G DATA VALIDATION REPORT

APPENDIX H MONTHLY ADMISSIONS REPORT

Monthly Admissions Report

APPENDIX I DATA STATUS REPORT

Data Status Report

APPENDIX J POLICY FOR UNITS FALLING OUTSIDE THE CONTROL LIMITS
PICANet policy on PICUs lying outside the control limits of the mortality ratio funnel plots (PICANet November 2005)

Background - mortality ratios and funnel plots

PICANet is required by the Department of Health to report on the mortality outcomes of all children admitted for paediatric intensive care. The PICANet Clinical Advisory Group and Steering Group recommended that the mortality outcomes from each PICU be adjusted for the illness severity of the child at admission using the Paediatric Index of Mortality (PIM).¹ PICANet reports the unadjusted mortality outcome from all PICUs and a mortality ratio based on the ratio of observed mortality in each PICU to the expected mortality calculated using PIM. From 2005, revised coefficients for PIM have been used derived from the recently completed United Kingdom Paediatric Intensive Care Outcome Study.² PIM2³ has been used for risk-adjustment in this report for 2006 only and will be used in future reports as the data become available.

Earlier work published by members of PICANet team⁴ has highlighted the problems of attempting to rank PICUs on their annual mortality, whether unadjusted or adjusted. PICANet, however, has also recognised the need to identify units which appear to have outcomes very different to other units. Consequently, PICANet has published a funnel plot of the observed to expected mortality ratio of individual PICUs. The funnel plots are constructed in such a way that there is an approximately 5% chance of a PICU falling outside the control limits, if the distribution of the mortality ratios is random.

The mortality ratio is calculated for each PICU by dividing the expected number of deaths calculated using the published PIM algorithm by the observed number of deaths for each PICU. The mortality ratio is then plotted on the y-axis against the number of admissions to the PICU on the x-axis. In order to satisfy the condition that if the overall distribution of the mortality ratios is random there exists an approximately 5% chance of a PICU falling outside the control limits, then the upper and lower control limits constructed at an individual PICU level must represent not 95% confidence intervals, but 99.9% confidence intervals around a mortality ratio of 1 by number of admissions.⁵ This is analogous to increasing the confidence interval (or significance level) when correcting for multiple comparisons in data containing numerous groups.

Data outliers

A PICU whose mortality ratio lies outside of these control limits will be identified as having returned data that is markedly different to the other PICUs.
It is important to note that a PICU lying outside the control limits is not sufficient evidence to suggest a PICU has either markedly higher or markedly lower mortality than the other PICUs, it merely indicates that the data they have returned is different to that of other PICUs.
For those PICUs that do lie outside the control limits, the principals of clinical governance should apply:
- PICANet will raise the issue with the lead clinician of the PICU and the Trust Chief Executive
- PICANet will work with the PICU and the Trust, following the plan below until the issue is resolved.

In these circumstances, PICANet will:

Review the data to investigate whether there are data driven reasons for a PICU lying outside of the control limits (it is known that risk-adjustment tools can be unreliable when a PICU has a particularly high proportion of patients at either end of the bounds of the tool.)
Review the data quality of the PICU. The quality of the data is the PICUs' responsibility. PICANet will provide feedback from PICU visits and central validation procedures. PICUs will be expected to check the quality of individual data items.
Plot the data quality indicators over time to identify whether the anomaly can be traced to a certain data collection period.
Plot the mortality ratio over time to identify whether the anomaly can be traced to a certain data collection period.
Plot the observed mortality over time to identify whether the anomaly can be traced to a certain data collection period.
Plot the expected mortality over time to identify whether the anomaly can be traced to a certain data collection period.
Investigate the primary reason for admission to the PICU. If the PICU has a markedly high proportion of some primary reason of admission to the PICU compared with other PICUs this may suggest further refinements to the risk-adjustment method are required.
Produce a brief summary report of the above to be forwarded to the lead clinician and Chief Executive at the PICU concerned, together with an invitation to meet in person to review the data with the PICANet team.

Where reference is made to the Chief Executive, it is accepted that they may be represented by their clinical governance lead.

NOTE: Excess mortality in particular sub-groups of patients or associated with other aspects of service provision may be identified using different statistical methods. The process outlined above will be implemented wherever anomalous results/outliers are identified.

References:

Parry GJ, Gould CR, McCabe CJ, Tarnow-Mordi WO. Annual league tables of hospital mortality in neonatal intensive care: A longitudinal study. BMJ 1998; 316:1931-1935.
Brady AR, Harrison D, Black S, Jones S, Rowan K, Pearson G, Ratcliffe J, Parry GJ, on behalf of the UK PICOS Study Group. Assessment and Optimization of Mortality Prediction Tools for Admissions to Pediatric Intensive Care in the United Kingdom. Pediatrics 2006; 117: 733-742.
Shann F, Slater A, Pearson G. PIM 2: a revised version of the Paediatric Index of mortality. Intensive Care Med 2003; 29:278-285
Shann F, Pearson G, Slater A, Wilkinson K, Paediatric index of mortality (PIM): a mortality prediction model for children in intensive care. Intensive Care Med 1997; 23:201-207
Spiegelhalter D. Funnel plots for institutional comparison. Qual. Saf. Health Care, Dec 2002; 11: 390- 391.

APPENDIX K PUBLICATIONS/PRESENTATIONS

K.1 Presentations

Meeting/Conference	Venue	Date	Presentation Title	PICANet Team Attendees
Presentation to Glasgow PICU team	Glasgow	18/08/2003	PICANet	Sam Jones & Tricia McKinney
NW Paediatric Intensive Care Seminar (North West Specialised Commissioning Group)	Dunkenhalgh Hotel, Clayton-le-Moors, Lancashire	23/06/2004	PICANet: Results of national activity	Sam Jones & Roger Parslow
PICANet AGM	London	24/06/2004	Presentation of National report	PICANet Team
Welsh National Commissioning Advisory Board Meeting	Royal Welsh Showground, Builth Wells	28/07/2004	PICANet: Presentation of National and Welsh report	Liz Draper & Nicky Davey
Strategic Issues in Health Care Management, Sixth International Conference	University of St Andrews	02/09/2004	Collection of personally identifiable information for a national clinical database: how feasible is it to obtain signed consent?	Sam Jones
PICS SG	Cambridge University	09/09/2004	PICANet: How can it be used for research and audit?	Nicky Davey, Sam Jones, Roger Parslow & Krish Thiru
Confidential Enquiry into Maternal and Child Health	London	08/03/2005	National Paediatric Intensive Care Database (PICANet)	Liz Draper
Intensive Care National Audit & Research Centre (ICNARC): Eight Annual Meeting of the Case Mix Programme	Savoy Hotel, London	13/04/2005	Why is it important to include information on paediatric admissions in the new Case Mix Programme Dataset?	Sam Jones
Pan Thames Report Update: Commissioning Consortium	London	06/05/2005	PICANet: Update on Pan Thames data quality for commissioning	Krish Thiru & Sam Jones
Paediatric Intensive Care Study Day	Royal Manchester Children's Hospital	10/05/2005	The epidemiology of critical illness in children	Roger Parslow
Trent PIC commissioners	QMC, Nottingham	12/05/2005	PICANet: Presentation of National report 2003-2004	Liz Draper
Paediatric Intensive Care Trainee Meeting	Royal Liverpool Children's Hospital (Alder Hey)	13/05/2005	Role of PICANet and the relevance of the national audit to the clinical community	Nicky Davey & Sam Jones
PICANet AGM	London	24/05/2005	Presentation of National report	PICANet Team
NORCOM, TRENTCOM & LNR PIC commissioners	Leicester	13/06/2005	PICANet in LNR, Trent & South Yorkshire PCTs	Liz Draper
Health Protection Agency (HPA) annual conference	Warwick	12/09/2005	Mortality, deprivation and ethnicity of critically ill children in England and Wales: preliminary findings from the Paediatric Intensive Care Audit Network (PICANet)	Roger Parslow
Paediatric Critical Care Network Board (East leeds PCT)	Leeds	06/10/2005	PICANet: Presentation of national data and relevance to commissioning	Tricia McKinney
Welsh National Commissioning Advisory Board Meeting	Lamb and Flag Hotel, Llanwenarth, Abergavenny	11/10/2005	PICANet: Presentation of National and Welsh Report	Gareth Parry
PICANet AGM	Perinatal Institute, Birmingham	29/06/2006	Presentation of the National Report	PICANet Team
Pan Thames Commissioners Meeting	London	28/07/2006	Pan Thames PICANet Report 2004-2005	Krish Thiru, Tricia McKinney
Paediatric Intensive Care Society Scientific Meeting	Glasgow	16 & 17/11/06	PICU Health Informatics	K Thiru, P Ramnarayan, S Rowe on behalf of the pan Thames Health Informatics Group
University of Leicester	Department of Health Sciences. University of Leicester	14/03/2007	The UK Paediatric Traumatic Brain Injury Study	Roger Parslow
Paediatric Intensive Care Society Study Group	Cambridge	21 & 22/03/07	PICU Health Informatics: Clinical Information Systems	K Thiru, P Ramnarayan, S Rowe on behalf of the pan Thames Health Informatics Group
Pan Thames Commissioners PbR Roadmap	ASIA House	14/06/2007	PICANet and the PCCMDS	Roger Parslow
Exploiting Existing Data for Health Research	University of St Andrews	19/09/2007	Privacy preserving record linkage	Tom Fleming
PICANet AGM	Leeds University Business School	04/07/2007	Presentation of the National Report	PICANet Team

K.2 Publications

Journal	Title	Authors
Pediatrics (2004) 113 1653-1657	Trends in the incidence of severe retinopathy of prematurity in a geographically defined population over a 10-year period	Hameed B, Shyamanur K, Kotecha S, Manktelow B, Woodruff G, Draper ES & Field D
Archives of Disease in Childhood (2005) 90 380-387	Neuropsychological and educational problems at school age associated with neonatal encephalopathy	Marlow N, Rose AS, Rands CE & Draper ES
Archives of Disease in Childhood (2005) 90 1182-1187	Epidemiology of traumatic brain injury in children receiving intensive care in the UK	Parslow RC, Morris KP, Tasker RC, Forsyth RJ & Hawley C
British Medical Journal (2005) 330 43 (1 January)	Paediatric cardiac surgical mortality after Bristol: details of risk adjustment tools were not given (letter)	Parry GJ, Draper ES & McKinney P
British Medical Journal (2005) 330 877-879 (16 April)	A feasibility study of signed consent for the collection of patient identifiable information for a national paediatric clinical audit database	McKinney PA, Jones S, Parslow R, Davey N, Darowski M, Chaudhry B, Stack C, Parry G, Draper ES for the PICANet Consent Study Group
European Journal of Obstetrics, Gynecology & Reproductive Biology (2005) 118 272-274	Presentation of the European project models of organising access to intensive care for very preterm births in Europe (MOSAIC) using European diversity to explore models for the care of the very preterm babies.	Zeitlin J, Papiernik E, Breart G, Draper E & Kollee L
Prenatal Diagnosis (2005) 25 286-291	Population based study of the outcome following the antenatal diagnosis of cystic hygroma	Howart ES, Draper ES, Budd JLS, Konje J, Kurinczuk JJ & Clarke M
Emergency Medical Journal (2006) 23 519-522	Emergency access to neurosurgery in the United Kingdom	Tasker RC, Morris KP, Forsyth RJ, Hawley CA, Parslow RC, on behalf of the UK Paediatric Brain Injury Study
Intensive Care Medicine (2006) 32(9) 1458	Organ donation in paediatric traumatic brain injury	Morris KP, Tasker RC, Parslow RC, Forsyth RJ, Hawley CA
Intensive Care Medicine (2006) 32(10) 1606-1612	Monitoring and management of intracranial pressure complicating severe traumatic brain injury in children	Morris KP, Forsyth RJ, Parslow RC, Tasker RC, Hawley CA on behalf of the UK Paediatric Traumatic Brain Injury Study Group and the Paediatric Intensive Care Society Study Group
Lancet (2006) 367 1080-85	Outcome after neonatal continuous negative-pressure ventilation: follow-up assessment	Telford K, Waters L, Vyas H, Manktelow BN, Draper ES, Marlow N
Pediatrics (2006) 117 733-742	Assessment and optimisation of mortality prediction tools for admissions to paediatric intensive care in the United Kingdom	Brady AR, Harrison D, Black S, Jones S, Rowan K, Pearson G, Ratcliffe J, Parry GJ; UK PICOS Study Group
Archives of Disease in Childhood. Fetal and Neonatal Edition (2007) 92 356-360	Mortality patterns of very preterm babies: a comparative analysis of two European regions in France and England	Draper ES, Zeitlin J, Field DJ, Manktelow BN, Truffert P
Paediatric Intensive Care Medicine, (2008) 9(1) 8-14	Prediction of raised intracranial pressure complicating severe traumatic brain injury in children: implications for trial design	Forsyth RJ, Parslow RC, Tasker RC, Hawley CA, Morris KP. On behalf of the UK Paediatric Traumatic Brain Injury Study Group and the Paediatric Intensive Care Society Study Group (PICS SG)
British Medical Journal (2008) 336 7655	Survival of extremely preterm babies in a geographically defined population: prospective cohort study of 1994-9 compared to 2000-5	Field DJ, Dorling JS, Manktelow B, Draper ES
American Journal of Epidemiology, (2008) 167 485-491	Recreational drug use: a major risk factor for gastroschisis?	Draper ES, Rankin J, Tonks A, Abrams K, Field DJ, Clarke M, Kurinczuk JJ

K.3 Abstracts

Abstract	Title	Authors
Health Protection Agency (HPA) Annual Conference, 12-15 September 2005, Warwick (oral presentation)	Mortality, deprivation and ethnicity of critically ill children in England and Wales: preliminary findings from the Paediatric Intensive Care Audit Network (PICANet)	Parslow RC, Tasker RC, Chater T, Davey N, Draper ES, Jones S, Parry GJ & McKinney PA.
European Society for Paediatric and Neonatal Intensive Care (ESPNIC) annual conference, 15-17 September 2005, Antwerp (oral presentation)	Mortality, deprivation and ethnicity of critically ill children in England and Wales: preliminary findings from the Paediatric Intensive Care Audit Network (PICANet)	Parslow RC, Tasker RC, Chater T, Davey N, Draper ES, Jones S, Parry GJ, Thiru K & McKinney PA.
Developmental Medicine and Child Neurology (2005) 47 (Suppl 101) 4	Design of randomized controlled trials of the management of raised intracranial pressure in paediatric traumatic brain injury	Forsyth RJ, Morris K, Parslow RC, Hawley C & Tasker RC
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (oral presentation)	Infants admitted to paediatric intensive care with acute respiratory failure in England and Wales	Parslow RC, McKinney PA, Draper ES, O'Donnell R
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (poster presentation)	Collecting national data for clinical audit: The Paediatric Intensive Care Audit Network in Great Britain	Parslow RC, McKinney PA, Draper ES, Thiru K
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (poster presentation)	Admission to PICU with severe bronchiolitis and acute respiratory failure after preterm birth is associated with a longer duration of stay and a higher incidence of apnoeas but not mortality	O'Donnell DR, Parslow RC, McKinney PA, Draper ES
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (poster presentation)	Severe bronchiolitis is associated with the annual UK winter increase in PICU admissions and prolonged stay compared with other diagnoses	O'Donnell DR, Parslow RC, McKinney PA, Draper ES
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (poster presentation)	Hyperglycaemia and insulin therapy in UK paediatric intensive care units	Nayak P, Morris KP, Parslow RC
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (oral presentation)	The effect of missing data on PIM-predicted SMR	Emsden S, Baines P, McClelland T, Parslow RC
5th World Congress on Pediatric Critical Care, 24-28 June 2007, Geneva, Switzerland (poster presentation)	Clinical information system utilisation in paediatric intensive care: A UK perspective	Ramnarayan P, Thiru K, Rowe S on behalf of pan Thames Health Informatics Group
The 15th Annual Public Health Forum, Edinburgh International Conference Centre, 28-29 March 2007, Edinburgh, UK (poster presentation)	Using Data to Inform Commissioning of Paediatric Intensive Care	Sidhu S, Rowe S & Thiru K
HSRN and NIHR SDO Programme joint annual conference. 4 & 5 June 2008, Manchester University Conference Centre (oral presentation)	Workforce wellbeing in paediatric intensive care units with and without extended nursing roles	Coleby D, Tucker J, Draper E, Parry G, McKee L, Skatun D, Davey N, Darowski M

APPENDIX L THE STRUCTURE OF THE NHS IN THE UK

The Structure of the NHS in the UK

APPENDIX M GLOSSARY

The following abbreviations / terms are used within the text of this report:

A&E	Accident and Emergency Department
AIC	Adult Intensive Care
AICU	Adult Intensive Care Unit
ANZPICS	Australian and New Zealand Paediatric Intensive Care Registry
CAG	Clinical Advisory Group
CATS	Children's Acute Transfer Service
CT3	Clinical Terms 3
ECMO	Extra corporeal membrane oxygenation
ENB	English National Board
GB	Great Britain
GOSH	Great Ormond Street Hospital
HB	Health Board
HQIP	Healthcare Quality Improvement Partnership
IC	Information Centre for health and social care
ICNARC	Intensive Care National Audit & Research Centre
ICP device	Intracranial pressure device
Invasive ventilation	Any method of ventilation delivered via an endotracheal tube, laryngeal mask or tracheotomy tube
IQR	Interquartile Range
IV	vasoactive therapy Intravenous drug therapy to support blood pressure and heart rate
LVAD	Left ventricular assist device to support cardiac function
NPfIT	National Programme for Information Technology
NSPD	National Statistics Postcode Directory
NHS	National Health Service
NHSIA	National Health Service Information Authority
NHSnet	A secure wide area network connecting NHS organisations which enables units to transfer data electronically to PICANet
Non-invasive ventilation	Any method of ventilation NOT given via an endotracheal tube, laryngeal mask or tracheostomy tube
PbR	Payment by Results
PCCEWG	Paediatric Critical Care Expert Working Group
PCCMDS	Paediatric Critical Care Minimum Dataset
PCO	Primary Care Organisations
PIAG	Patient Information Advisory Group
PIC	Paediatric Intensive Care
PICANet	Paediatric Intensive Care Audit Network
PICNET	Paediatric Intensive Care Network
PICS	Paediatric Intensive Care Society
PICS SG	Paediatric Intensive Care Society Study Group
PICU	Paediatric Intensive Care Unit
PIM	Paediatric Index of Mortality
PIM 2	Paediatric Index of Mortality version 2
READ Codes	Clinical terminology used to describe clinical conditions, symptoms and observations
RSV	Respiratory syncytial virus
SCT	See SNOMED CT^®
SHO	Senior House Officer
SG	Steering Group
SNOMED CT^®	SNOMED CT^® is a clinical terminology - the Systematised Nomenclature of Medicine. It is a common computerised language that will be used by all computers in the NHS to facilitate communications between healthcare professionals in clear and unambiguous terms
SMR	Standardised mortality ratio
SHA	Strategic Health Authority
SWACIC	South West Audit of Critically Ill Children
UK PICOS	United Kingdom Paediatric Intensive Care Outcome Study

http://www.picanet.org.uk/
picanet@leeds.ac.uk

University of Leeds	University of Leicester	Pan-Thames Co-Ordinator
Patricia McKinney Roger Parslow Thomas Fleming Sarah Skinner	Elizabeth Draper Caroline Lamming	Krish Thiru
PICANet Paediatric Epidemiology Group Centre for Epidemiology & Biostatistics The Leeds Institute of Genetics, Health and Therapeutics University of Leeds Worsley Building Leeds LS2 9JT	PICANet Department of Health Sciences University of Leicester 22-28 Princess Road West Leicester LE1 6TP	PICANet Cardiorespiratory & Critical Care Division Room 8086, Level 8 - Nurses Home Great Ormond Street Hospital for Children Great Ormond Street London WC1N 3JH
r.c.parslow@leeds.ac.uk 0113 343 4856	crl4@le.ac.uk 0116 252 5414	thiruk1@gosh.nhs.uk 020 7762 6713

Bookmarks

The background

Our experience

9 A CLINICIAN'S COMMENTARYDr Michael Marsh

9 A CLINICIAN'S COMMENTARY
Dr Michael Marsh