Statistical Publication Notice

27 July 2010

Click here for Audit of Critical Care in Scotland 2010.

Introduction

The Scottish Intensive Care Society Audit Group (SICSAG) is a voluntary audit that exists to improve the quality of care delivered to Critical Care patients by monitoring and comparing activities and outcomes of patients admitted to adult general Intensive Care Units (ICUs) and High Dependency Units (HDU) in Scotland. In order to examine this comprehensively we collect information including demographics, activity, severity of illness, diagnosis, interventions and outcomes electronically via a bespoke computer system called WardWatcher.

ICUs now collect further data on ICU associated infections which will be published in collaboration with Health Protection Scotland later this year.

Information is reported on 9760 ICU admissions and 22648 HDU admissions to Scottish hospitals during 2009. 

Data were collected prospectively from all general ICUs and over 90% of general High Dependency Units. The report focuses on a graphical presentation of the data, with accompanying text to alert the reader to points of interest.

Key Points

• Significant out of hours workload: 44% of admissions to ICU and 30% of admissions to HDU occur between 8pm and 8am, reflecting the 24/7 nature of Critical Care.  
• New measures: early discharges and readmission within 48 hours of discharge: Both early discharge and readmission within 48 hours can be a marker of insufficient resources, although this is not the only factor. It is recommended that units whose measures are above the norm, examine the reason for this. 
• Delayed discharges and Out of hours discharges: Delays continue to increase (although this may be partly due to improvements in the data collection method). The main reason for both ICU and HDU delays is delay in obtaining downstream beds. Out of hours discharges remain around 15%, reasons may be due to overall shortage of beds or difficulties in finding a downstream bed.
• ICU mortality continues to improve: ICU Standardised mortality ratio has reduced to 0.84. (APACHE II methodology).
• H1N1 planning and preparation: Preparation and planning for pandemic influenza was a major excersise for Critical Care which will serve Scotland well for the future. 
• Quality improvement measures/processes: Appendix 2 and 3 show that quality improvement measures have been adopted in all units.  

Interpretation

Funnel Plots

To show the differences between units, many statistics are shown in the form of funnel plots (a form of statistical quality control chart ? see glossary). Two control limits are shown on these charts. Units, which fall outside the outer control limit for a particular statistic, are considered to be different from the majority of other units. A unit?s divergence from the norm should be interpreted with caution. Although the degree of difference is highlighted in these charts, the reason for that difference is not explained and no judgements regarding ?good? or ?bad? practice are made. Differences may arise from many sources: differences in data accuracy, case-mix, service provision or practice. Sometimes a difference is just a random difference (?caused? by chance alone).

APACHE II

The outcome measure used by SICSAG is the patient?s survival status (alive or dead) when they finally leave hospital (even if this is not the original hospital). Patients admitted to ICUs are at significant, but varied, risk of death. Simply comparing the proportion of patients who die in each unit can give a misleading impression because the severity of their illnesses is different. To overcome this, we use the APACHE II system to adjust the outcome measure for case-mix.

APACHE II produces an expected mortality rate for a unit, which can be compared to the actual observed mortality rate to give a standardised mortality ratio (SMR). It is important to interpret SMRs with caution. It should be appreciated that whilst the system adjusts for case-mix, it does not do so perfectly. Many units admit a relatively small number of patients each year and the confidence intervals around the SMR are therefore wide for such units.

Detailed Findings

Section1: Activity

Number of Admissions

After several years of ICUs admitting more than 10,000 patients, in 2009 there has been a reduction (9760). This is mainly due to the effects of a change in the database at Wishaw General Hospital (this combined unit split into a separate ICU and HDU for audit purposes). A cohort of 17 units who have collected data for ten years show that admissions are stable at around 6000 admissions per year since 2006 for this cohort. 

The number of admissions to HDU has increased (22648) mainly due to the addition of seven more HDUs joining the audit. A cohort of 22 units who have collected data for the last five years shows numbers are stable at around 16000 admissions per year since 2007 for this cohort. 

Critical Care remains a key support for both emergency and planned care in NHS Scotland.

Occupancy

The average Scottish Critical Care bed occupancy (real time recording) remains stable at about 75%.

Length of Stay

Whilst median length of stay for both ICU and HDU is around 2 days, the mean is close to 5 days for ICU and 3 for HDU.
Variations in length of stay between units should generate debate as to whether they reflect differences in case-mix, service provision or practice.

Time of admission

44% of ICU admissions and 30% of HDU admissions occur between 8pm - 8am, reflecting the 24 hour nature of Critical Care.

Time of discharge

Around 15% of ICU and HDU (live) discharges are between 8pm and 8am. Units with a high level of ?out of hours? discharge rates may wish to look further into the reasons for it. Out of hours discharges from ICU have been suggested to be associated with increased mortality (Goldfrad, 2000).

Delayed Discharge

Delayed discharges have increased over the last few years to over 15% of patients in both ICU and HDUs. This may be partly accounted for by this data being better collected in the most recent version of the database (WardWatcher).

Early Discharge

This data is being reported for the first time. 4% of discharges from ICU and 2% of discharges from HDU are defined as ?early?. In these case clinical staff have made the decision to transfer a patient to another area due to pressure on resources and it is not in the best interest of the patient. This may be a sign of insufficient resources and therefore the units with a higher rate of early discharges should review this along with other factors.

Readmissions to Critical Care

This data is being reported for the first time. Readmissions within 48 hours can also be a marker of early discharge. Around 1% of patients are readmitted to the same unit within 48 hours of discharge. Discussion of individual cases may highlight ways of reducing this.

Source of admission

The continuous rise in patients admitted to ICU from Emergency Departments over the last ten years appears to have halted. Transfers between hospitals for ICU continues to increase. Admissions from theatre remains the most common single source of admissions.

Chronic Health

The number of patients with very severe cardiovascular disease has fallen to its lowest level for five years, and there has been a steady increase in the number of patients admitted with liver related chronic health problems.

Organ Donation

After steadily increasing since 2006/7 the number of organ donors fell in 2009/10. NHS Blood and Transplant are now collecting data on number of referrals, actual donors and reasons for organs not being used. In 2009/10 193 patients were referred for organ donation of which 63 became actual donors.  

Section 2: Interventions In Critical Care

Invasive ventilation

The proportion of patients admitted to ICU who receive invasive ventilation via an endotracheal tube or tracheostomy is almost 70%. Invasive ventilation was the primary reason for the development of Intensive Care and is almost exclusively delivered in ICU rather than HDU.

Non-invasive ventilation (NIV) and Continuous Positive Airway Pressure (CPAP)

The use of ventilation using a mask or hood rather than a tracheal tube in ICU has decreased from 13% in 2008 to 9% in 2009. This may be due to the way this data is collected or a change in practice.

NIV and CPAP rates in HDU show wide variation around the mean of 6.4%. This is due to many units (most surgical and specialist HDUs) not providing this intervention.

Cardiovascular support

The use of cardiovascular support (vasoactive and/or antiarrhythmic drugs) in ICU has increased slightly to 46% from 42%. This may be due to changing the definition to include antiarrhythmic drugs. The use of vasoactive and/or antiarrhythmic drugs in HDU is much less common (9%) however considerable inter-unit variation reflects the wide variety of patients who are admitted to HDU.

Renal replacement therapy (RRT)

The proportion of ICU patients who undergo renal replacement therapy, in the form of either dialysis or haemofiltration, has been stable at about12% over the last 4 years. A number of units do not offer this treatment, planning to transfer patients to other units when it is required.

Levels of care

There is a clear distinction between ICUs and HDUs in the level of care they provide (see glossary for definition). ICUs predominantly provide Level 3 care, while HDUs provide Level 2 care. Combined units admit both.

Section 3: Outcomes

Following admission to a Scottish intensive care bed in 2009, 26.7% of patients died before hospital discharge reflecting the severity of illness in this population. The consistent downward trend continues with crude mortality being the lowest recorded. It is also lower than expected by the APACHE II case-mix adjusted standardised mortality ratio, which has fallen from 1.02 in 2002 to 0.84 in 2009.

There is variation through time within units, and there is also variation between units. Two units outside the 2 standard deviations (SD) in 2009 were notified in April 2010 and have both reviewed data and processes to try and account for this difference. It should be recognised that comparison of 25 units has a considerable chance of an outlier at the 2 SD and that SICSAG continue to seek review at a lower level than other similar audits (ICNARC, SPSP).

Both units have been asked to summarise their investigation to SICSAG by July 23rd prior to publication of the report in the public domain.

Section 4: H1N1

Data included following the outbreak of H1N1 over 2009/2010. Critical care admitted 173 patients with confirmed H1N1. This is 33.3 cases per million inhabitants based on the General Registrars figures.

Main Contacts

Angela Kellacher
(Clinical Coordinator, SICSAG)
0131 275 6895
angela.kellacher@nhs.net

Brian Cook
(Chair, SICSAG)
0131 2421000
Page #6652 
brian.cook@luht.scot.nhs.uk

Diana Beard
(Project Manager, Quality Improvement Team)
0131 275 6333 
diana.beard@nhs.net

Glossary

APACHE II

A validated system, which takes account of both the patient?s acute condition and their chronic health, to express the severity of illness of patients in critical care. Based on the APACHE II score and diagnosis, patients are assigned a probability of dying, which is the expected mortality (see also SMR).

Case-mix adjusted mortality

Case-mix refers to the population of patients for which a unit provides critical care. Because different units provide care for different populations one can only compare mortality after an adjustment for case-mix has been made (see also SMR).

Confidence interval (for Standardised Mortality Ratios)

SMRs are usually estimated with numbers from samples. They are therefore subject to some variation from one sample to another. Confidence intervals are used to indicate the reliability of the estimated SMR. With a 95 percent confidence interval there is a 95% chance that the true SMR falls inside the confidence interval of the estimated SMR. And a 5% chance that the true SMR falls outside that interval.

Continuous Positive Airway Pressure

A form of basic respiratory support in which airway pressure is maintained above the normal air pressure to enhance breathing.

Funnel plot

A graph where a performance indicator is shown on the y-axis, while the number of admissions is shown on the x-axis. There is a data point for every unit in the funnel plot. Furthermore, the plot shows the mean as a horizontal line. The funnel plot also shows control limits for the performance indicator. The control limits are confidence intervals, calculated as 2 standard deviations and 3 standard deviations from the Scottish mean. Because the confidence intervals get smaller as the numbers of admissions get larger, the shape of a funnel appears.

Levels of care

A daily level of care is calculated based on the type of interventions/support the patient receives. It categorizes patients from level 3 to level 0. Level 3 patients require the highest level of care such as advanced respiratory support, or two or more organ systems being supported. Level 2 patients require support of one organ system. Level 1 patients need more observation than is available on a general ward and level 0 patients are able to be cared for in a ward.

Non-invasive ventilation

A form of basic respiratory support in which positive pressure from a ventilator is provided via a mask (face/nasal) or a hood to enhance breathing.

Occupancy

The length of time a bed is occupied as expressed as a percentage of the total time.

Out of hours

Admissions or discharges between 8pm and 8am, Monday to Friday.

Standardised Mortality Ratio (SMR)

The SMR is a measure of case-mix adjusted mortality. It is the ratio between the observed mortality and the expected mortality. The observed mortality is the number of patients who die before hospital discharge. According to SICSAG?s methodology the observed mortality is based on the ultimate hospital mortality. The expected mortality is calculated according to the APACHE II system.

Vasoactive drugs

These are drugs used to support the cardiovascular system by increasing blood pressure.

Pre-Release Access

Under terms of the "Pre-Release Access to Official Statistics (Scotland) Order 2008", ISD are obliged to publish information on those receiving Pre-Release Access ("Pre-Release Access" refers to statistics in their final form prior to publication). The standard maximum Pre-Release Access is five working days. Shown below are details of those receiving standard Pre-Release Access and, separately, those receiving extended Pre-Release Access.

Standard (five day) Pre-Release Access

Scottish Government Health Department (Analytical Services Division)
NHS Board Chief Executives
NHS Board Communication leads
Medical Directors of all participating Health Boards
Medical or Associate Medical Directors of all participating hospitals
Lead Audit Consultants (ICU) and Lead Nurses of all participating units/hospitals
SICSAG Steering group
Lead Audit Consultants (ICU) and Nurses (HDU), graphs/tables only for QA

Extended Pre-Release Access

Scottish Government Health Department (Analytical Services Division)
This extended Pre-Release Access is given to a small number of named individuals in the Scottish Government Health Department (Analytical Services Division). This Pre-Release Access is for the sole purpose of enabling that department to gain an understanding of the statistics prior to briefing others in Scottish Government (during the period of standard Pre-Release Access).

History of this Publication

Last Published: 28/07/2009
Next Due: 2011
Data Available Since: First report containing 1995-1997 data published in 1998