section 10: critical care services

199The Royal College of Anaesthetists nn Raising the Standard: a compendium of audit recipes

Section 10: Critical care services

Estimation of demand for critical care beds

Complications of central venous cannulation

Tracheal tube cuff pressure

Percutaneous tracheostomy

Glucose control in critically ill patients

Intracranial pressure monitoring

Care bundles in ventilated patients

Hand hygiene

Communication with relatives or visitors ofICU patients

Quality of handovers of patients in intensivecare units by medical staff

Substandard care after discharge as a factorin early mortality

Readiness of patients for transfer

Training for inter- and intra-hospital transfers

Evidence-based transfusion practice inintensive care patients

10.1

10.2

10.3

10.4

10.5

10.6

10.7

10.8

10.9

10.10

10.11

10.12

10.13

10.14

Edited byDr Giles Morgan

200

10 Critical care services

The Royal College of Anaesthetists nn Raising the Standard: a compendium of audit recipes

Estimation of demand forcritical care bedsDr A P Daykin

Why do thisaudit?

Best practice:research evidenceor authoritativeopinion

Suggestedindicators

Proposedstandard or targetfor best practice

In 1999 the Audit Commission recommended that Hospital Trusts should conduct audits todetermine the need for intensive care beds, particularly to determine whether use of thesebeds was appropriate. It identified a number of ways to estimate demand for critical care beds.1

Methodology to calculate critical care bed need is included in the accompanying CD-ROM.

Comprehensive critical care2 recommends that the concept of intensive care and high dependencybeds is replaced by four levels of patient dependency, regardless of the patient’s location or thedesignation of beds.3 Nonetheless, the designation of levels to beds is commonplace.

For many years it has been arbitrarily recommended that between 1 to 2% of a hospital’s acutebeds are Level 3 (intensive care), in units of between 4 and 8 beds, and that occupancy shouldrun at about 70%. Persistent occupancy of more than 70% suggests a unit is too small.4

Occupancy of 80% or more is likely to result in non-clinical transfers, with associated risks.Over provision is wasteful, but failure to admit in a timely manner increases the risk of deathand may increase length of stay, with increased cost. Too few beds may increase mortality ornon-clinical transfers, while extra capacity may not decrease occupancy if patients of a lowerdependency level are admitted, or there are problems discharging patients to other locations.5

Assessing the demand for beds for Level 2 (high dependency) care requires audit of patients onacute wards who fulfil the criteria for that level of care and can be done by the outreach teamusing early warning systems.6 Mathematical modelling has been used to simulate hypotheticalcritical care services and can predict requirements more accurately than has previously beenpossible.7 The demands for Levels 1, 2 and 3 care are inter-related entities because individual patientschange dependency while remaining in the same location. They can be audited separately or together.

Level 3 care

Number of patients requiring Level 3 care per day.Occupancy of Level 3 beds.% of appropriate admissions refused due to lack of beds.% of patients discharged prematurely, for non-clinical reasons.% of patients readmitted.% of non-clinical transfers.% of planned admissions whose elective surgery is deferred due to lack of beds.Level 2 care

Number of patients in critical care beds and acute wards fulfilling criteria for Level 2 care.% of appropriate referrals to ICU refused due to lack of beds.Number of deferred elective operations.% of premature discharges from and readmissions to ICU.

Level 3 care

100% of patients requiring Level 3 care are in intensive care.Less than 80% bed occupancy in intensive care.0% appropriate admissions refused.0% patients prematurely discharged.0% surgery deferred for non-clinical reasons.Less than 5% readmission rate.0% non-clinical transfers.

10.1

10

201

10.1


Critical care services

Suggested data tobe collected

Common reasonsfor failure toreach standards

References

Select a study period that reflects seasonal and other local variations in demand for critical care.

Bed occupancy may be provided at intervals by the ICNARC Casemix Programme to subscribingunits. Otherwise, occupancy may be calculated at a specific time daily or on change of nursingshift by expressing the number of occupied beds as a percentage of the total bed spaces oroperational bed spaces. The provision of nurses may vary on a shift-to-shift basis, which may havean effect on the number of operational beds as opposed to the number of bed spaces.

In the ICU:nn number of critical care bed spaces nn number of operational critical care bedsnn number of occupied critical care bedsnn calculated intensive care bed occupancynn number of appropriate patients denied intensive carenn number of planned surgical cases cancelled because of non-availability of critical care bedsnn number of patients discharged prematurelynn number of patients discharged at nightnn number of non-clinical inter-hospital transfersnn number of patients readmitted to intensive care.

Numbers of patients fulfilling the requirements for Level 2 and 3 critical care in:nn intensive care unitsnn high dependency unitsnn acute wards.

Insufficient critical care beds.

Poor bed management.

Lack of Level 1 beds.

Lack of outreach service.

1 Audit Commission. Critical to success. The place of efficient and effective critical care services withinthe acute hospital. Audit Commission, London 1999 (see: www.audit-commission.gov.uk/).

2 Department of Health. Comprehensive critical care. A review of adult critical care services. DH,London 2000 (see: www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidanceArticle/fs/en?CONTENT_ID=4006585&chk=BXvlex).

3 Intensive Care Society. Levels of critical care for adult patients. ICS, London 2002 (see: www.ics.ac.uk/).

4 Intensive Care Society. Standards for intensive care units. ICS, London 1997 (see: www.ics.ac.uk/).

5 Parker A,Wyatt R, Ridley S. Intensive care services; a crisis of increasing expressed demand.Anaesthesia 1998;5533:113–120.

6 Intensive Care Society. Guidelines for the introduction of outreach services. ICS, 2002 (see:www.ics.ac.uk/).

7 Costa A et al. Mathematical modelling and simulation for planning critical care capacity. Anaesthesia2003:5588;320–327.

Level 2 care

100% of patients requiring Level 2 care in appropriate beds.0% of patients requiring Level 3 care in Level 2 beds.0% appropriate referrals refused.0% of patients prematurely discharged.Less than 5% readmission rate.

http://www.ics.ac.uk/



http://www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidanceArticle/fs/en?CONTENT_ID=4006585&chk=BXvlex

http://www.dh.gov.uk/PublicationsAndStatistics/Publications/PublicationsPolicyAndGuidance/PublicationsPolicyAndGuidanceArticle/fs/en?CONTENT_ID=4006585&chk=BXvlex

http://www.audit-commission.gov.uk/

202



Complications of central venouscannulationDr G Anderson, Dr C G Powell

Why do thisaudit?

Best practice:researchevidence orauthoritativeopinion

Suggestedindicators

Proposedstandard ortarget for bestpractice

Central venous cannulation is a technique used widely for haemodynamic monitoring, theadministration of drugs, including inotropes, and total parenteral nutrition (TPN). It is also used tosecure vascular access for continuous renal replacement therapy and transvenous cardiac pacing.However, there are risks associated with the insertion and subsequent management of thesedevices. Early complications include pneumothorax, incorrect placement and vascular damage,while later complications include thrombosis and catheter related infection. Pulmonary arteryflotation catheter (PAFC) insertion may cause dysrhythmias, pulmonary artery rupture, heart valveor sub-endocardial damage and thromboembolism. This audit of the complications associatedwith the use of central venous catheters examines quality of practice in an environment wherestandards of supervised instruction should be high.

In 2002, the 1996 Hospital Infection Control Practices Advisory Committee (HIPAC) guidelines forthe prevention of infections relating to intravascular devices1 were replaced by the recommendationsof the Centers for Disease Control and Prevention (CDCP).2 The work was supervised by theSociety of Critical Care Medicine.The recommendations reflected the consensus of HIPAC and theother professional organisations represented in the working group. They include:

nn The use of maximal sterile barrier precautions during catheter placement.

nn The use of 2% chlorhexidine preparation for skin antisepsis.

nn Avoidance of routine (time-based) replacement of central venous catheters as a strategy toprevent infection.

nn Consider use of antiseptic or antibiotic impregnated short-term central venous catheters if therate of infection remains high despite adherence to these strategies.

In September 2002, the National Institute for Clinical Excellence (NICE) published guidelinesrecommending the use of ultrasound devices to assist with the placement of central venouscatheters.3 Practice guidelines for pulmonary artery catheterisation were published by the AmericanSociety of Anesthesiologists in 2003.4 Training and educational programmes for those who insert andmaintain catheters should include didactic and interactive components and should be documented.1

Documented evidence of educational programmes.

% Catheter insertions which adhere to CDCP guidelines.

% Catheter insertions adhering to the NICE guideline on the use of ultrasound.

% Complications at insertion.

% Subsequent catheter colonisation and catheter related infection.

Documented evidence of training for those who insert and maintain catheters.

100% catheter insertions should adhere to the CDCP guidelines.

Central line insertions should adhere to the NICE guidelines (2002) on the use of ultrasound:nn 100% of patients undergoing elective central line insertion should have ultrasound guided

catheter insertion in the internal jugular vein.nn 100% of central venous catheters should be inserted by healthcare practitioner trained in use

of 2D ultrasound imaging

The complication rate should be:

Pneumothorax: < 5% subclavian approach, < 3 % internal jugular approach.

Haematoma due to arterial puncture: < 1% subclavian approach, < 3% internal jugular approach.

Incorrect placement: < 4%.

10.2

10

203

10.2



Suggested datato be collected

Commonreasons forfailure to reachstandards

References

Documentation on evidence of training as above.At insertion:nn date of insertion and anticipated removal of catheternn grade and experience of operatornn inserted as an emergency or a planned procedurenn catheter purpose, e.g. drugs, nutrition, venous accessnn operator supervised or unsupervisednn ultrasound guided or landmark methodnn insertion site nn details of barrier precautions and skin preparationnn length of catheter inserted into patientnn type of catheter (number of lumens, antibiotic coating, length and diameter)nn immediate complicationsnn radiographic confirmation of situation of catheter.Subsequent observations:nn appearance of insertion site, e.g. inflamed, infectednn record of dressing changesnn antibiotic therapynn duration of cannulation, date of removalnn % bacterial colonisation of catheter after culturenn % catheter related blood stream infection (CRBSI)nn number of days catheter remained in place when no longer necessarynn number of days catheter remained in place in excess of days anticipatednn indications for PAFC insertion and duration of time used.

Insertion of catheter in emergency.Substandard antiseptic technique.Poor knowledge of anatomy and anatomical landmarks.Ultrasound device not available.Operator not trained in the use of 2D ultrasound.Lack of supervised training.Inappropriate indication for PAFC insertion.Requirement for PAFC data to assist management for longer than 72 h.Irritable myocardium.Poor care of insertion site and catheter.Inappropriate purpose of catheter, e.g. venous access.

1 Pearson ML. Guideline for the prevention of intravascular device-related infections. Am J InfectionControl 1996;2244((44)):262–277.

2 Centers for Disease Control and Prevention. Guidelines for the prevention of intravascular catheter-related infections. MMWR 2002;5511((NNoo RRRR––1100)):1–33.

3 National Institute of Clinical Excellence. Guidance on the use of ultrasound locating devices for placingcentral venous catheters; Technology appraisal guidance. NICE, London September 2002.

4 Practice guidelines for pulmonary artery catheterization: an updated report by the American Society ofAnaesthesiologists Task Force on Pulmonary Artery Catheterisation. Anesthesiology 2003;9999((44)):988–1014.

5 Shah KB. A review of pulmonary artery catheterisation in 6245 patients. Anesthesiology 1984;6611:271–275.

For PAFCs dysrhythmias at insertion: Moderate < 33%; major < 10%99% PAFCs should be removed within three days of insertion.5

Incidence of catheter colonisation should be < 25% and catheter related infection < 5%.



The auditor should visit the ICU without warning. The cuff pressure is measured and comparedto the minimum pressure required to prevent a leak. Associated factors may be recorded, e.g.patient recently arrived from elsewhere, shortly after intubation, etc. The ventilator settings andpeak airway pressure should be noted.

Ignorance of need to measure cuff pressure.

Failure to measure cuff pressure regularly.

Anatomical defects, e.g. tracheal dilatation.

204



Tracheal tube cuff pressureDr P Nightingale, Mr H Griffiths, Ms J A Clayton

Why do thisaudit?


Suggestedindicators


It is well known that a high pressure in the cuff of a tracheal tube can damage the trachealmucosa. In intensive care practice, where intubation is frequently prolonged, it is commonpractice to check the cuff pressure at frequent intervals. However, if this is performed withoutactually measuring the pressure in the cuff, high pressures may still result, even when the cuff isapparently only inflated enough to just stop any auscultated leak.

It has been shown that excessive cuff pressures can cause ischaemia and necrosis of thetracheal mucosa despite the introduction of large volume, low pressure cuffs.1 Also, largevolume cuffs can exert a high pressure on the tracheal wall if inflated past their resting volume.2

Excessive pressures must be prevented in all patients by objective measurement.3 Where thecuff pressure required to prevent a leak is high, options include changing the size or type oftube, changing the mode of ventilation to reduce peak airway pressure, placing a tracheostomybelow the area of dilation or accepting a small leak.

% cuffs inflated to the correct pressure when checked at random by the auditor. The correctpressure is the minimum pressure required to prevent a leak but no cuff pressure shouldexceed 35 cm H2O.

% cuff pressures measured and documented at the start of each nursing shift.

100% cuffs should be correctly inflated.

100% cuff pressures should be measured and documented at the start of each nursing shift.

10.3

10

205

10.3



References 1 Seegobin RD, van Hasselt GL. Endotracheal cuff pressure and tracheal mucosal blood flow:endoscopic studies of of effects of four large volume cuffs. Br Med J 1984;228888:965–968.

2 Guyton D, Banner MJ, Kirby RR. High-volume, low-pressure cuffs. Are they always low pressure?Chest 1991;110000:1076–1081.

3 Fernandez R et al. Endotracheal tube cuff pressure assessment: pitfalls of finger estimation and needfor objective measurement. Crit Care Med 1990;1188:1423–1426.

206



Percutaneous tracheostomyDr G J Bellingan

Why do thisaudit?


Suggestedindicators


Tracheostomy is frequently performed on patients in the ICU. The decision to form atracheostomy depends on the individual patient but is commonly made when mechanicalventilation is expected to be prolonged. Percutaneous tracheostomy has gained popularity inrecent years. It can be performed easily at the bedside, is minimally invasive and fewcomplications are claimed. Although this procedure is regarded as safe, it is clear thatcomplications can occur, especially when trainees are learning the procedure. The requirementfor anaesthesia in a critically ill patient adds further hazards. The optimal timing fortracheostomy is also unclear and a national study on this commenced in 2005 (TracMan study,see: www.tracman.org.uk). Audit of complications is therefore important.

The following four reviews give evidence of the expected complication rate:

Petros, 1997.1 Review of 137 patients:

Acute complications: 11% (one severe bleeding).

Postoperative in hospital complications: 5%.

Mortality: nil.

Marx, 1996.2 Review of 254 patients:

Minor complications: 6.5%.

Major complications: 1.5%.

Mortality: 1 (0.004%).

Barrachina, 1996.3 Review of 44 patients:

Immediate complications: one paratracheal cannula insertion.

Postoperative complications: 16%.

Mortality: nil.

There are two meta-analyses of percutaneous vs surgical tracheostomies.4,5

% patients in whom Trust consent protocols were followed.

% percutaneous tracheostomy procedures at which two anaesthetists/intensivists are presentwho are Specialist Registrar (SpR) level or above, one to do the procedure and one to look afterthe patient. One of them should be a consultant or a SpR Year 4 specialising in intensive care.

% patients in which minimum monitoring used: pulse oximetry and end tidal CO2.

% patients in which a fibreoptic bronchoscope was used.

% patients in which a note of the procedure and complications was made.

% patients in which a complication occurred.

% tracheostomy related mortality.

The first five indicators should be achieved in 100% patients.

The % cases in which a complication occurred should be less than quoted in the above reviews.

There should be 0% tracheostomy related mortality.

10.4

http://www.tracman.org.uk

10

207

10.4





References

Name and status of operator, and anaesthetist and assistants if present.

Name of doctor gaining consent if different from operator.

Indication for tracheostomy.

Timing of tracheostomy from date of admission to intensive care.

Timing of tracheostomy from date of intubation.

Relevant laboratory data including blood gas analysis and coagulation status.

Operative details.

Anaesthetic details.

Monitoring equipment available.

Bronchoscope availability and use.

Early and late complications.

Mortality.

Problems relating to availability of personnel, training or equipment.

Need for separate anaesthetist.

Value of and requirement for bronchoscopy not appreciated.

1 Petros S, Engleman L. Percutaneous dilatational tracheostomy in a medical ICU. Intens Care Med1997;2233:630–634.

2 Marx W H, Ciaglia P, Graniero KD. Some important details in the technique of percutaneousdilatational tracheostomy via the modified Seldinger technique. Chest 1996;111100:762–766.

3 Barrachina F et al. Percutaneous dilatational cricothyroidotomy outcome with 44 consecutive patients.Intens Care Med 1996;2222:937–940.

4 Dulguerov P et al. Percutaneous or surgical tracheostomy: a metanalysis Crit Care Med1999;2277((88)):1617–1625.

5 Freeman BD et al. A meta-analysis of prospective trials comparing percutaneous and surgicaltracheostomy in the critically ill. Chest 2000;111188((55)):1412–1418.

208



Glucose control in critically illpatientsDr S L Saunders, Dr L C Chicote-Hughes, Dr J J Paddle

Why do thisaudit?


Suggestedindicators


Hyperglycaemia is common in critically ill patients, even in those without diabetes mellitus.Carefully regulated insulin therapy to maintain lower blood glucose levels has been associatedwith a reduction in morbidity and mortality in critically ill patients.1–3 It is thought that control ofglucose levels, rather than absolute levels of exogenous insulin administered, is responsible forthis effect.2 Since these studies, protocols for tighter control of blood glucose in critically illpatients have been introduced successfully.4 However, tight control of blood glucose isassociated with hypoglycaemic episodes and these are associated with a poor outcome.5

Auditing control of blood glucose in critically ill patients is, therefore, important for optimisingpatient outcome and monitoring adverse effects.

A study of 1,548 patients in a predominantly cardiac surgical intensive care unit demonstratedthat control of blood glucose between 4.4 and 6.1 mmol/l during intensive care stay wasassociated with a reduction in mortality at 12 months, from 8.0% in the control group to 4.6%in the treatment group. Blood glucose in the control group was 10–11.1 mmol/l. Hospitalmortality was reduced by 34%. The incidence of acute renal failure requiring filtration or dialysiswas reduced by 41%, red cell transfusion by 50% and critical illness polyneuropathy by 44%.Prolonged mechanical ventilation and length of stay in intensive care were also reduced.1

This study has since been repeated by the same investigators in a cohort of medical patientswith worse severity of illness as reflected by APACHE score (median APACHE II score 24).3

Intensive care mortality was reduced from 38.1% to 31.3% and hospital mortality from 52.5%to 43.0%. Reductions in length of stay were also seen.

However, tight control of blood glucose must be practised with care because levels lower than2.2 mmol/l are associated with a poor outcome. In the VISEP study5 28-day mortality forpatients who had episodes of hypoglycaemia was 37.1% compared to 25.1% in patients whodid not.

Use of a local protocol to guide glycaemic control is associated with better control.1,4

An example protocol is contained on the accompanying CD-ROM.

A local protocol is used to guide therapy.

% patients with blood glucose > 7 mmol/l started on insulin therapy.

Time to starting insulin after blood glucose > 7 mmol/l.

% of recordings of blood glucose < 2.2 mmol/l.

% of patients with blood glucose < 2.2 mmol/l.

100% of eligible patients started on an intensive insulin therapy protocol.1,4

Incidence of hypoglycaemia (blood sugar < 2.2 mmol/l) < 2%.

10.5

10

209

10.5





References

Blood sugar levels recorded for all patients at regular, precise intervals.

Insulin infusion rates for all patients.

Duration of insulin infusion for all rates.

Lack of training or resources.

Overly complex protocol.

1 Van Den Berghe G et al. Intensive insulin therapy in the surgical intensive care unit. N Engl J Med2001;334455:1417–1418.

2 Finney J et al. Glucose control and mortality in critically ill patients. J Am Med Assoc2003;229900((1155)):2041–2047.

3 Van Den Berghe G et al. Intensive insulin therapy in medical intensive care patients improvesoutcome (unpublished data).

4 Laver S et al. Implementing intensive insulin therapy: development and audit of the Bath insulinprotocol. Anaesth Intens Care 2004;3322((33)):311–316.

5 Reinhart K et al. Volume therapy and intensive insulin therapy in sepsis (VISEP study – unpublisheddata).

210



Intracranial pressure monitoringDr J J Paddle, Dr G Morgan.

Why do thisaudit?


Suggestedindicators


Brain injury is common and disabling. Amongst the strategies to reduce morbidity and mortalityin brain injured patients intracranial pressure (ICP) monitoring has been shown to be effectiveand safe in adults1–3 and children.4,5 Despite this, over 40% of neurosurgical referral centres donot routinely monitor ICP,6 even in severe traumatic head injury, for which there is evidence ofbenefit.7

The Brain Trauma Foundation7 has undertaken a comprehensive review of the management oftraumatic brain injury, producing guidelines recommending ICP monitoring in severe brain injury(GCS ≤ 8) with an abnormal CT scan (or with a normal scan if two of the following criteria aremet – age > 40 years, motor posturing, systolic BP < 90 mmHg).

Evidence of benefit from ICP monitoring in other patient groups is less clear. CT abnormalitiesare found in over 40% of patients with moderate head injury (GCS 9–13),8 and monitoring thisgroup is suggested as an option.7 ICP monitoring in non-traumatic brain injury is becomingmore common, though there is only limited evidence of benefit in encephalitis9 and Reye’ssyndrome,10 and equivocal evidence in fulminant hepatic failure11 and strokes.12,13

There is a growing consensus that ICP monitoring is appropriate in general ICUs,14 wheresurgical intervention is not needed, and this is becoming established in units around the country.Fibreoptic ICP monitoring is relatively free of complications. From selected series1,2,5 theincidence of meningitis was < 1%, local wound infections 0–2.5%, intracranial haematoma0–0.3%. Device failure was up to 4.1% although 13% in a smaller series of children.4 Othermethods of monitoring have a higher proportion of complications.2

% of brain injured patients who meet the Brain Trauma Foundation guidelines criteria for severebrain injury should have ICP monitoring, regardless of the type of ICU in which they aremanaged.

% patients receiving a fibreoptic ICP monitor who have complications as listed below.

100% of patients fulfilling the Brain Trauma Foundation guideline criteria for severe brain injuryshould have ICP monitoring. Patients for whom the prognosis was thought to be hopeless maybe excluded at the discretion of the auditor. There are no agreed standards as to how thisshould be judged.

Targets for complication rates of insertion of a fibreoptic ICP monitor:

nn meningitis < 1%

nn local infection < 2%

nn intracranial haematoma < 1%

nn device failure < 4%.

10.6

10

211

10.66





References

Numbers of brain injured patients and whether ICP was monitored.

Aetiology of brain injury.

CT findings.

Glasgow Coma Score (GCS) – best and worst.

For each insertion of a fibreoptic monitor suggested data to be collected are:

nn grade of operator

nn complications

nn interventions based on ICP

nn outcome.

Failure to adhere to guidelines in traumatic brain injury.

Inappropriate indications for monitoring in non-traumatic brain injury.

Inexperienced operators.

1 Eddy VA et al. Aggressive use of ICP monitoring is safe and alters patient care. Am Surg1995;6611((11)):24–29.

2 Bochicchio M et al. Bedside burr hole for intracranial pressure monitoring performed by intensivecare physicians. A 5-year experience. Intens Care Med 1996;2222:1070–1074.

3 Marshall LF, Smith RW, Shapiro HM. The outcome with aggressive treatment in severe head injuries.Part 1:The significance of intracranial pressure monitoring. J Neurosurg 1979;5500:20–25.

4 Jensen RL, Hahn YS, Ciro E. Risk factors of intracranial pressure monitoring in children with fibreopticdevices: a critical review. Surg Neurol 1997;4477:16–22.

5 Pople IK et al. Results and complications of intracranial pressure monitoring in 303 children. PediatrNeurosurg 1995;2233((22)):64–67.

6 Matta B, Menon D. Severe head injury in the United Kingdom and Ireland: a survey of practice andimplications for management. Crit Care Med 1996;2244((1100)):1743–1748.

7 Bullock R et al. Guidelines for the management of severe head injury. Brain Trauma Foundation.J Neurotrauma 1996;1133((1111)):641–734.

8 Stein SC, Ross SE. Moderate head injury: a guide to initial management. J Neurosurg 1992;7777:562–564.

9 Barnett GH, Ropper AH, Romeo J. Intracranial pressure and outcome in adult encephalitis. JNeurosurg 1988;6688((44)):585–588.

10 Glasgow JF et al. Reye’s syndrome: assessment of intracranial monitoring. Br Med J 1987;229944:337–338.

11 Blei AT et al. Complications of intracranial pressure monitoring in fulminant hepatic failure. Lancet1993;334411:157–158.

12 Ropper AH, King RB. Intracranial pressure monitoring in comatose patients with cerebral hemorrhage.Arch Neurol 1984;4411((77)):725–728.

13 Schwab S et al. The value of intracranial pressure monitoring in acute hemispheric stroke. Neurology1996;4477((22)):393–398.

14 Waldmann CS,Thyveetil D. Management of head injury in a district general hospital. Care of theCritically Ill 1998;1144:65–70.

212



Care bundles in ventilatedpatientsDr D Adams, Dr J J Paddle

Why do thisaudit?


Suggestedindicators


Care bundles have been adopted by the NHS Modernisation Agency (see:wwwwww..mmooddeerrnn..nnhhss..uukk//ccrriittiiccaallccaarree//) as a means of grouping elements of care which, individually,have a sound research base and which together can improve patient outcomes. Mechanicalventilation in the intensive care unit (ICU) is a common and life saving intervention, though it isnot without associated morbidity and mortality. The concept of a ventilator care bundle wasintroduced to minimise complications of mechanical ventilation. The elements of the bundleare: gastric ulcer prophylaxis, deep venous thrombosis prophylaxis, semi-recumbent positioningand a daily sedation hold for each ventilated patient. An audit of the use of the ventilator carebundle allows an assessment of whether best practice is being adopted, and whether localguidelines are being followed.

Gastrointestinal haemorrhage occurs in 9% of ventilated patients, with clinically importantbleeding in 2%.1 Prophylaxis with drugs or enteral feeding reduces this incidence.2–4

Thromboembolic disease contributes to the deaths of up to 12% of ICU patients,5 withadditional morbidity. Low dose heparin, and low molecular weight heparins can reduce theincidence of deep venous thrombosis by 50%6,7 and 45%8 respectively in the general ICUpopulation.

The supine position is a factor in the development of ventilator associated pneumonia.9 Thesemi-recumbent position (45 degrees head up) reduces the incidence of clinically suspected ormicrobiologically proven pneumonias (relative risk reduction of 26% and 18% respectively).10

Continuous intravenous sedation may prolong the duration of mechanical ventilation and lengthof stay in the ICU.11 A daily sedation hold has been shown to reduce the duration ofmechanical ventilation by more than 2 days, and total ICU stay by 3.5 days.12

The introduction of ventilator care bundles has been shown to reduce both ventilator days andlength of stay in intensive care. Every ICU should have guidelines for the application of carebundles to mechanically ventilated patients.

% of patient episodes in which ventilator care bundles were applied.

% of patient episodes in which ventilator care bundles were adhered to.

Establish a written guideline for the administration of the care bundle for ventilated patients.

100% of mechanically ventilated patients should receive care bundles.

10.7

http://www.modern.nhs.uk/criticalcare/

10

213

10.7





References

For every ventilated patient, every day, collect documentary evidence that prophylaxis againstgastric ulcer and deep venous thrombosis are administered and that semi-recumbentpositioning and daily sedation holds are being applied as per unit policy.

The incidence of omissions of individual elements of the care bundle.

Deviations from policy and the reasons for this.

Absence of a clear written guideline relating to the care bundle.

Ignorance among staff of local guidelines.

Inadequate system for monitoring compliance.

1 Cook DJ et al. Incidence of clinically important bleeding in mechanically ventilated patients. J IntensCare Med 1991;66:167–174.

2 Cook DJ et al. Stress ulcer prophylaxis in the critically ill: a meta-analysis. Am J Med 1991;9911:519–527.

3 Cook D et al. A comparison of sucralfate and ranitidine for the prevention of upper gastrointestinalbleeding in patients requiring mechanical ventilation. N Eng J Med 1998;333388:791–797.

4 Pingleton SK, Hadzima SK. Enteral alimentation and gastrointestinal bleeding in mechanically ventilatedpatients. Crit Care Med 1983;1111:13–16.

5 Geerts W, Shelby R. Prevention of venous thromboembolism in the ICU. Chest 2003;112244:357S.

6 Attia J et al. Deep vein thrombosis and its prevention in critically ill patients. Arch Int Med2001;661111:1268–1279.

7 Fraisse F et al. Nadroparin in the prevention of deep vein thrombosis in acute decompensatedCOPD. Am J Respir Crit Care Med 2000;116611:1109–1114.

8 Cook DJ et al. Prevention of venous thromboembolism in critically ill surgical surgery patients: a crosssectional study. J Crit Care 2001;1166:101–166.

9 Kollef M. The prevention of ventilator associated pneumonia. N Eng J Med 1999;334400:627–634.

10 Drakulovic MB et al. Supine body position as a risk factor for nosocomial pneumonia in mechanicallyventilated patients: a randomized trial. Lancet 1999;335544:1852–1858.

11 Kollef MH et al. The use of continuous IV sedation is associated with prolongation of mechanicalventilation. Chest 1998;111144:541–548.

12 Kress JP et al. Daily interruption of sedative infusions in critically ill patients undergoing mechanicalventilation. N Eng J Med 2000;334422:1471–1477.

214



Hand hygieneDr S Knight

Why do thisaudit?


Suggestedindicators


Nosocomial infection remains a significant cause of morbidity and mortality in current practice.Pathogens can be passed from patient to patient by medical and nursing staff. Simple handhygiene measures can reduce the incidence of nosocomial infections.1

Hands should be washed and treated with alcohol hand rub in the following circumstances.2

nn Before and after direct contact with a patient

nn Before donning sterile gloves prior to aseptic procedures such as central venous catheter orurinary catheter insertion

nn When moving from a contaminated body site to a clean body site

nn After contact with inanimate objects in the immediate vicinity of the patient

Hands should be washed:

nn before eating

nn after using the toilet.

Hands should be dried with a disposable paper towel and the towel used to turn the taps off.

Alcohol hand rub should be available at each patient’s bedside, or staff should have personaldispensers with them.

Appropriate educational programmes and visual reminders of the importance of hand hygiene,including hand washing technique, should be provided for staff.

% of staff following hand hygiene guidelines.

% of bed areas provided with alcohol rub dispensers.

% of staff provided with personal alcohol rub dispensers.

% of sinks with nearby hand washing technique advice.

100% of staff contacts with patients should involve correct hand hygiene.

100% of bed areas should have alcohol rub dispensers.

100% of staff should have personal alcohol rub dispensers provided.

100% of sinks should have hand hygiene advice nearby.

10.8

10

215

10.8





References

Number of episodes of patient contact associated with correct hand hygiene.

Number of episodes of patient contact associated with incorrect hand hygiene.

Professional group of each staff member having patient contact.

Compliance of intensive care staff with hand hygiene guidelines.

Compliance of visiting staff with hand hygiene guidelines.

Workload of the unit at the time of the audit.

Availability of alcohol hand rub, soap, paper towels, and hand creams.

Evidence of hand hygiene educational materials in the clinical workplace, staff rest areas andtoilets.

Staff knowledge of hand hygiene guidelines.

Work pressures: bed occupancy, limited capacity, high patient turnover.

High incidence of emergency interventions.

Non-availability of alcohol hand rub.

Poor educational standards.

1 Pittet D et al. Effectiveness of a hospital-wide programme to improve compliance with hand hygiene.Lancet 2000;335566((99223388)):1307–1312.

2 Boyce JM, Pittet D. Guidelines for hand hygiene in healthcare settings. Recommendations of theHealthcare Infection Control Practice Advisory Committee and the HICPAC/SHEA/APIC/IDSAHygiene Task Force. Centre for Diseas Control and Infection, 2002: 51 (No. RR–16, see:www.cdc.gov/handhygiene).

http://www.cdc.gov/handhygiene

216



Communication with relatives orvisitors of ICU patientsDr V Jayaprakash, Professor G B Smith

Why do thisaudit?


Suggestedindicators


High quality communication with the relatives, partners or carers of ICU patients is importantfor several reasons. It introduces the staff caring for the patient and allows a description of thework of the ICU, treatment and equipment in use and the facilities for visitors. Goodcommunication serves to reduce the family’s anxiety, provides reassurance, facilitates betterunderstanding,1 gives information and answers questions. It is also essential for correctunderstanding of the patient’s diagnosis, prognosis and current condition. It assists the processof making difficult ethical decisions and of obtaining assent to treatment or its limitation orwithdrawal. Frequent iteration of issues is important. Good communication can minimise thelikelihood of complaints or litigation.

The ‘families’ of ICU patients frequently suffer confusion, bewilderment, denial, guilt, concern, fear,insecurity and stress.1–3 Although much of this is related to the patient’s severity of illness, it canbe modified. Relatives of patients in ICUs have highlighted the need for frequent re-assuranceand the provision of sufficient information on a regular basis.1 Communication between healthworkers and visitors can be deficient at times4 and often results in complaints to the NHSOmbudsman. Studies have shown that in some cases over 50% of information pertaining todiagnosis, prognosis and treatment is not assimilated.5 No data exists regarding standards forthe frequency and content of interviews by ICU staff with relatives or visitors of ICU patients,but there is an increasing body of information validating the perceived needs of the family.6,7

% of families acknowledged within 2 min of their arrival by an ICU nurse.

% of families seen within 15 min of arrival by the nurse caring for the patient.

% of families interviewed within 30 min of initial arrival by an ICU doctor.

% of days in which a formal interview took place between an ICU doctor and the family ofeach patient.

% of interviews documented in the patient’s medical and nursing notes.

% of interviews where there is consistency between the facts recorded in the nursing andmedical notes following an interview with the patient’s family.

100% of families should be acknowledged by one of the ICU nursing staff within 2 min of theirinitial arrival at the ICU.

100% of families should be seen by the nurse caring for the patient within 15 min of their initialarrival at the ICU. A general description of the patient’s condition should be given, andexplanation that further detailed communication will occur at the earliest opportunity.

100% families should be interviewed by the ICU doctor within 30 min of their arrival.

On 100% of days, there should be a formal interview between an ICU doctor and the family.On 75% of days this should be the consultant.

The content of all (100%) of interviews with the family should be documented in the nursingand medical notes. There should be 100% consistency in the record details.

10.9

10

217

10.9





References

Time of events listed above. During the daytime, an observer, such as the ward clerk, mightcollect this information.

Insufficient medical and nurse staffing on ICU. Medical or nursing staff may be unavailable dueto the immediate care of the patient concerned or the management of other emergencies orthere may be insufficient consultant or trainee sessions in intensive care.

The ‘family’ may be unable to travel to the hospital.

1 Quinn S, Redmond K, Begley C. The needs of relatives visiting adult critical care units as perceived byrelatives and nurses. Part 1. Int Crit Care Nurs 1996;1122:168–172.

2 McGauughey J, Harrison S. Understanding the pre-operative information needs of patients and theirrelatives in intensive care units. Int Crit Care Nurs 1994;1100:186–194.

3 Plowright CI. Needs of visitors in the intensive care unit. Br J Nurs 1995;44:1081–1083.

4 Personal view. Br Med J 1986;229933:1432.

5 Azoulay E et al. Half the families of intensive care unit patients experience inadequate communicationwith physicians. Crit Care Med 2000;2288:3044–3049.

6 Molter NC. Needs of relatives of critically ill patients: a descriptive study. Heart and Lung1979;88:332–339.

7 Bijttebier P et al. Needs of relatives of critical care patients: perceptions of relatives, physicians andnurses. Intens Care Med 2001;2277:160–165.


Construct an audit form to identify the domains of information to be discussed and transmittedto the next shift. An example is included on the accompanying CD-ROM.

Before the handover, a consultant familiar with the patients should outline to the auditor whichof the domains of information are relevant to a particular patient by ticking the ‘relevant’column on the audit form. Decide whether to audit all or a limited selection of the patients.Not all domains of information are relevant for all patients.

The auditor attends the handover and ticks the domains of information actually handed over.Time the duration of the handover for each patient and the total handover time.

Score the comprehensiveness of handover by the proportion of relevant data actually handedover.

218



Quality of handovers of patients inintensive care units by medical staffDr P McQuillan, Dr G Morgan, Dr S Redauceaneau

Why do thisaudit?


Suggestedindicators


This audit should be done to ensure that information about the condition and continuing careof the critically ill is transferred effectively and efficiently between the teams caring for them ona shift system. It is based on an audit carried out at the Department of Critical Care, QueenAlexandra Hospital, Portsmouth.1

Information transferred at handovers must be succinct, comprehensive and easily understood.The quality of information provided at handovers influences patient safety and is the basis forthe investigation and treatment of individual patients. It supports a forum for education andevidence-based practice. Time should be allocated within shift patterns for handovers whichshould take place at fixed times and locations, and with identified staff and clear leadership.

Handovers may be conducted at the bedside or in a quiet meeting room. Each location has itspros and cons.

All team members should be aware of the importance of effective handover systems.2,3

% key staff attending and reasons for non-attendance.

% of relevant information handed over.

% of handovers of whole unit completed within the designated time.

% of handovers beginning at appointed time.

% of handovers free from interruption.

100% of key staff present throughout handover.

100% of whole unit handovers completed within the designated time.

80% of handovers free from interruption.

100% of handovers commenced at appointed time.

100% of patients have ongoing plan for the day.

10.10

10

219

10.10




References

Interruptions of handovers for urgent patient needs.

Inadequate priority given to handover.

1 Longmore D, Morgan P, McQuillan PJ. Audit of the comprehensiveness and quality of handoverinformation in critical care. Department of Critical Care Medicine, Queen Alexandra Hospital,Portsmouth 2004 (unpublished data).

2 British Medical Association. Safe Hhandover: Safe patients. Guidance on clinical handover for cliniciansand managers. BMA, London 2004 (see: www.bma.org.uk/ap.nsf/content/handover).

3 Cheah L-P et al. Electronic medical handover: towards safer medical care. Med J Australia2005;118833((77)):369–372 (see: www.mja.com.au/public/issues/183_07_031005/che10928_fm.html).

Define key staff who should be present and any reasons for non-attendance

The auditor, who is not taking part in the handovers, should attend all handovers during a seriesof days and record:

nn incidence and relevance of completion of domains in information sheet (see example on theaccompanying CD-ROM)

nn duration of whole unit handovers

nn duration of individual patient handovers

nn number of interruptions per whole unit handover

nn number of late starts

nn number of late finishes

nn incidence of failure to complete information domains

nn number of patients with documented ongoing strategy.

Use a spreadsheet to record the data and hence calculate the % of transmitted domains foreach patient and thus all patients. It is easy to see which domains are handed over well andwhich domains are not.

http://www.mja.com.au/public/issues/183_07_031005/che10928_fm.html

http://www.bma.org.uk/ap.nsf/content/handover


This audit will be done by retrospective review of case notes.

For all patients admitted to ICU, gather the following information:

nn % of patients discharged from ICU who survive to hospital discharge.

For patients discharged from ICU who did not survive to hospital discharge, gather thefollowing information:

nn reason for ICU admission

nn diagnosis on ICU admission

nn age on ICU admission

nn APACHE II score as an indicator of severity of illness

nn maximum level of organ system support in ICU or level of dependency

nn length of stay in ICU

nn premature ICU discharge

nn discharged at night

220



Substandard care after dischargeas a factor in early mortalityDr P J Pridmore, Dr J R Sinclair, Dr A J Harvey

Why do thisaudit?


Suggestedindicators


The aim of this audit is to identify the incidence of substandard care delivered to patientsdischarged from intensive care units and its influence on mortality.

The most recent data from the Casemix Programme (2003–2004) of the UKs Intensive CareNational Audit & Research Centre (ICNARC) shows that 10.8% of ICU survivors die beforehospital discharge.1 There may be a number of reasons for this. Patient status before and duringintensive care influences post-ICU mortality.2 Severity of illness and age have been identified asindependent prognostic factors for mortality after discharge from ICU.3 In hospital locationsoutside intensive care, an adverse event rate of 10.8% has been identified among in-patients, ofwhich half may be preventable.4 Factors contributing to suboptimal care before admission tointensive care5 may continue to contribute to suboptimal care when the patient is discharged.A recent NCEPOD report on the acute care of medical patients identified, among other things,that 27% of hospitals did not have an early warning system and 44% of hospitals did not havean outreach service, both of which are considered important for the recognition of thepotentially critically ill.6

% of patients discharged from ICU who survive to discharge from hospital (casemix riskadjusted).

% of patients who suffer unexpected deterioration following discharge from the ICU.

% of patients whose deterioration was associated with substandard care.

Suggested elements of substandard care are listed in the section on data to be collected.

Fewer than 11% of patients discharged from ICU should die before discharge from hospital.

No patient should receive substandard care.

10.11

10

221

10.11




References

Lack of staff or facilities on general wards.

Pressure on ICU beds leading to early discharge.

Lack of a high dependency facility.

Lack of track and trigger system.

Lack of outreach service.

nn readmitted to ICU during same hospital admission

nn length of stay in hospital after discharge from ICU

nn cause of death

nn post-mortem carried out.

Evidence of substandard care in the days before death, e.g:

nn change in patient’s location without clinical indication

nn patient’s location inappropriate for dependency level

nn failure of attendants to respond to a clinical deterioration

nn frequency of recording clinical observations

nn use and completeness of Early Warning Scoring system on ward

nn frequency of attendance by medical staff

nn frequency of attendance by outreach team or re-referral to ICU

nn frequency of attendance by other professionals, e.g. physiotherapist, dietician

nn frequency of critical incidents

nn opinion of independent assessor(s) of standard of care.

1 Intensive Care National Audit and Research Centre. Case Mix Programme Database. ICNARCLondon (see: www.icnarc.org/).

2 Azoulay E et al. Determinants of post intensive care unit mortality: a prospective multi-centre study.Crit Care Med 2003:3311:428–432.

3 Ridley S et al. Long term survival after intensive care. Br Med J 1990;330011:1127–30.

4 Vincent C, Neale G,Woloshynowych M. Adverse events in British hospitals: preliminary retrospectiverecord review. Br Med J 2001:332222:517–519.

5 McQuillan P et al. Confidential inquiry into quality of care before admission to intensive care.Br Med J 1998;331166:1853–1858.

6 Cullinane M. An acute problem. NCEPOD May 2005 (see: www.ncepod.org.uk).

http://www.ncepod.org.uk

http://www.icnarc.org/

222



Readiness of patients for transferDr G Morgan

Why do thisaudit?


Suggestedindicators


Moving a patient from the intensive care unit to another location may be associated withcardiorespiratory destabilisation. Equipment used during transfer may not perform as well asthat in the intensive care unit and is prone to malfunction or failure unless regularly andfrequently checked. Personnel accompanying patients during transfer may have vaying levels ofexpertise. It may take an hour to prepare a patient for transfer to another department, even inthe same hospital.

The Intensive Care Society has published guidelines for the safe transport of the critically ill.1

The Advanced Life Support Group’s Safe transfer and retrieval2 provides a practical guide tomoving the critically ill from place to place. Both provide the basis for compilation of a check-list and in-transit documentation of physiological parameters so that transport of the critically illis undertaken in a safe environment by competent staff and with less likelihood of physiologicaldestabilisation or critical incident.

Provision of a check-list to ensure that all equipment is available and functioning.

Availability of training for staff undertaking transfers.

Level of monitoring during transfer.

Documentation of physiological variables during transfer.

Stability of patient on transfer equipment.

Readiness for transfer at designated departure time.

Communication between departure point and destination.

100% transfers with evidence of complete patient transfer document and data collection sheet.

100% transfers with evidence that equipment was checked and functional.

100% transfers include at least one accompanying person who has completed a transfer courseor other specific training.

100% of transfers include evidence of appropriate communication with receiving unit includinginformation on time of departure and estimated time of arrival.

100% patients are stabilised on transport ventilator 30 min before departure time.

100% patients have continuous monitoring of ECG, BP, oxygen saturation in arterial blood(SaO2),end tidal carbon dioxide (ETCO2) and temperature throughout transfer.

100% patients are ready for transfer at arranged departure time or reason for variance.

10.12

10

223

10.12





References

Patient transfer documentation.

Documentation on equipment checking.

Status, qualification and experience of accompanying personnel.

Ventilatory status of patient 30 min before departure time.

Blood gas analysis of patient 30 min before departure time.

Cardiovascular status of patient 30 min before departure time.

Monitoring of ECG, BP, SaO2, ETCO2 during transfer.

Level of communication between departure point and destination staff.

Inadequate time devoted to stabilisation of patient on transfer equipment.

Equipment failure or malfunction.

Inadequately trained staff.

Poor communication between departure point and destination.

1 The Intensive Care Society. Guidelines for the transport of the critically ill adult. ICS, London 2002(see: www.ics.ac.uk).

2 Advanced Life Support Group. Safe transfer and retrieval – The practical approach. BMJ Books,London 2002. (see: www.bmjbookshop.com).

http://www.bmjbookshop.com

http://www.ics.ac.uk

224



Training for inter- and intra-hospital transfersDr S J Tanser, Dr M Cordinly, Dr B Phypers

Why do thisaudit?


Suggestedindicators


Transfer of critically ill patients poses significant clinical risk. To minimise complications it isimportant that personnel involved in the transfer of patients should have received adequatetraining in the equipment used and potential complications.

The AAGBI’s Recommendations for the transfer of patients with acute head injuries to neurosurgical units1

recommends the patient should be: ‘accompanied by a doctor with at least 2 years ofexperience in an appropriate specialty (usually anaesthesia) ... familiar with ... the drugs andequipment they will use, working in the confines of an ambulance and have received supervisedtraining in the transfer of patients with head injuries’.

The Intensive Care Society’s guidelines2 state that the medical practitioner should:

nn have had appropriate training in intensive care medicine, anaesthesia or other acute specialty

nn have previous experience of transport in a supernumerary capacity

nn have demonstrated competencies in transport medicine

nn be familiar with the transport equipment.

The Intercollegiate Board for Training in Intensive Care Medicine also defines competenciesrequired, though does not specify a minimum acceptable time to be spent in any particularspecialty.3

% of medical staff having at least 2 years experience in an acute speciality prior to undertakingtheir first neurosurgical transfer.

% of staff having undergone specific training in the transfer of the critically ill.

% of staff having completed at least one supervised transfer prior to a solo transfer.

% of staff being able to demonstrate familiarity with transfer equipment used in their hospital.

100% of medical staff having at least 2 years experience in an acute speciality prior toundertaking their first neurosurgical transfer.

100% of staff having undergone specific training in the transfer of the critically ill.

100% of staff having completed at least one supervised transfer prior to a solo transfer.

100% of staff being able to demonstrate familiarity with transfer equipment used in theirhospital.

10.13

10

225

10.13





References

Time in acute speciality prior to first inter- and intra-hospital transfer.

Number of staff having undergone training in the transfer of the critically ill.

Number of transfers performed accompanied prior to first solo transfer.

Knowledge of ventilator used for transfers (battery life, gas consumption etc) and battery life ofpumps and monitor used locally.

Lack of training courses available locally.

Work pressures.

Lack of knowledge of standards set.

1 The Neuroanaesthesia Society of Great Britain and Ireland and The Association of Anaesthetists ofGreat Britain and Ireland. Recommendations for the transfer of patients with acute head injuries toneurosurgical units. NASGBI and AAGBI, London 1996 (see: www.aagbi.org/pdf/26doc.pdf).

2 Intensive Care Society. Guidelines for the transport of the critically ill adult. ICS, London 2002 (see:www.ics.ac.uk).

3 Intercollegiate Board for Training in Intensive Care Medicine. The CCST in Intensive Care Medicine:Competency based training and assessment; Part I: pp 23. IBTICM, London 2001 (see:www.rcoa.ac.uk/ibticm/docs/CBTPart1.pdf).

http://www.rcoa.ac.uk/ibticm/docs/CBTPart1.pdf

http://www.ics.ac.uk

http://www.aagbi.org/pdf/26doc.pdf

226



Evidence-based transfusionpractice in intensive care patientsDr T Walsh

Why do thisaudit?


Suggestedindicators

About 40% of all patients admitted to UK intensive care units receive blood transfusions. Ofthese, 20–40% are associated with bleeding and 60–80% occur in non-bleeding patients.1

Evidence from a randomised controlled trial suggests that restrictive use of blood transfusions isat least as safe as more liberal use, and is safer for less sick and younger patients.2 Accumulatingevidence points to possible harmful effects from blood transfusions in some patients so thattransfusions should only be given when clinically indicated.

Many non-randomised cohort studies have found associations between blood transfusions andincreased infection rates, prolonged ICU length of stay, and increased mortality among criticallyill patients.

From the literature cited below,1–4 in the following categories of patients, transfusion to a targetrange may be of benefit when the haemoglobin (Hb) falls to a trigger level:

Patients with established critical illness and no ischaemic heart disease (IHD):a transfusion trigger Hb of 70–80 g/l and a target range 70–90 g/l during ICU stay.

Patients with established critical illness and chronic IHD:a transfusion trigger Hb of 70–80 g/l (according to severity) and target range 70–90 g/l.

Patients with acute coronary syndrome (ACS):a transfusion trigger Hb of 80–90 g/l and target value > 90 g/l.

Patients with early sepsis and evidence of inadequate oxygen delivery (ScvO2 < 70%):a target Hb of 100 g/l during the early resuscitation phase.

Blood transfusions are given to treat bleeding in 20–40% of cases and to elevate Hb levels innon-bleeding cases in 60–80% of cases. During bleeding the Hb is an unreliable indicator of redcell mass and many factors influence the timing of transfusions. The Hb level may fluctuatesignificantly. For simplicity, therefore, we suggest that transfusion practice is audited in non-bleeding patients.

Method

Use some standard definitions, for example:

nn transfusion event: a prescription for a single or series of red cell units

nn transfusion trigger : the closest pre-transfusion Hb value to the start of the transfusion event

nn beeding: loss from any combination of sites of > 1 unit of red cells (300–400 ml) during the24 h before the transfusion

nn history of ischaemic heart disease: any record of angina, heart failure, or previous myocardialinfarction in the patient notes

nn post transfusion Hb: the next available Hb value.

Estimate the duration of the audit from the number of transfusion events to be audited.Typically, 40% of Level 3 patients receive transfusions in ICU. Expect about 80 transfusionevents per 100 ICU admissions, i.e. 50–60 transfusion events not associated with bleeding.1,5

Use an audit form to record information about each transfusion event (an example is availableon the accompanying CD-ROM).

10.14

10

227

10.14





Related audits

References

Number of patients admitted.

Number of patients transfused.

Number of patients transfused for bleeding.

Number of patients transfused not bleeding.

Number of patients with chronic IHD and/or ACS at the time of transfusion.

Number of patients undergoing early goal directed therapy for sepsis.

Number of transfusion events not associated with clinical bleeding per 100 admissions.

Hb transfusion trigger in non-bleeding patients.

Number of red cell units prescribed per transfusion event.

Post transfusion Hb.

Clinical uncertainty regarding the quality of the evidence.

Lack of consensus, especially for certain subgroups, e.g. IHD, head injury (higher triggers usuallyused), ‘slow weaning’ patient, COPD patient.

Traditional prescription of at least 2 red cell units per transfusion event.

2.9 – Safe red cell transfusion

1 Walsh TS et al. Red cell requirements for intensive care units adhering to evidence-based transfusionguidelines. Transfusion 2004;4444:1405–1411.

2 Hebert PC et al. A multicenter, randomized, controlled clinical trial of transfusion requirements incritical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care TrialsGroup. N Engl J Med 1999;334400:409–417.

3 Hebert PC, Fergusson DA. Do transfusions get to the heart of the matter? J Am Med Assoc2004;229922:1610–1612.

4 Rivers, E et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl JMed 2001;334455:1368–1377.

5 Chohan SS et al. Red cell transfusion practice following the transfusion requirements in critical care(TRICC) study: prospective observational cohort study in a large UK intensive care unit. Vox Sang2003;8844:211–218.


For non-bleeding patients with no IHD, excluding the ‘early sepsis group’:

nn 100% of Hb transfusion triggers should be ≤ 80 g/l

nn 100% of post transfusion Hb values should be ≤ 90g/l

nn 100% of transfusion prescriptions for patients with Hb > 70 g/l should be single unitsfollowed by reassessment before further prescription.

For non-bleeding patients with chronic IHD and/or an ACS the evidence to support bestpractice is less certain. A local guideline should be agreed to determine the cut off Hb valuefor audit. The values presented above represent the published consensus of experts familiarwith the available literature.

The auditor should be independent of the individuals making decisions about transfusions.

All patients should be assessed each day during the audit period and data extracted from thecharts.

Create a spreadsheet with each of the domains documented on the audit form.

228The Royal College of Anaesthetists nn Raising the Standard: a compendium of audit recipes

section 10: critical care services

Health & Medicine

level of care

care number of patients

intensive care beds

intensive care patients

critical care beds dr

intensive care units

critical care bed need

concept of intensive