Transfusion Outcomes Research Collaborative

1
Transfusion Outcomes Research Collaborative
2
Transfusion Outcomes Research Collaborative

• A partnership between Monash University and the
  Australian Red Cross Blood Service
• draws on
• Monash DEPM experience in research
• ARCBS clinical and laboratory expertise and
  extensive networks
• allows for the addition of new collaborators
• incorporates a variety of projects

3
Current Projects
4
Registries

• Have outcomes data
• Collect variables to allow risk adjustment
• eg. age, gender, acuity, co-morbidities, etc
• But they dont collect information on transfusion
  (or enough info)
• We want to add the transfusion information so
  that we can see what is the effect of different
  transfusion practices on outcomes

5
How do we add transfusion information?

• Ask the registries to collect more information
• We know this is difficult
• It has been tried before
• It would be expensive
• Alternatives?
• Collect this information for a limited time
• Collect the information in a different way

6
Transfusion Information

• Where else can we get transfusion information?
• Hospital blood banks have similar information
• Who a unit is issued to
• When it is issued
• Is this close enough?
• If this is consistent enough it could be used as
  a surrogate measure of transfusion

7
Time to Transfusion Study

• To determine the length of time between blood
  product release from hospital blood bank to the
  initiation of transfusion
• two hospitals
• St Vincents Hospital, Melbourne
• Alfred Hospital, Melbourne
• Information on 24 hours transfusion, 7 days of
  the week

8
Time to Transfusion Study

• Every unit was followed up in the patient notes
• Information Captured when matching transfusion
  located in patient notes
• Time of transfusion initiation
• Location

9
Results (1)
10
Conclusion (1)

• If the hospital blood bank says a patient
  received a blood product is this true?
• Yes (with 596 events 95 CI 0-0.006)
• What about the reverse? Is it possible that a
  patient received a blood product that is not
  recorded in the hospital blood bank?
• Not here
• But if there were units unaccounted for......

11
Results (2)
12
(No Transcript)
13
Time to Transfusion
14
Hospital
No significant difference between the two
hospitals
15
Day of the Week
16
Weekend versus Weekday
17
Shift
18
Location
19
Conclusion (2)

• Can we use hospital blood bank data to estimate
  the time of transfusion?
• Yes approximate issue time 20 minutes
• How accurate is this estimate?
• probably not very

20
Conclusion (2)

• Time of transfusion written in patient medical
  records is patently unreliable
• To improve estimates of actual transfusion time
  would require electronic capture of data at
  bedside
• In the meantime, time of issue (20 min) is
  likely the best estimate of time of transfusion

21
Conclusion (final)

• The hospital blood bank computer records are
  appropriate sources of blood product information
  for linking with other data sources
• It is probably as accurate (if not more so) than
  what is written in the patient notes and is
  certainly more complete

22
ASCTS Phase One

• Data being collected about transfusion
• Number of RBC, FFP, Cryo, Platelet Units
• Only intra and post-operative
• No laboratory results
• Some relevant adverse events
• Numbers of patients transfused
• Amount of transfusion being received
• Outcomes of transfused vs non-transfused patients
• Variation across institution

23
ASCTS

• Data from 6 Victorian Hospitals
• Alfred, Austin, Geelong, Monash Medical Centre,
  Royal Melb, St Vincents
• Analysis of cases from 2005 on
• this was when transfusion information began to be
  collected
• 9363 patients

24
Hospitals
25
Transfusion Variation all patients
cases
26
Surgery Type
27
Transfusion Variation CABG only
of cases
28
ASCTS Phase Two

• Link with further information
• Pre-op transfusion
• Laboratory results
• Relevant medications
• Look at influence of pre-op medications eg
  clopidogrel
• Triggers for transfusion (eg haemoglobin or
  platelet level)

29
Current Projects
30
What are the issues?

• TTP NAIT are uncommon diseases
• even major centres encounter few cases
• associated with significant morbidity and
  mortality
• difficult to establish optimal management
• significant variation in practice
• disease rarity hampers definitive clinical trials
• resource implications
• incl provision of specialised blood products

31
NAIT Questions

• role of screening
• optimal antenatal management
• role of pre-emptive management in anticipated
  cases
• role of adjunctive agents
• diagnostic strategies
• long term outcomes

32
TTP Questions

• role of diagnostic tests
• different clinical settings
• pregnancy, BMT, HIV
• optimal schedules ofplasma exchange
• role of adjunctive agents
• tapering therapy
• long term outcomes

33
Why a Registry?

• In both TTP and NAIT, the number of cases and the
  heterogeneity of individual circumstances make
  the design and conduct of clinical trials
  difficult.
• Registries offer a methodology that allows
  aggregation of the full range of community
  experience and comparison of treatment variation
  and outcome.

34
Aims of the Registries

• to better define NAIT TTP in Australia
• incidence, natural history and clinical outcome
• to provide information on the range of
  therapeutic strategies currently being employed
  in treatment
• to explore factors influencing clinical outcomes
• to better define optimal management
• to inform and inspire future research

35
Recruitment Enrolment

• patients will be identified either
• by the treating clinician at the participating
  hospital, or
• by ARCBS clinicians as a result of referral for
  specialised testing or provision of blood
  components or intravenous immunoglobulin for
  therapy.
• patient liaison and data collection will take
  place in participating hospitals primarily
  through the treating clinical team

36
Current Projects
37
Pandemic Project

• Looking at how to plan for the circumstances (eg
  pandemic or sustained natural disaster) where
  ARCBS inventory is compromised
• How could we reduce demand for blood?
• How would we ration blood?
• Where is blood currently used?
• What would happen to this patient if this
  transfusion was not available?
• Working with ID specialists and Modelling
  experts as well as ARCBS

38
Human Factors

• Under the lead of Dr Shelly Jeffcott
• Human factors is a broad science that deals with
  many factors influencing human performance
• e.g. physical environment, task characteristics,
  individual characteristics and organizational or
  management systems
• i.e. factors that affect how people perform,
  think, communicate, and interact with technology
  in complex socio-technical systems

39
WBITs

• Thanks to programs such as STIR we know when they
  occur (at least some of them)
• But we have limited understanding of the
  complexity of events behind WBITs
• We know about a variety of strategies undertaken
  to limit WBITS
• But we dont know which strategies work and why
  (or why not)

40
TORC

• Transfusion Outcomes Research
  Collaborative
• We are interested in what you are doing
• We are there to help if you are interested in
  research projects
• We would like and need you to be involved

41
Contact

• Louise Phillips
• 9903 0051
• 0417 722 449
• louise.phillips_at_med.monash.edu.au