Pathways in Aged Care PIAC

1
Pathways in Aged Care (PIAC)

• Phil AndersonCommunity Services Integration and
  Linkage Unit

2
Background

• NHMRC grant 2006 (622,000 over 3 yrs)
• From ACAP assessment to service outcomes Care
  Pathways for Older Australians with Dementia, CVD
  and Arthritis (D Gibson, AACU, CSILU, LaTrobe,
  QU)
• Method
• Identify service use by ACAP 0304 cohort by
  linking to aged care data sets and deaths,
• Longitudinal data analysis

3
Aged care system at a glance
4
Why look at pathways?

• What DO people use over their lifetime?
• What ARE the common pathways?
• What works? What doesnt?
• Does community care delay entry to RAC?
• Does respite care matter?
• What affects take-up of care?
• What effect do health conditions have?

5
What we linked
ACAP 03-04
200304
200405
200506
200203
HACC
VHC
CACP
RAC
ACAP 04-05
Deaths (NDI)
6
Data linkage steps linking Cohort (ACAP 0304)
to 8 program data sets and NDI

• RAC and CACP SLK-581, pc (April 2007)
• RAC/CACP to NDI SLK-581(s), pc(s), DOD
• Cohort to RAC/CACP/NDI SLK-581(s), pc(s), ACAT
  info
• Unlinked Cohort to unlinked NDI SLK-581, pc
• Cohort to ACAP 0405 SLK-581(s), pc(s)
• Cohort to HACC (4 years separately) SLK-581(s),
  pc(s)
• Cohort to VHC SLK-581(s), pc(s), DOD (June 2008)

7
The data linkage method

• Linkage data
• SLK-581 (5 letters of name, DOB, sex)
• Postcode(s)
• Other dates where available (assessment, death)
• ACAT ID
• Method stepwise deterministic matching
• Picks best matches first
• Allows some variation in SLK-581
• Stopping rules based on uniqueness of data
  combination

8
Linkage stopping rules

• Uniqueness gt 97.5 for both data sets
• Estimated false match rate lt 0.5
• Estimated trade-off of truefalse matches gt 2
  (compared with more detailed match

9
Linkage method example of data for stopping
rules (ACAPRAC)
10
Linkage method example of match results
(ACAPRAC)
11
The resource links to the cohort
12
The data linkage quality

• Matches using clients with a complete SLK-581
  gt97.5
• Matches made with supplementary data 310
• Matches to cases with duplicate key lt0.1
• Cohort member with link to inconsistent
  death lt0.2

13
The resource the view from the data
what was used
what is then used
ACAP
?
200506
200203
200304
200405
HACC
VHC
CACP
RAC
ACAP
Deaths
14
The first move to programs needing an ACAT
15
Case studies home?package?permanent RAC
16
Example 1 People with an approval for
residential respite care time to first use of
respite RAC
17
Example 2 are people having unnecessary
assessments?
18
Example 2 are people having unnecessary
assessments?
19
The PIAC team

• Chief InvestigatorsStephen Duckett (UQ), Diane
  Gibson (UC)
• Associate Investigators AIHW Phil Anderson,
  Peter Braun, John Goss, Rose Karmel, Ann
  PeutLaTrobe Yvonne Wells
• Research team members (AIHW)Evon Bowler, Joanna
  Forster-Jones