The rise and rise of chronic disease in Far North

Report
The rise and rise of chronic disease
in Far North Queensland
A new Centre for Chronic Disease Prevention
at JCU Cairns
Snapshot of past, current and future work
Robyn McDermott
MBBS, FAFPHM, MPH, PhD.
Director CCDP, JCU Cairns
CBH Grand Rounds Friday 28 March 2014.
Block “A” Lecture Theatre 12.15-1.30pm
Today
• Brief background and selected past and
current descriptive work in far north
Queensland
• Approach of the CCDP
• Interventions
• Where are we heading?
Some “political arithmetic of crowd disease” in Australia:
CVD Death rates, 2007-8
Source: AIHW 2011, Age-standardised deaths per 100,000
CVD hospitalisation rates, 2007-8
Source AIHW 2011: Age standardised hospitalisations per 100,000
Prevalence of diabetes, Indigenous NQ (WPHC)
and Australia (AusDiab),
1999-2000
60
50
40
Non-Indigenous
Aboriginal
Torres Strait Islander
30
20
10
0
15-24 25-34 35-44 45-54 55-64
65+
Age standardised rates for “ACS” avoidable
admissions by Queensland Health District, 2003-6
Source: QHAPDC, 2007 (rates per 100,000)
Ambulatory Care Sensitive (ACS) avoidable hospitalisations for
selected chronic diseases, Queensland, 1999-2006.
Source: QAPDC, 2007, rates per 100,000
Potentially Preventable Hospitalisations in SA
(2007-9) - Top 15
Adjusted incidence rate ratios for CHD events in FNQ
Aboriginal and TSI adults, 2000-7 (n=1706)
Source: McDermott et al, MJA, 2011
Measure
IRR
95% CI
Obesity
1.7
1.01-2.8
High BP (>140/90)
1.5
1.01-2.3
Smoking
1.4
0.9-2.2
Low HDL (<1.0mmol/l)
1.3
0.9-1.9
High TG (>=2.0 mmol/l)
1.9
1.3-2.7
IFG (FBG 5.5-6.9 mmol/l)
1.3
0.8-2.2
Diabetes (FBG >=7.0)
2.4
1.6-3.6
Micro-albuminuria
1.4
0.9-2.3
Macro-albuminuria
4.6
2.9-7.1
Glycemia and albuminuria, especially when
combined, predict much of the “gap” in CHD
incidence
• Baseline prevalence of high glycemia is >25%
• Baseline prevalence of albuminuria (>3.4 mmol/l) =
33.5%
• Those with diabetes at baseline were 5.5 (4.2-7.3)
times more likely to have albuminuria than those
without diabetes
• Adjusted CHD IRR for both diabetes and albuminuria
= 5.9 (3.4-10.1)
Risk accumulation along the care continuum
Low birth
weight
Maternal
diabetes in
pregnancy
Epigenetics
Adolescent
adiposity
Poor nutrition
Smoking
High BP
Lipids
Glycaemia
etc
Screening
and
Secondary
prevention
in primary
care
Death
Hospitalisation
for
complications
“pushback” – CCDP preventive approach
Rehab
Improve preventive systems for CD
management
Cluster Randomised Trial of HW-managed diabetes care
system improvement in the Torres Strait, 1999-2001.
Study Design and Patient Recruitment
21 eligible clinics
Baseline data collection
Random allocation
8 intervention sites
(250 patients: mean
age 52.1, SD 13.1 yr)
51 patients
added to regs
13 control sites
(305 patients: mean
age 52.4, SD 13.9 yr)
Diabetes outreach team
1. Diabetologist
2. Nutritionist
3. Diabetes healthcare
worker
4. Podiatrist
121 patients
added to regs
Intervention
 Recall and
reminder system
 Health worker
training
 Regular
phonecalls
 Newsletter
 Workshop
19 patients lost
to follow-up
Follow-up data collection (282 patients)
30 patients lost to
follow-up
Follow-up data collection (396 patients)
Hospitalisation of people with diabetes, Torres Strait, 19992002 (n=921), Cape York 2002-3 (n=240):
Proportion of diabetics hospitalised for avoidable conditions in previous 12 months
30
25
20
DR HospTorres
Other hosp Torres
DR Hosp Cape
Other Hosp Cape
15
10
5
0
1999
2000
2002
2003
Can improved care processes be sustained with
rising caseloads and current workforce
configuration?
Snapshot from Island in the Central Group Torres, 2009.
Incident cases 3%, younger ages, increasing obesity
Source: Forbes et al, 2012.
Measure
2004, n=34
2009, n=67
ANDIAB 2009
Age
54
52.4
56.8
Median HbA1c
9.35
9.53
8.0
Current smokers (%)
29%
30%
10%
% “good” glycemic control
(A1c<7%)
16
20
26
% taking insulin
16%
32%
35%
% without albuminuria
25%
33%
67%
Mean weight, kg (BMI)
96.14 (34.7)
101.74 (35.9)
N/A (30.2)
Getting Better at Chronic Care (GBACC)
in North Queensland: a cluster RCT of
community health worker care coordination in remote FNQ settings
Robyn McDermott, Barbara Schmidt, Vickie Owens, Cilla
Preece,
Sean Taylor, Adrian Esterman
“Getting better at chronic care”
Cluster RCT of health-worker led case management
for high risk clients
Aim: Test if HW-led care for high risk poorly managed
adults with complicated T2DM would improve care
processes (checks, referrals, self management) and
outcomes
Primary outcome: improved HbA1c
Secondary outcomes: Improved QoL, reduced CVD risk
factors and complications (avoidable hospitalisations)
Mixed methods evaluation in 3 phases
NHMRC Partnership Project, 2011-2015
GBACC: mixed methods evaluation in 3
phases
Phase 1 (Intervention period: March 2012 – Sept
2013)
• Randomised controlled trial of intensive case management by
IHWs
Phase 2 (Nov 2013 – Feb 2014)
• Review of lessons learned
• Implementation plan
Phase 3 (May 2014 – June 2015)
• Economic analysis
• Rollout of model
12 Participating Communities
*Intervention sites in phase 1 (randomly allocated)
Torres and NPA HHS
• Badu*
• Bamaga
• Injinoo*
• New Mapoon
• Seisia
• Umagico*
Cape York HHS
• Kowanyama*
• Mapoon*
• Mareeba (Mulungu)
Cairns and Hinterland HHS
• Mossman Gorge (ACYHC)*
• Napranum
• Yarrabah (GYHS)
PHASE 1:
COCONSORT DIAGRAM: GBACC, 2012-14,
2012-14RCT)
Enrolment: 12 sites recruited and 327 patients assessed as eligible
Baseline data collected, n=213
Excluded: 114 patients declined to participate
Group randomisation: 12 sites
Intervention: 6 sites
(n=100 patients)
Received intervention, n=100
Lost to follow-up (n=16)
•
•
Allocated to waitlist group: 6 sites
(n=113 patients)
Allocation
Follow up
Lost to follow up (n=6)
•
•
•
Moved away (12)
Died (4)
Moved away (3)
Died (2)
Withdrew from study (1)
Analysis
Analysed for primary outcome,
n= 84 (84%)
Analysed for primary outcome,
n=108 (96%)
Clinical care processes at baseline and follow up (%)
Baseline
Foot check%
Endpoint (excluding 22 loss of follow up)
Control n=113
Intervention n=100
Control n=107
intervention n=84
No
No
% (95% CI)
No
% (95% CI)
No
% (95% CI)
31.0 (21.8-40.2)
38
35.5 (26.3-44.7)
26
31.0 (20.9-41.0)
50
% (95% CI)
44.2 (35.0-53.5) 31
Seen by DM educator 46
%
Seen by dietician %
22
40.7 (31.6-49.9)
52
52.0 (42.1-61.9)
41
38.3 (29.0-47.6)
44
52.4 (41.6-63.2)
19.5 (12.1-26.8)
30
30.0 (20.9-39.1)
21
19.6 (12.0-27.2)
37
44.0 (33.3-54.8)
Dentist check %
20
17.7 (10.6-24.8)
13
13.0 (6.3-19.7)
9
8.4 (3.1-13.7)
15
17.9 (9.6-26.5)
ECG check%
37
32.7 (24.0-41.5)
42
42.0 (32.2-51.8)
34
43.9 (34.4-53.4)
35
40.5 (29.8-51.1)
Eye check %
54
47.8 (38.5-57.1)
42
42.0 (32.2-51.8)
56
52.3 (42.8-61.9)
37
44.0 (33.3-54.8)
Smoker %
38
34.5 (25.6-43.5)
34
35.1 (25.5-44.7)
33
31.2 (22.4-40.4)
34
41.5 (30.7-52.2)
Blood sugar selfmonitor %
45
40.9 (31.6-50.2)
46
46.0 (36.1-55.9)
63
59.4 (50.0-68.9)
44
52.4 (41.6-63.2)
Taking insulin%
55
48.7 (39.4-58.0)
40
40.0 (30.3-49.7)
47
43.9 (34.4-53.4)
40
47.6 (36.8-58.4)
Dyslipidemia %
83
73.5 (65.2-81.7)
84
84.0 (76.7-91.3)
91
85.0 (78.2-91.9)
76
90.5 (84.1-96.8)
Taking lipid lowering
medicines%
5
4.4 (0.6-8.3)
3
3.0 (-0.4-6.4)
3
2.8 (-0.4-6.0)
5
6.0 (0.8-11.1)
Albuminuria and
taking ACEi or ARB
drugs
Adherent to all
medicines
Had Fluvax
46
88.5 (79.6-97.3)
47
88.7 (80.0-97.4)
58
82.9 (73.9-91.8)
51
89.5 (81.4-97.6)
53
46.9 (37.6-56.2)
55
55.0 (45.1-64.9)
57
53.3 (43.7-62.8)
41
48.8 (38.0-59.6)
50
44.2 (35.0-53.5)
66
66.0 (56.6-75.4)
51
47.7 (38.1-57.2)
50
59.5 (48.9-70.2)
HbA1c measures at baseline and follow-up by group, absolute
values: GBACC Phase 1 trial results
11.5
11
10.5
Control
10
Intervention
9.5
9
8.5
Baseline
Endpoint
FNQ Hospital Avoidance Trial
Cairns, Innisfail, Mareeba
2014-16
Health Innovation Fund Project
Overview
Funded by QH (CARU)
Neil Beaton, Mary Streatfield, Robyn McDermott
Aim: to evaluate a new approach to community-based
management of “frequent flyers” in FNQ hospitals –
Hospital Avoidance Trial, 2013-16
Background: Pilot HAP in Cairns showed a dramatic reduction in ED and inpatient
episodes in 68 frequent flyers using a nurse-led case management approach.
• Pragmatic RCT of intensive community-based case management of frequently
hospitalised adults with chronic conditions in 3 CHHHS sites
• 530 patients in 3 sites randomly assigned to
• 265 Intervention: usual care plus shared electronic record including CDM tool,
close case management (caseload for each care co-ord =<40) and selfmanagement training and support
• 265 “controls”: usual care (referral to a medical home with offer of shared
record)
• Eligibility criteria: 8 or more ED/inpatient episodes in the previous 12 months
• Evaluation endpoints: Avoidable ED visits or hospital admissions over 18
months, care processes (GPMP, referrals, self management training),
intermediate clinical indicators (HbA1c, BP, Lipids, UACR/eGFR), disease
progression, quality of life
• Economic (DRGs and AQoL) and process evaluation
2012-13 FY ED and Separations
(patients)
Number of Visits
Cairns
>=5
>=8
ED
Inpatient
Total
1,105
543
2,979
324
187
1,006
ED
Inpatient
Total
751
122
1,077
235
40
352
ED
Inpatient
Total
369
95
682
104
32
234
ED
Inpatient
Total
2,225
760
4,738
663
259
1,592
Mareeba
Innisfail
Total of three sites
FNQ HAT Trial design
Patient recruitment 3 sites, n=530
Baseline interviews + data collection
Randomisation
Control group: n=265
Usual best practice care
GPMP, cdmNet audit & feedback
Intervention group: n=265
GPMP, cdm tool audit & feedback
+ Case manager
Process evaluation
including fidelity of
implementation
Follow up data collection:
Interviews, ED & inpatient episodes
Cdm tool audit, HIC/PBS, costings
Follow up data collection:
Interviews, ED & inpatient episodes
Cdm tool audit, HIC/PBS, costings
The patient journey, FNQ HAT
Patient identified as
eligible by EDIS/HBCISand
invited to participate in the
trial
Consent not obtained
Not in trial, usual care
Consent obtained
Care co-ordinator conducts
baseline assessment and interview,
arranges GP referral and GP
consent to be in trial
Randomisation
Usual care group:
Offer of shared record,
Referrals to AHPs
Data capture
and QI reports
to GPs from
ED/IP and CDM
tool
Intervention group:
GPMP, referrals, CDM tool, Care coordination, self management training
and support
Self management
training
GPMP and
referrals, care
co-ordinator
Allied health and
medical specialists
Other services as
required
Hospital admissions and ED visits
Why a Randomised Controlled Trial Design?
•
RCT is the most robust study design which will give the highest level of evidence:
all previous published studies looking at hospital avoidance (a complex
intervention in a complex environment) were uncontrolled before-and-after
designs – weak evidence for policy change and unable to be properly evaluated
economically
•
Controls provide the counterfactual for robust clinical and economic analysis
•
Randomisation deals with selection/allocation bias
•
Controls deal with secular trends in exposures and outcomes, regression to the
mean and changes in the policy and fiscal environment.
•
Good pilot data gives a clear effect size so a robust power calculation (sample size)
will ensure the question can be clearly answered without (too much) statistical
error
•
Will be publishable and in the public domain, not sit on the shelf
•
High scientific quality will be competitive for matching NHMRC Partnership Project
Grant funding
Expanding the impact of our research
Source: Duryea, Hochman, Parfitt. Research Global: Feb 2007.
Traditional
quality domain
Research impact scope
Research
outputs:
Research
Transfer:
Research
Outcomes:
Research
Impact:
eg
Discoveries
Publications
Patents
Engagement
with end
users
New
products
or services
Value
added,
Improvements
achieved
National
benefits
Association between PHC resourcing (staff) and costs of
hospitalisation among diabetics in FNQ remote communities, 2001-5
(Gibson, Segal, McDermott 2011)
4000
Diabetes-related hospital admissions> and PHC Staffing^
(All FTE staffing levels)
3000
13
2000
10
2
21
5
4
1
1000
11
8
12
15
20
6
14
17 7
0
18
.005
.01
.015
.02
Mean phc staffing per person
Communities
Fitted values
>Jan2001-Dec2005.^2003/04-Dec2005.Source:QldHealth. C-Coeff: -0.6862*(0.05sig)
.025
ACKNOWLEDGEMENTS
The CCDP is supported by QH Senior Clinical Research Fellowship and the Australian Primary
Health Care Research Institute (APHCRI) as a PHC Centre for Research Excellence (CRE)
GBACC is supported by NHMRC Partnership project grant 570149
FNQ HAT is funded by QH (CARU)
CCDP and CRE team includes:
Admin: Jacqui Lavis and Sally McDonald
Clinical Epidemiology: Sandy Campbell*, Robyn McDermott*, Klaus Gebel, Linton Harriss
Biostatistics/informatics: Haider Mannan, Arindam Dey
Community-based prevention studies group: Alan Clough*, Caryn West*
PhD students: Ashleigh Sushames, Sean Taylor*, Barb Schmidt, Jan Robertson, Dympna
Leonard, Russell Hayes, Richard Turner, Malcolm Forbes* (Masters)
Health Economics: Kenny Lawson
Clinical Research Associates: Vickie Owens, Cilla Preece
Collaborating institutions: QH, UniSA, SAHMRI, UQ, Melbourne University, Baker-IDI, Menzies
School of Health Research, Apunipima CYHC, Gurinny, Mulungu, AHCSA, QAIHC, UNSW
*Receiving NHMRC or NHF Fellowship support

similar documents