Transcript BUPA Presentation Template

Bupa Health Dialog
Our Experience of how English
NHS bodies are using The
Combined Predictive Model
PRISM Learning Workshop
24th February 2010
Ian Manovel, Associate Director of Analytics
07826 557040
[email protected]
Combined Predictive Model
Combination of datasets = (IP, OP, A&E + GP)
Case
Management
NHS
0.5%
Disease
Management
4.5%
Supported
Self Care
15%
Prevention &
Wellness Promotion
80%
Bupa Health Dialog
2
Clinical Commissioning Analytics Model
Integrated Care Manager
Shared Decision Making
Telehealth
Virtual Wards
Clinical & Analytic Lens
Commissioning
Organisation
• Vision
• Objectives
• Deliverables
Case Interventions
Intelligent
Commissioning
Implementation
• Data driven
• Forward looking
• Evidence-based
• Locally informed
• Integrated
• Demand & Supply
Focussed Interventions
Service Design
Commissioning OA
Clinical Claims Mgt
Benchmarking
Root Cause Analysis
Pathway Redesign
Bupa Health Dialog
3
Evaluation
• ROI
•PROMS
•Randomised Controlled
Trial
• Benchmarking
Examples of three NHS bodies that deployed the
Combined Predictive Model and tools
West Midlands SHA
Lambeth PCT
17 PCTs
2 years experience
2 years experience
5.4 million people
All 50 GP practices
About half of GP
Practices
Some of 990 practices
Case Finding
Information Governance
Challenges
Case Finding
Service Design
Case Finding
Service Design
Virtual Ward
Health Coaching
Bupa Health Dialog
4
West Kent PCT
Gaps in care for Long Term Condition patients
Most NHS bodies target LTC patients due to high cost and QOF targets
Bupa Health Dialog
5
Gap rates as of Sept 2009
20%
10%
0%
Bupa Health Dialog
6
No microalbuminuria
test (no
nephropathy)
No Lipid Rx (with
CHD)*
Not more than one
HbA1C test
No HbA1C test
No Eye exam
min
Most recent
cholesterol level not
below 5 mmol/l
overall rate
No cholesterol
check
No ACEI/ARB Rx
(with proteinuria or
microalbuminuria)*
No smoking
assessment
No smoking advice
Most recent HbA1C
value not below
10.0
Most recent HbA1C
value not below 7.4
Most recent blood
pressure not below
145/85
No blood pressure
check
Diabetes gaps in care – PCT view
Improve care for diabetics by focusing on HbA1c
PCT Diabetes effective care measures
max
100%
90%
80%
70%
60%
50%
40%
30%
Diabetes gaps in care – Practice based Commissioning
The cluster with the highest average risk score for diabetics has the lowest gap
Diabetes most recent HbA1C test not below 7.4
% with gap
max
average risk score of Diabetes patients
60%
0.14
55%
0.135
50%
0.13
45%
0.125
40%
0.12
35%
0.115
30%
0.11
25%
0.105
20%
0.1
A
Bupa Health Dialog
C
B
7
D
average risk score
min
overall rate
Medicines Management
Practice variation in % of high risk patients prescribed 10 or more drugs per
month. Polypharmacy associated with adverse health outcomes.
Top 5 Percent Risk - Polypharmacy 10+
Polypharmacy 10+ rate
average risk score for Top 5%
25%
0.4
0.35
20%
0.25
15%
0.2
10%
0.15
0.1
5%
0.05
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
0%
practice
Bupa Health Dialog
8
Average Risk Score
% of patients
0.3
Dementia Polypharmacy
Increased risk of incorrect medicines usage and adverse health outcomes including
emergency admissions.
Dementia - Polypharmacy 5+
120%
0.6
100%
0.5
80%
0.4
60%
0.3
40%
0.2
20%
0.1
0%
0
practice
Bupa Health Dialog
9
Average Risk Score
% of dementia patients
Polypharmacy 5+
AVG(RISK_SCORE)
“London Cosmopolitan Cluster” PCT benchmarking
Variation in rate and cost of OP attendances compared to peers
Outpatient Attendances for Trauma
& Orthopaedics 2008/09
Attendances Per 1000
Average Cost per Patient Attended
160
£350
140
£300
120
£250
Average
attendance
£200 per 1000 (98)
Cost
Attendances
100
80
£150
60
£100
40
£50
20
0
£0
LAMBETH PCT
BRENT TEACHING CITY AND HACKNEY
PCT
TEACHING PCT
HARINGEY
TEACHING PCT
HEART OF
BIRMINGHAM
TEACHING PCT
LEWISHAM PCT
NEWHAM PCT
A cluster is a group of geographic areas (Supergroup) where the local populations have similar characteristics. This information
is derived from census data used by the Office of National Statistics.
Bupa Health Dialog
10
SOUTHWARK PCT
Variation in OP attendances – Trauma & Orthopaedics
PCT has very high variation in referral rates among practices
Bupa Health Dialog
11
Variation in OP attendances – Trauma & Orthopaedics
Higher than expected rate and cost associated with greater than three
attendances.
Attendances and Costs for Trauma and Orthopaedics
12000
£1,200,000
10000
£1,000,000
8000
£800,000
6000
£600,000
4000
£400,000
2000
£200,000
0
Total Attendances
Total OP Tariff
Only One Attendance
3723
£604,539
Bupa Health Dialog
Up to 3 Attendances
8903
£1,062,316
12
More Than 3 Attendances
11055
£1,080,507
£0
Next year’s predicted cost analysis
Top 5% highest risk patients this year are responsible for 32% of predicted year
PbR costs and 49% of emergency admission costs
Bupa Health Dialog
13
Shared Decision Making
Increased opportunity for intervention and impact
Other
Single LTC
PSC Only
Single LTC + PSC
Comorbid LTC
100%
90%
N = 20,130
16% of
patients
N = 3,593
N = 3,655
6.9 x Average
4.2 x Average
N = 19,150
1.8 x Average
80%
70%
2.5 x Average
60%
50%
N = 249,241
40%
0.7 x Average
30%
20%
10%
0%
% of Total Patients
Bupa Health Dialog
% of Total Cost
14
42% of next
year’s cost or
£40.3M
Extended Definitions
The extended diabetes definition
Patients with a diagnosis of diabetes, complications of diabetes,
procedures related to diabetes or on medication for diabetes using
codes in GP (Read code) and Acute setting (ICD-10 codes) in IP data
Including :
•
Those who currently take insulin or oral anti-diabetic medication
•
Those with diet controlled diabetes
Excluding :
•
Those with gestational or steroid-induced diabetes within the last 12 months
•
Those whose GP’s last Read code indicates diabetes resolved or diabetes excluded
•
Those on metformin alone without insulin or other anti-diabetic medications
Bupa Health Dialog
15
Extended Definitions
The extended diabetes definition outcome
GP identified
11,253
=
Diabetes
population
11,815
Extended
definition
562
Bupa Health Dialog
16
Analysis of PCT population
(344,168) saw an increase of 575
patients
5% increase from the GP registry
method currently used.
Extended Definition: Current and Predicted Costs
Increased opportunity for intervention and savings
The extended definition of diabetes identifies
additional patients who have a higher initial cost
and have greater cost increases. We see £647 in
additional total costs in the year after
identification (36% increase).
The cost in the year AFTER standard identification
shows an increase of £234 per person in average
annual costs (26% increase).
Current Diabetes QOF Definition
Extended Diabetes Definition
Ave rage Cost Pe r Pat ie nt
£1,147
£1,200
£1,000
( Excluding Curre nt De finit ion)
£3,000
£1,400
£913
£800
£2,000
£547
£600
£599
£400
£200
Y 1: Pre-Identification
Bupa Health Dialog
£1,030
£498
£500
£9
£7
£37
£0
£0
A ve. Out-Patient
Cos t
£1,284 £1,369
£1,000
£2
A ve. In-patient
Cos t
Ave rage Cost Pe r Pat ie nt
£1,789
£1,500
£538
£365
A ve. Total Cos t
£2,436
£2,500
A ve. Total Cos t
A ve. A &E Cos t
A ve. In-patient Cos t A ve. Out-Patient
Cos t
Y 1: Pre-Identification
Y 2: Pos t-Identification
17
Y 2: Pos t-Identification
A ve. A &E Cos t
A whole person approach to LTC management is critical:
For the 3,180 diabetics at the end of Year 1, only 9% of admissions and 12% of
costs in Year 2 were directly or indirectly related to diabetes
Admissions for diabetics
Diabetes Related
Complications of Diabetes
Other Conditions
100%
90%
80%
70%
60%
83.5%
87.5%
50%
40%
30%
20%
10%
5.7%
11.1%
6.7%
5.4%
% Admits
% Total Cost
0%
Bupa Health Dialog
18
Community Service Redesign ROI
Top 5% high risk and very high risk segments predicted cost £40m
Assuming a 20% reduction in acute services £8m budget for community service
Bupa Health Dialog
19
Bupa Health Dialog
We still have a long way to go
to design patient centered
services that are cost effective
and deliver the same quality of
care but at least we have the
evidence base to make
informed decisions.
Ian Manovel, Associate Director of Analytics
07826 557040
[email protected]