The Impact of Co-morbidity
ACC Service Utilisation & Costs
2012 - 2025
Dr Barry Gribben
CBG Health Research
Dr John Wren
Principal Research Advisor
ACC
Lauren Prosser
Senior Policy Advisor
ACC
2nd ACHRF
Auckland, New Zealand
8 November 2012
The Questions
 WHAT is the effect of a health co-morbidity on ACC clients ?
˃ injury treatment claim rates (utilisation)
˃ duration of claim
˃ costs over time
 WHAT are the cost effects on an aging population ?
 WHAT are the policy implications ?
The Process
 BUILT on the pilot studies reported in 2010 (Wren & Mason)
 LINKED Primary Health Care data (GP Practice) with
Ministry of Health & ACC data using New Zealand NHI
˃ Random sample of 337,665 people
˃ Sample representative of the New Zealand population
 Descriptive & Multivariate Statistical Analysis
Co-morbidities
 Asthma
 Chronic obstructive pulmonary disease
 Ischaemic heart disease
 Heart failure
 Diabetes mellitus
 Mental health condition
 Cancer diagnosis
 Osteoarthritis
 Hypertension
Variables & Interactions
 Age
 Sex
 Ethnicity
 Socio-economic status (New Zealand social deprivation index)
 Treatment utilisation
 Claims duration
 ACC Costs
Headline Statistical Results
Multivariate Model Analysis
 Presence of a health co-morbidity was found to have a strong
statistically significant (95%) association with:
˃ increased service utilisation
˃ higher costs
 The effects were independent of, and additional to, normal
health cost effects typically associated with age, gender,
ethnicity & socio-economic status
Headline Statistical Results
Presence of one or more health co-morbidities showed …
˃ 28% more claims
˃ 346% higher lump sum payments
˃ 59% higher medical treatment costs
˃ 39% more weekly compensation costs
OVERALL 59% more total ACC cash costs
across all cost categories
Cost relativity
… stronger relativity for some than others
Average total cost per person per year (95% CI)
Average total cost per annum
NZD
1400
1200
1000
800
600
400
200
0
Yes
No
Osteo-arthritis
Yes
No
Stroke
Yes
No
Hyper-tension
Yes
No
Mental Health
Yes
No
IHD
Yes
No
Heart Failure
Yes
Cancer
No
Yes
Asthma
No
Yes
No
Diabetes
Claim Utilisation
Claims Utilisation vs. Number of Co-morbidities
2.50
2.00
Claims
per
annum
1.50
1.00
0.50
0.00
0
1
2
3
4
Number of co-morbidities present
5
6
7
Medical Treatment Cost
Medical Treatment Costs vs. Number of Co-morbidities
$500
$450
$400
$350
Medical $300
Treatment
$250
Cost
$200
$150
$100
$50
$0
1
2
3
4
Number of co-morbidities present
5
6
7
Total Annual Cost
Total Annual Cost vs. Number of Co-morbidities
$3,000
$2,500
Total
Annual
Cost $2,000
$1,500
$1,000
$500
$-
0
1
2
3
4
Number of co-morbidities present
5
6
7
Based on the analysis
10.7%
Total Annual ACC Cost is
attributable to
presence of co-morbidities in the
New Zealand population
276 million
$
(NZD, 2011)
Ageing Effects
Mean $ NO co-morbidity
Average total cost per person
Age group co-morbidity vs. no co-morbidity
Mean $ co-morbidity
2000
Mean cost per annum per person
1800
1600
1400
1200
1000
800
600
400
200
Excess cost is the area
of the gap between the
two lines – largest gap is
in the working age
population
0
Age group
Modelling Ageing Effects to 2025
Population by age group
with co-morbidity vs no co-morbidity
160,000
Without co-morbidity
140,000
With co-morbidity
100,000
80,000
Expect this area to
get bigger
60,000
40,000
20,000
Age group
+
90
5-
9
10
-1
4
15
-1
9
20
-2
4
25
-2
9
30
-3
4
35
-3
9
40
-4
4
45
-4
9
50
-5
4
55
-5
9
60
-6
4
65
-6
9
70
-7
4
75
-7
9
80
-8
4
85
-9
0
0
04
Number of people
120,000
Costs Attributable to Co-morbidities
14.0%
Percentage of total ACC costs
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
no change
1% growth
2% growth
2012
10.66%
10.71%
10.77%
2015
10.67%
10.83%
11.00%
2020
10.96%
11.43%
11.91%
2025
11.25%
12.06%
12.72%
Conclusions
 Presence of a wide range of health co-morbidities in the
population has a real effect on injury compensation
treatment utilisation volumes and costs
 To 2025, 10.7% to 12.7% of total annual ACC costs is
estimated to be attributable to presence of co-morbidities in
the population
 It appears that although aging of the population means more
people have co-morbidities, this is counterbalanced by
relatively fewer people being in the age groups where the
cost differences are greater
 Results are consistent with recent research from NCCI in
America about working age effects
Policy response
 Confirms and quantifies our assumptions – ACC appears to be paying
more than required to meet a person’s injury-related needs
 But how big is the problem?
 Within current legislative constraints? ACC is liable for injury costs unless an unrelated comorbidity is ‘wholly or substantially’ the cause of the person’s ongoing incapacity
 Where does the cost burden fall in the ACC Accounts?
 What is the impact on liability?
 Policy questions:
 How can ACC be smarter at managing the additional costs associated with co-morbidities?
 Should the costs be shared with others, eg individuals or other agencies?
 How can ACC continue to deliver a client-centred service?
 No silver bullet – distinguishing injury and non-injury related needs is
difficult, particularly in the context of different funding systems and
philosophies
Working through the policy issues
and options
Potential responses could include:
 cost-sharing arrangements
 targeted risk and claims management
 integrated assessment and services
 injury prevention initiatives
 reviewing assessment of individual entitlements
 status quo?
For further information
 Gribben, B. & Wren, J. ( 2012) The Impact of Health
Comorbidities on ACC Injury Treatment and Rehabilitation
Utilisation and Costs, and cost estimate to 2025 of effects in
an aging population.
CBG Health Research and ACC Research, Sep 2012.
 Wren, J. & Mason, J. 2010. Results of Three Pilot Studies
Exploring & Quantifying Health Co-morbidity Effects on ACC
Injury Treatment Utilisation and Costs.
ACC.
Additional Slides
Health Literature
 Pre-existing health co-morbidity effects on increased health
service utilisation well-documented in recent World Health
Organisation (WHO) reports
˃ Injured people are different from the non-injured population in terms
of pre-existing morbidity
˃ Patients with higher numbers of co-morbidities utilise injury services
more than patients with lower co-morbidities.
Cameron, Prudie, Kliewer et al., 2005)
Health co-morbidity (ICD9-CM Chapter)
Source: Adapted from Cameron et al,
2005. Tables 4 and 5 respectively.
Rate Ratios* Injured/ Non-Injured
*Adjusted for age, sex and place of
residence *
Hospital Admissions
Physician
Claims per 1000 person years
Mental Health disorders
9.31
3.50
Injury and poisonings
3.68
2.72
Blood diseases
3.36
1.53
Endocrine and metabolic
2.79
1.38
Musculoskeletal disorders
2.61
1.76
Nervous system diseases
2.35
1.42
Respiratory diseases
1.98
1.38
Circulatory diseases
1.70
1.21
Health Literature
Role of Mental Health, Alcohol and Psychological Traits
“There appears to an aetiological link between mental health
conditions and injury, particularly in relation to risk-taking
behaviours, alcohol misuse, and psychological traits such as
impulsivity, sensation-seeking, and risk-perception.”
(Cripps & Harrison, 2008. Briefing report for the Australian Institute of Health and Welfare)
Workers Compensation Literature
 Increased injury risks, higher medical treatment costs
(including pharmaceutical services), workers compensation
costs, and poor work performance (presenteeism) have
consistently been associated with specific lifestyle risk
factors such as tobacco use (current and previous), obesity,
stress, and lack of regular physical activity among working
people in a variety of settings
(Studies published by Health Management Research Centre, and Others)
Workers Compensation Literature
 Considerable confidence the excess risk from health comorbidities accounts for at least 25% to 30% of medical
costs per year across a wide variety of companies,
regardless of industry or demographics
 The biggest cost factors are the cost of extra treatment
utilisation, and medical costs associated with the
complications of a co-morbidity
(Studies published by Health Management Research Centre)
ACC Claims Costs
Highly skewed
All Results Significant at 95%
Multivariate Analysis
Future Cost Calculation
The proportion of ACC costs attributable to chronic
illness in any given out year is a function of:
 the population structure (the matrix Nij)
 the number of years from our baseline, n.
 Pij, r, $ccij and $nccij are all constants calculated
earlier, or assumed.
P  f ( N , n) 
n
N
P
(
1

i
)
($ccij  $nccij )
 ij ij
ij
n
n
N
($
cc
P
(
1

i
)

$
ncc
(
1

P
(
1

i
)
)
 ij ij ij
ij
ij
ij