Integrating community based addiction treatment services in a community hospital setting

1
Integrating community based addiction treatment
services in a community hospital setting

• Addictions Ontario Annual Conference
• June 6, 2006
• Michel Larose, M. Sc
• Glenn Barnes, M.H.A., LL.B.
• Addiction Services of Eastern Ontario

2
ASEOs Bailiwick

• 5 counties of eastern Ontario

3
(No Transcript)
4
North Dundas Population Industry

• Population 11014
• Land area 503.18 square km
• Census 2001
• Private dwelling characteristics
• Owned 80.1
• Rented 19.9
• Population projection growth of 1060 in 15 years

5
WDMH Vision

• Provides core services to meet the current and
  evolving needs
• Develops a range of services to meet the evolving
  health care needs of the community through
  strengthening linkages with other local and
  regional service providers

6
Winchester District Memorial HospitalQuick Facts

• Service area includes the municipalities of
  Dundas (South North,) Stormont (South North,)
  North Grenville, Edwardsburg/Cardinal, Russell,
  and the Southern portion of Ottawa (Osgood and
  Metcalfe).
• Population within the geographic area serviced is
  estimated to be 94,000.
• The Hospitals referral population is estimated
  at 27,000 and growing.
• Fiscal Year 2005/06
• 17,740 ER visits
• 17,995 Outpatient Clinic Visits
• 2,914 Surgical Cases
• 2,310 Separations
• 280 employees 32 active physicians, including 6
  specialists

7
WDMH - Core services

• 24-Hour Emergency Care
• Family Medicine
• General Medicine
• Obstetrics / Gynecology
• General Surgery
• Complex Continuing Care
• Rehabilitation
• Laboratory Medicine
• Diagnostic Imaging

8
Office facilities

• In 2005 moved from co-habitation with CMHA and/or
  Parole
• Occupied on-site detached building used for
  visiting medical staff clinics and laboratory
  facilities
• Introduction to Medical Staff thru CEO and Chief
  of Staff vocal support critical

9
Change in ASEO treatment delivery

• Traditional ASEO client caseload is
• Abuse cases with/out negative consequences
• Dependence cases - moderate to chronic
• Walk-in clientele
• SHIFT TO
• earlier intervention (a proactive approach in
  hospital setting)
• Assertive continuing care
• Applications in other environments (primary
  health care clinics, etc)

10
North Dundas Population Industry
11
North Dundas Population Industry
12
North Dundas Population Industry
13
North Dundas Population Industry
14
North Dundas Population Industry
15
Referral Source
16
Age
.0012
17
Language of Preference
NS .8140
18
Gender
NS .6795
19
Presenting Problem
NS .4082
20
Employment
.0000
21
Legal Status
.0000
22
Relationship Status
NS .3563
23
Non-Medical IDU
NS .4605
24
Educational Status
NS .7620
25
Income Source
.0046
26
Address Effective Duration
NS .9262
27
Mental Health Diagnosis
.0092
28
Psychiatric Hospitalization
NS .5225
29
Concurrent Services
.0389
30
Prescribed Medication
NS .0915
31
Prescribed Medication
NS .5100 NS .0984 NS .6156
32
Prescribed Medication
.0104 NS .1966 .0500
33
Physiological Impairments
NS .9066 NS .9444 NS .3469
34
Physical Conditions
NS .2976 NS .5804 NS .4274
35
Physical Conditions (Cont.)
NS .3302 NS .1431 NS .6416
36
Physical Conditions (Cont.)
NS .3991 NS .9702 NS .7286
37
Physical Conditions (Cont.)
NS .3550 NS .6514 NS .8497
38
Physical Conditions (Cont.)
NS .2902 NS .3401 NS .2675
39
Victim of Abuse
NS .8950 NS .9350 NS .9604
40
Patient Drug History Questionnaire Alcohol
(Beer, liquor, wine)
NS .1668
41
Patient Drug History Questionnaire Cocaine/Crack
NS .1069
42
Patient Drug History Questionnaire
Amphetamines/Other Stimulants
NS .3206
43
Patient Drug History Questionnaire Cannabis
NS .1789
44
Patient Drug History Questionnaire
Benzodiazepines
.0162
45
Patient Drug History Questionnaire Barbiturates
NS .8957
46
Patient Drug History Questionnaire Heroin/Opium
NS .0744
47
Patient Drug History Questionnaire Prescription
Opiates
NS .4229
48
Patient Drug History Questionnaire Codeine
NS .6220
49
Patient Drug History Questionnaire Hallucinogens
NS .4011
50
Patient Drug History Questionnaire Glue/Other
inhalants
NS .5074
51
Patient Drug History Questionnaire Tobacco
NS .8003
52
Adverse Consequences of Substance UseProblems
with physical health (e.g., overdose)
.0002
53
Adverse Consequences of Substance UseBlackouts
or memory problems, forgetting, confusion,
difficulty thinking
.0083
54
Adverse Consequences of Substance UseMood
changes, personality changes, substance-related
psychoses, flashbacks when using
.0001
55
Adverse Consequences of Substance UseProblems
in relationships (including friendships, family
of origin, partner/spouse, etc.)
.0002
56
Adverse Consequences of Substance UseVerbally
or/and physically abusive when using
.0021
57
Adverse Consequences of Substance UseSchool
or/and work problems (performance affected or
loss of job/expelled from school)
.0017
58
Adverse Consequences of Substance UseLegal
problems (substance-related charges)
.0158
59
Adverse Consequences of Substance UseFinancial
problems (due to substance use)
.0094
60
SOCRATES - Alcohol
.0000 .0000 .0045
61
SOCRATES Drug
.0004 .0013 .0451
62
SOCRATES Drug (other)
.0390 .0191 .0093
63
Treatment Entry Questionnaire
.0000 .0001 .0027
64
Drug-Taking Confidence Questionnaire - Alcohol
.0003 NS.1855 NS.0980 .0001 .0003
.0005 .0003 .0003 .0001
65
Drug-Taking Confidence Questionnaire - Drug
NS.0817 NS.0912 NS.1522 NS.0631 .0197
NS.3702 .0003 .0009 .0045
66
BASIS-32 (Behaviour And Symptom Identification
Scale)
.0005 .0023 .0000
.0011 .0028 .0000 .0003
67
Integrating research into practice
68
Results questioned

• Are we reaching everyone we can?
• Are we reaching them early enough?
• Could we do better if we could apply some of the
  research findings to the clinical scene?

69
Methods of economic analysis

• Cost analyses
• Describes cost of treatment (direct cost)
  monetary value of problem to society (indirect
  cost)
• Cost-effectiveness analysis (aka cost benefit
  ratio analysis)
• Compares 2 or more Tx alternatives in terms of
  both cost and effectiveness
• Cost-benefit analysis
• Cost of Tx compared to monetary value of outcomes
• All make assumptions from clinical research
  findings

70
Search for Silent Markers

• The Canadian scene
• Cost analyses impact studies 1992 2002
  assumptions
• WHO studies
• Risk factor oriented
• European studies
• Diagnostic category risk factors by gm/day
• United States
• Cost ratio benefit approach

71
Our approach

• GOAL - find the silent markers for an early
  intervention program
• CAUTION - Canadian environment substantially
  different from US
• Best literature source is in cost benefit ratio
  analysis area
• Reviews impact SUD on hospital usage
• Blends economic with clinical analysis
• WHY not Canadian data?

72
2002 CCENDU report on costs

• CCSA study on impact of substance abuse /
  dependence on hospitalization rates
• PROBLEM - Restricted diagnostic categories
• 1 most responsible diagnosis (Type M codes)
• total hospital admissions at 56,161
• 2- responsible to some extent (Type 1
  codes) 137,429 hospitalizations / male female

Refer to handout Table 1
73
What are diagnostic category codes?

• The International Classification of Disease
  coding system

74
ICD codes

• Canadian system different from US
• Canadian system uses Case Mix Grouper (CMG) to
  measure case intensity weight
• Canadian coding distinguishes
• Type M (diagnostic code that is the direct link
  for the admission )
• Type 1 (pre-admit co-morbidity diagnoses or
  diagnoses directly affecting length of stay)
  called responsible to some extent by CCENDU
• Type 2 (post admit co-morbidity diagnoses)
• Type 3 (secondary diagnoses that does not
  directly affect that length of stay)
• US hospitals use Diagnostic Related Grouper (DRG)
  which includes all diagnoses (Types M, 1, 2 and 3)

75
Put things in perspective!

• WDMH had 2300 discharges 16,092 discharge days
  in 2004-5
• Applying 1992 CCSA protocol to Type M category
  yields
• 1.74 of admissions / 1.73 of admission days
• Applying 1992 CCSA protocol to Type M and Type 1
  3 categories yields
• 2.9 of admissions / 2.73 of admission days

76
Is that the true picture of direct health care
cost of substance abuse/dependence?
77
Review of US CBR literature

• CAUTION
• Avoid inner city studies
• Avoid homelessness studies
• Avoid Veterans Affairs studies
• Restrict search to populations that most closely
  resemble Canadian profiles
• Settled on Sacramento Kaiser Permanente HMO
  population

78
3 major stages
79
Offset study - findings

• Slight non significant trend to higher cost for
  group receiving primary care within addiction
  treatment program
• However patients in study group with certain
  diagnostic conditions more likely to be abstinent
  and there was a slight cost effectiveness factor
  to the delivery of primary care in the addiction
  treatment program

80
3 major stages
81
3 major stages
82
SUD label study 95 CI Odds Ratiosfor
Substance Abuse Medical Conditions
See handout Table 2
83
The Diagnostic Category Study

• Examines impact on health care utilization and
  cost by aiming substance abuse treatment to high
  odds diagnostic categories determined in Offset
  SUD studies

84
Diagnostic category study Patient acceptance
criteria

• New adult patients entering KP Chemical
  Dependency Recovery Program between Mar 97 Dec
  98 12 mts
• Follow up on impact on substance abuse related
  medical conditions (SMAC) reported in previous
  articles
• Excluded previous patients with known substance
  abuse or psychiatric diagnoses or known service
  utilization
• Divided into two groups
• Integrated Care group receiving substance abuse
  treatment the study group (n318)
• Independent Care group the control (no
  substance abuse treatment) (n336)

85
Diagnostic category studydemographics

• Characteristics of overall group
• Mean age of 37 (SD10)
• 11 retired (refer to NYS OASAS handout
  subsequent slide)
• 73 caucasian
• 44 female
• 60 employed
• 41 income gt 40K
• 41 married
• 51 some college and above
• 11 employer mandated treatment

86
3 major stages
See handouts - Table 3
87
SAMC study Integrated Care results

• Inpatient days decreased from 114.2/1000 members
  months to 39.5/1000 member months (p0.05)
• Hospitalization rates declined from 26.6/1000 m.
  m. to 11.7/1000 m. m. (p 0.04)
• Average monthly inpatient cost DOWN 204 to 43
  (p 0.08)
• ER visits DOWN from 0.10 to 0.07 visits /m. m. (p
  0.03)
• ER costs DOWN 52 to 30 / member month (p
  0.02)

88
SAMC Integrated Care study resultsfor Concurrent
sub-set

• Inpatient days decreased from 164.3/1000 to
  45.9/1000 member months (p0.05)
• Hospitalization rates declined from 36.3/1000 m.
  m. to 14.5/1000 m. m. (p 0.05)
• Average monthly inpatient cost DOWN 291 to 43
  (p 0.08)
• ER visits DOWN from 0.11 to 0.07 visits/m. m. (p
  0.07)
• ER costs DOWN 55 to 35 (p 0.13)
• Total medical costs (inpatient outpatient) DOWN
  566 to 222

89
Verification

• Kaiser Permanente tracked same utilization for 5
  year period
• Patient characteristics that affect utilization
• Medical severity at intake
• Women have higher inpatient costs at entry but
  higher decrease in primary care costs
• Older patients have smaller declines in hospital
  days inpatient cost over time
• Parthsarathy, S. Weisner, CM. Five year
  trajectories of health care and cost in a drug
  and alcohol treatment sample. 2005 Drug and
  Alcohol Dependence Vol 80, 231-240

90
A new perspective!

• Applying SAMC protocol for Type M codes
  identifies
• 19 of admissions (437 cases) compare to 1.74
• 19.6 of admission days compare to 1.73
• with high correlation to abuse / dependence
  lifestyle

See Table 4 for details

• Applying SAMC protocol for Types M, 1, 2 and 3
  flags identifies
• 35,7 of admissions (821 cases) 2.9
• 47.2 of admission days 2.72
• with high correlation to abuse / dependence
  lifestyle

AND
And we now have a readily available early
detection flag
91
Consistency with other reports

• Corrao, G et al., A meta-analysis of alcohol
  consumption and the risk of 15 diseases. 2004
  Preventive Medicine Vol 38, pp. 613-619
• Rehm, J et al., Alcohol-Related Morbidity and
  Mortality. 2003 Alcohol Research Health Vol
  27(10), pp. 39-51
• Reynolds, K et al., Alcohol Consumption and Risk
  of Stroke A Meta-Analysis. 2003 JAMA Vol
  289(5), pp. 579-588

92
Corrao
Relative risk alcohol consumption vs abstainers
13.8 g alcohol 1 standard Ontario drink / 1 5
oz.glass 14 red wine
93
Reynolds

• 35 cohort control studies reviewed
• Stroke risk factor _at_ 95 statistical confidence
  intervals

94
CCSA 2006New kid on the block!

• Reverted back to Type M codes only
• Diagnostic categories broken down by age/sex with
  attribution fraction for each
• Closely resembles KP findings?
• Not using same parameters or assumptions
• BUT hospital admissions now up from 56,161
  (1992 report) to 358,199 (2006 report)
• WHY? Rehm uses diagnostic categories based on
  attribution fractions to AOD

95
Attribution fractionsCAUTION

• Despite the broad consensus on many health
  consequences of alcohol consumption, further
  research is needed to clarify the conditions that
  are caused by alcohol consumption, magnitudes of
  causal relationships, and effects of different
  patterns of consumption and individual
  characteristics.
• Bloss, G. Measuring the Health Consequences of
  Alcohol Consumption Current Needs and
  Methodological Challenges. 2005 Dig Dis Vol. 23,
  pp 162-169

96
Proposed project with WDMH

• Identify admitted patients with high odds ratios
  diagnostic categories
• Screen for substance abuse / dependence min.
  460 (Type M coding alone) to 850 cases (Types
  M, 1, 2 3)
• If screening positive (estimate 25 115 cases),
  offer substance abuse treatment while in-patient
  and after discharge
• Follow-up post discharge assertive continuing
  care approach
• Detailed analysis of ER visits ( 4,425
  non-admitted SAMC - age 16 over - visits in
  2004-5) to determine cost effective flagging
  protocol

97
Comment from articleWeisner, CM et al 2003 Arch
Intern Med Vol. 163, at p. 2515

• there is evidence that many physicians are
  unaware that their patients have alcohol or drug
  problems this lack of awareness may be
  particularly problematic for patients on
  prescription medications, or when prescription
  medications may interact with alcohol and other
  substances of abuse to cause adverse effects.
  The findings also indirectly highlight the
  importance of screening to detect individuals
  with alcohol or drug problems, especially in
  emergency departments where the prevalence of
  alcohol and drug problems is particularly high.

98
Expected benefits

• Reduction in LOS for certain diagnostic
  categories (SMAC)
• Reduction in re-admission rates for SAMC
• Some downward impact on ER visits (still in
  development)
• Application to primary health care setting
• Earlier intervention better quality of life
  intercept chronicity pathway

99
Re-orienting addiction services delivery

• What should our relationship to community
  hospitals be?
• To teaching hospitals?
• What should our relationship to CCAC be? A
  services vendor?
• What role should addiction treatment play in the
  new primary health care projects?

100
Public policy decided by what data?

• What if CCNEDU data determined provincial funding
  priorities?
• BUT what if KP data determined provincial funding
  priorities?
• How can we encourage funding policies to reflect
  2006 CCSA report (attribution fractions)?