Developing a Resource Guide for CANS Data Analysis

Report
Developing a Resource Guide for
CANS Data Analysis and Reporting
Vicki Sprague Effland, Ph.D.
Youth Improved!
Dimension Scores
18
16
14
12
10
8
6
4
Intake
2
Discharge
0
Item Scores - Life Domain Functioning
18
16
14
12
10
Intake
8
Discharge
6
4
2
0
Living
Situation
Permanence Placement
Stability
Attendance
Behavior
Achievement
Item Scores - Child Behavioral and Emotional Needs
16
14
12
10
8
6
4
2
0
Intake
Discharge
Item Scores - Child Risk Behaviors Needs
6
5
4
3
Intake
Discharge
2
1
0
Suicide Risk
Self
Other Self Danger to
Sexual
Runaway Fire Setting
Mutilation
Harm
Others Aggression
Item Scores - Caregiver Functioning
16
14
12
10
8
6
Intake
Discharge
4
2
0
Any Improvement
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Improved - Life Domain Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Living
Situation
Permanence
Placement
Stability
Attendance
Behavior
Achievement
% Youth Improved - Child Behavioral and Emotional
Needs
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Improved - Child Risk Behaviors
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Suicide Risk
Self Mutilation Other Self Harm Danger to Others
Sexual
Aggression
% Youth Improved - Caregiver Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Reliable Change
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Actionable Needs
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Intake
Discharge
# Needs Met
6.0
5.0
4.0
3.0
2.0
1.0
0.0
% Needs Met
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Met Need(s) in Dimension
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Met Needs - Life Domain Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Living
Situation
Permanence
Placement
Stability
Attendance
Behavior
Achievement
% Youth Met Needs - Child Behavioral and Emotional
Needs
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Met Needs - Child Risk Behaviors
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Suicide Risk
Self Mutilation Other Self Harm Danger to Others
Sexual
Aggression
% Youth Met Needs - Caregiver Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Youth Improved!
Did Youth Improve Enough?
Need for Resource Guide
• Standardize methodology for CANS data
analysis
• Establish benchmarks for various data analysis
methods
• Develop guidelines for reporting CANS results
Introduction to Choices
Choices, Inc.
• Non profit care management entity created in
1997
• Developed around a community need: “high
cost youth”
• Blended system of care principles with
wraparound values and managed care
technology.
Choices Care Management
•More than 220 employees
• $35 million annual budget
• More than 1300 youth served in
child and family teams daily
• Working across ALL child serving
systems – 60% child welfare
Indiana Choices – Since 1997
Maryland Choices – Since 2005
DC Choices – Since 2008
Louisiana Choices – Since 2012
Choices, Inc.
• Adopted CANS in 2006
– Comprehensive version
– 12 Life Domains
• Outcomes Champion – Agency in 2007
Outcomes Monitoring
• Internal
– Program effectiveness
– Quality improvement
• External
– Adherence to contract requirements
– Marketing to new partners and communities
Successes
• Have lots of CANS data
• Multiple resources to analyze and report
data
– Outcomes and evaluation
– Software development
– Communications
• Ability to look at trends over time
Challenges
• Difficult to compare our performance to
others
– Multiple versions of the CANS
– Variation in how CANS is analyzed
– Multiple tools used across communities
• Need to establish meaningful performance
expectations
– Minimum levels of change
– % youth expected to improve
Important Points about the CANS
Critical Elements of Communimetrics
Measures
1.
2.
3.
4.
5.
6.
Partner Involvement
Malleable to the Organization
Just Enough Information Philosophy
Meaningfulness to Decision Process
Reliability at Item Level
Utility of Measure Based on its
Communication Value
“Unlike psychometric measures in which clinical
significance is a more rigorous standard than
statistical significance, any change on the
CANS is clinically significant.”
- Lyons (2009), Communimetrics: A Communication Theory
of Measurement in Human Service Settings
Total Clinical Outcomes Management
Decision
Support
Family &
Youth
Program
System
Service
Planning
Eligibility
Resource
Management
Accreditation
Transformation
Evaluation
Performance
Contracting
Case
Quality
Management
Improvement & Supervision
Outcome
Monitoring
Service
Planning &
Celebrations
Methods for Analyzing the CANS
• Dimension-Level Analyses
• Item-Level Analyses
Dimension-Level Analyses
Change in Dimension Scores
• Analysis Steps
1.
2.
3.
4.
Sum items in a specified dimensions
Divide by the number of valid responses
Multiply by 10
Conduct statistical analysis
Change in Dimension Scores
• Reporting Results
– Intake and discharge means
– Results of statistical analysis
– Statistically significant change in scores between
intake and discharge
• Benchmarks
– Accepted statistical criteria
– None available for clinical significance
Dimension Scores
18
16
14
12
10
8
6
4
Intake
2
Discharge
0
Change in Dimension Scores
• Advantages
– Uses well known statistical methods
– Statistical significance has a commonly
understood meaning
• Disadvantages
– Statistical significance not always indicative of
clinical significance
– Does not communicate results in terms of number
of youth showing improvement
Any Improvement in Functioning
• Analysis Steps
1. Calculate intake and discharge mean scores
2. Identify youth with lower scores at discharge
•
Intake Mean Score > Discharge Mean score
3. Divide by # youth in sample
Any Improvement in Functioning
• Reporting Results
– % of youth with any improvement in functioning
• Benchmarks
– N/A
Any Improvement
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Any Improvement in Functioning
• Advantages
– Simple to analyze
– Easy to explain methodology
• Challenges
– Lack of established benchmarks
– Difficult to communicate that change is clinically
meaningful
Reliable Change
• Equation
– RCI = 1.28 * SD * SQRT(1 – Reliability)
• Analysis Steps
1. Compute the RCI
2. Calculate change in intake and discharge mean
scores
3. Identify youth with change in scores >= RCI
4. Divide by # youth in sample
Reliable Change
• Reporting Results
– % of youth with a reliable improvement in
functioning
• Benchmarks
– 60-80% of youth expected to improve in at least
one of the dimensions measured
– 20-40% of youth expected to improve in a
specific dimension
Reliable Change
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Reliable Change
• Advantages
– Clearly defined method
– Available benchmarks
• Challenges
– Difficult for program staff to interpret and
communicate results
– Results include youth with no needs at intake
Actionable Needs
• Analysis Steps
1. Count the number of needs rated as a 2 or 3
within each dimension
2. Compare needs at Intake and Discharge
Actionable Needs
• Reporting Results
– Average number of needs at intake and discharge
across dimensions
• Benchmarks
– N/A
Actionable Needs
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Intake
Discharge
Actionable Needs
• Advantages
– Easy to display graphically
– Simple for audiences familiar with the CANS to
understand
• Challenges
– Requires additional explanation if audience
includes individuals not familiar with the CANS
– Lack of established benchmarks
Met Needs
• Analysis Steps
1. Identify youth with ratings of 2 or 3 on
individual items at Intake
2. Determine whether item ratings decreased to
a 0 or 1 by Discharge
3. Compute the number and percent of items
met within each dimension
4. Calculate the percent of youth who met at
least one (or more) needs within the
dimension
Met Needs
• Reporting Results
– Average number of needs met by dimension
– Percent of needs met
– Percent of youth who met at least one need
• Benchmarks
– N/A
# Needs Met
6.0
5.0
4.0
3.0
2.0
1.0
0.0
% Needs Met
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Met Need(s) in Dimension
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Met Needs
• Advantages
– Effective way to communicate improvement
– Simple for audiences familiar with the CANS to
understand
– Several options for reporting
• Challenges
– Requires additional explanation if audience
includes individuals not familiar with the CANS
– Lack of established benchmarks
Dimension-Level Analyses
•
Questions?
•
Additional Methods?
•
Thoughts?
Item-Level Analyses
Item Score
• Analysis Steps
1. Mean item score for all youth at Intake
and at Discharge
2. Multiply by 10
Item Score
• Reporting Results
– Graph of intake and discharge scores
• Benchmarks
– N/A
Item Scores - Life Domain Functioning
18
16
14
12
10
Intake
8
Discharge
6
4
2
0
Living
Situation
Permanence Placement
Stability
Attendance
Behavior
Achievement
Item Scores - Child Behavioral and Emotional Needs
16
14
12
10
8
6
4
2
0
Intake
Discharge
Item Scores - Child Risk Behaviors Needs
6
5
4
3
Intake
Discharge
2
1
0
Suicide Risk
Self
Other Self Danger to
Sexual
Runaway Fire Setting
Mutilation
Harm
Others Aggression
Item Scores - Caregiver Functioning
16
14
12
10
8
6
Intake
Discharge
4
2
0
Item Score
• Advantages
– Easy to present graphically
• Disadvantages
– Does not communicate results in terms of number
of youth showing improvement
Any Improvement
• Analysis Steps
1. Identify youth with ratings of 2 or 3 at Intake
2. Identify youth with lower scores at Discharge
• Intake Rating > Discharge Rating
3. Compute mean number of youth showing
improvement
• Note that need does not have to be met to count
in this analysis
Any Improvement
• Reporting Results
– % of youth with any improvement in functioning
• Benchmarks
– N/A
% Youth Improved - Life Domain Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Living
Situation
Permanence
Placement
Stability
Attendance
Behavior
Achievement
% Youth Improved - Child Behavioral and Emotional
Needs
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Improved - Child Risk Behaviors
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Suicide Risk
Self Mutilation Other Self Harm Danger to Others
Sexual
Aggression
% Youth Improved - Caregiver Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Any Improvement
• Advantages
– Simple to analyze
– Allows for any improvement in functioning to be
reflected
• Challenges
– Lack of established benchmarks
Met Needs
• Analysis Steps
1. Identify youth with ratings of 2 or 3 on
individual items at Intake
2. Determine whether item ratings decreased to
a 0 or 1 by Discharge
3. Calculate the percent of youth who met the
item
Met Needs
• Reporting Results
– Percent of youth who met individual needs
– Results for individual needs within a dimension
• Benchmarks
– N/A
% Youth Met Needs - Life Domain Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Living
Situation
Permanence
Placement
Stability
Attendance
Behavior
Achievement
% Youth Met Needs - Child Behavioral and Emotional
Needs
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
% Youth Met Needs - Child Risk Behaviors
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Suicide Risk
Self Mutilation Other Self Harm Danger to Others
Sexual
Aggression
% Youth Met Needs - Caregiver Functioning
100.0%
90.0%
80.0%
70.0%
60.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Met Needs
• Advantages
– Effective way to communicate improvement
– Simple for audiences familiar with the CANS to
understand
• Challenges
– Requires additional explanation if audience
includes individuals not familiar with the CANS
– Lack of established benchmarks
Item-Level Analyses
•
Questions?
•
Additional Methods?
•
Thoughts?
Establishing Benchmarks
Grab those pens and pencils!
Establishing Benchmarks
1.
2.
3.
4.
5.
6.
Youth
Service models
CANS versions
Availability of data
Analysis methods
Reporting results
Establishing Benchmarks
1. Youth
a. Age
b. Race/ethnicity
c. Strengths and needs prior to intervention
Establishing Benchmarks
2. Service models
a. Wraparound
b. Residential treatment
c. Crisis intervention
d. Outpatient therapy
e. Detention
Establishing Benchmarks
3. CANS versions
a. Comprehensive
b. Mental health
c. Juvenile justice
d. Child welfare
e. Education
f. Crisis
Establishing Benchmarks
4. Availability of data
a. Number of youth served annually
b. Method for completing CANS
c. Data management
d. Willingness/ability to share data
Establishing Benchmarks
5. Analysis methods
Dimension-Level
a. Dimension scores
b. Any improvement
c. Actionable Needs
d. Met Needs
Establishing Benchmarks
5. Analysis methods
Item-Level
a. Item scores
b. Any improvement
c. Met Needs
Establishing Benchmarks
6. Reporting results
a.
b.
c.
d.
e.
f.
g.
Youth demographics
Service context
Amount, frequency and/or duration of services
Sample size
Length of stay
CANS version used
Data analysis method used
Next Steps
1. Compile your survey responses
2. Share survey with other CANS users
3. Form CANS Benchmarking Workgroup
a. John Lyons
b. Volunteers?
c. Nominations
4. Develop action plan
5. Provide updates on progress
Thank You!
Vicki Sprague Effland, Ph.D.
Director, Outcomes and Evaluation
Choices, Inc.
4701 N. Keystone Ave., #150
Indianapolis, IN 46205
[email protected]

similar documents