Transcript Research Presentations 2013 Summer Student Research and

University of Wisconsin
School of Medicine and Public Health
Department of Family Medicine
Evaluating Clinical Response to
Electronic Health Record
Surveillance of Childhood Obesity
Brittney Golbach
University of Wisconsin School of Medicine and Public Health
Summer Student Research and Clinical Assistantship Program
Summer 2013
At a Glance
 I. Background
 II. Goals
 III. Methods
 IV. Results
 V. Discussion
 VI. Future directions
Background: In the Literature
 From 2003-2007 obesity prevalence increased by
10% for all U.S. children 10-17 years old (Singh, 2010)
 2007-2008 NHANES found 28.8% of children 2-19
to be obese (OB) and 31.7% overweight (OV)
(Ogden, 2010)
 The probability of obesity as an adult is >50% for
OV/OB children over 6 years old (Whitaker, 1997)
Background: Work in DFM
 UW E-Health Public Health Information Exchange
(PHINEX); survey of 25 UW Family Medicine
residency clinics (Pillai, 2012)



30% of children 2-17 OV/OB
2% of those children coded in Epic as OV
<1% coded as OB
 2011: UW Health EHR begins
auto-coding visit BMI percentiles
by category
Background: Current Guidelines
 2007 Expert Committee Recommendations for the
Prevention, Assessment and Treatment of Childhood
and Adolescent Overweight and Obesity (Barlow, 2007)
 Healthcare Effectiveness Data and Information Set
(HEDIS) Measures
Project Goals
 1. Determine frequency at which UW Family
Medicine providers adhere to best practice
recommendations for OV/OB prevention and
assessment
 2. Offer targeted feedback to family practice
providers
 3. Establish a baseline for measuring future progress
in response to a QI intervention
Methods: Sample & Setting
 Patients: children 2-19 in well-child checks (WCC’s)
 Providers: family medicine resident and faculty
physicians, physician assistants
 Settings: Northeast, Verona and Belleville UWDFM
residency clinics
 Quality improvement (QI) study
Methods: Nutrition & Activity Measures
 Electronic appointment evaluation form completed
by researcher (me) during/following each WCC
 Evaluation form criteria reflect the Expert
Committee Recommendations for WCC’s for all
pediatric patients
 Compare to relevant Epic Checklist counseling items
Methods: Nutrition & Activity Measures
EPIC
EXPERT COMMITTEE
 I. Nutrition
 Sugar-Sweetened Beverages
 Fruits/Vegetables
 Breakfast
 Eating Out
 Family Meals
 Portions
 Fruit Juice
 Meal Frequency/Snacking
 II. Physical Activity
 III. Sedentary Behavior
 TV in Bedroom
 Screentime
 I. Nutrition
 Food Groups
 Eating Patterns/Preferences
 Family Meals
 II. Physical activity
 III. Sedentary Behavior
 TV in Bedroom
 Screentime
 Sleep
Methods: Additional Measures
 I. Prevention & Assessment
 Family History
 Physical Exam & Review of Systems
 Labs
 II. Treatment
 Referral
 III. Office Routines
 BMI Documentation & Growth Charts
 OV/OB Diagnosis
 Language & Motivational Interviewing
Methods: Procedures
 1. Observation of WCC appointments
 2. Reference to EHR patient charts
 3. Qualitative observations
 4. Analysis of frequencies with Pivot Tables
Results: Demographics
N = 26 cases
Results: Counseling Practices
Results: Counseling Practices
Results: Qualitative Observations
 Appropriate language
 Family Medical History for risk factors
 FMH generally incomplete and/or vague
 Complete absence of obesity in FMH
 Problem list diagnosis vs. “Today’s visit” diagnosis
Results: Qualitative Observations
 Review of Systems: max. 4/7 items covered
 6 patients in OV/OB categories were asked no OV/OB-related
ROS questions
 Recommended set of lab tests were not ordered for
any OV/OB patients
 Referrals: 1 referral to physical therapy in this study
Discussion: Limitations
 Sample size
 Relying on EHR for Epic criteria completion
 Hawthorne Effect
Discussion: Findings
 1. BMI documentation is ahead of other primary care
facilities (Brandt, 2013)
 2. A sensitive approach
 3. Documentation of diagnosis & FMH (Whitaker, 1997)
 4. Epic vs. Expert Committee
 5. Assessing vs. Counseling
Future Directions
 1. Best practice alert & modify auto-coding
(Taveras, 2013)
 2. Assess provider-perceived barriers
 3. Fit Epic to Expert Committee Rec’s (Taveras, 2013)
 4. Future QI intervention (?)
Acknowledgements
 Brian Arndt, MD
 Clinic Staff at Verona,
 Larry Hanrahan, PhD,MS
Northeast and
Belleville
 PHINEX Team
 Jon Temte, MD, MS, PhD
 Aman Tandias
 Melissa Behren
 Peter Capelli
 Tracy Flood, MD,PhD
 Emily Tomayko, PhD, RD
 Aaron Carrel, MD
 Alex Adams, MD, PhD
Final Thoughts & Questions
Climate Change and Health in
Primary Care: an exploratory study
Temte, Jonathan, MD/PhD; Holzhauer, John, BS; Kushner, Kenneth, PhD
University of Wisconsin School of Medicine and Public Health
Summer Research and Clinical Assistantship
July 19, 2013
Overview
• Purpose
• Background
• Methods
• Results
• Discussion
• Acknowledgements
Purpose
• Determine recognition rate of climate change within
convenient population
• To assess whether adult primary care patients perceive any
effects of climate change on their health
• Evaluate any association between acknowledgement of
climate change and depression and/or anxiety
• Encourage discussion and future investigation into subject
Background
• Human health effects of climate change
– Direct
•  temperature
• Injury/death from extreme weather events
•  air pollution/aeroallergens
– Indirect
• Changes in infectious disease
• Changes in food production/delivery
McMichael 2003, 2011; Patz 2000, 2005
Background
• Mental health effects 3 likely ways
– Directly
• Experience of extreme weather event/novel weather
patterns
– Disruptions in social, economic and environmental
determinates of health
• Displacement
• Loss of connection
– Emotional distress and anxiety about the future
Fritze et al. 2008
Methods
• 23 Question survey – anonymous
• Convenience samples
• 4 University of Wisconsin-MadisonDepartment of Family Medicine affiliated
clinics
Methods
• The survey
– capture data on symptoms, attitudes, and
experiences associated with climate change
– PHQ-2 (depression) and GAD-2 (anxiety); validated
screening instruments
– demographic information
Methods
• 9 Questions regarding awareness/effect of
climate change
– Scored 0- “Not at all” to 5- “All the time”
– Response rates were analyzed
• Individually
• “Composite”
• 4 questions of PHQ-2/GAD-2
– Scored 0- “Not at all” to 3 “Nearly every day”
– Results analyzed
• Individually
• combined into “Dysphoria” score
Results
• 728 people approached
– 157 refusals
– 571 surveys attempted (78%)
• Common reasons for refusal
– Unwilling
– Too sick
– No glasses
– No English
Results
Total
Clinic
N
1 (Urban)
199
2 (Urban)
106
3 (Suburban)
146
4 (Rural)
120
571
Results
Age
Asian
3%
Mean
46.8
StDev
17.2
Range
18-96
Native
Pacific
Black American Islander White
13%
1%
0%
80%
Other
3%
Sex
% Male
34
% Female
66
Parent
Yes (%)
73
No (%)
27
Results
$19,000- $34,000- $54,000<$19,000 $33,999 $53,999 $82,999 >$84,000
24%
16%
19%
21%
20%
College
Post-Bac.
< high
High School/
Some
Degree
Study
school
GED
College BA/BS
or Degree
5%
21%
32%
27%
15%
Very Moderately Slightly
Slightly
Moderately
Very
Liberal Liberal Liberal Moderate Conservative Conservative Conservative
16%
22%
10%
30%
7%
11%
4%
Results
• Is global climate change occurring?
N
546
Yes
88%
No
12%
• Is global climate change due to human
activities?
N
525
Yes
83%
No
17%
Political leaning = only significant
predictor of responses. (p < .001)
Results
• Highest mean “Composite” questions
– “Paying more attention to changes in climate”
– “Are you troubled by the lack of action on climate
change by leaders”
• Lowest mean…
– “Have you noted an health effects in you or your
family members from climate change?”
Results
Selected “Composite” Means
1.2
"Attention"
2.5
"Troubled"
"Health"
2.6
-1
0
1
2
3
4
Rarely Sometimes Frequently Most of the time
5
Results
• Rank Correlation: Dysphoria, Composite
– R= 0.345
– P < 0.001
Discussion
Nationwide:
• 64% Cautious,
Concerned,
Alarmed
• 47%
Anthropogenic
• Majority expect
direct health
impacts
• Less than half
expect indirect
Discussion
• Determining causality between climate change
and dysphoria?
• Does it matter?
• Acceptance of climate change acts as force
multiplier for mental health problems?
• Patients are aware and topic should be discussed
• % of “climate change accepters” makes it a
good population to study effects of accepting
climate change
Acknowledgements
• Drs. Temte, Kushner
• Staff at DFM affiliated clinics
• Department of Family Medicine- University of
Wisconsin School of Medicine and Public
Health (Summer Research and Clinical
Assistantship)
Questions?
Current Practices: Miscarriage
Management in Wisconsin
K. Hope Wilkinson, MS
Jess Dalby, MD
Background
• Miscarriage is not a rare event
• Multiple studies have validated that there are
four safe treatment options for women
experiencing a miscarriage prior to 12 weeks
gestational age.
• These options are: expectant waiting, medical
management, uterine aspiration in the office,
and uterine aspiration in the operating room.
Methods – Chart Review
ICD09 Codes
related to
Miscarriage
83
• Between
December 1,
2011 and
November 30,
2012
Exclude
women who
don’t qualify
65
• Did not miscarry (12)
• Received prenatal care
elsewhere (3)
• Had a still birth (3)
Collected
Data
• Gravida
• Previous Miscarriage
• Counseling Received
• Treatment Received
• ER visits
What options are women offered?
50%
46%
37%
40%
30%
20%
10%
3%
0%
Expectant
Surgical
Medical
What happens in miscarriage?
Elective D&C
20%
Passed
Products of
Conception
without
intervention
69%
Emergent
D&C
11%
How many women go to the ER?
90%
80%
70%
*
60%
50%
40%
30%
20%
10%
0%
Total
MD
CNM
Previous No Previous Old (25+) Young (<25)
Miscarraige Miscarriage
Conclusions
Patients lack education about
miscarriage natural history.
Methods – Provider Survey
What options are safe?
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
98%
91%
87%
70%
Expectant
Medical
Office OR uterine
uterine aspiration
aspiration
How many choices are women
offered?
45%
Referrals only
40%
35%
All four options
30%
Expectant Waiting +
Medical + Surgical
25%
20%
Expectant Waiting +
Surgical
15%
Expectant Waiting +
Medical
10%
5%
Surgical
0%
1 option
2 options
3 options
4 options
Referral
Does the number of choices offered
vary by the number of years in
practice?
100% 100%
90% 90%
80%
80%
70%
70%
60%
60% 50%
50% 40%
40% 30%
30% 20%
20%
10%
0%
10%
0%
Currently in <5 years 5 - 10 years 10 - 20
Residency
years
0 - 10 Years
10 - 20 Years
> 20 years
No Options (1 treatment
offered)
Options
No Options
offered
(1 treatment
offered)
No options
Wallace
Optionset
offered
al. No
Options
Options
Wallace et al. Options
Provided
20 years
What do providers recommend?
80%
70%
60%
50%
Male
40%
*
Female
30%
Total
20%
10%
0%
Expectant
Medical
Office uterine
aspiration
OR uterine
aspiration
Patient
Preference
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
What are the barriers to
comprehensive miscarriage
management?
88%
56%
54%
Facilities
Staffing
49%
Training
Too little
patient
demand
Conclusions
Providers lack training about
miscarriage treatment options.
Conclusions
Current best practices are not the
standard of care in Wisconsin.
Conclusions
Providers lack training about
miscarriage treatment options.
Conclusions
Patients lack education about
miscarriage natural history.
Acknowledgements
• Jess Dalby, MD
• Ronnie Hayon, MD
• University of Wisconsin - Department of
Family Medicine
• Access Community Health Centers
PATIENT ACTIVATION
AND HIGH UTILIZERS
Dan Ziebell
Medical Student
UW School of Medicine and Public Health
Introduction
• Patients who take a more active role in their health care
tend to have health better outcomes.
• The Patient Activation Measure (PAM) is a tool that
evaluates an individual’s knowledge, skill and confidence
in managing one’s health.
• High utilizers of health care resources are very costly to
society and themselves.
• If high utilizers can be identified before they need to resort
to using costly health services, money can be saved while
improving outcomes.
Background
• Health Care spending
• 2.6 trillion dollars in 2010
• 50% of that was hospital and physician/clinic services
• Who accounts for all this money?
• 5% account for 50% 1
• 1% account for 25% 1
1 (dept. of health and human services)
Goals
• Assess the relationship between patient activation and
physician reported over-utilization by patients
• Assess the relationship between patient activation and
UW complexity scale from Dr. Brian Arndt (to be
completed later)
Methods
• PAM surveys were distributed to patients at Wingra Clinic.
• Once a patent filled out this survey his/her PCP was given
a questionnaire in regards to complexity and utilization.
Patient Activation Measure
• Scored from 0-100
• 0 being least activated
• 100 being most activated
• Activation Levels:
• Level 1:
0
• Level 2:
47.1
• Level 3:
55.2
• Level 4:
67.1
-
47
55.1
67
100
Patient Activation Measure (PAM)
• Multiple studies have shown that higher PAM scores
correlate with increased likelihood to engage in
preventative behaviors.
• Highly activated people are more likely to engage in
eating healthy and getting regular exercise.
• Chronically ill patients who are highly activated are more
likely to adhere to treatment regimens.
Physician Survey
• Complexity questions taken from (Huyse et al., 2001)
Utilization
• What exactly defines it?
• ER visits
• Clinic visits
• Procedures
• Physician contacts
• Others
• What defines over-utilization?
Methods
Wingra Clinic Patients
- 18 years old
- PCP is faculty member
- English or Spanish speaker
- Physically/Mentally capable
to fill out the survey
Inclusion Criteria
Total Patients who qualified
N= 59
Patient Refusal or incomplete
survey
N= 176
Total Surveys Completed
N= 117
Demographics
Age
Number
Percent
18-30
17
14.53%
31-50
36
30.77%
51-64
45
38.46%
65+
19
16.24%
Sex
Number
Percent
Female
71
60.68%
Male
46
39.32
Demographics
Race
Number
Percent
American Indian
1
0.85%
Asian
1
0.85%
Black
22
18.80%
Hispanic
10
8.55%
White
82
70.09%
Decline
1
0.85%
Education
Number
Percent
Grade School
3
2.56%
High School
39
33.33%
Tech School
15
12.82%
College
39
33.33%
Masters
14
11.97%
Doctorate
7
5.98%
Results
Variable
Mean
PAM Score
62.90
Utilization Score
4.036
Question 1 (challenge)
4.018
Question 2 (cooperation)
5.153
Question 3 (social)
4.723
Question 4 (insurance)
5.509
Results
PAM Score vs. Utilization
7
Utilization Score
6
5
4
y = -0.023x + 5.4723
R² = 0.0547
3
2
1
0
0
20
40
60
80
PAM Score
Correlation Coefficient
-0.234
P-Value
0.013
100
Summary
• Statistically significant correlation between PAM score and
physician reported utilization score.
• Overall utilization is related to many complex physical,
social and psychological variables.
• Patient activation plays a small but important role.
Next Steps
• How to improve PAM score?
• Patient education
• Improve Dr./Patient communication
• Improve health literacy
• Focus on patient compliance and social support
Possible uses of PAM
• Identify high utilizers
• Can be used by health care systems?
• What do you do?
• More aggressive preventative care?
• Focus more strongly on compliance and social support
• Focus care to help prevent unnecessary costly utilization
Acknowledgments
• Jonas Lee, MD
• Jon Temte, MD, PhD
• Providers and staff at Wingra Clinic
Questions?