Bozena J Katic, PhDc MPH1, Richard A. Jackson, MD2

Transcription

Bozena J Katic, PhDc MPH1, Richard A. Jackson, MD2
Patient-Centered Outcomes in Diabetes Care: A Study of A1C Awareness and Diabetes Distress
Bozena J Katic, PhDc MPH , Richard A. Jackson, MD
1
2
For correspondence, email: richard.jackson@joslin.harvard.edu or research@patientslikeme.com
Abstract
Patient-centered outcomes are increasingly recognized as an important component
of diabetes (DM) care, and the daily fears and distress that accompany DM are
important patient outcomes. We queried members of the online patient community,
PatientsLikeMe, who had DM on the 17-item diabetes distress scale (DDS-17), along
with questions on their A1C numbers and interpretation. DDS item, subscale and total
scores were calculated. Summary statistics and bivariate statistics were computed on the
sample; ANOVA and correlation coefficients were used on continuous/ordinal variables.
A total of 594 members responded to the survey; 558 completed all survey measures.
The majority of respondents had type 2 diabetes (88%), more than half (59%) were
female, and were on average 52 years old. The overall distress score was 2.43 (SD:
1.01), indicating moderate diabetes distress among the sample. Actual A1C numbers
were highly correlated to what people with diabetes (PWD) perceived their A1C
number to mean (from very good to poor; r =0.70), and PWD reporting A1C numbers
>7 had significantly higher mean distress scores compared to patients with A1C
numbers in healthy ranges (2.72 vs. 2.12; p<0.0001). There were higher correlations
between what PWD thought their A1C numbers meant and individual distress items
than with their actual A1C numbers. Items causing significant distress, such as “feeling
that I am failing at my diabetes regimen” or “ending up with complications, no matter
what I do” (Means = 2.76 and 3.15, respectively) were strongly correlated with the
perception that one’s A1C number meant they were doing poorly (r = 0.49 and r =
0.38.) Actual A1C number and patients’ interpretation of their number was significantly
related to total distress score, and A1C interpretation was correlated to feelings of
failure or doom regarding one’s diabetes care. Allowing PWD a more active role in
choosing their DM regimen may address pervasive feelings of failure and complications
and help decrease diabetes-related distress.
Background
Figure 1
Results
Study Population
Results
Diabetes Distress Scale
Emotional Burden Subscale
2884 memebers
with DM invited to
participate
Regimen-related Distress
Energy
200
Controls
Complications
Overwhelmed
2290 non-responses
594 responses
(no response or opt-outs)
Failing regimen
250
Angry
150
Reported A1C & perceived A1C
Not confident
Meal plan
200
Not motivated
Meal plan
Not motivated
Not confident
150
100
Failing regimen
100
50
558 completes
50
36 incompletes
0
Not a
problem
Slight
problem
Moderate
problem
Somewhat
serious
problem
Physician-related Distress
Results
Demographics
400
10%
Very
serious
problem
Doctor not knowledgeable
No clear directions
350
Doctor not concerned
300
2%
Serious
problem
No regular doctor
Male
Type 2
Type 1
Unsure
88% had Type 2 diabetes,
10% had Type 1 diabetes,
and ~2% were unsure
of the type of diabetes
they had
More than half of those who
completed the DDS survey
were female (59%)
• Patient-centered outcomes are increasingly recognized as an important component
of diabetes (DM) care
Interpersonal Distress
300
Serious
problem
Very
serious
problem
“My A1C number means...”
Self care
Diabetes difficult
No support
Objective
• To assess diabetes distress among a real-world online patient sample
• To identify differences between one’s reported A1C number and what they perceive
their A1C number means about how they are doing with their diabetes management
Methods
• Members of an online patient community, Patientslikeme (PLM) with diabetes
mellitus (DM) were invited to participate in a online survey on A1C awareness
and diabetes distress
• 2884 active members with DM were invited to participate
• Active members were those who had logged on to the PLM website at least once
in the previous 90 days
Statistical Analyses
• Summary statistics and bivariate statistics were computed on the sample of DM
patients; the ANOVA procedure was used to compare group means on continuous
variables
• Overall F-statistic and tests for multiple comparisons were computed
• Correlation coefficients were used to determine the correlation between
continuous/ordinal variables
© 2014 Copyright PatientsLikeMe. All rights reserved.
1
PatientsLikeMe, Cambridge, MA, USA; 2Joslin Diabetes Center, Boston, MA, USA
(A1C number <=7.5)
(A1C number= 7.0-8.5)
<=7.5: 97% accuracy
>7.5: 3% overestimated A1C
<=7.0: 56% underestimated A1C
7.0-8.5: 39% accuracy
Not a
problem
Slight
problem
Moderate
problem
Somewhat
serious
problem
Serious
problem
Very
serious
problem
0
“I’m not doing ok” n=77
“I’m doing very poorly”n=46
(A1C number=8.0-9.5)
(A1C number >=9.0)
8.0-9.5: 28% accuracy
Not a
problem
Slight
problem
Moderate
problem
Somewhat
serious
problem
Serious
problem
Very
serious
problem
• Mean DDS total score: 2.43 (SD: 1.01), indicating moderate diabetes distress among the PLM population (n=594)
• Variation on individual subscales: PLM patients with diabetes had the highest scores on emotional burden (2.74) and
regimen-related distress (2.62), and the lowest scores on physician-related distress (1.84) and interpersonal distress (2.37)
<=9.0: 46% underestimated A1C
>=9.0: 54% accuracy
>9.5: 18% overestimated A1C
DDS score
Correlation
perceived A1C(r)
Feeling that I will end up with long term complications,
no matter what I do.
3.15
0.38
Feeling that I am not sticking closely enough to a good meal plan
2.96
0.33
Feeling that I am often failing with my diabetes regimen.
2.76
0.49
Feeling that diabetes controls my life.
2.72
0.45
Feeling that diabetes is taking up too much of my physical/mental
energy each day.
2.70
0.40
Feeling overwhelmed by the demands of living with diabetes.
2.62
0.38
DDS Item*
*r values for item response options ordered by levels of increasing agreement from 1 (strongly disagree) to
5 (strongly agree), such that stronger levels of agreement correspond to increasing distress
• Items causing significant distress, such as “ending up with
long-term complications, no matter what I do”, "feeling that
I am not sticking closely enough to a good meal plan" or
“feeling that I am often failing with my diabetes regimen”
(DDS means: 3.15, 2.96 and 2.76 respectively) were
correlated with the perception that one’s A1C number meant
they were doing poorly (r = 0.38, 0.33 and 0.49 respectively.)
• Items denoting doom and failure with diabetes day to day
management, such as "feeling that diabetes controls my life"
and "feeling that I am often failing with my diabetes regimen"
were most strongly correlated with perceived A1C (r= 0.45
and r=0.49 respectively).
• The correlation of perceived A1C to acceptable A1C ranges
was not altogether accurate; many people thought they were
doing worse than their reported A1C score would indicate
— There was no difference in diabetes distress score among
those who accurately and inaccurately interpreted their A1C,
however.
200
N
Results
150
Diabetes Distress and A1C Number
100
Correlation of Reported and Perceived A1C
%
DDS Score by Grouped A1C Number
8th grade
Some
HS/HS
Some
college
College
degree
Post-grad
• However, many people interpret their A1C as being worse than it is
The majority of those who completed the survey report having
‘some college.’
DDS and A1C
• Both reported A1C number and peoples’ perception of their number were significantly related to total distress score
• However, perceived A1C correlated more closely with DDS than did reported A1C
N
200
• Those people who reported not knowing what their number meant experienced significantly more distress than
those who reported doing ‘very well’ or ‘okay’
150
N
100
Very well
OK
Not OK
Very poorly
20
21-27
and under
28-44
45-52
53-59
60-87
The largest number of completed surveys were contributed by
those aged 60-87 years. Mean: 53; SD: 10.5
Accurate understanding of their A1C could lead to better diabetes control and less distress for people with diabetes.
I don’t know
50
0
Conclusions
A1C interpretation
• There is a reasonable concordance between what people with diabetes think that there A1C means and what their
reported A1C number is
50
0
Study Population
“I’m doing ok” n=142
<8.0: 53% underestimated A1C
• Patient centered outcomes include feelings about and perception of one’s illness
• The daily fears and distress that accompany diabetes are important patient outcomes
“I’m doing very well” n=81
>8.5: 5% overestimated A1C
50
50
Female
Somewhat
serious
problem
100
100
0
Moderate
problem
Correlation of perceived A1C
to acceptable A1C ranges
150
150
88%
Slight
problem
• The emotional burden and regimen-related distress
subscales were more highly correlated with perceived A1C
than were the interpersonal distress and physician-related
distress subscales
200
200
59%
Not
a problem
250
250
41%
0
• Item correlations for perceived A1C were stronger than for
reported A1C across all subscales
Item Correlations and
Perceived A1C
Perceived A1C and Reported A1C score are highly correlated
(r= 0.702)
Patients reporting A1C numbers >7 have significantly higher
mean distress scores (2.72, SD 0.98) compared to patients with
A1C numbers in healthy ranges (2.12, SD 0.86); p<0.0001
Disclosures
• B. Katic is an employee of PatientsLikeMe and owns stocks/stock options in the company.
• Dr. Jackson is an occasional consultant for PatientsLikeMe.
www.patientslikeme.com