Epidemiology/Health Services/Psychosocial Research

O R I G I N A L A R T I C L E

The Confidence in Diabetes Self-Care

Scale
Psychometric properties of a new measure of diabetes-specific self-efficacy
in Dutch and U.S. patients with type 1 diabetes
NICOLE C.W. VAN DER VEN, MSC1,2 HERMAN ADÈR, PHD5 Self-efficacy beliefs are specific to be-
KATIE WEINGER, EDD3,4 HENK M. VAN DER PLOEG, PHD1,5 haviors and the situations in which they
JOYCE YI, BA3 FRANK J. SNOEK, PHD1,2,5 occur, affecting the courses of action peo-
FRANS POUWER, PHD5 ple choose to take, the amount of effort
invested, how long they will persevere,
their resilience to adversity, and what
they ultimately accomplish (2). The value
OBJECTIVE — To examine psychometric properties of the Confidence in Diabetes Self-Care of self-efficacy in predicting self-care be-
(CIDS) scale, a newly developed instrument assessing diabetes-specific self-efficacy in Dutch and haviors and outcomes in patients with di-
U.S. patients with type 1 diabetes. abetes is supported by several studies, in
RESEARCH DESIGN AND METHODS — Reliability and validity of the CIDS scale
which self-efficacy was associated with
were evaluated in Dutch (n ⫽ 151) and U.S. (n ⫽ 190) outpatients with type 1 diabetes. In self-reported adherence in adults (3–9)
addition to the CIDS scale, assessment included HbA1c, emotional distress, fear of hypoglycemia, and adolescents (10,11), glycemic control
self-esteem, anxiety, depression, and self-care behavior. The Dutch sample completed additional (8,12,13), and better perceived general
measures on perceived burden and importance of self-care. Test-retest reliability was established health, mental health, and social func-
in a second Dutch sample (n ⫽ 62). tioning (14).
In contrast to more stable personality
RESULTS — Internal consistency (Cronbach’s ␣ ⫽ 0.86 for Dutch patients and 0.90 U.S. characteristics, self-efficacy is a dynamic,
patients) and test-retest reliability (Spearman’s r ⫽ 0.85, P ⬍ 0.0001) of the CIDS scale were changeable belief, which may be en-
high. Exploratory factor analysis showed one strong general factor. Spearman’s correlations hanced by behavioral interventions, re-
between the CIDS scale and other measures were moderate and in the expected directions, and
high HbA1c levels were associated with low CIDS scores in the U.S. sample only. Low CIDS
sulting in an increased motivation for
scores were positively associated with self-care but not with glycemic control in the original behavioral efforts (2). In diabetes, such
samples. CIDS scores in the U.S. and Dutch samples did not show any statistically significant interventions have been successful in en-
differences. U.S. men had higher CIDS scores than U.S. women. hancing specific aspects (15) or more gen-
eral self-efficacy beliefs, along with
CONCLUSIONS — The CIDS scale is a reliable and valid measure of diabetes-specific self- improved HbA1c (16) and self-care be-
efficacy for use in patients with type 1 diabetes. High psychometric similarity allows for cross- havior (17). To assist patients in optimiz-
cultural comparisons. ing their self-care behavior, it may be
Diabetes Care 26:713–718, 2003 useful to assess self-efficacy specific to
self-care behaviors. When identified, sub-
optimal levels of self-efficacy can be tar-
geted by tailored behavioral interven-

E
ffective self-regulation of diabetes is the individual’s confidence in his or her
not just based on simple adherence own ability to perform specific tasks re- tions. Several instruments to assess self-
to a prescribed regimen but requires quired to reach a desired goal (1). To cope efficacy specific to self-care behavior have
active behavioral involvement of patients effectively with the complex demands of been used in adult patients with diabetes.
on a day-to-day basis. A key factor in at- the diabetes treatment regimen, a suffi- In some studies, a single item is used (18);
taining behavioral goals is self-efficacy— cient sense of self-efficacy is required. in other studies, data on reliability and
validity are lacking (3,8) or not available
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● in English (9). In some cases, the instru-
From the 1Department of Medical Psychology, Vrije Universiteit Medical Center, Amsterdam, the Nether- ment relies heavily on diet-related items
lands; the 2Research Institute for Endocrinology, Reproduction and Metabolism, Vrije Universiteit Medical (4,15,19) or is concerned with psychoso-
Center, Amsterdam, the Netherlands; the 3Behavioral and Mental Health Research Section, Joslin Diabetes cial issues instead of self-care behavior
Center, Boston, Massachusetts; 4Harvard Medical School, Boston, Massachusetts; and the 5Institute for (20). Measures for use in type 2 diabetic
Research in Extramural Medicine, Vrije Universiteit Medical Center, Amsterdam, the Netherlands.
Address correspondence and reprint requests to Nicole C.W. van der Ven, Vrije Universiteit Medical patients (12,21) and children and adoles-
Center, Department of Medical Psychology, Van der Boechorststraat 7, 1081 BT Amsterdam, the Nether- cents (22) are generally well validated and
lands. E-mail: [email protected]. widely used by others (13,11,16). Ad-
Received for publication 5 July 2002 and accepted in revised form 11 November 2002. vancing on these existing scales, we set
Abbreviations: CIDS, Confidence in Diabetes Self-Care; HFS, Hypoglycemia Fear Survey; PAID, Problem
Areas in Diabetes; SMBG, self-monitoring of blood glucose.
out to develop a short instrument to as-
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion sess self-efficacy in adults with type 1 di-
factors for many substances. abetes, in Dutch as well as U.S. patients.

DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003 713

The CIDS scale

The objective of this study was to examine Study 1: Evaluation of psychometric ● In the U.S. sample, self-esteem was as-
the psychometric properties of this Con- properties. A sample of 200 Dutch type sessed by the Rosenberg Self-Esteem
fidence in Diabetes Self-Care (CIDS) 1 diabetic patients randomly selected questionnaire (29) with good reliability
scale. from 3,000 patients taking part in a large (Cronbach’s ␣ ⫽ 0.77– 0.88) (30). In
survey (24) (randomly selected of the the Dutch sample, the self-esteem sub-
40,000 members of the Dutch Diabetes scale of the Dutch Personality Inven-
RESEARCH DESIGN AND Association) received a set of self-report tory with satisfactory reliability
METHODS questionnaires, including the CIDS scale, (Cronbach’s ␣ ⫽ 0.74) (31) was used.
by mail. Written consent was obtained to ● To assess anxiety and depression, the
Development of the CIDS scale retrieve the most recent HbA1c level from anxiety and depression subscales of the
The CIDS scale was designed as a short the treating physician. A second sample of Symptom Check List 90-R (32) were
20-item self-report questionnaire (23) as- 192 U.S. type 1 diabetic patients attend- used in the U.S. sample, with good re-
sessing self-efficacy, the perceived ability ing a scheduled appointment at the out- liability (Cronbach’s ␣ ⫽ 0.90 for de-
to perform diabetes self-care tasks, in pa- patient clinic of the Joslin Diabetes Center pression and 0.84 for anxiety). The
tients with type 1 diabetes. The CIDS completed the self-report questionnaires Dutch sample filled out the Hospital
scale was constructed in Dutch and U.S. and gave permission to obtain HbA1c Anxiety and Depression Scale (Cron-
English simultaneously by two of the au- results. bach’s ␣ ⫽ 0.93 for anxiety and 0.90 for
thors (K.W. and F.J.S.). After exploring Study 2: Test-retest reliability. A third depression) (33,34).
existing measures referred to in the intro- sample of 62 type 1 diabetic patients vis- ● Diabetes self-care behaviors were as-
duction, items were constructed to cover iting the outpatient diabetes clinic of the sessed using the Self-Care Inventory
all domains of self-care (e.g., following Vrije Universiteit Medical Center com- (35) in the U.S. sample, with good reli-
recommendations about food, exercise, pleted the CIDS scale at the clinic. They ability (Cronbach’s ␣ ⫽ 0.87), and sim-
foot care, insulin administration, and self- received the CIDS scale for a second as- ilar items were used in the Dutch
monitoring of blood glucose [SMBG]; and sessment by mail. Because self-efficacy is sample (questions regarding frequency
self-regulation of blood glucose, e.g., ad- considered a dynamic construct, a short of taking into account dietary recom-
justing insulin and detecting and treating time interval (2 weeks) was chosen. mendations [1 ⫽ “never” and 5 ⫽ “al-
high and low levels of blood glucose). So- The study was approved by the Joslin ways”], SMBG [1 ⫽ “never” and 5 ⫽
cial skills (e.g., asking friends/relatives for Diabetes Center Committee on human “ⱖ5 times a day”], performing the pre-
help) were included because they are con- subjects and the Medical Ethics Commit- scribed number of daily insulin injec-
sidered an essential part of an active prob- tee of the Vrije Universiteit medical cen- tions [1 ⫽ “never” and 5 ⫽ “always”],
lem-solving approach to diabetes. Items ter, and written informed consent was frequency of exercise [1 ⫽ “never” and
were judged on content validity and obtained before participation. 5 ⫽ “⬎7 times a week”], adjusting in-
adapted if necessary by three additional sulin in special situations [1 ⫽ “never”
research psychologists working in the Measures and 5 ⫽ “always”], and inspection of
field of diabetes. Comprehensibility and In addition to the CIDS scale, patients in feet [1 ⫽ “never” and 5 ⫽ “⬎20 times a
ease of use were tested and found to be study 1 filled out several self-report ques- month”]) (36).
satisfactory in a small sample of Dutch tionnaires: ● For the Dutch sample, an overall diabe-
patients (n ⫽ 11). tes treatment self-efficacy rating was in-
Each item is preceded by, “I believe I ● Sociodemographical and clinical char- cluded, using the question, “I believe I
can. . . ,” with the strength of this belief acteristics, including age, sex, years of can. . .manage my diabetes well over-
rated on a 5-point Likert scale ranging education, duration of diabetes, and all,” rated on a scale from 1 (“No, I am
from 1 (“No, I am sure I cannot”) to 5 number of complications. sure I cannot”) to 5 (“Yes, I am sure I
(“Yes, I am sure I can”). An example item ● Diabetes-related emotional distress was can”) preceding the CIDS scale. In ad-
is, “I believe I can. . .adjust my insulin for assessed by the Problem Areas in Dia- dition, the Dutch sample rated the per-
exercise, traveling, or celebrations.” A to- betes (PAID) scale, a 20-item measure ceived burden and importance to
tal CIDS score is calculated by summation assessing a broad range of feelings re- prevent future complications of each
of all item scores and then transformed to lated to living with diabetes and its behavior on a 5-point scale (1 ⫽ “not
a 0 –100 scale (see STATISTICAL ANALYSIS), treatment, including guilt, anger, frus- burdensome/not important” and 5 ⫽
with higher scores indicating higher self- tration, depressed mood, worry, and “very burdensome/very important”).
efficacy. fear. The PAID scale proved to have
high internal consistency in U.S. (Cron- Glycemic control was assessed by de-
bach’s ␣ ⫽ 0.95) (25) as well as Dutch termining the percent of glycosylated he-
Subjects and procedures samples (Cronbach’s ␣ ⫽ 0.93– 0.95) moglobin. For the U.S. sample, one
For this study, three distinct samples were (26). laboratory in Boston was used (high-
used to examine the psychometric prop- ● Fear of hypoglycemia was assessed by performance liquid chromatography ion
erties of the CIDS scale. Inclusion criteria the Hypoglycemia Fear Survey (HFS) capture method; Tosoh Medics, San Fran-
for all samples were 1) age ⬎18 years and worry scale, with good reliability cisco, CA; reference: 4.0 – 6.0%). Because
2) having type 1 diabetes, defined as onset (Cronbach’s ␣ ⫽ 0.89 – 0.96) and va- multiple laboratories were used for the
of diabetes before age 40 years and treated lidity in U.S. (27) and Dutch samples Dutch sample, correlations with Diabetes
with insulin from diagnosis. (Cronbach’s ␣ ⫽ 0.92) (28). Control and Complications Trial target

714 DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003

 van der Ven and Associates

values were calculated and HbA1c values Table 1—Patient characteristics

were adjusted accordingly. Patients in
study 2 (test-retest reliability) filled out Dutch sample U.S. sample
the CIDS scale only.
n 151 190
Statistical analysis Age (years) 43.2 ⫾ 13.4 42.6 ⫾ 13.1
Statistical analyses were performed using Years of education 14.7 ⫾ 3.4 15.3 ⫾ 2.2
SPSS 9.0 for Windows (37). Values are Duration of diabetes (years) 21.8 ⫾ 13.0 22.2 ⫾ 13.5
expressed as means ⫾ SD. For ease of Sex (M/F) 73/78 75/115
comparison, all total scores on question- % with ⱖ1 complication 40 38
naires were transformed to a 0 –100 scale HbA1c (%) 8.1 ⫾ 1.3 8.3 ⫾ 1.5
(using the Medical Outcome Survey scor- Data are n, means ⫾ SD, or %. HbA1c was available for 118 and 145 patients in the Dutch and U.S. samples,
ing techniques, in which the patient raw respectively.
score minus the lowest possible score is
divided by the possible score range and
multiplied by 100 [38]). Missing values to obtain the most recent HbA1c level spectively (57.8% in total). Principal
on the CIDS scale (one missing: n ⫽ 15, from their physician. A total of 21 physi- components analysis with two to four
four missing: n ⫽ 1) and PAID scale (max- cians did not respond to this request, components showed most items loaded
imum of two missing: n ⫽ 9) were cor- leaving HbA1c values for 119 patients high on the first principal component
rected using the same Medical Outcome (59.5%). HbA1c results were available for (0.32– 0.72), with only two items loading
Survey scoring techniques (38). Cases 145 patients (76%) in the U.S. sample. ⬍0.40 (“exercise two to three times
with missing values on other question- Data from three participants (one Dutch, weekly” and “perform the prescribed
naires were excluded from the respective two U.S.) could not be used because of number of daily insulin injections”).
analyses. Analyses included Student’s t incomplete data on the CIDS scale. Data To assess whether items clustered
tests and ␹2 tests. Because scores on the were analyzed for 151 Dutch patients and into further meaningful underlying fac-
CIDS are not normally distributed, Spear- 190 U.S. patients. tors, forced factor solutions with two,
man’s correlation coefficients were esti- Sociodemographic and clinical char- three, and four varimax rotated factors
mated to determine associations between acteristics of both samples are displayed were examined, accounting for 40.3,
variables and for test-retest reliability. The in Table 1. 47.2, and 52.7% of total variance, respec-
Cronbach’s ␣ coefficient was determined The Dutch and U.S. samples had sim- tively. No meaningful interpretable fac-
for internal consistency. P ⬍ 0.05 was ilar characteristics and comparable mean tors could be distinguished. The initial
considered to be statistically significant. scores on the HFS worry scale (Dutch unrotated factor solution in the U.S. sam-
For comparisons on item level, P ⬍ 0.01 28.0 ⫾ 16.8, U.S. 30.7 ⫾ 20.3). The ple showed a similar pattern of high load-
was considered statistically significant. mean PAID score was higher in the U.S. ings on the first component, with “keep
Exploratory factor analysis with varimax sample (36.0 ⫾ 24.0 vs. 21.5 ⫾ 16.0, P ⬍ daily records of my blood glucose” and
rotation was performed to examine the 0.001), confirming earlier findings (26). “keep my medical appointments” loading
factor structure of the CIDS scale in the Reliability: internal consistency. In- ⬍0.40. Because analysis of reliability
Dutch and U.S. samples. To reduce skew- ternal consistency (standardized Cron- showed that Cronbach’s ␣ was not com-
ness and kurtosis caused by nonnormal- bach’s ␣) of the 20-item CIDS scale was promised by any of the items, it was de-
ity, CIDS scores were transformed before high in both samples (Dutch 0.86, U.S. cided to retain all 20 items in one single
analysis by squaring variables. Forced 0.90). Deletion of any of the items would scale.
four-, three-, two-, and one-factor solu- not result in an increase of Cronbach’s ␣ The item “performing the prescribed
tions were considered in the Dutch sam- ⬎0.01, indicating that the CIDS is a ho- number of insulin injections” did not load
ple to identify meaningful factors. Missing mogeneous scale. Item-total correlations substantially on any factor. This is proba-
values in this procedure were substituted were all positive and ranged from 0.22 to bly because of the very skewed scoring
by the mean. 0.62 (Dutch sample) and 0.32 to 0.67 distribution, with 94% (U.S. 86%) indi-
(U.S. sample). cating, “Yes, I am sure I can do this.”
RESULTS Factor structure. The CIDS was de- Validity. Mean CIDS scores and scoring
signed to measure the same construct of distributions were similar in both samples
Study 1: Evaluation of psychometric self-efficacy across a range of self-care be- (Dutch 0.83 ⫾ 11.5, U.S. 85.0 ⫾ 12.5),
properties haviors. Exploratory factor analysis was with some small but statistically signifi-
In the Dutch sample, questionnaires were performed on the Dutch sample to assess cant differences between U.S. and Dutch
returned by 152 of 200 patients (76%). whether one general underlying factor respondents (Table 2). Scoring distribu-
The reason for nonresponse was known could be identified. Inspection of the tions were very skewed, with 80% scoring
for only six patients (physical condition scree plot of the initial unrotated factor above 72.0 (Dutch) and 75.0 (U.S.), re-
did not allow completing the question- solution showed five components with spectively, indicating high levels of self-
naires [two patients], moved to an un- eigenvalues ⬎1 (6.4, 1.7, 1.4, 1.1, and efficacy. Items with highest mean scores
known address [three patients], question- 1.0), with a sharp elbow between the first were “perform the prescribed number of
naire was sent in too late [one patient]). two components, accounting for 31.9, daily injections” (Dutch sample, mean
Of the 152 responders, 140 gave consent 8.4, 6.9, 5.5, and 5.2% of variance, re- score 4.88 of 5) and “keep my medical

DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003 715

The CIDS scale

Table 2—Item content and means of the CIDS scale self-esteem (Dutch 0.16, P ⫽ 0.05; U.S.
0.35, P ⬍ 0.0001) and self-care behavior
Dutch sample U.S. sample (Dutch 0.44, U.S. 0.42, both P ⬍
0.0001). CIDS scores were associated
n 151 190 with HbA1c in the U.S. sample only (U.S.
I believe I can. . . ⫺0.25, P ⫽ 0.003; Dutch ⫺0.09, NS).
plan my meals and snacks according to dietary 4.24 ⫾ 0.78 4.31 ⫾ 0.93 In the Dutch sample, the positive as-
guidelines. sociation between CIDS scores and per-
check my blood glucose at least two times a day. 4.48 ⫾ 0.98 4.67 ⫾ 0.81 ceived importance of self-care behavior
perform the prescribed number of daily insulin 4.88 ⫾ 0.55 4.82 ⫾ 0.52 (0.37, P ⬍ 0.0001) and the negative asso-
injections. ciation with perceived burden of self-care
adjust my insulin for exercise, traveling, or celebrations. 4.78 ⫾ 0.53 4.52* ⫾ 0.79 behavior (⫺0.39, P ⬍ 0.0001) were also
adjust my insulin when I am sick. 4.44 ⫾ 0.85 4.34 ⫾ 0.88 in the expected directions.
detect high levels of blood glucose in time to correct. 4.28 ⫾ 0.90 4.34 ⫾ 0.92
detect low levels of blood glucose in time to correct. 4.17 ⫾ 0.89 4.21 ⫾ 0.94 Study 2: test-retest reliability
treat a high blood glucose correctly. 4.50 ⫾ 0.67 4.54 ⫾ 0.73 In the test-retest sample, response to the
treat a low blood glucose correctly. 4.45 ⫾ 0.68 4.59 ⫾ 0.71 second assessment after 2 weeks was
keep daily records of my blood glucose. 4.05 ⫾ 1.25 4.44* ⫾ 0.91 69%. Patients in this sample were
decide when it’s necessary to contact my doctor or 4.51 ⫾ 0.62 4.56 ⫾ 0.65 younger than the first Dutch sample
diabetes educator. (mean age 36.3 vs. 43.2 years, P ⫽
ask my doctor questions about my treatment plan. 4.72 ⫾ 0.54 4.69 ⫾ 0.63 0.001), whereas mean CIDS score and
keep my blood glucose in the normal range when 3.77 ⫾ 0.98 3.66 ⫾ 1.05 duration of diabetes were comparable.
under stress. Spearman’s correlation coefficient be-
check my feet for sores or blisters every day. 4.01 ⫾ 1.09 4.54* ⫾ 0.83 tween test and retest was 0.85 (P ⫽
ask my friends or relatives for help with my diabetes. 3.80 ⫾ 1.17 4.01 ⫾ 1.13 0.0001), indicating the CIDS scale is a sta-
inform colleagues/others of my diabetes, if needed. 4.42 ⫾ 0.79 4.45 ⫾ 0.91 ble short-term measure.
keep my medical appointments. 4.68 ⫾ 0.66 4.84† ⫾ 0.41
exercise two to three times weekly. 3.84 ⫾ 1.29 4.19† ⫾ 0.98 CONCLUSIONS — Results of this
figure out what foods to eat when dining out. 4.37 ⫾ 0.82 4.32 ⫾ 0.89 study support the reliability and validity
read and hear about diabetes complications without 4.05 ⫾ 1.08 4.02 ⫾ 1.07 of the CIDS scale in patients with type 1
getting discouraged. diabetes. Moderate correlations in the ex-
Total scale 83.0 ⫾ 11.5 85.0 ⫾ 12.5 pected directions support construct va-
Data are n or means ⫾ SD. Scoring range: from 1 (“No, I am sure I cannot”) to 5 (“Yes, I am sure I can”). *P ⬍ lidity while indicating that the CIDS scale
0.001, †P ⬍ 0.01. reflects a unique construct. The CIDS
scale demonstrated high psychometric
similarity in Dutch and U.S. samples,
appointments” (U.S. sample, mean score glucose in the normal range when under allowing for cross-cultural comparisons.
4.84 of 5). Items with lowest mean scores stress” (4.00 vs. 3.34, P ⬍ 0.0001). In the Examination of the factor structure
were similar in both samples: “keep my Dutch sample, men also had a higher showed that the CIDS scale is best consid-
blood glucose in the normal range when mean score on this last item (3.99 vs. ered and used as a unidimensional scale.
under stress,” with mean item scores of 3.56, P ⬍ 0.01) and a lower mean score Because all items relate to different as-
3.77 (Dutch) and 3.66 (U.S.), respec- on “keep my medical appointments” pects of diabetes self-management and in-
tively. This finding is consistent with (4.48 vs. 4.86, P ⬍ 0.0001). ternal consistency is not compromised by
former findings (39), in which “maintain- Total CIDS score correlated moder- any of them, redundancy does not seem
ing normal blood glucose levels when un- ately (0.51, P ⬍ 0.0001) with the overall to be a major concern. Research has con-
der stress” was considered a serious rating of diabetes treatment self-efficacy sistently found that diabetes self-care be-
problem by a large proportion (44%) of in the Dutch sample. haviors are relatively independent of one
the study participants and “injecting insu- Spearman correlation coefficients be- another and that assessing the specific ar-
lin at regular intervals before meals” and tween the CIDS scale and other measures eas is important in determining the need
“injecting outdoors” was considered no were all in the expected directions. CIDS for further support (40). The value of re-
problem by ⬃80% of the study scores were negatively associated with di- taining the present items is underlined by
participants. abetes-related emotional distress (PAID the moderate correlation between the
When exploring sex differences, scale) (Dutch ⫺0.44, U.S. ⫺0.52, both full-scale and the single general diabetes
mean score of U.S. men appeared higher P ⬍ 0.0001), fear of hypoglycemia (HFS treatment self-efficacy item. However, re-
than that of U.S. women (87.6 ⫾ 11.2 vs. worry scale) (Dutch ⫺0.16, P ⫽ 0.05; vision of some of the items might further
83.4 ⫾ 13.0, P ⫽ 0.02). On an item level, U.S. ⫺0.38, P ⬍ 0.0001), anxiety (Dutch enhance the utility of the CIDS scale. The
there were small differences. Self-efficacy ⫺0.24, P ⫽ 0.003; U.S. ⫺0.26, P ⬍ low discriminating power of the item con-
ratings were higher in men on the items 0.0001), and depression (Dutch ⫺0.17, cerning injection of insulin was recog-
“treat a low blood glucose correctly” (4.79 P ⫽ 0.04; U.S. ⫺0.26, P ⬍ 0.0001). CIDS nized earlier by other researchers (6).
vs. 4.46, P ⬍ 0.0001) and “keep my blood scores were positively associated with However, because correct administration

716 DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003

 van der Ven and Associates

of insulin is crucial for good control, we and surpass their usual level of accom- 147, 1991
consider this information valuable. Re- plishment (41,42). In a recent study, pa- 5. Senécal C, Nouwen A, White D: Motiva-
wording the item in more specific terms, tients with type 1 diabetes with more tion and dietary self-care in adults with
e.g., “I believe I can. . .always inject my positive efficacy expectancies did use diabetes: are self-efficacy and autono-
mous self-regulation complementary or
insulin on the correct time/. . .outdoors/ more adaptive coping and reported better competing constructs? Health Psychol 19:
. . . when at work,” etc., may enhance the mental health than patients holding a less 452– 457, 2000
variance on this item. The item “. . .read optimistic view (43). 6. Hurley CA, Shea CA: Self-efficacy: strat-
and hear about diabetes complications However, beliefs are not the only de- egy for enhancing diabetes self-care. Dia-
without getting discouraged” is referring terminants of behavioral change. High betes Educ 18:146 –150, 1992
to mood rather than behavior and might levels of self-efficacy regarding a particu- 7. Glasgow RE, Toobert DJ, Riddle M,
be better represented on a scale assessing lar behavior do not automatically mean a Donelly J, Mitchell DL, Calder D: Diabe-
mood and emotional problems, such as person will engage in that behavior: pa- tes-specific social learning variables and
the PAID scale. Whereas the topic of tients do not perform several blood glu- self-care behaviors among persons with
stress is relevant to the management of cose tests per day simply because they feel type II diabetes. Health Psychol 8:285–
303, 1989
diabetes for most patients, the item “I be- able to do so (13). Appropriate skills, ca- 8. Kavanagh DJ, Gooley S, Wilson PH: Pre-
lieve I can. . .keep my blood glucose in pabilities, knowledge, and incentives are diction of adherence and control in diabe-
the normal range when under stress” does required as well (44). tes. J Behav Med 16:509 –522, 1993
not clearly state whether stress manage- Although self-efficacy is not the only 9. Aalto A, Uutela A: Glycemic control, self-
ment or self-management skills are ad- explanatory factor, it can add substan- care behaviors, and psychosocial factors
dressed. This item might also be tially to our understanding of self-care be- among insulin treated diabetics: a test of
rephrased in future studies. havior: whereas self-judgements of an extended health belief model. Int J Be-
Data suggest that self-efficacy expec- efficacy may not presume individuals hav Med 4:191–214, 1997
tations are not automatically the same for readiness to follow treatment recommen- 10. Littlefield CH, Craven JL, Rodin GM,
men and women. In addition to the dations, they must be present for treat- Daneman D, Murray MA, Rydall AC: Re-
lationship of self-efficacy and bingeing to
higher mean score of men in the U.S. sam- ment adherence to occur (13). Enhancing adherence to diabetes regimen among ad-
ple, U.S. as well as Dutch men had a self-efficacy may increase patients’ moti- olescents. Diabetes Care 15:90 –94, 1992
stronger belief in their own capability to vation for—and success with— behav- 11. Ott J, Greening L, Palardy N, Holderby A,
keep their blood glucose within the nor- ioral efforts. DeBell WK: Self-efficacy as a mediator
mal range when under stress. These find- To assess self-efficacy beliefs in adult variable for adolescents’ adherence to
ings may be linked to the specific patients with type 1 diabetes, the CIDS treatment for insulin-dependent diabetes
difficulties women experience with unex- scale proves to be a valuable, reliable, and mellitus. Child Health Care 29:47– 63,
pected blood glucose fluctuations due to easy-to-administer instrument, leaving its 2000
hormonal changes. predictive value and responsiveness to 12. Talbot F, Nouwen A, Gingras J, Gosselin
The usefulness of the CIDS scale as a change to be demonstrated in future M, Audet J: The assessment of diabetes-
related cognitive and social factors: the
screening tool to identify patients with research. multidimensional diabetes questionnaire.
suboptimal levels of self-efficacy remains J Behav Med 20:291–312, 1997
open to question. Although glycemic con- 13. Griva K, Myers LB, Newman S: Illness
trol was associated with self-efficacy in Acknowledgments — This project was par- perceptions and self efficacy beliefs in ad-
the U.S. sample, self-efficacy was found tially supported by a grant from the National olescents and young adults with insulin-
to be more strongly associated with be- Institutes of Health (DK-42315) and the dependent diabetes mellitus. Psychol
Dutch Diabetes Research Fund (99.005). Health 15:733–750, 2000
havior than with outcome, with moderate
We thank Dr. C. Weykamp, Streekzieken- 14. Aalto A, Uutela A, Aro AR: Health related
correlations between CIDS scores and huis Koningin Beatrix, Winterswijk, the Neth-
self-care. This result is in line with our quality of life among insulin-dependent
erlands, for his cooperation in standardizing diabetics: disease-related and psychoso-
expectations because not all CIDS items Dutch HbA1c values to Diabetes Control and cial correlates. Patient Educ Couns 30:
have a direct link with blood glucose (e.g., Complications Trial standards. 215–225, 1997
foot inspection), and glycemic control is 15. Schlundt DG, Flannery ME, Davis DL,
determined by factors other than behav- Kinzer CK, Pichert JW: Evaluation of a
ior (e.g., genetics, treatment prescrip- References multicomponent, behaviorally oriented,
tions, etc.). However, discordance still 1. Bandura A: Self-Efficacy: The Exercise of problem-based “summer school” pro-
exists between what people feel they are Control. New York, W.H. Freeman, 1997 gram for adolescents with diabetes. Behav
able to do and what they actually do. 2. Maibach E, Murphy DA: Self-efficacy in Modif 23:79 –105, 1999
The overall high levels of self-efficacy health promoting research and practice: 16. Grey M, Boland EA, Davidson M, Yu
reported reflect that patients believe conceptualization and measurement. C, Sullivan-Bolyai S, Tamborlane WV:
strongly in their capabilities. It is assumed Health Educ Res 10:37–50, 1995 Short-term effects of coping skills training
3. McCaul KD, Glasgow RE, Schafer LC: Di- as adjunct to intensive therapy in adoles-
that self-efficacy beliefs need not be accu- abetes regimen behaviors: predicting ad- cents. Diabetes Care 21:902–908, 1998
rate to be adaptive, and they operate par- herence. Med Care 25:868 – 881, 1987 17. Rubin RR, Peyrot M, Saudek CD: The ef-
tially independent of actual skills (41). 4. Padgett DK: Correlates of self-efficacy be- fect of a diabetes education program in-
High levels of self-efficacy have been con- liefs among patients with non-insulin de- corporating coping skills training on
sidered adaptive because they stimulate pendent diabetes mellitus in Zagreb, emotional well-being and diabetes self-ef-
people to set higher goals and persevere Yugoslavia. Patient Educ Couns 18:139 – ficacy. Diabetes Educ 19:210 –214, 1993

DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003 717

The CIDS scale

18. Connell CM, Davis WK, Gallant MP, book of Psychology and Diabetes: A Guide to with poorly controlled insulin-dependent
Sharpe PA: Impact of social support, so- Psychological Measurement in Diabetes Re- (type 1) diabetes. Patient Educ Couns 45:
cial cognitive variables, and perceived search and Practice. Bradley C, Ed. Chur, 143–148, 2001
threat on depression among adults with Switzerland, Harwood Academic, 1994, 37. SPSS: SPSS Base 9.0 for Windows Users
diabetes. Health Psychol 13:263–273, 1994 p. 133–155 Guide. Chicago, SPSS, 2000
19. Hurley AC: Diabetes Health Beliefs and Self- 28. Snoek FJ, Pouwer F, Mollema ED, Heine 38. International Resource Center (IRC) for
Care of Individuals Who Require Insulin. RJ: The Dutch version of the Hypoglyce- Health Care Assessment: How to Score the
Boston, MA, Boston University, 1989 mia Fear Survey: internal consistency and MOS 36-Item Short-Form Health Survey
20. Anderson RM, Funnell MM, Fitzgerald validity. (De Angst voor Hypoglycemie (SF36). Boston, MA, New England Medi-
JT, Marrero DG: The diabetes empower- Vragenlijst [AHV]: Interne consistentie en cal Center Hospitals, 1991
ment scale: a measure of psychosocial validiteit.) Gedrag Gezondheid 24:287–
39. Mollema ED, Snoek FJ, Heine RJ: Assess-
self-efficacy. Diabetes Care 23:739 –743, 292, 1996
ment of perceived barriers in self-care of
2000 29. Rosenberg M: Conceiving the Self. New
21. Bijl JV, Poelgeest-Eeltink AV, Shortridge- York, Basic Books, 1979 insulin-requiring diabetic patients. Pa-
Baggett L: The psychometric properties of 30. Blascovich J, Tomaka J: Measures of self- tient Educ Couns 29:277–281, 1996
the diabetes management self-efficacy esteem. In Measures of Personality and 40. Ruggiero L, Glasgow RE, Dryfoos JM,
scale for patients with type 2 diabetes Social Psychological Attitudes. 3rd ed. Rossi JS, Prochaska JO, Orleans CT,
mellitus. J Adv Nurs 30:352–359, 1999 Robinson JP, Shaver PR, Wrightsman LS, Prokhorov AV, Rossi SR, Greene GW,
22. Grossman HY, Brink S, Hauser ST: Self- Eds. Ann Arbor, MI, Institute for Social Reed GR, Kelly K, Chobanian L, Johnson
efficacy in adolescent girls and boys with Research, 1993, p. 115–160 S: Diabetes self-management: self-re-
insulin-dependent diabetes mellitus. Dia- 31. Luteijn F, Starren J, van Dijk H: Manual to ported recommendations and patterns in
betes Care 10:324 –329, 1987 the Dutch Personality Questionnaire a large population. Diabetes Care 20:568 –
23. Weinger K, Ven NV, Yi J, Snoek F: Con- (Handleiding bij de NPV). Lisse, the 576, 1997
fidence in Diabetes Selfcare (CIDS) scale: Netherlands, Swets en Zeitlinger, 1975 41. Bandura A: Self-regulation of motivation
first psychometric evaluation in USA and 32. Derogatis LR, Rickels K, Rock A: The SCL- and action through goal systems. In Cog-
Dutch type 1 diabetes patients (Abstract). 90-R: Administration, Scoring and Proce- nitive Perspectives on Emotion and Motiva-
Diabetes 49 (Suppl. 1):A1327, 2000 dures Manual II. Baltimore, MD, Clinical tion. Hamilton V, Bower GH, Frijda NH,
24. Pouwer F, Van der Ploeg HM, Adèr HJ, Psychometric Research, 1983 Eds. Dordrecht, the Netherlands, Kluwer
Heine RJ, Snoek FJ: The 12-item Well- 33. Snaith RP, Zigmond AS: The Hospital Anx- Academic, 1988, p. 37– 61
Being Questionnaire (WBQ-12): an eval- iety and Depression Scale Manual. Wind- 42. Maddux JE, Lewis J: Self-efficacy and ad-
uation of its validity and reliability in sor, U.K., NFER-Nelson, 1994 justment: basic principles and issues. In
Dutch people with diabetes. Diabetes Care 34. Spinhoven P, Ormel J, Sloekers PPA, Self-Efficacy, Adaptation, and Adjustment:
22:2004 –2010, 1999 Kempen GIJM, Speckens AEM, van Theory, Research, and Application. Maddux
25. Polonsky WH, Anderson BJ, Lohrer PA, Hemert AM: A validation study of the
JE, Ed. New York, Plenum Press, 1995, p.
Welch G, Jacobson AM, Aponte JE, Hospital Anxiety and Depression Scale
37– 68
Schwartz CE: Assessment of diabetes-re- (HADS) in different groups of Dutch sub-
lated distress. Diabetes Care 18:754 –760, jects. Psychol Med 27:363–370, 1997 43. Fournier M, de Ridder DTD, Bensing JM:
1995 35. Greco P, LaGreca AM, Ireland S, Wick P, How optimism contributes to the adapta-
26. Snoek FJ, Pouwer F, Welch GW, Polon- Freeman C, Agramonte R, Gutt M, Skyler tion of chronic illness: a prospective study
sky WH: Diabetes-related emotional dis- J: Assessing adherence in IDDM: a com- into the enduring effects of optimism on
tress in Dutch and U.S. diabetic patients: parison of two methods (Abstract). Diabe- adaptation moderated by the controllabil-
cross-cultural validity of the Problem Ar- tes 40 (Suppl. 1):A657, 1990 ity of chronic illness. Pers Indiv Differ 33:
eas in Diabetes Scale. Diabetes Care 23: 36. Snoek FJ, van der Ven NCW, Lubach C, 1163–1183, 2002
1305–1309, 2000 Chatrou M, Adèr HJ, Heine RJ, Jacobson 44. Bandura A: Self-efficacy: toward a unify-
27. Irvine A, Cox DJ, Gonder-Frederick L: AM: Effects of Cognitive Behavioural ing theory of behavioral change. Psychol
The Fear of Hypoglycemia Scale. In Hand- Group Training (CBGT) in adult patients Rev 84:191–215, 1977

718 DIABETES CARE, VOLUME 26, NUMBER 3, MARCH 2003