OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 1

O RIGINAL A RTICLE

Development and validation of a diabetes

knowledge questionnaire
CA Eigenmann*, T Skinner, R Colagiuri

Introduction
Improving methods and strategies ABSTRACT
for evaluating diabetes patient An Australian National Consensus Position on Outcomes and Indicators for Diabetes
education interventions has been Education identified knowledge and understanding as the outcomes most directly
increasingly advocated by both affected by diabetes education. A subsequent literature review failed to identify a
validated, suitable questionnaire for measuring knowledge. Consequently, we aimed to
provider and consumer organisa-
develop a minimum diabetes knowledge questionnaire (DKQ) suitable for people with
tions globally.1–3 both type 1 and type 2 diabetes.
Evaluation efforts have tradition- Content validity was established through literature review, Delphi survey of 52
ally been hampered by lack of formal opinion leaders and a workshop of Australian Diabetes Educators (n ≥300). The resulting
agreement about goals and systemati- instrument was tested for internal consistency on 129 and for reliability on 57 people with
cally derived benchmarks4–9 and type 1 and type 2 diabetes, respectively.
lack of standardised and validated The final questionnaire contains: 12 multiple choice questions common to type 1 and
evaluation measures.10–12 To address type 2 diabetes, e.g. normal blood glucose levels, complications, diet, exercise, self-
this problem, Diabetes Australia monitoring of blood glucose, annual check-ups, support services, and sick-days; two
developed an Australian National questions for people on oral medication/insulin only; and one question (sick-days) for
people with type 1 diabetes only.
Consensus Position on Outcomes
For the first 12 questions, the internal consistency was good (Cronbach’s α=0.73);
and Indicators for Diabetes Patient with the additional item for type 1 diabetes, the internal consistency was slightly better
Education (O&I).13,14 The O&I con- (α=0.79) as it was with the additional items for people on medication/insulin (α=0.76). No
sensus identified four patient-centred particular item seemed to adversely affect the overall consistency of the questionnaire.
key outcome areas of knowledge and Comparing test-retest pilots, total scores showed good reliability with no evidence of
understanding, self-management, change over time (t=1.73; df=56; p<0.85), and a correlation of 0.62.
self-determination and psychological The DKQ is now ready to use for evaluating knowledge outcomes of diabetes
adjustment expressed in the order in education. Copyright © 2011 John Wiley & Sons.
which each area is most influenced by Practical Diabetes Int 2011; 28(4): 166–170
diabetes education.
A subsequent study identified KEY WORDS
outcome; measure; knowledge; questionnaire; diabetes education
and evaluated available tools on their
ability and suitability for measuring
changes in the four key outcomes.15 interventions.18–20 As some questions longer reflect current guidelines
While three knowledge assessment do not reflect current Australian and standards of care (e.g. they refer
tools were identified, none met all of guidelines (e.g. a free food is to urine sugar testing which is no
the systematically derived quality not defined as having less than longer recommended) and, like the
appraisal criteria. The first of the 20 calories per serving), and uses DKT, they contain many questions
three tools, the diabetes knowledge terminology not used in the regarding insulin therapy which did
test (DKT), was developed and vali- Australian context (e.g. an insulin not fit our criteria for a generic tool
dated in the mid-1980s by the reaction), the DKT is not suitable for that is applicable to all people with
Michigan Diabetes Research and use in Australia. type 1 or 2 diabetes. The third iden-
Training Centre to address the need In 1984, a series of three diabetes tified validated knowledge question-
for a valid and reliable diabetes knowledge assessment scales naire, the ADKnowl, was developed
specific knowledge instrument that (DKNA, DKNB and DKNC, each of and tested in the UK.22 It consists of
could be used by diabetes educators 15 items) were developed and vali- 23-item sets with a total of 104 ques-
and researchers.16,17 Later scales dated for the Australian environ- tions/items which makes it a more
have been adapted from the ment.21 None of the DKN scales has comprehensive, thus a more oner-
DKT to suit particular groups and since been updated and they no ous and resource intensive tool for

Cecile A Eigenmann, RN, CDE, MPH, Policy & Management, Sydney University – 26 Arundel St, Sydney, New South Wales
Australian Diabetes Council, Sydney, New Health and Sustainability Unit – Menzies 2037, Australia; email:
South Wales, Australia Centre for Health Policy, Sydney, New [email protected]
Timothy Skinner, BSc Hons, PhD, Flinders South Wales, Australia
University Rural Clinical School, Mt Received: 9 February 2011
Gambier, South Australia, Australia *Correspondence to: Cecile A Accepted in revised form: 21 March 2011
Ruth Colagiuri, RN, BEd, Grad Cert Health Eigenmann, Australian Diabetes Council,

166 Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 2

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

application in a clinical setting, and as to whether it should be included profit organisation located in the
hence did not meet our criteria for a in the final questionnaire (from Sydney central business district; and
brief tool. strongly agree to strongly disagree), (c) from one private diabetes educa-
Given the lack of an up-to-date, whether the question should be tor setting. Health professionals
validated instrument, we aimed to included if the phrasing was modi- (primarily diabetes educators) were
develop, pilot and validate a generic, fied (from strongly agree to strongly asked to distribute the pilot ques-
brief DKQ capable of measuring disagree) and, if they had to choose tionnaires to consecutive attendees
knowledge change following a dia- between two questions referring at regular clinic visits or education
betes education intervention and to the same domain, which one programmes.
which would be suitable for people they would prefer to see included in Reliability testing (test-retest)
with both type 1 and 2 diabetes. the questionnaire. was carried out by random selection
of 300 people with type 1 or 2 dia-
Methods Diabetes educators’ consultation betes from the DANSW member-
Ethical approval for the study was workshop. To further test content ship database. Letters of invitation
obtained from the University of and face validity, the 20-item draft were mailed out and consenting
Sydney’s Human Ethics Research questionnaire was presented at a respondents were sent identical
Committee. Although the definition consultation workshop comprising questionnaires on two different
of knowledge is a matter of on-going members of the Australian Diabetes occasions with a two-week interval.
debate, for the purpose of this Educators Association (ADEA) dur- Instructions accompanied the initial
project the following definition was ing a national diabetes conference questionnaire requesting partici-
adopted: ‘knowledge is the confi- in 2007.24 Participants were asked to pants not to take part in diabetes
dent understanding of a subject, discuss the questionnaire in small education nor consult books,
potentially with the ability to use it groups, and indicate on a standard websites or colleagues, within the
for a specific purpose’. survey, for each of the 20 questions, two-week interval.
whether the questions should be
Design of the questionnaire included in the final questionnaire Statistical analysis
Delphi survey. A purposive sample (yes or no response), and any com- Questionnaires were scored, with
of 52 nationally recognised diabetes ments they had with regard to the each correct answer worth 1 point,
educators, dietitians, endocrin- phrasing of the questions. each ‘unsure’ answer worth 0.5 and
ologists, podiatrists, psychologists, an incorrect answer received no
primary care physicians, and Responses from Delphi surveys and points. This scoring system allows
researchers were invited to partici- the ADEA workshop were coded, for all responses to be added in the
pate in an email Delphi survey. The analysed and collated into a pilot total and is based on the premise
Delphi method is widely used in the questionnaire. that it is preferable for the partici-
development of research scales and The readability of the question- pant to recognise that they are
questionnaires as a systematic, inter- naire was assessed by using the ‘unsure’ of the answer, than for
active method where a panel of Flesch Reading Ease score.25 The them to ‘think’ they know the
experts answers questionnaires in test is based on the average number correct answer when in fact they
two or more rounds.23 In the first of syllables per word and words are incorrect.
round, participants were asked to per sentence with scores ranging To determine internal consis-
indicate, on a standard question- from 0–100. The higher the tency, item-to-total correlations and
naire, their opinion on the most score the easier a document is to Cronbach’s coefficient alpha (α)
important topics to be included in comprehend, with 60–70 consid- were used. For these reasons, it is
the questionnaire. The second ered an acceptable score for suggested that Cronbach’s α should
round consisted of 20 questions with literate adults. be above 0.70 but not higher than
five multiple choice options, devel- 0.90.26 Pearson’s r correlation co-
oped by the researcher (CAE) based Testing the questionnaire efficient was used to examine the
on the domains that were answered For all pilot tests, a detailed study relationship between the total
as ‘very important’ by ≥60% of information sheet was provided to knowledge score (calculated by
first-round survey respondents. All prospective participants and written summing the scores of all items) and
questions were based on current consent was obtained. age, type and duration of diabetes.
Australian clinical and/or educa- People with type 1 and 2 diabetes Descriptive statistics were used to
tional guidelines for the care of were recruited from: (a) three dia- describe the sample.
people with type 1 and 2 diabetes. betes centres attached to large Stability (test-retest reliability)
Additional domains identified were metropolitan teaching hospitals in was assessed using Pearson’s product-
only included if they were listed by Sydney; (b) attendees of group moment correlation coefficient
≥15% of respondents. Comments education programmes held at and paired t-tests to examine consis-
were incorporated into the second- Diabetes Australia – New South tency and stability of responses.
round questionnaires. Wales (DANSW [since July 2010 Commonly cited minimal standards
Participants were asked to indi- trading as Australian Diabetes for reliability coefficients are 0.70 for
cate their opinion of each question Council]), a charitable, not for group data.26

Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons 167
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 3

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Table 1. Results from the first round Table 2. Consensual comments expressed by ADEA workshop participants
of Delphi – domains to be included
in the knowledge questionnaire • Don’t use ‘false’ options; keep questions positive
• Simplify language to accommodate for people with low literacy level
Domains identified as ‘very • A question is needed with regard to carbohydrate food
important’ by ≥60% of Delphi • Separate questionnaire or section for type 1 and type 2 diabetes in
respondents particular with regard to sick-day management (including ketone testing),
1. Normal range of blood glucose hypoglycaemia treatment and medications (i.e. insulin)
level
2. Duration/chronic nature Table 3. Pilot diabetes knowledge questionnaire: question topics
3. Consequences short- and
long-term 12 questions common to people 12. Support services – National
4. Nutrition with type 1 and 2 diabetes: Diabetes Services Scheme
5. Physical activity 1. Ideal blood glucose levels (NDSS – a government
6. Foot care 2. HbA1c subsidised support scheme)
7. Medication taking 3. Chronic nature (no cure)
8. Sick-day management 4. Dietary guidelines Two questions to be completed by
9. Self-monitoring of blood glucose 5. Benefits of physical activity people on oral medication and/or
10. Appropriate attendance for 6. Frequency of physical activity insulin only:
medical care 7. General diabetes long-term 13. Diabetes medication
11. Support services complications 14. Hypoglycaemia
8. Diabetes foot complications
For people on insulin therapy 9. Self-monitoring of blood glucose One question to be completed by
and/or insulin stimulating tablets 10. Sick-day management people with type 1 diabetes only:
12. Hypoglycaemia 11. Annual check-ups 15. Sick-day management
13. Medication
14. Sick-day management emphasise the need for ketone test- The final questionnaire, incorpo-
ing as a self-care requirement rating responses from the Delphi
Additional domains identified as unique to type 1 diabetes. survey and the ADEA workshop,
‘very important’ by ≥15% of Two additional topics (i.e. family contains 15 questions: nine ques-
Delphi respondents adjustment/support and mental tions eliciting a ‘one answer’
15. Family adjustment/support health), indicated as very important response to five multiple choice
16. Effects on and management of by ≥15% of respondents, were also options, and six questions eliciting
mental health incorporated into the round-two an ‘as many as apply’ to six multiple
questionnaire. choice options. An ‘unsure’ option
All responses were collated, Twenty questions with five multi- was included for each question.
coded and analysed by using the ple choice options, eliciting either Table 3 lists the question topics:
Statistical Package for the Social ‘one false answer’ or ‘one correct 12 topics common to people with
Sciences (SPSS 13.0, SPSS Inc, answer’ response, were developed by type 1 and 2 diabetes, two questions
Chicago, IL, USA). the researcher based on the 16 for people taking diabetes medica-
agreed topics in round one. tion/insulin, and one question for
Results Synthesis and analysis following those with type 1 only.
Content validity round-two surveys resulted in the Seven demographic questions
Two rounds of the Delphi survey deletion of the items measuring were added including: age, gender,
were conducted for the develop- mental health and family support as duration and type of diabetes, type
ment of questionnaire items. The these were difficult to phrase and of medication (if any) and previous
response rate for the first round was received negative feedback from the visit to a diabetes educator and/
71% (37/52) and 62% for round majority of respondents. or dietitian.
two (23/37). A number of questions were Flesch test readability score for
In round one, 14 of 20 domains rephrased to incorporate respon- the total questionnaire was 66.5.
were identified as ‘very important’ dents’ constructive suggestions for
by ≥60% of respondents (see Table simplifying the language and short- Pilot testing
1). A large number of key opinion ening the questions. In all, 129 people completed the
leaders commented that only people The ADEA workshop was pilot test of whom 47% were male
treated on diabetes medication attended by over 300 people. A total and 39% had type 1 diabetes; 10%
and/or insulin needed to know of 64 questionnaires were returned, had never seen a diabetes educator
about hypoglycaemia and medica- each resulting from a group discus- and 13% had never seen a dietitian.
tion. Similarly, respondents indi- sion of between three and seven
cated that sick-day management participants. Results of summarised Internal consistency. Analysis of the
should include a separate question consensual comments are illustrated 129 pilots showed good internal con-
for people with type 1 diabetes to in Table 2. sistency with a Cronbach’s α of 0.73

168 Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 4

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

for the first 12 questions, common

to all individuals. Appendix 1 (avail- Key points
able online at www.practicaldiabetes
international.com) lists for each • An Australian National Consensus Position on Outcomes and Indicators for
question the scale’s mean and Diabetes Patient Education identified knowledge and understanding as the
variance if the item is deleted. With key outcomes most directly influenced by diabetes education
the additional question for type 1 • A subsequent literature review failed to identify a validated, suitable
diabetes regarding sick-day manage- questionnaire for measuring diabetes knowledge
ment, the internal consistency was • We developed, piloted and validated a generic, brief diabetes knowledge
slightly better (α=0.79). Similarly, questionnaire for application before and after education interventions and
Cronbach’s α was slightly better at suitable for people with both type 1 and type 2 diabetes
0.76 with the two additional ques- • Methods to establish content validity included a literature review, a Delphi
tions regarding medication and survey of national key opinion leaders and a workshop of Australian
hypoglycaemia. Diabetes Educator Association members
No question seemed to particu- • The resulting 15-item diabetes knowledge questionnaire (DKQ), with an
larly adversely affect the overall additional seven demographic questions, was pilot tested for internal
consistency of the questionnaire. consistency and for stability (test-retest) on 129 and 57 people with type 1
Although seven items would and type 2 diabetes, respectively
improve the consistency in respond- • Results showed good reliability and internal consistency and a highly
ing if removed, none would increase acceptable ‘Flesch Reading Ease’ score, hence making the tool applicable
the alpha by more than 0.01, and so to people with a low literacy level
do not warrant removing. • The DKQ can now be applied by health professionals as part of quality
assurance and improvement for assessing their own diabetes education
Test-retest reliability. Comparing the practice and as a research tool
first and second rounds of the 57
completed pilots, total scores making.27–31 Others advocate that highly acceptable, hence making the
showed good reliability with no evi- acquiring adequate knowledge of tool applicable to people with a low
dence of change over time (t=1.73; this chronic illness is the corner- literacy level.
df=56; p<0.85), with a test-retest stone for enabling people towards The importance of consumer
correlation of 0.62. diabetes self-management and input has been increasingly advo-
hence prevention of ill health.20 The cated by consumer and health care
Construct validity. In support of the O&I Consensus Position13,14 identi- provider organisations. Although
tool’s construct validity, there was no fied knowledge as the outcome the Delphi survey did not specifi-
significant difference between males mostly affected by diabetes educa- cally obtain input from people with
and females (t=1.07; df=97; p>0.28) tion. Nonetheless, a further three diabetes, the draft questionnaire
and no significant association with key outcomes, i.e. self-management was pilot tested on a small sample of
age (r=0.04) or duration of diabetes behaviours, self-determination and people with diabetes, and their feed-
(r=0.12). However, as would be psychological adjustment, were iden- back was sought and incorporated
expected, individuals with type 1 dia- tified and recommended to form before the large scale pilot test was
betes scored slightly higher (mean part of the outcome assessments of conducted. Further, a number of
29 [SD 3.7]) than individuals with diabetes education interventions – Delphi and ADEA workshop partici-
type 2 diabetes (mean 26 [SD 5.0]; albeit with appropriate validated pants who responded in their
t=2.48; df=94; p<0.01). instruments.15 professional capacity also have dia-
The DKQ is unique in the way in betes themselves.
Discussion which it includes separate questions Valid outcome data can only be
The purpose of this study was to for people not taking diabetes med- achieved if education providers
develop a valid and reliable question- ication (12 items), people taking assure adequate item to content cov-
naire capable of assessing the effect diabetes medication and/or insulin erage, i.e. all questions in the DKQ
of a diabetes education intervention (two items) and an additional item need to be covered during an educa-
on knowledge of diabetes and its for people with type 1 diabetes only tion intervention.
self-management requirements in (total 15 items), therefore making it Further testing for the scale’s
people with type 1 and 2 diabetes. a more widely applicable tool. responsiveness/sensitivity to change
Although it is well acknowledged The DKQ is a brief, 15-item through an education session is
that knowledge acquisition does not knowledge questionnaire with seven now needed. This is currently
readily translate into behaviour additional demographic questions underway at the Australian Diabetes
change, a myriad of studies discuss added if required, taking between Council where the DKQ is
the need for assessing knowledge as 5–15 minutes to complete – hence applied to assess the effect of an
an important measure of effective- making it feasible to apply in a busy ongoing diabetes group education
ness of diabetes educational inter- clinical setting. programme on type 2 diabetes
ventions and as a prerequisite The readability, using the ‘Flesch patients’ knowledge before and
for informed health decision Reading Ease’ test, was considered after the programme.

Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons 169
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 5

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Limitations Heritage Foundation for Medical ing outcomes of diabetes education?

Although we do not claim that the Research, 2001: IHTA 23 Series A. Diabet Med 2009;26:425–36.
DKQ will be applicable globally, it is 2. Peeples M, et al. The Conceptual 16. Hess GE, Davis WK. The validation of
important to note that the question- Framework of the National Diabetes a diabetes patient knowledge test.
naire is based on international3 Education Outcomes System Diabetes Care 1983;6(6):591–6.
(NDEOS). Diabetes Educ 2001; 17. Fitzgerald JT, et al. The reliability and
and Australian evidence-based 27(4):547–62. validity of a brief diabetes knowledge
guidelines which have been derived 3. International Diabetes Federation. test. Diabetes Care 1998;21(5):706–10.
from the international peer International Standards for Diabetes 18. van den Arend IJM, et al. Education
reviewed literature.6 Education, 3rd edn. Brussels: IDF, 2009. integrated into general practice care
In the face of ongoing changes in 4. Majumdar SR, et al. A Canadian for type 2 diabetic patients results in
diabetes standards/guidelines of Consensus for the Standardized sustained improvement of disease
care and technologies, the DKQ will Evaluation of Quality Improvement knowledge and self-care. Diabet Med
need regular revision to assure con- Interventions in Type 2 Diabetes. 2000;17:190–7.
tent accuracy and re-validation if Can J Diabetes 2005;29(3):220–9. 19. Garcia AA, et al. The Starr County
changes are made. 5. Sigurdardottir AK, et al. Outcomes of Diabetes Education Study – develop-
educational interventions in type 2 ment of the Spanish-language dia-
diabetes: WEKA data-mining analysis. betes knowledge questionnaire.
Conclusion Patient Educ Couns 2007;67(1–2): Diabetes Care 2001;24(1):16–21.
The Diabetes Knowledge Question- 21–31. 20. Persell SD, et al. Relationship of dia-
naire has undergone rigorous valida- 6. Colagiuri R, et al. National Evidence betes-specific knowledge to self-man-
tion and has shown good reliability Based Guidelines for Patient agement activities, ambulatory pre-
and internal consistency; it can now Education in Type 2 diabetes. ventative care, and metabolic out-
be applied by health professionals Diabetes Australia and the NHMRC, comes. Prev Med 2004;39:746–52.
as part of quality assurance and Canberra 2009. Available from www. 21. Dunn SM, et al. Development of the
improvement for assessing their own nhmrc.gov.au/_files_nhmrc/file/ diabetes knowledge (DKN) scales:
diabetes education practice and as a publications/synopses/di16- forms DKNA, DKNB, and DKNC.
diabetes-patient-education.pdf [last Diabetes Care 1984;7(1):36–41.
research tool. accessed 12 Jan 2011]. 22. Speight J, Bradley C. ADknowl: iden-
7. Glasgow RE, Osteen VL. Evaluating tifying knowledge deficits in diabetes
Appendices 2 and 3 (available online at diabetes education. Are we measur- care. Diabet Med 2001;18:626–39.
www.practicaldiabetesinternational.com) ing the most important outcomes? 23. Duffield C. The Delphi technique.
show the text of the final Diabetes Diabetes Care 1992;15(10):1423–32. Aust J Adv Nurs 1988;6:41–5.
Knowledge Questionnaire and scoring 8. Home P, et al. Health Outcome 24. Eigenmann C, et al. Measuring out-
instructions, respectively. Indicators: Diabetes. Report of a comes of diabetes education (MODE).
A formatted version of the full ques- working group to the Department of A Consultation Workshop towards a
tionnaire is also available from http:// Health. National Centre for Health National Consensus. Australian
www.australiandiabetescouncil.com/ Outcomes, Oxford, 1999. Diabetes Society/Australian Diabetes
9. Naqib J. Patient education for effec- Educators Association Annual
Resources/PDFs/Diabetes-Knowledge- tive diabetes self-management: Scientific Meeting, Christchurch, New
Questionniare-DKQ-2009.pdf. report, recommendations and exam- Zealand, Sept 2007.
ples of good practice. London: 25. Flesch RF. A new readability yard-
Acknowledgements Diabetes UK, 2002. stick. J Appl Psychol 1948;32:221–33.
We would like to thank the people 10. Scientific Advisory Committee of the 26. Fitzpatrick R, et al. Evaluating patient-
with diabetes who participated in the Medical Outcomes Trust. Assessing based outcome measures for use in
pilot testing, the members of the health status and quality-of-life instru- clinical trials. Health Technol Assess
Delphi expert panel and the diabetes ments: attributes and review criteria. 1998;2(14):i–iv, 1–74.
educators who participated in the Qual Life Res 2002;11:193–205. 27. Tomky D, et al. Diabetes education
workshop for their valuable input. We 11. Ellis SE, et al. Diabetes patient educa- outcomes: what educators are doing.
tion: a meta-analysis and meta-regres- Diabetes Educ 2000;26(6):951–4.
are also very grateful for the assis- sion. Patient Educ Couns 2004; 28. Koopman DJEM, van der Bijl JJ. The
tance of diabetes educators from dia- 52(1):97–105. use of self-efficacy enhancing meth-
betes centres and health professionals 12. Mokkink LB, et al. Protocol of the ods in diabetes education in the
at the Australian Diabetes Council COSMIN study: COnsensus-based Netherlands. Sch Inq Nurs Pract 2001;
who assisted with the pilot testing. Standards for the selection of health 15(3):249–57.
Measurement INstruments. BMC 29. Mulcahy K, et al. Diabetes self-manage-
Declaration of interests Med Res Methodol 2006;6:2. ment education core outcomes meas-
This project was partially funded 13. Eigenmann C, Colagiuri R. Outcomes ures. Diabetes Educ 2003;29(5):768–70.
by a 2006 Australian Diabetes and Indicators for Diabetes Education: A 30. Mensing C, et al. National standards
Society–Servier National Action National Consensus Position. Canberra: for diabetes self-management educa-
Diabetes Australia, 2007. tion. Diabetes Care 2003;26(Suppl
Plan Grant. 14. Colagiuri R, Eigenmann C. A 1):S149–56.
national consensus on outcomes and 31. Heisler M, et al. The relationship
References indicators for diabetes patient educa- between knowledge of recent HbA1c
1. Corabian P, Harstall C. Patient dia- tion. Diabet Med 2009;26:442–6. values and diabetes care understand-
betes education in the management of 15. Eigenmann C, et al. Are current psy- ing and self-management. Diabetes
adult type 2 diabetes. Alberta: Alberta chometric tools suitable for measur- Care 2005;28(4):816–22.

170 Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 6

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Appendix 1. Diabetes Knowledge Questionnaire pilot test (n=129): results of internal consistency

Question Question topic Corrected item: Scale variance Scale mean Cronbach’s α
(item) total correlation if item deleted if item deleted if item deleted
no.

1 Ideal range of blood glucose level (BGL) 34.55 18.273 0.197 0.725
2 HbA1c result indicating lowest risk 34.64 17.855 0.128 0.734
3 Diabetes chronic nature/no cure 34.64 18.255 0.131 0.729
4 Which statement re diabetes diet is true 34.55 17.473 0.519 0.712

Benefits of physical activity (PA):

5a Controls BGLs 34.64 18.455 0.073 0.732
5b Lowers blood pressure 34.82 16.564 0.500 0.705
5c Affects mood 35.27 16.818 0.570 0.704
5d Skin cancer risk 34.45 18.873 0.000 0.730
5e Lowers cholesterol 34.91 15.291 0.810 0.679
6 How often exercise or PA 34.45 18.873 0.000 0.730

Well-controlled diabetes reduces risk of:

7a Kidney damage 34.64 16.855 0.558 0.705
7b Blindness 34.45 18.873 0.000 0.730
7c Melanoma 34.45 18.873 0.000 0.730
7d Heart disease 34.73 17.018 0.425 0.711
7e Foot ulcers 34.64 16.855 0.558 0.705
7f Unsure 34.45 18.873 0.000 0.730

Foot problems most at risk of:

8a Poor circulation 34.55 17.473 0.519 0.712
8b Loss of feeling in the feet 34.55 18.073 0.277 0.722
8c Foot ulcers 34.91 16.891 0.398 0.712
8d Hammer toe 34.55 18.873 -0.035 0.735
8e Infections 34.64 18.255 0.131 0.729
8f Unsure 34.45 18.873 0.000 0.730
9 Why advised to self-monitor BG 34.82 16.564 0.500 0.705

What to do if ill:
10a Check BGLs 34.55 18.273 0.197 0.725
10b Stop all diabetes medications/insulin 34.45 18.873 0.000 0.730
10c Drink lots of unsweetened fluids 35.09 19.691 -0.240 0.755
10d Seek medical attention if very unwell 34.55 18.073 0.277 0.722
10e Exercise to lower BGLs 34.64 18.255 0.131 0.729
10f Unsure 34.45 18.873 0.000 0.730
11 Frequency of medical check-ups 34.82 18.764 -0.033 0.742

The National Diabetes Services Scheme:

12a Allows purchase of BG testing strips 34.64 16.655 0.622 0.701
12b Offers to provide free syringes 34.82 16.564 0.500 0.705
12c Is for low income earners only 34.45 18.873 0.000 0.730
12d Is for all types of diabetes 34.45 18.873 0.000 0.730
12e Is free to join 34.82 17.964 0.153 0.730
12f Unsure 34.55 17.473 0.519 0.712
13 Which statement re medications is true 34.45 18.873 0.000 0.730
14 Hypoglycaemia treatment 34.45 18.873 0.000 0.730

Type 1: if feeling unwell and unable to eat:

15a Check BG 34.64 18.855 -0.041 0.738
15b Drink carbohydrate fluid if BGL <15mmol/L 35.45 18.873 0.000 0.730
15c Hospital if vomiting/diarrhoea 34.82 18.964 -0.079 0.745
15d Stop taking all insulin 34.55 19.073 -0.110 0.738
15e Medical help to adjust insulin doses 34.73 19.418 -0.185 0.750
15f Unsure 34.45 18.873 0.000 0.730

For questions 1–4, 6, 9,11,13,14 – one correct answer only; all other questions – as many as apply answers.

Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons 170a
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 7

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Appendix 2. Text of the Diabetes Knowledge Questionnaire

DIABETES KNOWLEDGE QUESTIONNAIRE

TEST YOUR KNOWLEDGE AND UNDERSTANDING OF DIABETES AND ITS MANAGEMENT

YOUR NAME: ..........................................

Dear participant
Please read the instruction for each question carefully as the required responses change from
‘circle ONE answer only’ to ‘circle AS MANY as apply’.
Please assist us by answering ALL questions on EVERY page.

1. What is the ideal range for blood glucose (sugar) c. It can help to regulate a person’s mood
levels a person with diabetes should aim for? d. It can reduce the risk of skin cancer
Please circle ONE answer only e. It can lower cholesterol levels
a. 2 to 6mmol/L f. Unsure
b. 7 to 13mmol/L
c. 4 to 8mmol/L 6. How often should people with diabetes exercise or
d. 4.5 to 15mmol/L be physically active?
e. Unsure Please circle ONE answer only
a. Most days of the week for at least 30 minutes
2. A blood test called HbA1c (or A1c) measures the b. Once a week for at least 30 minutes
average blood glucose levels over the past 2 to 3 c. Once a month for one hour
months. What is the HbA1c result that indicates a d. At least every fortnight for two hours
lowest risk of developing long-term diabetes e. Unsure
complications?
Please circle ONE answer only 7. Well-managed diabetes decreases the risk of:
a. Less than or equal to 7% Please circle AS MANY as apply, or circle ‘Unsure’
b. Less than 8% a. Kidney damage
c. 9% b. Blindness
d. Less than or equal to 10% c. Melanoma
e. Unsure d. Heart disease
e. Foot ulcers
3. Diabetes is a condition that: f. Unsure
Please circle ONE answer only
a. Can be cured by adopting a healthy lifestyle 8. What foot problems are people with diabetes most
b. Can be cured with tablets and/or insulin at risk of?
c. Is currently not curable Please circle AS MANY as apply, or circle ‘Unsure’
d. Is always life threatening when first diagnosed a. Poor circulation
e. Unsure b. Loss of feeling in the feet
c. Foot ulcers
4. Which of the following statements about diabetes d. Hammer toe
and diet is true? e. Infections
Please circle ONE answer only f. Unsure
a. People with diabetes should eat a sugar free diet
b. It is OK to eat fried take away food three times a 9. Why are people with diabetes advised to test their
week own blood glucose (BG)?
c. Red meat is a carbohydrate food Please circle ONE option only
d. A diet which is low in fat, high in fibre, low in added a. To alert them to changes in BG level patterns
sugar is recommended for everyone with diabetes b. To help make decisions in relation to exercise,
e. Unsure treating ‘hypos’ (low BG) or sick-day management
c. It can make people more confident in looking after
5. Why is doing regular exercise or being physically their diabetes
active good for your health? d. All of the above
Please circle AS MANY as apply, or circle ‘Unsure’ e. Unsure
a. It can help to control blood glucose levels
b. It can lower blood pressure (continued on next page)

170b Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 8

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Appendix 2. Text of the Diabetes Knowledge Questionnaire (continued from previous page)

10. What should a person with diabetes do if s/he c. Immediately have some sugary food or drink
becomes ill (e.g. flu, gastric upset, infection)? (e.g. jelly beans, soft drink)
Please circle AS MANY as apply, or circle ‘Unsure’ d. Drink some diet soft drink
a. Check blood glucose level more frequently (every e. Unsure
2 to 4 hours)
b. Stop taking all diabetes tablets and/or insulin The next question (No 15) is to be completed by
c. Drink lots of non-sweet fluid if blood glucose levels people with or carers of people with type 1
are over 15mmol/L diabetes only
d. Seek medical attention if very unwell and unable to
check blood glucose
15. A person with type 1 diabetes feeling unwell and
e. Try to do as much exercise as possible to lower
unable to eat should:
blood glucose levels
Please circle AS MANY as apply, or circle ‘Unsure’
f. Unsure
a. Check blood glucose and ketone levels at least
every 2 hours
11. People with diabetes need a medical check-up of b. Drink carbohydrate containing (sugary) fluids if
their eyes, nerve and kidney function at least: blood glucose below 15mmol/L
Please circle ONE answer only c. Go to the hospital if persistent vomiting and/or
a. Every month diarrhoea
b. Six monthly d. Stop taking all insulin
c. Once a year e. Seek medical advice for adjusting insulin doses
d. Every two to three years f. Unsure
e. Unsure
16. Finally, we would like to ask you some questions
12. The National Diabetes Services Scheme (NDSS): about yourself. This questionnaire is strictly
Please circle AS MANY as apply, or circle ‘Unsure’ confidential. Please assist us by answering all
a. Allows members to purchase blood glucose testing questions.
strips at reduced price
b. Offers members free syringes and insulin pen needles
What is your age? ………… years
c. Is only available to people on low incomes
d. Is available to people with all types of diabetes What is your gender? Female n Male n
e. Is free to join
f. Unsure How long have you had diabetes?
….…. years or …..… months or …….. days
The following questions (13 and 14) are to be What type of diabetes do you have?
completed by people taking diabetes medication Type 1 n Type 2 n Unsure n
(i.e. blood glucose lowering tablets or insulin) Other n please specify .....……….............................
If you are not taking any diabetes medication please
go to question No 16 Do you take diabetes medication? Yes n No n
If yes n glucose lowering tablets
and/or n insulin
13. Which of the following statements about diabetes
medication is true?
If you ticked insulin how many injections per day?
Please circle ONE answer only
n1 n2 n3 n4
a. If blood glucose levels are normal for two months,
n Other, please specify
diabetes medication can be stopped
………….........................
b. Tablets for diabetes work by increasing blood
glucose levels If you ticked glucose lowering tablets, how many
c. Regular medical check-ups are necessary to assess different tablets?
the need for adjustments to diabetes medication n1 n2 n3
d. People taking diabetes medication do not need to n Other, please specify
worry about healthy eating ……………………………
e. Unsure
Have you ever seen a Diabetes Educator? Yes n No n
14. If a person with diabetes has a hypo (low blood Have you ever seen a Dietitian? Yes n No n
glucose level) reaction, s/he should:
Please circle ONE answer only Thank you very much for completing
a. Immediately take some insulin or diabetes tablets
this questionnaire!
b. Rest and wait until s/he feels better

Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons 170c
OA Eigenmann_Layout 1 5/3/11 12:53 PM Page 9

O RIGINAL A RTICLE
Development and validation of a diabetes knowledge questionnaire

Appendix 3. Diabetes Knowledge Questionnaire scoring instructions

For all questions with ONE answer: Correct answers

1 = correct Question Correct
0 = incorrect number answer(s)
0.5 = unsure
1 c
For all questions with AS MANY as
2 a
apply answers enter
for each option i.e. a. b. c. d. e.:
3 c
1 = correct
4 d
0 = incorrect
5 a, b, c, e
For Question 9:
Enter 0.3 if option a. or b. or c. 6 a
are circled individually or
1 = correct, i.e. option d. 7 a, b, d, e

8 a, b, c, e,

For people with type 2 diabetes 9 d

NOT taking diabetes medication
the total possible score is 26. 10 a, c, d

11 c
For people with type 2 diabetes
taking diabetes medication the 12 a, b, d, e
total possible score is 28.
13 c

14 c
For people/carers of people with
type 1 diabetes the total possible 15 a, b, c, e
score is 32.

170d Pract Diab Int May 2011 Vol. 28 No. 4 Copyright © 2011 John Wiley & Sons

View publication stats