|
|
|
|
Practical Information
|
Instrument Name:
|
Diabetes Knowledge Test (DKT)
|
Instrument Description:
|
The DKT was designed to be an efficient and inexpensive way to gain a general assessment of a patient’s knowledge about their diabetes and its care. The Michigan Diabetes Research and Training Center (MDRTC) developed it to be used by educators and researchers. A multidisciplinary diabetes expert panel identified the content areas to test (established by a Delphi-type decision making process with mailed surveys) and developed the test items during a 1-day conference. Test items were reviewed and edited by the expert panel, and then the MDRTC staff members performed item revision. The resulting items went through several cycles of testing, analysis, and refinement, resulting in the current DKT. (Ref: 1) It contains 23 items and two subscales: the general subscale (14 items, appropriate for adults with Type I or II Diabetes) and the insulin-use subscale (9 items, appropriate for adults with Type I Diabetes or Type II patients using insulin).
|
Price:
|
Free; Available on MDRTC website, see address below
|
Administration Time:
|
15 minutes for the entire 23-item test. (Ref: 1)
|
Publication Year:
|
1998
|
Item Readability:
|
The computed Flesch-Kincaid Grade Level suggested that a person with a 4th grade education could read and comprehend the scale. The authors report a readability of 6th grade level. (Ref: 1)
|
Scale Format:
|
Multiple choice, 3-4 response choices per item.
|
Administration Technique:
|
Interview or self-administered. (Ref: 1)
|
Scoring and Interpretation:
|
Each correct response receives a score of 1. Higher scores indicate higher levels of diabetes specific knowledge.
|
Forms:
|
Multiple languages, e.g. Navajo. (Ref: personal communication with the author).
|
|
|
Research Contacts
|
Instrument Developers:
|
James T. Fitzgerald and Robert M. Anderson, et al.
|
Instrument Development Location:
|
Dept. of Medical Education
University of Michigan Medical School
The Towsley Center, Room 1114
Box 0201
Ann Arbor, MI 48109
|
Instrument Developer Email:
|
tfitz@umich.edu
|
Instrument Developer Website:
|
www.med.umich.edu/mdrtc/survey/index.html
|
|
|
Annotated Bibliography
|
1. Fitzgerald JT, Anderson RM, Funnell MM, Hiss RG, Hess GE, Davis WK, Barr, PA. The reliability and validity of a brief diabetes knowledge test. Diabetes Care 1998;21(5):706-10.[PMID: 9589228]
|
Purpose:
To examine the reliability and validity of the DKT.
|
|
|
Sample:
1) n=312 self-selected participants treated by private health providers in four Michigan communities. 58% female, mean age=60 years, 89% Caucasian, 64% Type II Diabetes non-insulin using, 61% received past Diabetes education, 39% completed 12th grade and 26% completed 13-15 years of school; 2) n=499 patients admitted to the County Health Department Diabetes Program (four local health departments). 68% female, mean age=56 years, 70% Caucasian, 61% Type II Diabetes non-insulin using, 52% received past Diabetes education 35% completed 12th grade and 24% completed 13-15 years of school. |
|
|
Methods:
Patients in the health department sample completed the DKT before a scheduled health care visit; patients in the community sample completed the DKT during a nurse visit to their home.
|
|
|
Implications:
The DKT performed reliably and validly in both samples, suggesting it may be used in these two settings. Type I Diabetes patients in the community sample scored higher than patients with Type II on both components of the DKT. Patients with Type I Diabetes in the health department sample scored higher than Type II patients on the insulin-use subscale. As years of formal education increased, scores increased in both samples.
|
|
|
top
|
Factors and Norms
|
Factor Analysis Work:
|
No information found.
|
Normative Information Availability:
|
No normative information was found beyond the original samples (312 community patients and 499 health department patients). (Ref: 1)
|
|
|
Reliability Evidence
|
Test-retest:
|
No information found.
|
Inter-rater:
|
No information found.
|
Internal Consistency:
|
Cronbach’s alpha for the general subscale was 0.70 in a community sample, 0.71 in a health department sample, and 0.71 for the total. Cronbach’s alpha for the insulin-use subscale was 0.74 in the community sample, 0.76 in the health department sample, and 0.75 for the total. (Ref: 1) All item-total correlations ranged from 0.22 to 0.43 for the total population for the general subscale and 0.32 to 0.53 for the insulin-use subscale. (Ref: 1)
|
Alternate Forms:
|
No information found.
|
|
|
Validity Evidence
|
Construct/ Convergent/ Discriminant:
|
Individuals with Type I diabetes score higher on the DKT than individuals with Type II diabetes. This supports the construct validity of the DKT, because Type I diabetes is much more serious and severe than Type II, and thus requires a higher level of diabetes knowledge by the patient to be successfully controlled. Also in support of construct validity, patients who reported having received diabetes education scored higher on the DKT than those who had not received education. However, it was also found that patients with a higher overall level of education (i.e., not diabetes-specific, but level of schooling), also scored higher on the DKT than those with lower levels of education; this finding argues against discriminant validity. (Ref: 1)
|
Criterion-related/ Concurrent/ Predictive:
|
No information found.
|
Content:
|
Content validity was established by using an expert panel to identify content domains to be tested. Consensus was reached through a Delphi process. (Ref: 1)
|
Responsiveness Evidence:
|
No information found.
|
Scale Application in VA Populations:
|
No information found.
|
Scale Application in non-VA Populations:
|
Yes. (Ref: 1)
|
|
|
Comments
|
The DKT may be used to assess patients’ knowledge, to provide targeted education to patients in knowledge areas in which they perform poorly, as a measure of general diabetes knowledge for research purposes, for group comparisons, and for assessing knowledge over time. The authors advise against using the DKT to comprehensively test specific areas of diabetes knowledge or self-care behavior.
In general, the DKT demonstrates adequate, although not outstanding internal consistency. The authors’ use of more and less educated samples is a plus, and helps demonstrate the DKT’s ability to discriminate among individuals with different levels of knowledge about diabetes. However, without information on responsiveness, it’s impossible to know how useful the DKT would be as an outcome measure for educational interventions.
|
Updates
No information found.
|
|
|
