Families, Systems, & Health 2015, Vol. 33, No. 1, 28 –35
© 2015 American Psychological Association 1091-7527/15/$12.00 http://dx.doi.org/10.1037/fsh0000101
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Diabetes Knowledge in Young Adults: Associations With Hemoglobin A1C Joni K. Beck, PharmD, BC-ADM, CDE
Ying Zhang, MD, PhD
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
University of Oklahoma Health Sciences Center
Christina M. Shay, PhD
Cynthia A. Muhamedagic, PhD and Steven A. Sternlof, PhD
University of North Carolina at Chapel Hill
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
Kai Ding, PhD
Megan M. Short, BS
University of Oklahoma Health Sciences Center
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
Justin D. Dvorak, BA
James T. Lane, MD
University of Oklahoma Health Sciences Center
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
The purpose of this study was to quantify associations between hemoglobin A1C (A1C) and diabetes knowledge score using an assessment tool developed to evaluate the level of diabetes knowledge in young adults with Type 1 diabetes
This article was published Online First January 19, 2015. Joni K. Beck, PharmD, BC-ADM, CDE, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma; Ying Zhang, MD, PhD, Department of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center, College of Public Health; Christina M. Shay, PhD, Department of Nutrition, University of North Carolina at Chapel Hill, School of Public Health; Cynthia A. Muhamedagic, PhD, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Steven A. Sternlof, PhD, Department of Internal Medicine, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Kai Ding, PhD, Department of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center, College of Public Health; Megan M. Short, BS, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Justin D. Dvorak, BA, Department of Biostatistics and Epidemiology, University
of Oklahoma Health Sciences Center, College of Public Health; James T. Lane, MD, Department of Internal Medicine, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center. The authors have no financial disclosures or any conflicts of interest to report. Preliminary results were presented in a poster presentation at the American Diabetes Association 73rd Scientific Session and Annual Meeting, Chicago, Illinois, Diabetes 2013 June;62(suppl(1) A-179). The authors would like to acknowledge Barbara J. Anderson, PhD, Professor of Pediatrics, Associate Head, Psychology Section, Baylor College of Medicine, Houston, Texas for her review and contributions to the knowledge assessment questionnaires. Also, the authors would like to acknowledge Kayla Bloom, BS, Research Technician, University of Oklahoma, College of Medicine, Department of Pediatrics for her technical editing assistance. Correspondence concerning this article should be addressed to Joni K. Beck, PharmD, BC-ADM, CDE, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, 1200 Children’s Avenue, 4D, Oklahoma City, OK 73104. E-mail: [email protected] 28
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
DIABETES KNOWLEDGE IN YOUNG ADULTS AND A1C
29
(T1DM) and their parent/primary caregiver. Seventy-five participants with T1DM, ages 15–22 years, completed questionnaires. Two 25-item questionnaires were developed: one for patient and one for caregiver. Linear regression quantified associations between correct items on the tools and participant A1C and demographic characteristics. Mean age of participants was 16.7 ⫾ 1.7 years, diabetes duration 5.9 ⫾ 4.2 years, 46.7% male, 74.7% Caucasian, 69.3% on multiple daily injections, and 30.7% on continuous subcutaneous insulin infusion therapy; 78.7% of parents/caregivers completed the questionnaire. A significant interaction was observed between patient and caregiver scores with A1C by diabetes duration. Among patients with diabetes ⬍6 years, higher patient and caregiver scores were associated with lower A1C (⫺0.25 ⫾ 0.11, p ⫽ .03 and ⫺0.59 ⫾ 0.19, p ⫽ .005, respectively) accounting for age, gender, race, therapy, and insurance. Neither patient nor caregiver score was associated with A1C in patients with diabetes duration ⱖ6 years. Better performance on a diabetes knowledge assessment (for both patient and the caregiver) was found to be associated with more favorable levels of glycemic control among young adults with diabetes ⬍6 years. Additional evaluation of these questionnaires and novel interventions to enhance knowledge in this population are needed. Keywords: type 1 diabetes, young adults, knowledge assessment, transitioning care, hemoglobin A1c Supplemental materials: http://dx.doi.org/10.1037/fsh0000101.supp
Pediatric Type 1 diabetes (T1DM) is a prevalent chronic disease in children. When a pediatric patient is first diagnosed with diabetes, although included in the education process as age-appropriate, the parent/primary caregiver must be responsible for learning the skills necessary to achieve optimal hemoglobin A1C (A1C) control. National diabetes self-management guidelines exist that recommend nine curriculum content areas are taught to patients with diabetes (Funnell et al., 2007; Haas et al., 2012). The nine content areas are: describing the diabetes disease process and treatment options, incorporating nutritional management into lifestyle, incorporating physical activity into lifestyle, using medication(s) safely and for maximum therapeutic effectiveness, monitoring blood glucose and other parameters and interpreting and using the results for self-management decision making, preventing, detecting, and treating acute and chronic complications, and developing personal strategies to address psychosocial issues, promote health, and behavior change. Although this process can be overwhelming for the patient and entire family unit in a complex relentless disease such as T1DM, learning is first on the continuum of educational outcomes. As the child progresses to adolescence and ultimately moves into young adulthood, the knowledge and skills necessary for
caring for their disease must be transferred from the parent to the young adult. The importance of parental assistance and support for a child, adolescent, or young adult, especially during this transition phase, is crucial. Since parents are the responsible caregiver, any diabetes knowledge deficits would likely result in an adolescent or young adult knowledge deficit. If diabetes knowledge can be mastered, the information and skills can lead to subsequent behavior change and improvements in glycemic control. As children with diabetes reach young adulthood, numerous barriers have been documented, placing them at a higher risk for adverse diabetes-related outcomes, such as gaps in clinical care due to competing priorities in life, worsening glycemic control, and increased acute complications (Bryden, Dunger, Mayou, Peveler, & Neil, 2003; Garvey et al., 2013; Laing, Jones, Swerdlow, Burden, & Gatling, 2005; Lancaster et al., 2010; Lane et al., 2007; Sparud-Lundin, Ohrn, Danielson, & Forsander, 2008; Weissberg-Benchell, Wolper, & Anderson, 2007). Studies indicate that ongoing diabetes education is fundamental in improving patient self-care management, which, in turn, can lead to improvements in glycemic control ultimately helping to prevent or delay diabetes complications (Haas et al., 2012). Results from
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
30
BECK ET AL.
a meta-analysis within the Type I pediatric population indicate that greater treatment adherence is associated with better glycemic control (Hood, Peterson, Rohan, & Drotar, 2009). Markowitz et al. (2011) also have shown, in youth ages 9 –15, that greater adherence to diabetes-related tasks, as measured on a diabetes self-management questionnaire, was associated with lower A1C. Despite recommendations for pediatric and adult care providers, outcomes in this population continue to present a challenge, specifically regarding the best approaches to assist young adults throughout their emerging adult years (Garvey et al., 2013; Peters, Laffel, & the American Diabetes Association Transitions Working Group, 2011). Assessment of basic diabetes knowledge of parents and young adults is important prior to transitioning from pediatric diabetes clinical care since evidence suggests that diabetes outcomes are poorer when young adults do not know how to adequately manage their diabetes. Rovner et al. (2012) developed a T1DM nutrition knowledge survey that was tested in teens (13.3 ⫾ 2.9 years) and their caregivers. Although this survey was limited to nutrition knowledge it was determined that more favorable nutrition knowledge survey scores from caregivers were associated with lower patient A1C. Although some assessment instruments have been developed that assess specific areas of diabetes knowledge, no instruments currently exist that assess both patient and the caregiver levels of diabetes-related knowledge across the entire spectrum of content areas that have been indicated as valuable for successful glycemic control and prevention of diabetes related complications. To address this gap, we have designed an assessment tool that evaluates the level of diabetes knowledge among young adults with T1DM and their parent/primary caregiver. We hypothesized that those young adults (defined as 15–25) and caregivers with the highest assessment scores (which indicate higher levels of diabetes-related knowledge and subsequent application of diabetes knowledge) will have a lower A1C. Therefore, the primary purpose of this study was to quantify associations between A1C and diabetes knowledge score using the assessment tool developed to evaluate the level of diabetes knowledge among young adults with T1DM and their parent/primary caregiver.
Method Clinical Education Services The program where this study was completed is recognized by the American Diabetes Association (ADA) for its pediatric diabetes selfmanagement education (DSME) program. Over the entire time of the analysis, all patients, either at the time of diagnosis or when transferred into our program, received the same specialty education services from a multidisciplinary team (pediatric endocrinologists, diabetes educators from various disciplines, child life specialists, social workers, and pediatric psychologists). All patients and their parent/primary caregiver were taught all nine content areas, as recommended (Funnell et al., 2007; Haas et al., 2012). After completion of the DSME curriculum (typically about 2–3 months after diagnosis), review of unclear information, as determined by diabetes team members, was provided to the patient/ family at routine outpatient clinic visits. Development of the Knowledge Assessment Tools Two 25-item, multiple choice, and true/false questionnaires were developed: one for assessment of patient knowledge and one for assessment of parent/primary caregiver knowledge. The majority of the 25 items on the questionnaires were created in a multiple choice format to incorporate various types of questions, such as recall, application, and analysis. One was developed for patients on multiple daily injections (MDI) and one for patients on continuous subcutaneous insulin infusion (CSII) therapy. A separate questionnaire for each regimen, (available as supplemental materials) was necessary as the knowledge required varies by insulin regimen. The questions on the caregiver and young adult questionnaires were the same for each type of insulin therapy, except for grammatical adjustments. The items for the questionnaires were center-developed based on the nine core curriculum topics as previously defined, current at the time of instrument development and also is consistent with the current guidelines (Funnell et al., 2007; Haas et al., 2012). As an example, two of the multiple choice survey questions are: When do you take your longacting, basal insulin dose? Selections are: at the
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
DIABETES KNOWLEDGE IN YOUNG ADULTS AND A1C
same time every day, only if my blood glucose levels are above target, only when I eat a meal or have a snack, or none of the choices are correct. Another survey question is: Which of the following is the MOST appropriate treatment for a low blood glucose episode with a finger stick glucose value of 65 mg/dL? Answer options are: 2 medium pears, 1 chocolate candy bar, [1/2] cup apple juice, or 1 cup of orange juice with 2 packets of sugar. Both questionnaires are available as supplemental materials. After the questionnaires were developed, they were systematically reviewed and pretested by a multidisciplinary team (n ⫽ 13) including endocrinologists, pediatric nurses, dietitians, pharmacists, certified diabetes educators, and behavioral health specialists. To assess comprehension difficulty of the questionnaires, the Flesch-Kincaid readability score was determined. The Flesch-Kincaid readability score for the MDI questionnaire was 5.2 for the primary caregiver and 5.0 for the young adults. For the CSII questionnaire, the primary caregiver readability score was 5.1 and the young adult readability score was 4.9. After development of the questionnaires, they were piloted from a random patient sample from the study population. To check if questionnaires were appropriately understood by the participants and if they had any concerns about the questionnaires, 13 caregivers and young adults (separate from the current project) piloted the knowledge assessment tools (n ⫽ 7 on MDI; n ⫽ 6 on CSII). After the pilot testers completed the questionnaires, confusing or unclear questions and unreliable responses were reviewed and revisions were made to the questionnaire items prior to study implementation. For determination of questionnaire scores, scores for both groups are defined as the number of correct items. If a patient or caregiver did not answer a question it was counted as incorrect. A question unique to our practice, but could easily be modified by other practices, is our target blood glucose values. For target blood glucose range, two answers were accepted as correct, based on what target blood glucose value was prescribed for the unique patient. After the forms were completed, the results were entered into Velos eResearch Clinical Trial Management System (Velos eResearch, Fremont, CA) by study administrative support personnel and double-checked by a second study team member prior to data analysis.
31
Participants Young adult T1DM patients between the ages of 15–25 years in an outpatient academic pediatric endocrinology practice from April, 2012 through September, 2013 preparing to transition to adult care services, but still followed by the pediatric endocrinology team, were invited in consecutive order to complete the diabetes knowledge assessment questionnaires. Upon review of eligible patients ages 15–25, study personnel excluded those not diagnosed with T1DM at least 1 year, no legal guardian present to sign the consent, English was not their primary language, or concurrent medical conditions preventing autonomous care in the young adult. At the time of this analysis, 105 young adults and their self-identified parent/primary caregiver provided consent and/or assent to complete this knowledge assessment. Of those 105 patients, 75 had complete data available for analysis. Twenty-five patients had consented; however, at the time data was analyzed for this report, they had not yet completed this specific assessment tool. This assessment tool was one of eight documents presented to patients (other data collection included additional medical data and psychological assessments; not part of this report). The remaining five eligible participants did not correctly complete the forms appropriate for analysis (i.e., completed the wrong questionnaire for their insulin regimen, etc.). These errors were not identified at the time of the visit, thus could not be corrected or included in the group analysis. Participants and parents/ caregivers were asked to independently complete the diabetes knowledge assessment questionnaire. Instructions were provided to not discuss the questions until the document was completed, at which time, they could speak to a certified diabetes educator regarding any questions. The diagnosis of T1DM was confirmed by board certified pediatric endocrinologists using ADA standards (American Diabetes Association, 2012). Laboratory tests used to confirm the diagnosis included autoantibodies to glutamic acid decarboxylase, islet cell auto antibodies, antibodies to tyrosine phosphatase, and plasma C-peptide (as an index of residual  cell secretion). All patients included in this study were on MDI (basal/bolus regimen) defined as four or
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
32
BECK ET AL.
more injections/day or CSII, both with flexible carbohydrate counting. Medical data collection included insulin regimen and A1C, determined by point-of-care testing in clinic by DCA Vantage Analyzer (Siemens Health care Diagnostics, Inc., Tarrytown, NY). Demographic data specific to this questionnaire included: age, duration of diabetes, gender, race, payer source, and caregiver’s education level. The Institutional Review Board at the University of Oklahoma Health Sciences Center approved this study. Statistical Analysis Demographic characteristics have been summarized using mean (SD) for continuous measures, and count (percentage) for categorical measures. The association between patient/ primary caregiver score and A1C was assessed using linear regression models with and without adjusting for gender, duration of disease, race, insulin therapy, and insurance type. Due to a significant interaction between patient/primary caregiver score and duration of disease in the multivariate models, stratified analyses based on duration of disease (dichotomized at its median) were performed, adjusting for age, gender, race, therapy type, and insurance status. We were not able to include patients who had no insurance (n ⫽ 5) in the multivariate analysis due to the small sample size of this category after stratification. Two-sided tests with a significance level of 0.05 were used throughout. Results Patient and caregiver characteristics are displayed in Table 1. Mean age of participants was 16.7 (1.7) with a mean diabetes duration of 5.9 (4.2) years. Participants were equally distributed by gender (53.3% female, 46.7% male) and the majority (74.7%) were Caucasian and using MDI (69.3%). The majority of participants also used a private payer source (52.0%). Over twothirds of the patient’s fathers and mothers reported educational attainment of at least high school graduation. Caregivers completed the knowledge assessment 78.7% of the time. Regression coefficients quantifying the association between patient and primary caregiver diabetes knowledge assessment score with A1C
Table 1 Characteristics of Patients With Type 1 Diabetes and Their Caregiver N ⫽ 75 Age (years) 16.7 (1.7) Duration of diabetes (year) 5.9 (4.2) Insulin regimen (%) Multiple daily injections 52 (69.3%) Continuous subcutaneous insulin infusion 23 (30.7%) Race, n (%) White 56 (74.7%) Non-White 18 (24%) Male gender, n (%) 35 (46.7%) Insurance, n (%) Private 39 (52%) Nonprivate (Medicaid) 31 (41.3%) No insurance 5 (6.7%) Mother’s education, n (%)a ⬍ High school 5 (11.4%) ⱖ High school 39 (88.6%) b Father’s education, n (%) ⬍ High school 8 (21.1%) ⱖ High school 30 (78.9%) A1C (%) 9.6 (2.2) Note. DM: diabetes mellitus; A1C: hemoglobin A1C. a 31 missing values. b 37 missing values.
according to diabetes duration are displayed in Table 2. In patients with diabetes duration of ⬍6 years, higher patient and caregiver scores on the diabetes knowledge assessment were associated with significantly lower A1C levels; the strength of the association for the caregiver score was nearly three-times that of the association with the patient’s score. These associations remained statistically significant after accounting for age, gender, White race, insulin therapy, and insurance status. Specifically in patients, each additional correct answer on the diabetes knowledge assessment tool was associated with 0.25 lower A1C. Similarly for the caregiver score, each additional correct answer on the diabetes knowledge assessment tool was associated with 0.59 lower level of A1C. Neither the patient nor caregiver score was associated with A1C in patients with diabetes duration ⱖ6 years. We did not find any significant correlations related to deficits on specific knowledge content areas (data not shown); however, the most common content areas that the parents answered incorrectly (for both MDI and CSII) were medications, nutrition, and chronic complications, respectively. The young adults on CSII therapy
SE 0.15
 ⫺0.65

p-value
© 2015 American Psychological Association 1091-7527/15/$12.00 http://dx.doi.org/10.1037/fsh0000101
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Diabetes Knowledge in Young Adults: Associations With Hemoglobin A1C Joni K. Beck, PharmD, BC-ADM, CDE
Ying Zhang, MD, PhD
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
University of Oklahoma Health Sciences Center
Christina M. Shay, PhD
Cynthia A. Muhamedagic, PhD and Steven A. Sternlof, PhD
University of North Carolina at Chapel Hill
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
Kai Ding, PhD
Megan M. Short, BS
University of Oklahoma Health Sciences Center
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
Justin D. Dvorak, BA
James T. Lane, MD
University of Oklahoma Health Sciences Center
University of Oklahoma Health Sciences Center and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma
The purpose of this study was to quantify associations between hemoglobin A1C (A1C) and diabetes knowledge score using an assessment tool developed to evaluate the level of diabetes knowledge in young adults with Type 1 diabetes
This article was published Online First January 19, 2015. Joni K. Beck, PharmD, BC-ADM, CDE, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center, Oklahoma City, Oklahoma; Ying Zhang, MD, PhD, Department of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center, College of Public Health; Christina M. Shay, PhD, Department of Nutrition, University of North Carolina at Chapel Hill, School of Public Health; Cynthia A. Muhamedagic, PhD, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Steven A. Sternlof, PhD, Department of Internal Medicine, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Kai Ding, PhD, Department of Biostatistics and Epidemiology, University of Oklahoma Health Sciences Center, College of Public Health; Megan M. Short, BS, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center; Justin D. Dvorak, BA, Department of Biostatistics and Epidemiology, University
of Oklahoma Health Sciences Center, College of Public Health; James T. Lane, MD, Department of Internal Medicine, University of Oklahoma Health Sciences Center, College of Medicine, and Harold Hamm Diabetes Center. The authors have no financial disclosures or any conflicts of interest to report. Preliminary results were presented in a poster presentation at the American Diabetes Association 73rd Scientific Session and Annual Meeting, Chicago, Illinois, Diabetes 2013 June;62(suppl(1) A-179). The authors would like to acknowledge Barbara J. Anderson, PhD, Professor of Pediatrics, Associate Head, Psychology Section, Baylor College of Medicine, Houston, Texas for her review and contributions to the knowledge assessment questionnaires. Also, the authors would like to acknowledge Kayla Bloom, BS, Research Technician, University of Oklahoma, College of Medicine, Department of Pediatrics for her technical editing assistance. Correspondence concerning this article should be addressed to Joni K. Beck, PharmD, BC-ADM, CDE, Department of Pediatrics, University of Oklahoma Health Sciences Center, College of Medicine, 1200 Children’s Avenue, 4D, Oklahoma City, OK 73104. E-mail: [email protected] 28
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
DIABETES KNOWLEDGE IN YOUNG ADULTS AND A1C
29
(T1DM) and their parent/primary caregiver. Seventy-five participants with T1DM, ages 15–22 years, completed questionnaires. Two 25-item questionnaires were developed: one for patient and one for caregiver. Linear regression quantified associations between correct items on the tools and participant A1C and demographic characteristics. Mean age of participants was 16.7 ⫾ 1.7 years, diabetes duration 5.9 ⫾ 4.2 years, 46.7% male, 74.7% Caucasian, 69.3% on multiple daily injections, and 30.7% on continuous subcutaneous insulin infusion therapy; 78.7% of parents/caregivers completed the questionnaire. A significant interaction was observed between patient and caregiver scores with A1C by diabetes duration. Among patients with diabetes ⬍6 years, higher patient and caregiver scores were associated with lower A1C (⫺0.25 ⫾ 0.11, p ⫽ .03 and ⫺0.59 ⫾ 0.19, p ⫽ .005, respectively) accounting for age, gender, race, therapy, and insurance. Neither patient nor caregiver score was associated with A1C in patients with diabetes duration ⱖ6 years. Better performance on a diabetes knowledge assessment (for both patient and the caregiver) was found to be associated with more favorable levels of glycemic control among young adults with diabetes ⬍6 years. Additional evaluation of these questionnaires and novel interventions to enhance knowledge in this population are needed. Keywords: type 1 diabetes, young adults, knowledge assessment, transitioning care, hemoglobin A1c Supplemental materials: http://dx.doi.org/10.1037/fsh0000101.supp
Pediatric Type 1 diabetes (T1DM) is a prevalent chronic disease in children. When a pediatric patient is first diagnosed with diabetes, although included in the education process as age-appropriate, the parent/primary caregiver must be responsible for learning the skills necessary to achieve optimal hemoglobin A1C (A1C) control. National diabetes self-management guidelines exist that recommend nine curriculum content areas are taught to patients with diabetes (Funnell et al., 2007; Haas et al., 2012). The nine content areas are: describing the diabetes disease process and treatment options, incorporating nutritional management into lifestyle, incorporating physical activity into lifestyle, using medication(s) safely and for maximum therapeutic effectiveness, monitoring blood glucose and other parameters and interpreting and using the results for self-management decision making, preventing, detecting, and treating acute and chronic complications, and developing personal strategies to address psychosocial issues, promote health, and behavior change. Although this process can be overwhelming for the patient and entire family unit in a complex relentless disease such as T1DM, learning is first on the continuum of educational outcomes. As the child progresses to adolescence and ultimately moves into young adulthood, the knowledge and skills necessary for
caring for their disease must be transferred from the parent to the young adult. The importance of parental assistance and support for a child, adolescent, or young adult, especially during this transition phase, is crucial. Since parents are the responsible caregiver, any diabetes knowledge deficits would likely result in an adolescent or young adult knowledge deficit. If diabetes knowledge can be mastered, the information and skills can lead to subsequent behavior change and improvements in glycemic control. As children with diabetes reach young adulthood, numerous barriers have been documented, placing them at a higher risk for adverse diabetes-related outcomes, such as gaps in clinical care due to competing priorities in life, worsening glycemic control, and increased acute complications (Bryden, Dunger, Mayou, Peveler, & Neil, 2003; Garvey et al., 2013; Laing, Jones, Swerdlow, Burden, & Gatling, 2005; Lancaster et al., 2010; Lane et al., 2007; Sparud-Lundin, Ohrn, Danielson, & Forsander, 2008; Weissberg-Benchell, Wolper, & Anderson, 2007). Studies indicate that ongoing diabetes education is fundamental in improving patient self-care management, which, in turn, can lead to improvements in glycemic control ultimately helping to prevent or delay diabetes complications (Haas et al., 2012). Results from
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
30
BECK ET AL.
a meta-analysis within the Type I pediatric population indicate that greater treatment adherence is associated with better glycemic control (Hood, Peterson, Rohan, & Drotar, 2009). Markowitz et al. (2011) also have shown, in youth ages 9 –15, that greater adherence to diabetes-related tasks, as measured on a diabetes self-management questionnaire, was associated with lower A1C. Despite recommendations for pediatric and adult care providers, outcomes in this population continue to present a challenge, specifically regarding the best approaches to assist young adults throughout their emerging adult years (Garvey et al., 2013; Peters, Laffel, & the American Diabetes Association Transitions Working Group, 2011). Assessment of basic diabetes knowledge of parents and young adults is important prior to transitioning from pediatric diabetes clinical care since evidence suggests that diabetes outcomes are poorer when young adults do not know how to adequately manage their diabetes. Rovner et al. (2012) developed a T1DM nutrition knowledge survey that was tested in teens (13.3 ⫾ 2.9 years) and their caregivers. Although this survey was limited to nutrition knowledge it was determined that more favorable nutrition knowledge survey scores from caregivers were associated with lower patient A1C. Although some assessment instruments have been developed that assess specific areas of diabetes knowledge, no instruments currently exist that assess both patient and the caregiver levels of diabetes-related knowledge across the entire spectrum of content areas that have been indicated as valuable for successful glycemic control and prevention of diabetes related complications. To address this gap, we have designed an assessment tool that evaluates the level of diabetes knowledge among young adults with T1DM and their parent/primary caregiver. We hypothesized that those young adults (defined as 15–25) and caregivers with the highest assessment scores (which indicate higher levels of diabetes-related knowledge and subsequent application of diabetes knowledge) will have a lower A1C. Therefore, the primary purpose of this study was to quantify associations between A1C and diabetes knowledge score using the assessment tool developed to evaluate the level of diabetes knowledge among young adults with T1DM and their parent/primary caregiver.
Method Clinical Education Services The program where this study was completed is recognized by the American Diabetes Association (ADA) for its pediatric diabetes selfmanagement education (DSME) program. Over the entire time of the analysis, all patients, either at the time of diagnosis or when transferred into our program, received the same specialty education services from a multidisciplinary team (pediatric endocrinologists, diabetes educators from various disciplines, child life specialists, social workers, and pediatric psychologists). All patients and their parent/primary caregiver were taught all nine content areas, as recommended (Funnell et al., 2007; Haas et al., 2012). After completion of the DSME curriculum (typically about 2–3 months after diagnosis), review of unclear information, as determined by diabetes team members, was provided to the patient/ family at routine outpatient clinic visits. Development of the Knowledge Assessment Tools Two 25-item, multiple choice, and true/false questionnaires were developed: one for assessment of patient knowledge and one for assessment of parent/primary caregiver knowledge. The majority of the 25 items on the questionnaires were created in a multiple choice format to incorporate various types of questions, such as recall, application, and analysis. One was developed for patients on multiple daily injections (MDI) and one for patients on continuous subcutaneous insulin infusion (CSII) therapy. A separate questionnaire for each regimen, (available as supplemental materials) was necessary as the knowledge required varies by insulin regimen. The questions on the caregiver and young adult questionnaires were the same for each type of insulin therapy, except for grammatical adjustments. The items for the questionnaires were center-developed based on the nine core curriculum topics as previously defined, current at the time of instrument development and also is consistent with the current guidelines (Funnell et al., 2007; Haas et al., 2012). As an example, two of the multiple choice survey questions are: When do you take your longacting, basal insulin dose? Selections are: at the
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
DIABETES KNOWLEDGE IN YOUNG ADULTS AND A1C
same time every day, only if my blood glucose levels are above target, only when I eat a meal or have a snack, or none of the choices are correct. Another survey question is: Which of the following is the MOST appropriate treatment for a low blood glucose episode with a finger stick glucose value of 65 mg/dL? Answer options are: 2 medium pears, 1 chocolate candy bar, [1/2] cup apple juice, or 1 cup of orange juice with 2 packets of sugar. Both questionnaires are available as supplemental materials. After the questionnaires were developed, they were systematically reviewed and pretested by a multidisciplinary team (n ⫽ 13) including endocrinologists, pediatric nurses, dietitians, pharmacists, certified diabetes educators, and behavioral health specialists. To assess comprehension difficulty of the questionnaires, the Flesch-Kincaid readability score was determined. The Flesch-Kincaid readability score for the MDI questionnaire was 5.2 for the primary caregiver and 5.0 for the young adults. For the CSII questionnaire, the primary caregiver readability score was 5.1 and the young adult readability score was 4.9. After development of the questionnaires, they were piloted from a random patient sample from the study population. To check if questionnaires were appropriately understood by the participants and if they had any concerns about the questionnaires, 13 caregivers and young adults (separate from the current project) piloted the knowledge assessment tools (n ⫽ 7 on MDI; n ⫽ 6 on CSII). After the pilot testers completed the questionnaires, confusing or unclear questions and unreliable responses were reviewed and revisions were made to the questionnaire items prior to study implementation. For determination of questionnaire scores, scores for both groups are defined as the number of correct items. If a patient or caregiver did not answer a question it was counted as incorrect. A question unique to our practice, but could easily be modified by other practices, is our target blood glucose values. For target blood glucose range, two answers were accepted as correct, based on what target blood glucose value was prescribed for the unique patient. After the forms were completed, the results were entered into Velos eResearch Clinical Trial Management System (Velos eResearch, Fremont, CA) by study administrative support personnel and double-checked by a second study team member prior to data analysis.
31
Participants Young adult T1DM patients between the ages of 15–25 years in an outpatient academic pediatric endocrinology practice from April, 2012 through September, 2013 preparing to transition to adult care services, but still followed by the pediatric endocrinology team, were invited in consecutive order to complete the diabetes knowledge assessment questionnaires. Upon review of eligible patients ages 15–25, study personnel excluded those not diagnosed with T1DM at least 1 year, no legal guardian present to sign the consent, English was not their primary language, or concurrent medical conditions preventing autonomous care in the young adult. At the time of this analysis, 105 young adults and their self-identified parent/primary caregiver provided consent and/or assent to complete this knowledge assessment. Of those 105 patients, 75 had complete data available for analysis. Twenty-five patients had consented; however, at the time data was analyzed for this report, they had not yet completed this specific assessment tool. This assessment tool was one of eight documents presented to patients (other data collection included additional medical data and psychological assessments; not part of this report). The remaining five eligible participants did not correctly complete the forms appropriate for analysis (i.e., completed the wrong questionnaire for their insulin regimen, etc.). These errors were not identified at the time of the visit, thus could not be corrected or included in the group analysis. Participants and parents/ caregivers were asked to independently complete the diabetes knowledge assessment questionnaire. Instructions were provided to not discuss the questions until the document was completed, at which time, they could speak to a certified diabetes educator regarding any questions. The diagnosis of T1DM was confirmed by board certified pediatric endocrinologists using ADA standards (American Diabetes Association, 2012). Laboratory tests used to confirm the diagnosis included autoantibodies to glutamic acid decarboxylase, islet cell auto antibodies, antibodies to tyrosine phosphatase, and plasma C-peptide (as an index of residual  cell secretion). All patients included in this study were on MDI (basal/bolus regimen) defined as four or
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
32
BECK ET AL.
more injections/day or CSII, both with flexible carbohydrate counting. Medical data collection included insulin regimen and A1C, determined by point-of-care testing in clinic by DCA Vantage Analyzer (Siemens Health care Diagnostics, Inc., Tarrytown, NY). Demographic data specific to this questionnaire included: age, duration of diabetes, gender, race, payer source, and caregiver’s education level. The Institutional Review Board at the University of Oklahoma Health Sciences Center approved this study. Statistical Analysis Demographic characteristics have been summarized using mean (SD) for continuous measures, and count (percentage) for categorical measures. The association between patient/ primary caregiver score and A1C was assessed using linear regression models with and without adjusting for gender, duration of disease, race, insulin therapy, and insurance type. Due to a significant interaction between patient/primary caregiver score and duration of disease in the multivariate models, stratified analyses based on duration of disease (dichotomized at its median) were performed, adjusting for age, gender, race, therapy type, and insurance status. We were not able to include patients who had no insurance (n ⫽ 5) in the multivariate analysis due to the small sample size of this category after stratification. Two-sided tests with a significance level of 0.05 were used throughout. Results Patient and caregiver characteristics are displayed in Table 1. Mean age of participants was 16.7 (1.7) with a mean diabetes duration of 5.9 (4.2) years. Participants were equally distributed by gender (53.3% female, 46.7% male) and the majority (74.7%) were Caucasian and using MDI (69.3%). The majority of participants also used a private payer source (52.0%). Over twothirds of the patient’s fathers and mothers reported educational attainment of at least high school graduation. Caregivers completed the knowledge assessment 78.7% of the time. Regression coefficients quantifying the association between patient and primary caregiver diabetes knowledge assessment score with A1C
Table 1 Characteristics of Patients With Type 1 Diabetes and Their Caregiver N ⫽ 75 Age (years) 16.7 (1.7) Duration of diabetes (year) 5.9 (4.2) Insulin regimen (%) Multiple daily injections 52 (69.3%) Continuous subcutaneous insulin infusion 23 (30.7%) Race, n (%) White 56 (74.7%) Non-White 18 (24%) Male gender, n (%) 35 (46.7%) Insurance, n (%) Private 39 (52%) Nonprivate (Medicaid) 31 (41.3%) No insurance 5 (6.7%) Mother’s education, n (%)a ⬍ High school 5 (11.4%) ⱖ High school 39 (88.6%) b Father’s education, n (%) ⬍ High school 8 (21.1%) ⱖ High school 30 (78.9%) A1C (%) 9.6 (2.2) Note. DM: diabetes mellitus; A1C: hemoglobin A1C. a 31 missing values. b 37 missing values.
according to diabetes duration are displayed in Table 2. In patients with diabetes duration of ⬍6 years, higher patient and caregiver scores on the diabetes knowledge assessment were associated with significantly lower A1C levels; the strength of the association for the caregiver score was nearly three-times that of the association with the patient’s score. These associations remained statistically significant after accounting for age, gender, White race, insulin therapy, and insurance status. Specifically in patients, each additional correct answer on the diabetes knowledge assessment tool was associated with 0.25 lower A1C. Similarly for the caregiver score, each additional correct answer on the diabetes knowledge assessment tool was associated with 0.59 lower level of A1C. Neither the patient nor caregiver score was associated with A1C in patients with diabetes duration ⱖ6 years. We did not find any significant correlations related to deficits on specific knowledge content areas (data not shown); however, the most common content areas that the parents answered incorrectly (for both MDI and CSII) were medications, nutrition, and chronic complications, respectively. The young adults on CSII therapy
SE 0.15
 ⫺0.65

p-value
