DIABETES TECHNOLOGY & THERAPEUTICS Volume 17, Number 3, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/dia.2014.0224
ORIGINAL ARTICLE
Important Determinants of Diabetes Control in Insulin Pump Therapy in Patients with Type 1 Diabetes Mellitus Asma Deeb, MD,1 Samar Abu-Awad, BN,2 Salima Abood, MD,1 Mohamed El-Abiary, MD,2 Jamal Al-Jubeh, MD,2 Hana Yousef, BN,1 Laila AbdelRahman, BN,1 Ahlam Al Hajeri, BND,1 and Huda Mustafa, MD 2
Abstract
Background: Insulin pumps are equipped with advanced functions. Intensive training and adherence are required for optimum use of the technology. We aimed to assess the association of various key elements in insulin pump functions on blood glucose control. Patients and Methods: Patients on insulin pump therapy were enrolled. Insulin pumps were downloaded (CareLink Pro 3 software; Medtronic Minimed, Northridge, CA), and data were collected over an 8–12-week period. Glycemic control of patients was classified as controlled (hemoglobin A1c [HbA1c] level of 7.5% or less in adults and 8% or less in children) and uncontrolled based on HbA1c level at enrollment. Variables studied were use of sensors and duration, frequency of blood glucose monitoring, Bolus Wizard (Medtronic Minimed) use, frequency of correction boluses, and frequency of cannula changing. Results: Seventy-two patients were enrolled (50 children). Median age was 12 years for children and 27.5 years for adults. Respective median numbers of blood glucose checks were 4.4 and 3.2 for controlled and uncontrolled children (P < 0.021) and 3.1 and 2.8 for controlled and uncontrolled adults, respectively. Respective frequency of Bolus Wizard use per day showed a median of 6 and 4.15 for controlled and uncontrolled children (P < 0.001) and 3.8 and 3.5 for controlled and uncontrolled adults. Controlled children wore sensors for longer (5 vs. 2.9 days/week) and did more corrections (3.9 vs. 2.5). There was no difference in the frequency of changing the infusion cannula in children’s or adults’ groups. Conclusions: We conclude that the frequency of blood glucose monitoring and Bolus Wizard use have a favorable association with glycemic control. These observations were more significant in the children’s groups. Our data shows that patients with better control tend to bolus more for correction and wear sensors longer. Introduction
C
ontinuous subcutaneous insulin infusion (CSII) is considered the most physiological method to mimic the normal pancreatic function in terms of insulin profile.1 Among young type 1 diabetes patients, CSII showed a similar efficacy in reducing hemoglobin A1c (HbA1c) levels compared with multiple daily injections, with less hypoglycemia and fewer glycemic excursions, and was better in improving overall treatment satisfaction and the rate of perceived hyperglycemia and hypoglycemia.2 Internationally, insulin pump therapy is becoming more popular in the treatment of type 1 diabetes in children and adolescents3,4 and has been proven to be of benefit in this group of patients.5,6 1 2
Insulin pumps are equipped with advanced functions to control blood glucose; however, high level of understanding and adherence are required from the patient’s side to achieve optimal results. A high level of training is required for proper use of the technology. The use of advanced pump features was associated with a greater improvement in HbA1c.7 However, despite insulin pump treatment, metabolic control in adolescents is often unsatisfactory.8 Optimal diabetes treatment also depends on treatment satisfaction. In a study by Olinder et al.,9 it was found that patients were extremely satisfied with the treatment; they preferred this treatment to multiple daily injections, and this reflected into a better glycemic control. Efficacy and safety of technology devices are the main targets for their use. Equally important is the full
Paediatric Endocrinology Department, Mafraq Hospital, AbuDhabi, United Arab Emirates. Endocrine Department, Shaikh Khalifa Medical Center, AbuDhabi, United Arab Emirates.
166
INSULIN PUMP DETERMINANTS OF DIABETES CONTROL
167
understanding and the willingness of use by the end-user. Despite the advances seen in technology development for diabetes treatment, the technology has not been as well used as expected.10 Accordingly, proper education and understanding are crucial. The degree of adherence to the important daily tasks for insulin pump use is markedly correlated with improvement of glycemic control.11 Lack of compliance to the proper use of insulin pumps can be multifactorial. Lack of education is a major factor. However, diminished motivation, deliberate omission, and behavioral attitude can all be factors that necessitate comprehensive regular reviews at outpatient clinics. Deviation from recommended behaviors and lack of adherence to proper pump use tasks are common in adolescents and are found more commonly with increasing age and duration of CSII use.11 The aim of the present study was to assess variables and various key elements related to insulin pump self-management. We also aim to examine effect of adherence to recommended pump instructions and behavior on glycemic control in children and adults.
take), frequency of cannula changing, use of sensors, and duration of sensor use. All patients have had received standardized education at CSII initiation and on follow-up visits. Point-of-care HbA1c levels (DCA Vantage Analyzer; Siemens, Erlangen, Germany) were obtained on all patients on the same clinic visit at enrollment. The DCA machines were calibrated in accordance to the manufacturer’s instructions. The study received institutional ethical approval in accordance with the Declaration of Helsinki.
Patients and Methods
Results
Patients with type 1 diabetes (children and adults) on insulin pump therapy who presented to diabetes clinics for a follow-up were approached to participate in the study. Inclusion criteria were children and adults who had been diagnosed with type 1 diabetes mellitus for ‡ 1 year and who had been using Medtronic Minimed (Northridge, CA) insulin pumps for ‡ 6 months. The study took place in two diabetes centers in the city of AbuDhabi, United Arab Emirates, over a 6-month period ( July–December 2013). Patients who agreed to participate in the study had their insulin pumps downloaded using CareLink Pro 3 software (Medtronic MiniMed), and data were collected over an 8–12-week period. Patients were grouped into children and adults based on a cut limit of age of 18 years. Body mass index was calculated as weight/(height)2 and plotted versus the body mass index percentile chart (Centers for Disease Control and Prevention, Atlanta, GA). Children with a body mass index above the 85th percentile are considered overweight and those above the 97th percentile obese based on the World Health Organization classification. Adults with a body mass index above 25 kg/m2 are considered overweight and those above 30 kg/m2 obese based on the World Health Organization classification. Both groups were subcategorized based on HbA1c to controlled and uncontrolled groups. The cutoff level for the controlled group in children and adults was below 8% and 7.5%, respectively, according to American Diabetes Association guidelines. Children’s groups with controlled and uncontrolled diabetes were designated G1child and G2child, respectively. Similarly, adults with an HbA1c level of less than 7.5% were included in the G1adults group, whereas those with an HbA1c of 7.5% or higher were included in the G2adults group. Subjects’ diabetes duration, duration of insulin pump use, and total daily insulin doses/kg were recorded. Studied variables related to pump use were frequency of blood glucose monitoring, frequency of Bolus Wizard (Medtronic Minimed) use, frequency of correction boluses (with no food in-
Children
Statistical analysis
The independent-sample t test (parametric test) or Mann– Whitney U test (nonparametric test) was used to examine difference between the groups. Categorical data were compared using the Pearson v2 test, two-tailed. Results were expressed either as median values with range or as mean values with SD based on distribution. Data were analyzed using the Statistical Package for Social Sciences version 19.0 software (SPSS, Chicago, IL). A P value of < 0.05 was considered significant.
Fifty patients were enrolled (16 males). Age range was 2.3–17.1 years with a median of 12 years. Fourteen children (28%) were of normal weight, and 36 (72%) were overweight. No children were obese in this group. Mean HbA1c level was 8.5 – 1.4%. Twenty patients were assigned to the controlled group (G1child) versus 30 to the uncontrolled group (G2child). Eighteen (60%) of those in the G2child group were pubertal, whereas 12 (40%) were prepubertal (P = 0.43). This observation showed no statistical difference. Mean diabetes durations in the G1child and G2child groups were 5.50 – 4.07 and 5.47 – 3.15 years, respectively, and respective duration of pump use was 2.35 – 1.60 and 2.61 – 2.01 years. Mean total daily insulin requirement/kg was 0.84 – 0.18 and 0.88 – 0.12 units for the G1child and G2child groups, respectively (P = 0.41). Median number of blood glucose checks per day was 4.4 (2–11.4) and 3.2 (0.5–7.9) for the G1child and G2child groups, respectively. The difference was statistically significant (P = 0.021). Frequency of Bolus Wizard use per day showed a statistically significant difference between the two groups with a median of 6 (3.9–12.9) and 4.15 (0.6–9) for the G1child and G2child groups, respectively (P < 0.001). Patients in the G1child group did more corrections per week compared with G2child patients (3.9 vs. 2.5), but the difference was not statistically significant. There was no difference in the frequency of changing the infusion cannula in the two groups (3.5 days). Six patients (30%) in the G1child group and 12 (40%) patients in the G2child group used sensors during the study period, which showed no statistical difference between the two groups. The G1child patients used sensors for a longer period (mean, 5 days/week vs. 2.9 days/ week) (Table 1). Adults
Twenty-two patients were enrolled (three males). Age range was 20–48 years with a median of 27.5 years. Fourteen adults (64%) had an HbA1c level below 7.5% (G1adults
168
DEEB ET AL.
Table 1. Comparison of Selected Variables in the G1child (Controlled) and G2child (Uncontrolled) Groups Variable, group (n) Age (years) G1 child (20) G2 child (30) Duration of diabetes (years) G1child (20) G2child (30) Duration of pump use (years) G1child (20) G2child (30) Total daily insulin (units/kg) G1child (20) G2child (30) Blood glucose check (number/day) G1child (20) G2child (30) Bolus Wizard use (times/day) G1child (20) G2child (30) Sensor duration (days/week) G1child (6) G2child (12) Sensor use (differentiated by HbA1c) G1child No Yes G2child No Yes Correction boluses (number/week) G1child (8) G2child (19) Cannula changes (number/week) G1child (20) G2child (30)
Median (range) or mean – SD
P value 0.772
12.7 (2.3–17) 11.8 (6.1–17.1) 0.60 5.50 – 4.07 5.47 – 3.15 0.54 2.35 – 1.60 2.61 – 2.01 0.41 0.84 – 0.18 0.88 – 0.12 0.021 4.4 (2–11.4) 3.2 (0.5–7.9)
ORIGINAL ARTICLE
Important Determinants of Diabetes Control in Insulin Pump Therapy in Patients with Type 1 Diabetes Mellitus Asma Deeb, MD,1 Samar Abu-Awad, BN,2 Salima Abood, MD,1 Mohamed El-Abiary, MD,2 Jamal Al-Jubeh, MD,2 Hana Yousef, BN,1 Laila AbdelRahman, BN,1 Ahlam Al Hajeri, BND,1 and Huda Mustafa, MD 2
Abstract
Background: Insulin pumps are equipped with advanced functions. Intensive training and adherence are required for optimum use of the technology. We aimed to assess the association of various key elements in insulin pump functions on blood glucose control. Patients and Methods: Patients on insulin pump therapy were enrolled. Insulin pumps were downloaded (CareLink Pro 3 software; Medtronic Minimed, Northridge, CA), and data were collected over an 8–12-week period. Glycemic control of patients was classified as controlled (hemoglobin A1c [HbA1c] level of 7.5% or less in adults and 8% or less in children) and uncontrolled based on HbA1c level at enrollment. Variables studied were use of sensors and duration, frequency of blood glucose monitoring, Bolus Wizard (Medtronic Minimed) use, frequency of correction boluses, and frequency of cannula changing. Results: Seventy-two patients were enrolled (50 children). Median age was 12 years for children and 27.5 years for adults. Respective median numbers of blood glucose checks were 4.4 and 3.2 for controlled and uncontrolled children (P < 0.021) and 3.1 and 2.8 for controlled and uncontrolled adults, respectively. Respective frequency of Bolus Wizard use per day showed a median of 6 and 4.15 for controlled and uncontrolled children (P < 0.001) and 3.8 and 3.5 for controlled and uncontrolled adults. Controlled children wore sensors for longer (5 vs. 2.9 days/week) and did more corrections (3.9 vs. 2.5). There was no difference in the frequency of changing the infusion cannula in children’s or adults’ groups. Conclusions: We conclude that the frequency of blood glucose monitoring and Bolus Wizard use have a favorable association with glycemic control. These observations were more significant in the children’s groups. Our data shows that patients with better control tend to bolus more for correction and wear sensors longer. Introduction
C
ontinuous subcutaneous insulin infusion (CSII) is considered the most physiological method to mimic the normal pancreatic function in terms of insulin profile.1 Among young type 1 diabetes patients, CSII showed a similar efficacy in reducing hemoglobin A1c (HbA1c) levels compared with multiple daily injections, with less hypoglycemia and fewer glycemic excursions, and was better in improving overall treatment satisfaction and the rate of perceived hyperglycemia and hypoglycemia.2 Internationally, insulin pump therapy is becoming more popular in the treatment of type 1 diabetes in children and adolescents3,4 and has been proven to be of benefit in this group of patients.5,6 1 2
Insulin pumps are equipped with advanced functions to control blood glucose; however, high level of understanding and adherence are required from the patient’s side to achieve optimal results. A high level of training is required for proper use of the technology. The use of advanced pump features was associated with a greater improvement in HbA1c.7 However, despite insulin pump treatment, metabolic control in adolescents is often unsatisfactory.8 Optimal diabetes treatment also depends on treatment satisfaction. In a study by Olinder et al.,9 it was found that patients were extremely satisfied with the treatment; they preferred this treatment to multiple daily injections, and this reflected into a better glycemic control. Efficacy and safety of technology devices are the main targets for their use. Equally important is the full
Paediatric Endocrinology Department, Mafraq Hospital, AbuDhabi, United Arab Emirates. Endocrine Department, Shaikh Khalifa Medical Center, AbuDhabi, United Arab Emirates.
166
INSULIN PUMP DETERMINANTS OF DIABETES CONTROL
167
understanding and the willingness of use by the end-user. Despite the advances seen in technology development for diabetes treatment, the technology has not been as well used as expected.10 Accordingly, proper education and understanding are crucial. The degree of adherence to the important daily tasks for insulin pump use is markedly correlated with improvement of glycemic control.11 Lack of compliance to the proper use of insulin pumps can be multifactorial. Lack of education is a major factor. However, diminished motivation, deliberate omission, and behavioral attitude can all be factors that necessitate comprehensive regular reviews at outpatient clinics. Deviation from recommended behaviors and lack of adherence to proper pump use tasks are common in adolescents and are found more commonly with increasing age and duration of CSII use.11 The aim of the present study was to assess variables and various key elements related to insulin pump self-management. We also aim to examine effect of adherence to recommended pump instructions and behavior on glycemic control in children and adults.
take), frequency of cannula changing, use of sensors, and duration of sensor use. All patients have had received standardized education at CSII initiation and on follow-up visits. Point-of-care HbA1c levels (DCA Vantage Analyzer; Siemens, Erlangen, Germany) were obtained on all patients on the same clinic visit at enrollment. The DCA machines were calibrated in accordance to the manufacturer’s instructions. The study received institutional ethical approval in accordance with the Declaration of Helsinki.
Patients and Methods
Results
Patients with type 1 diabetes (children and adults) on insulin pump therapy who presented to diabetes clinics for a follow-up were approached to participate in the study. Inclusion criteria were children and adults who had been diagnosed with type 1 diabetes mellitus for ‡ 1 year and who had been using Medtronic Minimed (Northridge, CA) insulin pumps for ‡ 6 months. The study took place in two diabetes centers in the city of AbuDhabi, United Arab Emirates, over a 6-month period ( July–December 2013). Patients who agreed to participate in the study had their insulin pumps downloaded using CareLink Pro 3 software (Medtronic MiniMed), and data were collected over an 8–12-week period. Patients were grouped into children and adults based on a cut limit of age of 18 years. Body mass index was calculated as weight/(height)2 and plotted versus the body mass index percentile chart (Centers for Disease Control and Prevention, Atlanta, GA). Children with a body mass index above the 85th percentile are considered overweight and those above the 97th percentile obese based on the World Health Organization classification. Adults with a body mass index above 25 kg/m2 are considered overweight and those above 30 kg/m2 obese based on the World Health Organization classification. Both groups were subcategorized based on HbA1c to controlled and uncontrolled groups. The cutoff level for the controlled group in children and adults was below 8% and 7.5%, respectively, according to American Diabetes Association guidelines. Children’s groups with controlled and uncontrolled diabetes were designated G1child and G2child, respectively. Similarly, adults with an HbA1c level of less than 7.5% were included in the G1adults group, whereas those with an HbA1c of 7.5% or higher were included in the G2adults group. Subjects’ diabetes duration, duration of insulin pump use, and total daily insulin doses/kg were recorded. Studied variables related to pump use were frequency of blood glucose monitoring, frequency of Bolus Wizard (Medtronic Minimed) use, frequency of correction boluses (with no food in-
Children
Statistical analysis
The independent-sample t test (parametric test) or Mann– Whitney U test (nonparametric test) was used to examine difference between the groups. Categorical data were compared using the Pearson v2 test, two-tailed. Results were expressed either as median values with range or as mean values with SD based on distribution. Data were analyzed using the Statistical Package for Social Sciences version 19.0 software (SPSS, Chicago, IL). A P value of < 0.05 was considered significant.
Fifty patients were enrolled (16 males). Age range was 2.3–17.1 years with a median of 12 years. Fourteen children (28%) were of normal weight, and 36 (72%) were overweight. No children were obese in this group. Mean HbA1c level was 8.5 – 1.4%. Twenty patients were assigned to the controlled group (G1child) versus 30 to the uncontrolled group (G2child). Eighteen (60%) of those in the G2child group were pubertal, whereas 12 (40%) were prepubertal (P = 0.43). This observation showed no statistical difference. Mean diabetes durations in the G1child and G2child groups were 5.50 – 4.07 and 5.47 – 3.15 years, respectively, and respective duration of pump use was 2.35 – 1.60 and 2.61 – 2.01 years. Mean total daily insulin requirement/kg was 0.84 – 0.18 and 0.88 – 0.12 units for the G1child and G2child groups, respectively (P = 0.41). Median number of blood glucose checks per day was 4.4 (2–11.4) and 3.2 (0.5–7.9) for the G1child and G2child groups, respectively. The difference was statistically significant (P = 0.021). Frequency of Bolus Wizard use per day showed a statistically significant difference between the two groups with a median of 6 (3.9–12.9) and 4.15 (0.6–9) for the G1child and G2child groups, respectively (P < 0.001). Patients in the G1child group did more corrections per week compared with G2child patients (3.9 vs. 2.5), but the difference was not statistically significant. There was no difference in the frequency of changing the infusion cannula in the two groups (3.5 days). Six patients (30%) in the G1child group and 12 (40%) patients in the G2child group used sensors during the study period, which showed no statistical difference between the two groups. The G1child patients used sensors for a longer period (mean, 5 days/week vs. 2.9 days/ week) (Table 1). Adults
Twenty-two patients were enrolled (three males). Age range was 20–48 years with a median of 27.5 years. Fourteen adults (64%) had an HbA1c level below 7.5% (G1adults
168
DEEB ET AL.
Table 1. Comparison of Selected Variables in the G1child (Controlled) and G2child (Uncontrolled) Groups Variable, group (n) Age (years) G1 child (20) G2 child (30) Duration of diabetes (years) G1child (20) G2child (30) Duration of pump use (years) G1child (20) G2child (30) Total daily insulin (units/kg) G1child (20) G2child (30) Blood glucose check (number/day) G1child (20) G2child (30) Bolus Wizard use (times/day) G1child (20) G2child (30) Sensor duration (days/week) G1child (6) G2child (12) Sensor use (differentiated by HbA1c) G1child No Yes G2child No Yes Correction boluses (number/week) G1child (8) G2child (19) Cannula changes (number/week) G1child (20) G2child (30)
Median (range) or mean – SD
P value 0.772
12.7 (2.3–17) 11.8 (6.1–17.1) 0.60 5.50 – 4.07 5.47 – 3.15 0.54 2.35 – 1.60 2.61 – 2.01 0.41 0.84 – 0.18 0.88 – 0.12 0.021 4.4 (2–11.4) 3.2 (0.5–7.9)
