Correlation between centre size, metabolic variation and mean HbA1c in major paediatric diabetes centres.

AIM: To exploit a relatively homogeneous national health care context and a national diabetes database to address the questions: Is there an optimal clinic/centre size in determining outcomes?; and Can improvement in median centre outcomes be driven by reducing variability in outcome? METHODS: Using the Australasian Diabetes Database Network, data from seven tertiary hospital paediatric diabetes clinics for patients with type one diabetes from Australia were recorded from 6-month uploads: September 2017, March 2018, September 2018 and March 2019. Data from 25 244 patient visits included demographic variables, HbA1C, number of patient visits and insulin regimens. RESULTS: There was no association between centre size and median HbA1C. On the other hand, there was a significant association between or median absolute deviation of HbA1C outcomes and the median HbA1C result between centres. On average every two thirds of a median absolute deviation increase in clinic HbA1C was associated with a 1.0% (10.9 mmol/mol) increase in median clinic HbA1C. CONCLUSIONS: Our data have shown that it is likely difficult for centres to have a low median HbA1C if there is high variance of HbA1C's within centres or within centre treatment groups. This appears to be true regardless of centre size. These findings need to be carefully considered by teams who wish to lower their clinic median HbA1C.

Has subsidized continuous glucose monitoring improved outcomes in pediatric diabetes?

INTRODUCTION: In 2017, the Australian Federal Government fully subsidized continuous glucose monitoring (CGM) devices for patients under 21 years of age with T1D with the aim of reducing rates of severe hypoglycaemia (SH) and improving metabolic control. The aim of this study was to reports on metabolic outcomes in youth from a single tertiary centre. METHODS: The study design was observational. Data were obtained on youth who commenced CGM between May 2017 and December 2019. RESULTS: Three hundred and forty one youth who commenced CGM and had clinical outcome data for a minimum of 4 months. 301, 261, 216, 172, and 125 had outcome data out to 8, 12, 16, 20, and 24 months, respectively. Cessation occurred between 27.9% and 32.8% of patients 12 to 24 months after CGM commencement. HbA1c did not change in patients who continued to use CGM. In the 12 months prior to starting CGM the rate of severe hypoglycaemia events were 5.0 per 100 patient years. The rates of severe hypoglycaemia in those continuing to use CGM at 4, 8, 12, 16, 20, and 24 months, were 5.2, 5.1, 1.6, 6.1, 2.4, and 0 per 100 patient years, respectively. DISCUSSION: Our experience of patients either ceasing or underusing CGM is less than reported in other cohorts but is nonetheless still high. There may have been a reduction in rates of severe hypoglycaemia over the 24 months follow up period; however, the absolute numbers of events were so low as to preclude meaningful statistical analysis.

The clinician factor: Personality characteristics of clinicians and their impact upon clinical outcomes in the management of children and adolescents with type 1 diabetes.

OBJECTIVE: The purpose of this study was to estimate clinician qualities that influence metabolic outcomes in youth with type 1 diabetes. RESEARCH DESIGN AND METHODS: Data were gathered over two 3 month periods in a large tertiary diabetes center (1500 patients, 8 clinicians) from patients with type 1 diabetes who received continuous care from each clinician. Data included sex, age, diabetes duration, insulin regimen, body mass index (BMI), insulin dose and episodes of severe hypoglycemia. Clinician data included target blood glucose levels, target glycated hemoglobin (HbA1c), Diabetes Attitude Scale and Big 5 Personality Inventory Scale. Mean HbA1c per clinician was the primary outcome variable. RESULTS: The 8 clinicians saw a total of 464 patients during the first time period, and 603 in the second time period. Lowest to highest mean HbA1c per clinician varied by 0.7%. There were small but statistically significant differences between clinicians with their patients' age at diagnosis, duration of diabetes, age, gender, treatment type and BMI SD score. After controlling for these differences, the clinician characteristics that were associated with lower mean HbA1c were having no lower limit in target HbA1c and being self-reportedly "less agreeable." The impact of these clinician attitudinal traits was equivalent to the combined effects of patient characteristics and treatment type. CONCLUSIONS: There was a significant variation in metabolic outcomes between treating clinicians. After controlling for patient clinical differences, clinician mean HbA1c was associated with lower limit in target HbA1c and being "less agreeable." Clinicians who were more demanding and dogmatic appeared to have better outcomes.