In:
Diabetes, Obesity and Metabolism, Wiley, Vol. 20, No. 10 ( 2018-10), p. 2458-2466
Abstract:
To evaluate physicians' adjustments of insulin pump settings based on continuous glucose monitoring (CGM) for patients with type 1 diabetes and to compare these to automated insulin dose adjustments. Methods A total of 26 physicians from 16 centres in Europe, Israel and South America participated in the study. All were asked to adjust insulin dosing based on insulin pump, CGM and glucometer downloads of 15 patients (mean age 16.2 ± 4.3 years, six female, mean glycated haemoglobin 8.3 ± 0.9% [66.8 ± 7.3 mmol/mol]) gathered over a 3‐week period. Recommendations were compared for the relative changes in the basal, carbohydrate to insulin ratio (CR) and correction factor (CF) plans among physicians and among centres and also between the physicians and an automated algorithm, the Advisor Pro (DreaMed Diabetes Ltd, Petah Tikva, Israel). Study endpoints were the percentage of comparison points for which there was full agreement on the trend of insulin dose adjustments (same trend), partial agreement (increase/decrease vs no change) and full disagreement (opposite trend). Results The percentages for full agreement between physicians on the trend of insulin adjustments of the basal, CR and CF plans were 41 ± 9%, 45 ± 11% and 45.5 ± 13%, and for complete disagreement they were 12 ± 7%, 9.5 ± 7% and 10 ± 8%, respectively. Significantly similar results were found between the physicians and the automated algorithm. The algorithm magnitude of insulin dose change was at least equal to or less than that proposed by the physicians. Conclusions Physicians provide different insulin dose recommendations based on the same datasets. The automated advice of the Advisor Pro did not differ significantly from the advice given by the physicians in the direction or magnitude of the insulin dosing.
Type of Medium:
Online Resource
ISSN:
1462-8902
,
1463-1326
DOI:
10.1111/dom.2018.20.issue-10
Language:
English
Publisher:
Wiley
Publication Date:
2018
detail.hit.zdb_id:
2004918-3
Bookmarklink