SAN FRANCISCO – A UK-based audit on the use of continuous glucose monitoring (CGM) system, FreeStyle Libre (Abbott), in more than 4000 cases of type 1 diabetes, shows significant improvements in HbA1c and other measures of disease control, as well as reduced hospital admissions.
The results were presented by Dr Harshal Deshmukh, clinical lecturer in diabetes, endocrinology and metabolism, University of Hull, here at this year’s American Diabetes Association (ADA) 2019 Scientific Sessions .
Nationwide Audit Data
“FreeStyle Libre [FSL] use was found to be associated with a significant improvement in HbA1c, and in the GOLD score [related to awareness of hypoglycaemia]; a possible reduction in hypoglycaemia, diabetic ketoacidosis and hyperglycaemia hospital admissions; as well as improved diabetes-related distress,” reported Dr Deshmukh.
The wearable sensor, which is the size of a £2 coin and sits on the arm, avoids the need for inconvenient finger prick blood tests by relaying glucose levels to a smart phone or e-reader. In April this year, NHS England mandated that FSL be funded across England for eligible type 1 diabetes patients, ending the much criticised ‘postcode prescribing‘ of the devices.
The nationwide audit of FreeStyle Libre, reported here, supports widespread use of the CGM system. It aimed to explore the impact of FSL on HbA1c, hypoglycaemia awareness, resource utilisation, hospital admissions, user satisfaction, diabetes-related distress, discontinuation rate and its causes, and predictors of response.
Dr Viral Shah, assistant professor of paediatrics, at Barbara Davis Center for Diabetes , University of Colorado, in the US, moderated the session. “This is a real-life data and affirms the findings from randomised controlled trials [RCTs] of improved glycaemia, with almost all RCTs using CGM showing improvement in A1c by 0.5-1% depending on baseline A1c,” he said. “Any CGM, either real-time or FSL, has also been shown in studies to reduce hypoglycaemia in type 1 diabetes as well as insulin-treated type 2 diabetes. We also see great improvement in A1c over time with the use of CGM in type 1 diabetes in clinical practice, with the drop in A1c generally greater if A1c is higher than 8.5%.”
A total of 114 centres across the UK took part using a secure IT tool developed on the NHS computer network, N3, to make the process as easy and user-friendly as possible, explained Dr Deshmukh.
Clinicians were invited to submit data routinely collected using forms completed in clinic for later online submission, or alternatively the online tool could be completed in real time.
With respect to demographics and audit characteristics of the 4709 people living with diabetes, median age was 44 years, 99% had type 1 diabetes, 54% were female, duration of diabetes was an average of 20 years, baseline HbA1c was 8.2%, median body mass index (BMI) was 26 kg/m2, and median follow up was 6 months.
Dr Deshmukh analysed the reasons for initiating these people on FSL, and the indications for FSL use. “According to the audit, these reasons were, in order of the most frequent: replacement of self-monitoring of blood glucose (SMBG), high HbA1c, frequent hypoglycaemia, fear of hypoglycaemia, high glucose variability, hypoglycaemia unawareness, to avoid an insulin pump, and recurrent hospital admissions.”
Six Month Data: HbA1c Drops, Fewer Hypos, Hospital Admissions Fall
The median HbA1c after use of FSL was 7.6% (p<0.0001) representing a drop of 0.6% HbA1c, which was statistically significant. “For patients with a baseline HbA1c >8.5% the drop was more significant. The median pre-FSL HbA1c was 9.4%, and after using the FSL it was a median of 8.6% representing a drop of 0.8%, which was also statistically significant,” reported the researcher.
The effect of FSL on hypoglycaemia was measured by the GOLD score which relates to a questionnaire that asks the patient, ‘do you know when your hypos are commencing?’ The score ranges from one to seven, and a score of four or more suggests impaired awareness of hypoglycaemia.
“Use of FSL lead to a significant decline in median GOLD score with 9% reporting reversal of hypoglycaemic awareness (GOLD score drop below 4),” said Deshmukh. “Also, 79% reported that with use of FSL they were able to reduce the proportion of time in hypoglycaemia, 31% reduced the rate of hypoglycaemia, and 39% had reduced nocturnal hypoglycaemia.”
Turning to hospital admissions in relation to use of FSL, Deshmukh noted that with a median follow up of 6 months, hyperglycaemia and diabetic ketoacidosis (DKA) admissions dropped from 7.3% to 1.9%. Admissions due to hypoglycaemia dropped from 2.71% to 0.5%. “We do need more follow up data to do a more definite comparison.”
Diabetes-related distress was also assessed. “This score dropped from three to two at follow up and this was statistically significant,” remarked the researcher.
Discontinuations were minimal with only 1% of users discontinuing use of FSL at 6 months follow up, the majority due to a lack of funding or because funding had stopped being available. There was also a significant improvement in diabetes-related emotional score, added the researcher.
‘Improved Peace of Mind’
Dr Deshmukh used a machine-learning algorithm that predicted response to FSL and he found that baseline BMI, baseline HbA1c, followed by age and the number of scans per day, were the best predictors of response to the CGM system.
The audit also collected some qualitative data. Dr Deshmukh quoted some of the comments about use of FSL from people who had participated in the audit. FSL gives me, ‘Improved peace of mind’. Another user recalled, ‘I find my freestyle libre invaluable due to my lifestyle and hillwalking. For my sport I could be doing a blood glucose test up to 15 times a day or more.”
Finally, Dr Shah noted that more research would be beneficial. “This is not an RCT and therefore, selection bias cannot be excluded. For example, type 1 diabetes patients who are compliant and would benefit for FSL would have been initiated FSL. Longitudinal follow up of these subjects and economic analyses would provide valuable insight.”
Presented at the American Diabetes Association (ADA) 2019 Scientific Sessions on June 10th 2019. Abstract 299 OR.
COI: Dr Deshmukh has declared no conflicts of interest.
Dr Shah has no direct conflict but disclosed that he sits on advisory boards for Dexcom Inc, Sanofi US. Research support: NIH, Sanofi US, Dexcom Inc, NovoNordisk, EyeNuk and Center for Women’s Health Research Other: communication director, diabetes technology interest group at ADA, served as T1D Exchange Clinical Network Steering Committee.