GP-OSMOTIC Study: A randomised controlled trial (RCT) to determine the effect of retrospective continuous glucose monitoring on HbA1c in adults with Type 2 diabetes (T2D) in primary care

Talk Code: 
John Furler
David O'Neal, Jane Speight, Jo-Anne Manski-Nankervis, Sharmala Thuraisingam, Katie De La Rue, Rebecca Doyle, Louise Ginnivan, Elizabeth Holmes-Truscott, Kim Dalziel, Max Catchpool, Jason Chiang, Ralph Audehm, Mark Kennedy, Malcolm Clark, Alicia Jenkins, Andrzej Januszewski, Kamlesh Khunti, Jim Best
Author institutions: 
University of Melboune, Deakin University, Sydney University, Monash University, Latrobe University, Nanyang Technological University,


T2D is a progressive condition, requiring ongoing self-management and stepwise treatment intensification to achieve individualised and sustained glycaemic targets with the aim of reducing downstream complications. Observing retrospective continuous glucose monitoring (r-CGM) patterns may be better than HbA1c to guide personalised decisions to optimise management and achieve glycaemic targets. However, evidence for r-CGM use in T2D remains limited.


Eligible participants were aged 18 years and above, had HbA1c >7mmol/mol (0.5%) above individualised target and were prescribed >2 medications for glycaemia. Intervention participants attended a one-hour diabetes education session, and every three months: had a HbA1c assessment and wore the FreeStyle Libre Pro (Abbott) device for up to 14 days prior to attending a clinic assessment visit with their primary care physician (baseline, 3, 6, 9 and 12 months). Physicians received training in the use and interpretation of ambulatory glucose profiles. Control group participants wore the r-CGM device (blinded to results) at baseline and had 3-monthly visits based on usual care, including finger prick blood glucose monitoring as appropriate. The primary outcome was the difference in mean HbA1c at 12 months. Secondary outcomes were mean differences in time in glucose target range (4-10 mmol/L) at 12 months, diabetes-specific distress (PAID scores) at 12 months and HbA1c at 6 months. Analysis was conducted on an intention to treat basis, with adjustment for age and socioeconomic disadvantage.


Twenty-five primary care practices, 78 primary care physicians and 299 people with T2D participated. Participants were aged (mean(SD)) 60(10) years with diabetes duration (median(IQR)) of 12(8,20) years and mean(SD) HbA1c of 8.9(1.2)% (74.2(13.4) mmol/mol). Primary outcome: At 12 months, the between-group difference in mean HbA1c was 0.2% (95% CI -0.05% to 0.5%). Secondary outcomes: The estimated mean percentage time in target glucose range was 8.4% (95% CI 2.6% to 14.1%) higher in the intervention group than the control group (p=0.004). There was no evidence of a between-group difference in diabetes-specific distress (0.5; 95% CI -2.1 to 3.1). At 6 months, HbA1c was significantly lower in the intervention group; mean difference 0.5% (95% CI 0.2% to 0.8%). We found little evidence of changes in glycaemic medications in either group. There were few sensor dislodgements observed and no serious adverse events.


Use of 3-monthly r-CGM in people with T2D in primary care does not significantly improve HbA1c at 12 months. We showed a statistically and clinically significant reduction in HbA1c at 6 months in the intervention group, as well as significant improvements in time spent in glucose target range, with no change in diabetes distress levels. Our findings suggest the primary impact of r-CGM use on HbA1c in this setting is short term and not driven by increase in the number of diabetes medications.

Submitted by: 
John Furler
Funding acknowledgement: 
Australian National Health and Medical Research Council, Sanofi Australia, Abbott (in-kind)