RCT data may overemphasize real-world insulin treatment response
medwireNews: UK researchers have highlighted discrepancies in the response to basal-bolus and premixed insulin among individuals with type 2 diabetes in real-world versus randomized controlled trial (RCT) settings.
“While the present analysis confirms the expected relationship between baseline HbA1c [glycated hemoglobin] and HbA1c reduction with either a premixed or a basal-bolus insulin regimen, the strength of this relationship was considerably more apparent in the RCT compared with the real-world population,” write Iskandar Idris and colleagues from the University of Nottingham in Diabetic Medicine.
Idris and team found that, after 6 months of treatment, HbA1c fell by an average 0.33% from a baseline of 8.7% in individuals from a UK primary care dataset who were receiving basal-bolus insulin (n=7483) and by 0.25% from the same baseline level in those receiving premixed insulin (n=10,744).
In pooled data from eight RCTs, mean HbA1c was 8.9% and baseline and fell by 1.44% with basal-bolus insulin (n=1893) and by 1.34% with premixed insulin (n=1517).
And although the response to treatment increased with increasing baseline HbA1c, the effect was more pronounced in the RCT population, where baseline HbA1c accounted for 88% of the variability in the change in HbA1c. By contrast, just 20% of the variability in the change in HbA1c was attributable to HbA1c at baseline in the real-world population.
There were also discrepancies in weight gain between the two populations, with average gains of 0.27 kg in the real-world population versus 2.96 kg in the RCT population, but there were no significant differences between basal-bolus and premixed insulin within each group.
Further analysis showed that individuals with baseline weight above approximately 60 kg experienced overall weight gain according to RCT data but slight weight loss in the real-world population.
In spite of this, greater baseline weight was associated with greater weight reductions after both premixed insulin and, to a lesser degree, basal-bolus insulin treatment, which the researchers say provides “important reassurance that, in routine clinical practice, meaningful reductions in HbA1c can be achieved without significant detrimental effect on weight.”
They speculate that the discrepancies between real-world and RCT data could be due to fear of hypoglycaemia, weight gain, reduced adherence to basal-bolus insulin, or hesitance to uptitrate insulin because of lack of monitoring, patient education, or rigid guidelines in the real-world setting. The younger age of the RCT population and higher rate of comorbidities and selection bias in the real-world population may also play a part, they note.
Idris et al say that their findings highlight “the importance of patient-related factors that would determine the potential success of one insulin regimen compared with another.”
“Choice of insulin regimen should therefore be individualized according to patients’ personal, social and clinical characteristics,” they conclude.
By Laura Cowen
medwireNews is an independent medical news service provided by Springer Healthcare. © 2017 Springer Healthcare part of the Springer Nature group