We are just coming up to 100 years since insulin was first used to treat diabetes mellitus: clearly a dramatic turning point in the management of this disease. A perhaps less dramatic but, it could be argued, equally important development was the ability to monitor capillary glucose and thus adjust the insulin dose. Devices to measure capillary glucose first became available in the 1970s. Although not many of today’s clinicians will remember this, a significant number of our patients with type 1 diabetes will.

Capillary blood glucose testing has been the standard tool for monitoring and adjusting diabetes treatment for 50 years, but it would seem we are now at another turning point. Continuous glucose monitoring has now been available for over 20 years but it is only in very recent times that we have seen its widespread use. Having now achieved a certain momentum, we are seeing a very rapid rise in the uptake of this technology. In the UK, this is despite commissioning restrictions in place to try and limit the use.

We already know that the more information a person has about their glucose levels, the better control of glucose they achieve. It would seem intuitively obvious that having a continuous record of glucose levels should allow for more accurate adjustment of diabetes treatment than just having two or three moments per day when we know what the level is. The paper by Anita Jeyam and colleagues addresses the question of whether this benefit is restricted to certain groups of patients or whether we are likely to see benefit across the board. The answer from their real-world study strongly suggests that the benefit is universal.

There is a danger in the diabetes clinic that people who are already managing their condition well are the ones to be offered the latest technology to help them improve further, while those who are struggling, for whatever reason, can be left behind. It is possible that the current commissioning arrangement for both flash and continuous glucose monitoring is contributing to this health inequality. In the interest of achieving value for money, we may be penalising the group of patients who potentially have the most to gain. The data from this Scottish paper would support that view. It seems probable that the increasing use of sensor technology will reduce the disease burden of diabetes.

Regardless of the health-economic arguments, these devices are achieving a momentum of their own. The right thing to do is to widen their availability. Now would seem the time to do this.

Effects of widespread flash glucose monitoring use in Scotland

Flash glucose monitoring became eligible for NHS funding in Scotland in 2018. The present study sought to evaluate flash usage in Scotland between 2014 and 2020, and to examine the impact of the technology on HbA_1c and diabetes emergencies according to baseline age, sex, socioeconomic group, glycaemic control, insulin pump usage and education.

Using data from the Scottish Care Information – Diabetes Collaboration database, and linked to hospital and national records, a total of 14682 individuals with type 1 diabetes were evaluated. Use of flash monitoring grew rapidly after it became eligible for funding, rising from 3.1% in 2017 to 45.9% in mid-2020. Usage varied widely by age (from 64.3% in those aged <13 years to 32.7% in those aged ≥65 years) and by socioeconomic status (54.4% vs 36.2% in the least vs most deprived postcodes).

Overall, median HbA_1c decreased by 2.5 mmol/mol (0.2%) in the year following flash initiation; however, there was wide variation according to baseline HbA_1c, with a median reduction of 15.5 mmol/mol in those with HbA_1c over 84 mmol/mol (9.8%) at baseline compared with an increase of 1 mmol/mol in those with an initial HbA_1c <54 mmol/mol (7.1%). Significant reductions in HbA_1c were observed in all age groups, sexes and socioeconomic levels and regardless of insulin pump use, completion of structured diabetes education or early (self-funded) adoption of the technology. Almost all variation within these subgroups was due to the HbA_1c level at baseline.

Diabetic ketoacidosis (DKA) rates fell significantly after flash initiation, both in users as a whole and within all subgroups except for adolescents. Adjusting for rates prior to initiation, the DKA event rate ratio was estimated to be 0.59 (95% CI, 0.53–0.64) in the year after flash initiation compared with up to 5 years before.

The rates of hospitalisation for severe hypoglycaemia (HSH) also decreased overall following flash initiation. The reduction was particularly high in those with a prior HSH event in the previous 5 years (rate ratio, 0.25; 95% CI, 0.20–0.32).

The authors conclude that flash glucose monitoring use in Scotland has been associated with significant improvements in HbA_1c, especially in individuals with high HbA_1c at baseline. They argue that the lower usage in more deprived areas is a priority to overcome, given the striking reductions in DKA seen in all socioeconomic groups.

This study was funded by Diabetes UK and the Chief Scientist Office.