Benchmarking, i.e. comparing quality indicators (e.g. HbA1c, BMI, blood pressure, lipids) between diabetes care centers, has been common for quite some time in pediatric diabetology, being effective in improving care in some settings but not in others. The SWEET project (Better Control in Pediatric and Adolescent DiabeteS: Working to CrEate CEnTers of Reference) was initiated as a 3-year EU Public Health Program, aiming to unify pediatric type 1 diabetes care by means of establishing ‘centers of reference’ (CoR), usually larger centers with multidisciplinary care. The idea is to determine which factors make these CoR excel in their care, based on reliable collection of high-quality data. Examples of these factors are frequency of uploads of pump or CGM data, individual targets, targeted education programs, or assessment of patient-reported outcome measures included into daily practice. More recently focus has changed from establishing CoR to database development and quality improvement in high-, medium-, and low-income countries through collaboration of smaller centers.

In this publication, co-authored by Henk Veeze from Diabeter, selected quality indicators were assessed among 16,082 children and youth (<25 year) with type 1 diabetes between 2008–2010 (n=4,930) and 2016–2010 (n=13,654) from 22 centers around the world. In 2,502 participants with data available for both periods a subanalysis was done.

Key findings:

• HbA1c (unadjusted and adjusted) decreased
• Use of insulin pumps an sensors increased
• People using pumps showed lower HbA1c than people on injections
• Increase of overall number of people with diabetes >200 at a center seemed to be associated with improvement of HbA1c
• Achievement of target HbA1c was stable from 2008 to 2013 and then started to increase, preceded by increased pump use from 2008 onwards
• Almost no sensor use was observed, until a sudden large increase from 2015 onwards, to >50% of people in 2018

Although one of the strengths of this study is the large number of included partcipants, one of its limitations is the enormous difference in numbers between the time periods. Still, the subanalysis of participants with data available in both time periods confirmed the results of the analysis using the complete cohorts. Another limitation is the multinational approach, with differences in methods for data collection. Additionally the CoR cannot be considered representative for a country or region.

Concluding, the authors state

Please click here for the pdf file.