5.4.4. Data discussion

Information shown in the presentation section indicates an objective difficulty in collecting and analysing data at EU level. There are very large discrepancies between reports, all though with a unique common message: diabetes is an increasingly threatening disease that is growing at an alarming rate. The EUCID core indicators are definitely the best way to start collecting data for an analysis of the current health status and represent a fundamental framework for the future collection of complete data.

It has been demonstrated that the most important risk factor for prevention, i.e. body mass index, increases with age. About 50% of the general population in most countries are now overweight. The body mass index is measured by weight and height and calculated by weight in kilogram divided by the square of the height in meters. The outcome categories are: below 20 underweight, 20-25 ideal weight, 25-30 overweight, and the most important, obesity, being defined as a BMI equal or above 30. According to EUCID, obesity in Europe ranges between 12% (Netherlands) and 25% (UK), with a median value of 14%. The problem increases with age, with a peak between 65 and 74 yrs, where all countries approach at least a figure of 20%. Overweight and obesity, combined, add up to almost 50% in the countries with figures available for EUCID. However, data are based upon BMI; on the other hand, there is another approach that is gaining credibility from scientific evidence, i.e. the measuring of one’s waist circumference. However, this characteristic is still not measured routinely in clinical practice, thus we cannot provide consistent data. The emergency dictated by this figure should be rather obvious for European Governments: we know from scientific literature that the current social trends also push for worsened conditions, particularly for disadvantaged strata. It is time to act strongly on this front with proper prevention strategies.

Incidence of T1DM seems to be less of a problem overall, but it is highly variable across Europe, flagging attention towards particular geographical areas and high risk strata. Regardless of the type of diabetes, the data may become more alarming, as we know from scientific evidence that the rise of diabetes is increasingly affecting the younger generation with an impact on lifestyle behaviours.

The emergency is constantly underlined by prevalence results. Evidence shows that prevalence is rapidly increasing. However, the figures currently available are too variable between sources, and are often unstable across the years, thus all estimates must be taken with extreme caution. The three fold difference found between EUCID (median=3%) and IDF (median=8.7%) data is difficult to explain. A joint effort to enhance the quality of data, with the adoption of standardised criteria is absolutely necessary. Prevalence forecasts for 2025 must be really taken only as one among the many possible scenarios.

Among core outcomes, blindness is one of the most difficult to capture, as it can be not only difficult to directly link this problem to diabetes (in the absence of validated registers), but also prone to differential selection, as blind patients are harder to reach and visit their physician on a less regular basis. Complete data in this area will remain extremely difficult to obtain in the short term.

The situation is certainly different for what concerns dialysis, for which patients are unfortunately urged to make regular visits at clinical centres. Cross-country variations here are not easy to interpret (Scotland more than 30 times the dialysis/transplant incidence of Cyprus), so it is possible that some countries are using tracking systems that probably lead to incomplete and inconsistent estimates. The variation in the prevalence of dialysis/transplant among countries is even more striking, the range being 3-632, something that may arise the same sort of considerations as those above.

Secondary outcomes, including renal failure, amputations, stroke, myocardial infarction, reflect and perhaps amplify the same problems of high variations found in dialysis. Mortality rates are extremely important, but this data also is more biases related possibly due to the use different coding practices. The average figure obtained by EUCID should serve to inform policy makers that, on average, a town of Europe with 100 000 inhabitants would loose 30 citizens per year due to diabetes.

As far as secondary indicators are concerned, it is worth reminding that impaired glucose tolerance has a prevalence at least as high as that of diabetes. 10% of these subjects will develop diabetes in one year. Clinical management can be followed up through a long list of process indicators, showing that, despite the large variations in data between countries, which may be the result of reporting problems, there is still room for improvement for many procedures in most countries. Notably, fundus inspection is performed in 57% of countries submitting data, with information on retinopathy and laser treatment scarcely available, flagging a gap that requires a fast filling. Moreover, between 25-60% of the patients are poorly controlled for a range of important clinical measurements. An average of 20% smokers among diabetics is realistic and extremely alarming.

Overall, process and outcome indicators in diabetes seem to highlight that health systems in Europe are not optimally organized to deliver the results expected for a proper management of the disease. Although more complete and consistent data are necessary to confirm this picture, this issue must be addressed with urgency. Strong initiatives should be identified in order to push the entire health sector towards a tight control of diabetes.