Diabetes in all hospitalized cases in Germany 2015–2019 and impact of the first COVID-19 year 2020

Introduction

The prevalence of diabetes is increasing worldwide (1). The number of people with diabetes in Germany, which has been recently estimated as 6.2 million (about 10%) (2), is expected to reach more than 10 million people by 2040 (3). This considerable increase may impact the prevalence of diabetes not only in the outpatient care but also in the inpatient setting. However, comprehensive and reliable data on the frequency of diabetes in the hospital are scarce. A problem is that most hospital statistics only report diabetes in case of a diabetes-related hospitalization (diabetes as main diagnosis) but do not take diabetes as a preexisting condition (secondary diagnosis) into account. However, as we found in a previous study, for more than 90% of the hospitalizations with diabetes in Germany, diabetes is not documented as the ‘main diagnosis’ (i.e. the primary reason for the hospital admission), what is partly due to reimbursement rules (4). Therefore, reports taking only main diagnoses into account greatly underestimate the prevalence of diabetes in the hospital. Analyzing all main and secondary diagnosis in the German Diagnosis-Related Groups (DRG) statistics from the years 2015 to 2017, our last analysis revealed that around 18% of all inpatient cases had a diabetes diagnosis, regardless of the cause of the hospital admission (4).

This information is indispensable to estimate the amount of diabetological expertise needed for proper inpatient care. In fact, there is a large evidence that hospitalized patients with diabetes, compared to those without diabetes, have a higher risk of complications, longer hospital stays, as well as higher mortality rates (4, 5, 6). To optimize the management of these high-risk patients in the hospital and reduce the risk of readmission, qualified and individualized diabetes care including up-to-date pharmacotherapy and diabetes technology is needed, even when diabetes is not coded as the primary reason for hospital admission (5, 6). However, less than 20% of German hospitals have currently a certification for diabetology or a dedicated diabetologist/diabetes educator (7).

The aim of the present study was therefore to estimate the diabetes prevalence among hospitalized subjects in Germany in recent years, as well as in 2020, the first year of the COVID-19-pandemic. We, therefore, used the nationwide mandatory DRG statistics and compared the temporal trends by sex and age groups.

Methods

This analysis is based on a secondary database, the German DRG statistics. Since the introduction of the DRG accounting system in the early 2000s in Germany, all general hospitals (except rehabilitation and psychiatric/psychosomatic clinics) are obligated to transmit extensive data on inpatient care to the Institute for the Hospital Remuneration System (InEK) for reimbursement. The InEK then transmits legally defined parameters to the German Federal Statistical Office (Destatis, Wiesbaden), which is available against the fee for analysis via controlled remote execution through the Research Data Centers (FDZ, Forschungsdatenzentren) of the Statistical Office. The FDZ ensures data protection through a strict review prior to data release. In addition, the DRG statistics only report the number of hospital admissions (inpatient cases) and do not comprise any information that allows to derive the number of patients involved. According to German law, the present analysis, which is based on strictly anonymized administrative data (secondary data), does not require consent or ethical approval. This research complies with the Declaration of Helsinki.

For the present analysis, we used the DRG statistics of 2015–2020 (data source: FDZ of the German Federal and State Statistical Offices, DRG statistics 2015–2020). Among all hospitalizations of patients aged 20 years and older documented during this period in Germany, we searched for five types of diabetes in the main diagnoses (reasons for admission) or secondary diagnoses (comorbidities) based on the ICD-10 (German modification) codes: type 1 diabetes (E10); type 2 diabetes (E11); other specified diabetes mellitus, including pancreatic diabetes (diabetes in the context of diseases of the exocrine pancreas) (E13); rare types of diabetes (E12 or E14); and gestational diabetes (O24). We also included Impaired glucose regulation/Prediabetes (R73), as it also requires diabetological expertise for patient counselling and care. For the year 2020, we additionally identified all inpatient cases with a COVID-19 diagnosis using one of the following secondary diagnosis codes: U071 (COVID-19 confirmed by PCR-Test), U072 (COVID-19 symptoms after contact with a confirmed COVID-19 case), or B972 (coronavirus infectious agents identified as the cause of disease).

The type of diabetes coded as a secondary diagnosis was only considered in cases without diabetes as the main diagnosis. In the case of multiple coding of diabetes in the secondary diagnoses, we retained only the most frequent or plausible type of diabetes: type 1 diabetes or type 2 diabetes > gestational diabetes or other/pancreatic diabetes > rare diabetes types > prediabetes. Cases with both type 1 and type 2 diabetes as double secondary diagnoses, which were very few (about 0.03% of all cases with diabetes; n = 991 in 2015; n = 979 in 2016; n = 926 in 2017; n = 854 in 2018; n = 795 in 2019; n = 694 in 2020), were excluded because they could not be clearly assigned to a specific diabetes type. Cases with double secondary diagnoses gestational diabetes and other specified diabetes mellitus, including pancreatic diabetes (n = 8228 in 2015; n = 12,874 in 2016; n = 18,210 in 2017; n = 27,481 in 2018; n = 32,050 in 2019; n = 37,743 in 2020) were assigned to gestational diabetes, as they are suspected to result from a frequent incorrect coding of gestational diabetes in Germany (4). Cases of unknown age (n = 52 in 2017; n = 36 in 2016; n = 76 in 2015; n = 198 in 2018; n = 184 in 2019; n = 195 in 2020) were excluded. Cases with unknown sex (overall less than 0.02% of all inpatient cases: n = 285 in 2015; n = 175 in 2016; n = 1073 in 2017; n = 2790 in 2018; n = 0 in 2019; n = 0 in 2020) were assigned to the female cases since it was the largest group. In order to prevent cell blocking due to small numbers, cases with gestational diabetes aged ≥50 years (n = 11 in 2015; n = 16 in 2016; n = 12 in 2017; n = 33 in 2018; n = 35 in 2019; n = 30 in 2020) were excluded.

Age was classified into 10- (or 20)-year groups from 20–29 (or -39) years to ≥90 (or ≥80) years. The absolute and relative frequencies of inpatient cases were analyzed stratified by year, type of diabetes, sex, and age group. Chi-square tests were used to compare the proportion of cases with or without diabetes, by sex or between treatment years. The significance level (two-sided) was set at 0.01. All percentages are rounded to two decimal places. The analysis programs were created using SAS 9.4.

Results

Temporal trend since 2015

Of the 16.5 million inpatient cases aged ≥20 years documented each year between 2015 and 2019 in Germany, about 3 million (18%) had a main or secondary diagnosis of diabetes. The exact proportion of documented diabetes in inpatient setting increased during this period from 18.28 to 18.47% (P < 0.001, Fig. 1, Table 1). Over these years, the absolute number and percentage of diabetes documented as secondary diagnosis increased, while those of diabetes coded as main diagnosis decreased (all P < 0.001, Table 1).

View Full Size

Number of inpatient cases without and with diabetes in Germany between 2015 and 2020: Absolute frequency of all inpatient cases aged ≥20 years without and with diabetes (any type of diabetes as main or secondary diagnosis). *The proportion of diabetes in inpatient cases differed significantly in 2019 vs 2015, as well as in 2020 vs 2019 (all P < 0.001).

Citation: Endocrine Connections 12, 4; 10.1530/EC-22-0475

• Download Figure
• Download figure as PowerPoint slide

Between 2015 and 2019, females were more frequent than males in inpatient cases without diabetes (males vs females in 2015: 45.4 vs 54.6%; in 2019: 45.7 vs 54.3%, all P < 0.001, results are not shown here), but in cases with type 1 or type 2 diabetes, males were more numerous (males vs females in 2015: 53.2 vs 46.8%; in 2019: 54.9 vs 45.1%, all P < 0.001, results are not shown here). Male cases with diabetes were in all these years particularly predominant in the age category 60–79 years (males vs females in 2019: 59.9 vs 40.1%, Fig. 2). However, female cases with diabetes were more frequent than males in the age groups 20–39 and ≥80 years (males vs females in 2019, respectively: 42.1 vs 57.9% and 44.6 vs 55.4%, Fig. 2).

View Full Size

Number of inpatient cases with diabetes by sex and age groups between 2015 and 2020: Number of inpatient cases with diabetes type 1 or type 2. Change in percentage: *between 2019 and 2015; **between 2020 and 2019.

Citation: Endocrine Connections 12, 4; 10.1530/EC-22-0475

• Download Figure
• Download figure as PowerPoint slide

From 2015 to 2019, the number of inpatient cases with diabetes type 1 or 2 aged 60–79 years decreased (−9.6% in females, and −1.7% in males), whereas the number of those aged ≥80 years increased (+7.9% in females, and +21.8% in males, Fig. 2). To a lesser extent, the number of cases in the age group 20–39 increased, whereas those in the age group 40–59 decreased, in both females and males (Fig. 2).

The first COVID-19 year 2020

In the first year of the COVID-19 pandemic, the total number of all hospitalizations decreased sharply, from 16.4 million in 2015 and 16.6 million in 2019 to 14.5 million in 2020 (Table 1, Fig. 1). Similarly, the absolute frequency of inpatient cases with diabetes decreased (Table 1, Fig. 1). The largest decrease in cases with diabetes was for both sexes in the age group 60–79 years (−15.3% in females, and −12.2% in males, Fig. 2). Nevertheless, the decrease of inpatient cases with diabetes was less pronounced compared to the decrease in cases without diabetes, so that the proportion of cases with diabetes in the hospital increased from 18.47% in 2019 to 18.81% in 2020 (P < 0.001, Fig. 1). This increase is higher than expected based on the linear trend observed from 2015 to 2019 (expected proportion in 2020: 18.51% (95% CI: 18.47‒18.55%)).

Table 1

Frequency (%) of different types of diabetes among all inpatient cases in Germany between 2015 and 2020.

	2015		2016		2017		2018		2019		P values	2020
	Absolute frequency	Relative frequency	Absolute frequency	Relative frequency	Absolute frequency	Relative frequency	Absolute frequency	Relative frequency	Absolute frequency	Relative frequency		Absolute frequency	Relative frequency
Type 1 diabetes (E10)
MD	25,399	0.15	26,044	0.16	26,298	0.16	26,845	0.16	26,920	0.16		22,166	0.15
	Only SD	47,972	0.29	50,297	0.30	50,287	0.30	52,682	0.32	53,270	0.32		48,892	0.34
	Total	73,371	0.45	76,341	0.46	76,585	0.46	79,527	0.48	80,190	0.48	0.001	71,058	0.49
Type 2 diabetes (E11)
MD	164,567	1.00	160,716	0.96	157,793	0.95	155,086	0.94	152,462	0.92		128,030	0.88
	Only SD	2,636,979	16.04	2,668,883	16.01	2,684,884	16.12	2,666,324	16.32	2,695,259	16.39		2,385,267	16.45
	Total	2,801,546	17.04	2,829,599	16.97	2,842,677	17.07	2,821,410	17.06	2,847,721	17.11	0.001	2,513,297	17.33
Other/ Pancreatic diabetes (E13)
MD	4,717	0.03	4,996	0.03	4,901	0.03	4,843	0.03	4,687	0.03		4,227	0.03
	Only SD	35,247	0.21	35,916	0.22	36,285	0.22	36,404	0.22	37,260	0.22		33,589	0.23
	Total	39,964	0.24	40,912	0.25	41,186	0.25	41,247	0.25	41,947	0.25	0.001	37,816	0.26
Rare types of diabetes (E12/E14)
MD	690	<0.01	566	<0.01	500	<0.01	528	<0.01	528	<0.01		423	<0.01
	Only SD	11,820	0.07	11,220	0.07	10,950	0.07	10,009	0.06	9,539	0.06		7,799	0.05
	Total	12,510	0.08	11,786	0.07	11,450	0.07	10,537	0.06	10,067	0.06	0.001	8,222	0.06
Gestational Diabetes (O24)
MD	11,403	0.07	12,069	0.07	12,752	0.08	12,846	0.08	13,158	0.08		13,056	0.09
	Only SD	36,531	0.22	41,187	0.25	44,082	0.26	46,354	0.28	47,716	0.29		50,736	0.35
	Total	47,934	0.29	53,256	0.32	56,834	0.34	59,200	0.36	60,874	0.37	0.001	63,792	0.44
Pre-diabetes (R73)
MD	296	<0.01	276	<0.01	302	<0.01	288	<0.01	284	<0.01		190	<0.01
	Only SD	30,140	0.18	28,446	0.17	29,644	0.18	31,214	0.19	32,298	0.20		32,996	0.23
	Total	30,436	0.19	28,722	0.17	29,946	0.18	31,502	0.19	32,582	0.20	0.001	33,186	0.23
All (with Diabetes)
MD	207,072	1.26	204,667	1.22	202,546	1.21	200,436	1.21	198,039	1.19	0.001	168,092	1.16
Only SD	2,798,689	17.02	2,835,949	17.01	2,856,132	17.15	2,842,987	17.19	2,875,342	17.28	0.001	2,559,279	17.65
Total	3,005,761	18.28	3,040,616	18.23	3,058,678	18.36	3,043,423	18.40	3,073,381	18.47	0.001	2,727,371	18.81
Total(with and without diabetes)	16,445,040	100	16,674,901	100	16,656,336	100	16,541,702	100	16,643,630	100		14,498,498	100

aChi2-tests, 2019 vs. 2015. Relative and absolute frequency of different types of diabetes as main diagnosis (MD) or secondary diagnosis (SD) among all inpatient cases aged ≥20 years. All percentages are rounded. Of note, population in the age group ≥20 years in Germany in 2019 was 67,836,209 (source: www.destatis.de).

In 2020 as well as in the years before, the very large majority of cases with diabetes were type 2 diabetes (2020: 92.2% of all diabetes cases, 17.3% of all inpatient cases, Fig. 3). Of the other diabetes forms, type 1 diabetes (2020: 2.6% of all diabetes cases, 0.5% of all inpatient cases) and gestational diabetes (2020: 2.3% of all diabetes cases, 0.4% of all inpatient cases) remained the most common types in 2020 (Fig. 3). Between 2015 and 2020, the proportion of all diabetes types among inpatient cases (except the rare diabetes types) increased, especially the proportion of cases with gestational diabetes (+52% between 2015 and 2020).

View Full Size

Diabetes as main or secondary diagnosis in 14,5 million inpatient cases ≥20 years in Germany in 2020: on the left, relative and absolute frequency of all inpatients cases with and without diabetes. On the right: relative and absolute frequency of all inpatient cases with diabetes by diabetes type.

Citation: Endocrine Connections 12, 4; 10.1530/EC-22-0475

• Download Figure
• Download figure as PowerPoint slide

The proportion of cases with a COVID-19 diagnosis was significantly higher in men than in women regardless of the diabetes status, and significantly higher in inpatient cases with diabetes in each sex and age subgroup compared to cases without diabetes, except in males aged 20–29 years (non-significant difference, Fig. 4). The relative risk (RR) of a COVID-19 infection in inpatient cases with vs without diabetes was higher in females than in males, and highest for the age group 40–49 years (RR in females: 1.51; in males: 1.41).

View Full Size

Proportion of inpatient cases with a COVID-19 diagnosis in 2020 by sex and age group: Difference between cases with and without diabetes. *Significant (P < 0.01); n.s., not significant (Chi-square test).

Citation: Endocrine Connections 12, 4; 10.1530/EC-22-0475

• Download Figure
• Download figure as PowerPoint slide

Discussion

This analysis of the national mandatory DRG statistics provides comprehensive and reliable information on diabetes in the hospital in Germany in the period from 2015 to 2020. The proportion of diabetes in inpatient care increased slightly during this period and reached 18.8% in 2020, the first year of the COVID-19 pandemic. The International Diabetes Federation (IDF) estimates that in 2021 around 6.2 million adults aged 20–79 years with diagnosed diabetes lived in Germany, corresponding to a prevalence of about 10% (2). The nearly twofold prevalence of diabetes in the hospital compared to the prevalence in the general population may arise from a higher need for inpatient care in people with diabetes, possibly due to complications, and from a higher rate of hospitalization in older people, who are more likely to suffer from diabetes (4).

We found that the increase of the proportion of diabetes in the hospital in 2020 was slightly higher than expected based on the trend 2015–2019. We observed in particular that the increase in the proportion of inpatient cases with type 2 diabetes or gestational diabetes has accelerated in 2020 compared to the years before.This might be due to the higher morbidity and also probably to a higher proportion of hospitalizations with COVID-19 in this patient group, as found in this and other studies (8).

A study from the US reported that the hospitalization rates for adults with diabetes (regardless of the cause of the hospital admission) decreased from 2008 to 2016/2017 (9). Due to a lack of accurate data on the prevalence of diabetes in the general population between 2015 and 2020, it was not possible to correctly estimate hospitalization rates in Germany. Nevertheless, if we consider the estimates of the IDF (2), the number of individuals aged 20–79 years with diabetes increased in Germany from 5,022,200 in 2011 (5.3% of the general population) to 6,199,900 in 2021 (6.9% of the general population). This represents a mean increase of about 2% per year. Between 2019 and 2020, we found an increase in the proportion of diabetes in the hospital of 1.85%. This could indicate that the increase in diabetes in the hospital is similar to the mean increase observed in the general population. However, these are very rough estimations based on different observation periods and data sources (patient-related vs case-related), and thus, comparisons are difficult.

The preponderance of male inpatient cases in the age category 40–79 is consistent with previously published results for Germany (4) and also accords with the higher prevalence of diabetes in men of this age group (2, 10). In Brazil, where women were more frequently hospitalized, an inverse trend has been observed in the last years (11). The higher prevalence of female cases in the younger age group can be related to the higher prevalence of type 2 diabetes in young women (10), as well as to cases of gestational diabetes and of pregnancy with type 2 diabetes, which all became more frequent in the last years in Germany (12, 13), as also seen in our results. The higher prevalence of female cases in the elderly (≥ 80 years) may be due to the higher life expectancy in women compared to men (14). Regardless of sex, our findings reveal a regular increase in the number of hospitalizations in patients with diabetes aged ≥80 years as well as a decrease in cases aged 60–79 years. The increase of patients aged ≥80 years in the hospital concord with the aging of the general population (15) and might also be due to better treatment and higher life expectancy in diabetes. The decrease in inpatient cases aged 60–79 years can be a consequence of improved treatment, which may contribute to reduce the number of hospitalizations. In addition, it is possible that the increase in hospitalizations in older patients results in fewer beds available for the younger age groups. The increase of hospitalizations among younger adults (20–39 years) with diabetes, also observed in the US (9), may be related to the increase of prevalent cases with diabetes in this age group, partly resulting from more diabetes risk factors in this age group as well as the earlier age of diabetes diagnosis (16).

In line with our results, the first year of the COVID-19 pandemic 2020 has generally been associated with a considerable decrease in inpatient care, also in people with diabetes (17). As explained in several publications, this decline is probably due to both restrictive medical service during the lockdown and avoidance of patients that feared to get infected in the hospital (17). Many studies have reported a higher hospitalization risk due to a COVID-19 infection in patients with diabetes (18, 19, 20). Accordingly, we found a significantly higher prevalence of COVID-19 diagnosis in inpatient cases with diabetes compared to cases without, in both men and women.

Conclusions

The present analysis confirms the high prevalence of diabetes in the hospital (4), including during the first year of the COVID-19 pandemic. Although the total number of hospitalized patients decreased in the first COVID-19 year, the relative number of patients with diabetes increased. In addition, our data confirm that hospitalization associated with COVID-19 infection was more frequent in adult subjects with diabetes. The present study provides reliable and recent information that could help estimate the amount of diabetological expertise needed in inpatient care.

Declaration of interest

Prof. Fritsche received lecture fees and consultancy fees from Sanofi, Novo Nordisk, Astra Zeneca, and Boehringer Ingelheim. Prof Heni reports research grants from Boehringer Ingelheim and Sanofi (both to the University Hospital of Tübingen), advisory board for Boehringer Ingelheim, and lecture fees from Boehringer Ingelheim, Amryt, Sanofi, Novo Nordisk, and Eli Lilly. The other authors do not report any potential conflict of interest in relation with this manuscript.

Funding

This work was supported by the German Diabetes Association (DDG, Deutsche Diabetes Gesellschaft) and by the Robert Koch Institute (RKI) within the framework of the National Diabetes Surveillance. Further financial support was provided by the German Center for Diabetes Research (DZD, Deutsches Zentrum für Diabetesforschung; FKZ: 82DZD14E03) and the University of Tübingen.