Diabetes in the Hospital

A Nationwide Analysis of All Hospitalized Cases in Germany With and Without Diabetes, 2015–2017

Background: Comprehensive data on the frequency of diabetes mellitus among hospitalized patients in Germany have not been published to date.

Methods: Among all inpatient cases aged ≥20 years that were documented in the German DRG statistics for 2015–2017, we analyzed the frequencies of five types of diabetes (type 1, type 2, other/pancreatic diabetes, “rare diabetes” with an ICD code of E12 or E14, gestational diabetes) and of prediabetes, stratified by sex and age group. The presence of any of these conditions was ascertained from the corresponding ICD-10 code among the main diagnoses (reasons for admission) or secondary diagnoses. We also compared the length of hospital stay, in-hospital mortality, and the frequency of various categories of main diagnosis in cases with and without diabetes in each age group.

Results: In the period 2015–2017, approximately 18% of the 16.4 to 16.7 million inpatient cases carried a main or secondary diagnosis of diabetes (in 2017: type 2, 17.1%; type 1, 0.5%). Diabetes was more common in male cases than in female cases (in 2017: type 2, 19.7% vs. 14.8%; type 1, 0.5% vs. 0.4%). In 2017, the greatest difference in length of hospital stay between patients with and without diabetes was for patients with type 1 diabetes aged 40–49 (7.3 vs. 4.5 days), while the greatest difference in in-hospital mortality was for patients with type 2 diabetes aged 70–79 (3.7% vs. 2.8%). From the age of 30 (age category 30–39), diseases of the cardiovascular system, and from the age of 50 (age category 50–59), diseases of the respiratory or urogenital systems were more frequently listed as a reason for admission in cases with than in those without diabetes.

Conclusion: The fact that diabetes is twice as prevalent in hospitalized cases as in the general population underscores the high morbidity associated with the disease and the greater need of persons with diabetes for in-hospital care, as the population of multimorbid diabetes patients continues to grow older.

According to the International Diabetes Federation (IDF), an estimated 7.5 million adults had diagnosed diabetes in Germany in 2017 (1). Recent estimates suggests that by 2040 Germany will be the home of 10.7 to 12.3 million people with type 2 diabetes (2).

The care of patients with diabetes is not solely provided in an outpatient setting. According to a survey of the University Hospital Tübingen, 22% of its inpatients had diabetes in 2016 (3). However, due to the fact that most hospital statistics merely include the reasons for admission (main diagnoses) and do not take comorbidities (secondary diagnosis) into account—and diabetes is often only documented as a secondary diagnosis in the German DRG system—, the frequency of diabetes in the hospital may be underestimated (4).

Even if diabetes is not the primary reason for hospitalization, expertise in diabetology is essential (4, 5). Compared to persons without diabetes, persons with diabetes stay longer in the hospital for the same main diagnosis, experience complications more frequently and have higher mortality rates (3, 6, 7). For example, metabolic decompensation often develops as the result of inpatient treatment with psychiatric medications, oncology drugs or steroids and can lead to life-threatening conditions (3, 4). Patients undergoing elective procedures also require adequate pre-, peri- and postoperative diabetes treatment to prevent complications (4, 5). However, comprehensive statistical information about diabetes in the hospital setting is lacking, even though such data is required to accurately assess the need for diabetological expertise (8). Thus, the aim of our study was to describe the proportion of diabetes cases among hospitalized patients, based on the mandatory DRG statistics for 2015–2017.

Methods

All inpatient cases aged ≥ 20 years (with and without diabetes) in the DRG statistics for 2015–2017 (source: Research Data Centers [FDZ, Forschungsdatenzentren] of the German Federal and State Statistical Offices [Destatis, Statistische Ämter des Bundes und der Länder]) were included in the study. Five types of diabetes were identified in the main diagnoses or secondary diagnoses based on the corresponding ICD-10 codes:

• Type 1 diabetes (E10)
• Type 2 diabetes (E11)
• Other specified diabetes mellitus, including pancreatic diabetes (diabetes resulting from diseases of the exocrine pancreas) (E13)
• Rare types of diabetes (E12 or E14)
• Gestational diabetes (O24).

In addition, we considered the cases with prediabetes (R73) as they are also associated with a high cardiovascular risk and require diabetological expertise, too. The absolute and relative frequencies of inpatient cases were analyzed stratified by year of treatment, type of diabetes, sex, and age group. In addition, length of hospital stay (median and mean) and in-hospital mortality (proportion of inpatient cases in percent) were analyzed stratified by type of diabetes and age group. The Wilcoxon test was used for comparisons between the 2015 and 2017 results as well as between cases with and without diabetes; all p-values were adjusted using false discovery rate correction (Benjamini-Hochberg procedure). The significance level (two-sided) was set at 0.05.

Based on prevalence estimates of the Central Institute for Statutory Health Care (Zi, Zentralinstituts für die kassenärztliche Versorgung) (9) and the German population as on 31 December 2017 (10), we estimated the prevalent population with type 2 diabetes in the general population so that we could calculate the proportion of inpatient treatments among patients with type 2 diabetes in 2017, stratified by age group and sex. The frequencies of various categories of main diagnoses were compared in cases with diabetes as a secondary diagnosis and cases without diabetes in each age group (refer to the eMethods section for further information).

Results

Frequencies of diabetes as a main or secondary diagnosis in 2017

In 2017, a main diagnosis or secondary diagnosis of diabetes mellitus was recorded in 3 058 685 (18.4%) of the total 16 656 350 hospitalized cases (Table). The vast majority of cases had type 2 diabetes (17.1% of all inpatient treatments, n = 2 842 677) which was mostly coded as a secondary diagnosis (94.4% of cases). Only 0.5% of all hospitalized cases had type 1 diabetes (n = 76 585); in 65.7% of these cases, the disease was documented as a secondary diagnosis (Table).

Table

Diabetes as a main diagnosis or secondary diagnosis in hospitalized cases from the age of 20 years, 2015–2017

Diabetes from 2015–2017

In the period from 2015 to 2017, the proportion of all cases with diabetes as a main diagnosis or secondary diagnosis remained largely stable at about 18% of a total of 16.4 to 16.7 million inpatient treatments of patients aged ≥ 20 years. Over this three-year period, type 2 diabetes was coded less frequently as a main diagnosis and more frequently as a secondary diagnosis (Table). The increase in cases documented as other/pancreatic diabetes (E13) (Table) is mainly due to a sharp increase in this secondary diagnosis among female, pregnant cases in the age category 20–39.

Diabetes stratified by sex and age group

During these three years, male inpatient cases with type 1 diabetes or type 2 diabetes were more common than female cases, especially in the age category 40–79 (eTable 1). In 2017, 0.5% of all male hospitalized cases (n = 40 605) versus 0.4% of all female hospitalized cases (n = 35 980) had type 1 diabetes and 19.7% of all male cases (1 536 988) versus 14.8% of all female cases (1 305 689) had type 2 diabetes (eTable 1).

eTable 1

Frequency of hospitalized cases with type 1 diabetes, type 2 diabetes or without diabetes in the period 2015–2017, stratified by sex and age group

In 2017, and in the two years prior, the absolute frequency of hospitalized cases increased up to the age of 50 (age category 50–59) in patients with type 1 diabetes and up to the age of 70 (age category 70–79) in patients with type 2 diabetes (eTable 1). In the period from 2017–2019, the absolute number of cases with type 2 diabetes decreased in the age category 70–79 years, while it increased in the age category 80 and older (eTable 1). With growing age, the number of cases with type 1 diabetes decreased among all hospitalized cases, while the proportion of cases with type 2 diabetes or diabetes in general increased (Figures 1a–c).

Figure 1a

Type 1 diabetes: Relative frequency of hospitalized cases with type 1 diabetes in 2017 stratified by age group and sex

Figure 1b

Type 2 diabetes: Relative frequency of hospitalized cases with type 2 diabetes in 2017 stratified by age group and sex

Figure 1c

Diabetes (all types): Relative frequency of hospitalized cases with Diabetes (all types) in 2017 stratified by age group and sex

Proportion of inpatient cases in the estimated population with type 2 diabetes in 2017

Up to the age of 60 (age category 60–69), the estimated number of inpatient treatments per 100 patients with type 2 diabetes was higher among women than among men (Figure 2). The largest difference was found in the age group of 20–39 in which more than a third of all female cases (35.2%) had a main diagnosis from the ICD group “O“ (pregnancy, childbirth and the puerperium).

Figure 2

Estimated number of inpatient cases per 100 patients with type 2 diabetes in 2017 stratified by age group and sex

Length of hospital stay and in-hospital mortality

Overall, between 2015 and 2017, both mean length of hospital stay and in-hospital mortality decreased significantly across all age groups, in all hospitalized patients (with or without diabetes), except for cases with type 1 diabetes in the age group of 60–89 years (eTable 2). In this three-year period, the mean length of hospital stay was significantly higher in cases with diabetes compared to cases without diabetes in each age group, especially in cases in the age category 40–69 years as well as in cases with type 1 diabetes. The largest difference in comparison to cases without diabetes was found in the cases with type 1 diabetes in the age category 40–49 years: 7.3 days vs. 4.5 days in the cases without diabetes in 2017 (eTable 2).

eTable 2

Length of hospital stay and in-hospital mortality stratified by age group in hospitalized cases with and without diabetes, 2015–2017

eTable 2 continued

Length of hospital stay and in-hospital mortality stratified by age group in hospitalized cases with and without diabetes, 2015–2017

Apart from cases with type 1 diabetes from the age of 80 (age category 80+), in-hospital mortality was higher in cases with diabetes compared to cases without diabetes in each age group. The largest difference was found in cases with type 2 diabetes in the age category 70–79: 3.7% vs. 2.8% in the cases without diabetes in 2017 (eTable 2).

Main diagnosis category in hospitalized cases with and without diabetes in 2017

Already from age 30 (age category 30–39; noticeably from age 40 [age category 40–49]), a disease of the cardiovascular system was more frequently listed as a reason for admission in cases with than in those without diabetes (in the age category 50–59: 23.3% of cases with diabetes vs. 13.9% of cases without diabetes, eFigure). From the age of 50 (age category 50–59), diseases of the respiratory or urogenital systems were also more frequently listed as the main diagnosis in cases with diabetes. By contrast, in hospitalized cases without diabetes aged 20–39, the relative frequency of admissions for diseases of the digestive, respiratory or urogenital systems was higher than in cases with diabetes. No significant differences were found for diseases of the nervous system and neoplasms (eFigure). The same is true for mental disorders/behavioral disorders, e.g. among the 40– to 49-year-olds, the age group most commonly affected by these disorders: 2.2% and 3.0% of the main diagnoses in cases with diabetes and without diabetes, respectively.

eFigure

Proportions of specific main diagnoses in hospitalized cases with and without diabetes stratified by age group in 2017

Discussion

Our analysis provides a comprehensive picture of diabetes in the hospital in Germany in the period from 2015 to 2017. The proportion of documented diabetes among hospitalized cases was stable at about three million (18%) over this period of three years. Based on nationwide billing data of panel doctors in Germany, the Central Research Institute of Ambulatory Health Care (Zi) estimated the prevalence of diabetes mellitus in 2015 to be 9.8% overall (9). The fact that the proportion of diabetes among hospitalized cases is almost twice as high in our results may be explained by the higher rate of hospitalizations in the elderly and/or by the increased need for inpatient care among people with diabetes.

Unlike other studies (3, 11), only cases with known and documented diabetes were included in our analysis. Thus, it is likely that the prevalence of diabetes in the hospital would be higher if cases with undiagnosed diabetes were included. An international comparison based on statistical information released by the Organization for Economic Cooperation and Development (OECD) shows that the rate of hospitalization with diabetes as the main diagnosis is particularly high in Germany (eTable 3).

eTable 3

Number of hospital discharges with diabetes mellitus by country, 2016 (OECD statistics)

Several studies indicate an increase in the prevalence of type 2 diabetes (7, 12), especially among men and among the 80– to 85-year-olds where the prevalence of diabetes is the highest (7, 9). Our analysis shows that, in line with the aging of the general population, the number of persons with diabetes in the hospital increases from the age of 80 years.

Although most of the cases in the middle age group (age category 40–79) were male, up to age 69, women with type 2 diabetes were more frequently hospitalized. One of the reasons for this could be the relatively higher diabetes-related mortality among middle-aged women (12). In addition, the number of pregnancies in women with known type 2 diabetes, including pregnancies after bariatric surgery, have steadily increased up to 2017 (13).

In recent years, gestational diabetes has more frequently been diagnosed, especially since the launch of the gestational diabetes screening program in 2012 (7, 13, 14). In our results, the increase in the category “other/pancreatic diabetes” can be partly explained by double coding of E13/O24 and is likely the result of incorrect coding of gestational diabetes.

Consistent with other publications (3, 6), we found a significantly higher mean length of stay and significantly higher in-hospital mortality in cases with diabetes. The lower in-hospital mortality among cases with type 1 diabetes aged 80 years and older may be explained by multiple readmissions in this patient group (lower in-hospital mortality per case, but not per person).

From the age of 40 (age category 40–49), a disease of the cardiovascular system was significantly more frequently listed as a reason for admission in cases with than in those without diabetes. It is a known fact that persons with diabetes are at a significantly increased risk of cardiovascular disease (2 to 4 times for men and 6 times for women) (15). In addition, persons with diabetes have twice the risk of heart failure (16).

Diseases of the respiratory and urogenital systems are further common reasons for admission in cases with diabetes from the age of 50 (age category 50–59). Smoking is associated with the development of diabetes (17); both diabetes and smoking are more common among individuals of low socioeconomic status (7). At the same time, smoking is the most common cause of chronic obstructive pulmonary disease and lung cancer (18). In addition, about 42% of people with type 2 diabetes are diagnosed with kidney damage in Germany. Renal failure due to diabetic nephropathy is the most common reason for renal replacement therapy (19).

Thanks to including diabetes as a main diagnosis and as a secondary diagnosis in the analysis, our study provides important information on the real prevalence of diabetes in inpatient care. Except for psychiatric and psychosomatic hospitals and rehabilitation facilities, the mandatory nationwide German DRG statistics can be regarded as representative. However, the DRG statistics consist of secondary data which were not primarily collected for research, but for billing purposes. There are certain limitations that come with this: One of these is that clinical information is lacking, another that the quality of coding, e.g. for gestational diabetes, is not guaranteed. In addition, comorbidities may be underestimated. Systematic overcoding of diabetes for billing reasons is unlikely, because diabetes, which is mostly coded as a secondary diagnosis, has currently little effect on revenue. Another limitation is that this statistical information is case-related and as such, does not allow conclusions to be drawn about the actual number of patients.

Conclusion

Since diabetes is more commonly coded as a secondary diagnosis in the DRG statistics, there is a risk that the need for expertise in diabetology is underestimated. Only 17% of hospitals have a qualification in diabetology (4). The fact that more than three million patients with diabetes are hospitalized each year underscores the need for qualified diabetes care in the hospital. Because diabetes is associated with life-threatening hypoglycemia and hyperglycemia as well as certain comorbidities, there is a need for a flexible and individual management in the hospital (20) which can keep up with the rapid advances in the combination drug therapy of diabetes and the field of technological treatment options, such as insulin pumps and continuous tissue glucose monitoring. If this need is left unaddressed, it will likely have negative effects on the treatment of patients with diabetes in the hospital as well as adverse economic consequences for the hospitals, such as an increased length of hospital stay and higher surgical complication rates. This is why reliable and detailed information about diabetes in the hospital is indispensable.

This analysis describes the extent of inpatient diabetes care in the period 2015–2017 with a special focus on the high need of persons with diabetes for inpatient care, as the population of often multimorbid diabetes patients continues to grow older.

Data source

Research Data Centers (FDZ) of the German Federal and State Statistical Offices, DRG statistics 2015–2017, own calculations.

Funding

This project was supported by the German Diabetes Association (DDG, Deutsche Diabetes Gesellschaft) and was developed at 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, Deutsche Zentrum für Diabetesforschung; FKZ: 82DZD14A02) and the University of Tübingen.

Acknowledgement

We would like to thank Mr. A. Hungele (ZIBMT, Institute of Epidemiology and Medical Biometry, University of Ulm) and Mrs. J. Loske (Research Data Center of the German Federal Statistical Office, Destatis).

Conflict of interest

Prof. Fritsche received lecture fees and consultancy fees from Sanofi, Novo Nordisk, Astra Zeneca, and Boehringer Ingelheim.

PD Dr. Siegel received consultancy fees from Lilly Deutschland GmbH, Novo Nordisk and Boehringer Ingelheim.

Prof. Karges received consultancy fees from Lilly Deutschland GmbH.

The remaining authors declare no conflict of interest.

Manuscript
received on 14 September 2020; revised version accepted on 4 February 2021

Translated from the original German by Ralf Thoene, MD.

Corresponding author
Marie Auzanneau, MPH
Institut für Epidemiologie und Medizinische Biometrie
ZIBMT, Universität Ulm, Albert-Einstein-Allee 41, 89081 Ulm, Germany

Cite this as:
Auzanneau M, Fritsche A, Icks A, Siegel E, Kilian R, Karges W, Lanzinger S, Holl RW: Diabetes in the hospital—a nationwide analysis of all hospitalized cases in Germany with and without diabetes, 2015–2017. Dtsch Arztebl Int 2021; 118: 407–12. DOI: 10.3238/arztebl.m2021.0151

►Supplementary material

eMethods, eTables, eFigures:
www.aerzteblatt-international.de/m2021.00151

References

1. International Diabetes Federation. IDF Diabetes Atlas, 8th edition. www.diabetesatlas.org (last accessed on 29 June 2020).

2. Tönnies T, Rockl S, Hoyer A, et al.: Projected number of people with diagnosed Type 2 diabetes in Germany in 2040. Diabet Med 2019; 36: 1217–25

3. Kufeldt J, Kovarova M, Adolph M, et al.: Prevalence and distribution of diabetes mellitus in a maximum care hospital: urgent need for HbA1c-Screening. Exp Clin Endocrinol Diabetes 2018; 126: 123–9

4. Fritsche A: Diabetes mellitus in der Klinik: Mehr Strukturen schaffen. Dtsch Arztebl 2017; 114(41): [16]

5. Moghissi ES, Korytkowski MT, DiNardo M, et al.: American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Diabetes Care 2009; 32: 1119–31

6. Levy N, Dhatariya K: Pre-operative optimisation of the surgical patient with diagnosed and undiagnosed diabetes: a practical review. Anaesthesia 2019; 74 (Suppl 1): 58–66

7. Nationale Diabetes-Surveillance am Robert-Koch-Institut: Diabetes in Deutschland – Bericht der Nationalen Diabetes-Surveillance 2019.

8. Fritsche A, Lobmann R: Diabetes mellitus in der Klinik. Stellungnahme des Fachbeirates Diabetes des Ministeriums für Soziales und Integration Baden-Württemberg: 2018.

9. Goffrier B, Schulz M, Bätzing-Feigenbaum J: Administrative Prävalenzen und Inzidenzen des Diabetes mellitus von 2009 bis 2015. Versorgungsatlas-Bericht Nr. 17/03. Zentralinstitut für die kassenärztliche Versorgung in Deutschland (Zi) Berlin: 2017.

10. Statistisches Bundesamt (Destatis): Bevölkerung: Deutschland, Stichtag, Altersjahre, Nationalität, Geschlecht/Familienstand. www-genesis.destatis.de (last accessed on 4 March 2020).

11. Müller-Wieland D, Merkel M, Hamann A, et al.: Survey to estimate the prevalence of type 2 diabetes mellitus in hospital patients in Germany by systematic HbA1c measurement upon admission. Int J Clin Pract 2018; 72: e13273

12. Jacobs E, Rathmann W: Epidemiologie des Diabetes in Deutschland. In: Deutsche Diabetes Gesellschaft (DDG) und diabetes DE (ed.): Deutscher Gesundheitsbericht, Diabetes 2019. 1 st edition. Mainz: Kirchheim 2019; p. 9–20.

13. Kleinwechter H, Schäfer-Graf U: Diabetes und Schwangerschaft. In: Deutsche Diabetes Gesellschaft (DDG) und diabetesDE (ed.): Deutscher Gesundheitsbericht, Diabetes 2019. Mainz: Kirchheim: 2019; p. 150–7.

14. Melchior H, Kurch-Bek D, Mund M: The prevalence of gestational diabetes—a population-based analysis of a nationwide screening program. Dtsch Arztebl Int 2017; 114: 412–8

15. Tschöpe D, Ringelstein EB, Motz W: Diabetes mellitus – Herzerkrankungen – Schlaganfall. In: Deutsche Diabetes Gesellschaft (DDG) und diabetesDE (ed.): Deutscher Gesundheitsbericht, Diabetes 2019. Mainz: Kirchheim 2019; p. 57–65.

16. Schuett K, Marx N: Herzinsuffizienz bei Diabetes mellitus in Deutschland. In: Deutsche Diabetes Gesellschaft (DDG) und diabetesDE (ed.): Deutscher Gesundheitsbericht, Diabetes 2019. Mainz: Kirchheim 2019; p. 118–23.

17. Willi C, Bodenmann P, Ghali WA, et al.: Active smoking and the risk of type 2 diabetes: a systematic review and meta-analysis. JAMA 2007; 298: 2654–64

18. World Health Organization: WHO global report: mortality attributable to tobacco. Geneva: WHO 2012.

19. Merker L: Diabetes und Nierenerkrankungen. In: Deutsche Diabetes Gesellschaft (DDG) und diabetesDE (ed.): Deutscher Gesundheitsbericht, Diabetes 2019. Mainz: Kirchheim 2019; p. 76–80.

20. Breuer TG, Meier JJ: Inpatient treatment of type 2 diabetes. Dtsch Arztebl Int 2012; 109: 466–74