The epidemiology of vertebral fractures in the adult German population and an analysis of the socio-economic burden: Trends from 2009 to 2019.

Objective To present nationwide rates of hospitalized patients with vertebral fractures over one decade and a comprehensive analysis of treatment characteristics and incurred costs in 2019. Methods Trends in the incidence were quantied based on annual ICD-10 diagnosis codes from all German medical facilities between 2009 and 2019, provided by the Federal Statistical Oce (Destatis). The ICD-10 Codes “S12.0-2; S22.0; S32.0; S32.1-2” were evaluated. The relative change from 2009 through 2019 was determined and data on the secondary diagnoses, OPS-codes, treatment on an intensive care unit (ICU), in-hospital mortality, proportion of G-DRGs and cumulative costs were assessed based on the Institute for the Hospital Remuneration System (InEK) data for 2019. The incidence increased by 45.6% between 2009 and 2019 to 150.7 per 100,000 inhabitants. Lumbar spine fractures showed the highest incidence in 2019 with 70.5/100,000 inhabitants (46.8% of all vertebral fractures). The highest increases were seen in subaxial cervical fractures (+121.2%) and sacral fractures (+306.6%). 63.7% of vertebral fractures in 2019 were diagnosed in women and 69.0% in patients ages 70 years or older. Osteoporosis was documented in 17.9% of cases. In 10.1% of cases an ICU treatment was documented. The in-hospital mortality was 2.0%. The most frequently applied G-DRG code was I68D (33.3%). The total costs in 2019 were €589,205,715. The evaluation of 955,091 vertebral fractures showed a sharp increase in the incidence rate with incurred cost of approximately €590 million in 2019. The age and gender distribution suggests an increase in the diagnosis of osteoporotic fractures.


Introduction
Injuries of the spine make up a relevant share in trauma patients and thoracolumbar fractures rank among the severest injuries of the human skeleton 1,2 . Around 2-3% of all accident victims have an injury to the cervical spine 3,4 . The most common cause of traumatic vertebral injury are tra c accidents, followed by falls from low and higher heights (>2m) 5,6 . Injuries to the spine have been reported to be associated with spinal cord injuries (SCI) in up to 20% of cases, with this occurring more frequently in injuries to the lower cervical spine and upper to middle thoracic spine 6 . Further, it is reported that the quality of life of patients suffering thoracolumbar fractures is compromised, mostly independent from the treatment modalities, but seems to be determined by the severity of the injury [7][8][9] . Although many individuals who experience a vertebral fracture can be managed on an outpatient basis, a considerable number of patients must be hospitalized, resulting in substantial costs. However, current knowledge about the epidemiology of vertebral fractures is largely based on relatively small populations and only a few nation-wide analyses exist 2,10 . The in uence of geriatric injuries and osteoporotic spontaneous fractures on the prevalence of vertebral fractures is not su ciently clari ed. It is recognized that vertebral fractures are one of the most common skeletal fractures associated with low bone mass and other causes of skeletal fragility 11 . Nation-speci c epidemiological analyses are essential as incidences, the hospital supply structure and reimbursement systems vary country-wise: Discharge rates of hospitalised patients with vertebral fractures can vary by more than 4-fold between European countries 12 . Detailed analyses of epidemiologic data, current treatment maxims and cost incurred are a valuable resource for stakeholders in the health systems, which can assist decision making of health services and give insights into effects of current prevention and therapeutic strategies.
Therefore, the aim of this study was (1) to determine the development of the nationwide incidence of vertebral fractures as a function of age, gender and anatomical localization for Germany between 2009 through 2019 (2) to provide a comprehensive overview of the treatment characteristics based on documented OPS-codes, secondary diseases, in-hospital mortality rates and intensive care unit (ICU) treatments. (3) Lastly annual costs of vertebral fractures were to be evaluated for the German health care insurance system based on a G-DRG code analysis. (1) The development of the nationwide incidence of vertebral fractures from 2009 to 2019.

Results
In total 955,091 fractures of the spine were registered between 2009 and 2019. In comparison to the year 2009, in which 70,235 fractures were listed, the incidence increased by 45.6% from 105.8/100,000 inhabitants to 150.7/100,000 inhabitants ( Figure 1). Of the overall vertebral fractures more occurred in women (57%) and 69.0% % were calculated for patients aged 70 years or older (Table 1).  Figure 2D). Men in the age group 80-89 years suffered from lumbar spine fractures in 28.4% of all cases ( Figure 2D). Second, fractures of the thoracic spine were common with 28,057 hospitalized cases in 2019 and an incidence of 41.3 per 100,000 inhabitants. Of all the cases 63.6% were female and 36.4% male. Highest incidences were registered in women aged 80-89 years (38.6%), 70-79 years (25.9%) and female patients aged 60-69 years (11.5%), whereas 23.6% of all fractures occurred in men aged 80-89 years and 20.0% in men aged 70-79 ( Figure 2C). These were followed by fractures of the sacrum and coccyx, which together affected 14,748 patients with an incidence of 21.7 cases per 100,000 inhabitants in 2019. More women than men suffered from fractures at this site (81.9% vs. 18.1%), whereby female patients in the age group 80-89 years comprised the largest cohort (45.4%), followed by women aged 80-89 years (30.2%) and older than 90 years (15.3%). For male patients most of the cases occurred in the age group 80-89 years with 30.2% of all fractures ( Figure 2E). From all sacro-coccygeal fractures, when not differed by sex 78.8% of fractures were documented in patients 70 years or older.  (Table 2), the highest increase was found in fractures of the sacrum (+306.6%) and at the subaxial cervical spine (+121.2%), followed by atlas (+106.4%) and axis fractures (+82.5%). Also, numbers considerable heightened for thoracic spine fractures (+32.2%) and lumbar spine fractures (+21.7%). There was only a slight increase of the incidence coccygeal fractures (+4.4%). A detailed analysis of the incidences and total numbers of vertebral fractures from 2009 to 2019 broken down according to the anatomical location, sex and age is provided in    Table 5.
(3) The economic burden of vertebral fractures for inpatient cases based on a G-DRG analysis in 2019.
Considering all vertebral fractures, the mean hospital stay was 10.7 ± 9.6 days. In cases with sacrum fractures the longest mean hospital stay (13.9 ± 9.8 days per case) was documented. At average in 16.8% of all cases high clinical complexity levels (PCCL ≥ 3) were seen. In patients with Atlas & Axis fractures 24.4% of cases were classi ed PCCL ≥ 3, in patients with subaxial cervical fractures 24.7% and in patients with fractures of the sacrum 28.8%. Table 6 provides an overview of the mean hospital stay and the proportion of cases according to the PCCL for vertebral fractures.
The most frequently used G-DRG code in cases with vertebral fractures in 2019 was I68D (Diseases and injuries of the spine not treated surgically) in 33.3% of all cases. Second common was I41Z (Geriatric early rehabilitation and complex treatment for diseases and disorders of the musculoskeletal system and connective tissue) in 10

Discussion
In this cross-sectional study, the development of vertebral fracture incidence was determined as a function of age, gender, and differentiated by anatomical locations. An outstanding characteristic is that the analysis is based on registry data consisting of ICD-10 diagnosis codes from all German medical institutions showing the development of fracture rates over ten years. Additionally, nationwide data on secondary diagnoses, procedures (based on OPS-codes), and the G-DRG distribution were assessed based on the InEK data and report browsers for the year 2019. Whereas studies relying on data from single hospitals may yield skewed results, the ndings presented here are based on nationwide reports from the largest country of the European Union.
In 2019, a total of 102,285 fractures were registered depicting an increase by 45.6% since 2009. The incidence of all vertebral fractures was 151 per 100,00 inhabitants. We recently showed, that fractures of the thoracic and of the lumbar spine are among the ten most common fractures in Germany in 2019 13 . We demonstrated that most fractures can be found in the elderly population: The ratio of documented fracture cases in patients aged younger than 70 years over patients aged 70 years or older was 0.45. This indicates, that more than twice as many vertebral fractures to be 14% in X-ray, 88% in CT, and 100% in MRI 22 . They further gured out that additional pathologies were identi ed in MRI of the lumbar spine in 51% and pelvis in 18%, respectively 22 . They consequently advocated for extensive MRI diagnostic of the lumbar spine including the sacrum for symptomatic elderly patients 22 . In the current population in 60.5% of the cases a CT and in around 30% of the cases an MRI of the spine or the pelvis were applied in 2019.
The most documented secondary diagnoses were hypertension, osteoporosis, atrial brillation, type II diabetes and chronic kidney disease. These can be assumed to be indicative for the elderly German population 23,24 . We found 18,773 (17.9%) cases in which osteoporosis was documented as a secondary diagnosis. Noteworthy, it has not been screened for "M80. Interestingly concomitant vertebral fractures were documented as secondary diagnoses in a high share of cases. These fractures were seen at the lumbar spine in around 29% of cases and at the thoracic spine in around 18% of cases. Hypothetically these fractures partly can be accounted to poor bone quality in patients with osteoporosis: The incidence of concomitant vertebral fractures at rst contact in patients with osteoporotic vertebral fractures has been reported to be 26% in a register based epidemiological study 30 . Further, multilevel contiguous osteoporotic lumbar compression fractures have been described 31 . Also in non-osteoporotic patients multilevel vertebral injuries are common and incidences around 20-30% have been reported [32][33][34] . Especially occult, non-continuous vertebral injuries bare the risk of delayed diagnosis. Therefore some authors advocate for considering whole spine MRI in vertebral trauma patients 32,34 .
We found 1.1% cases of vertebral fractures to be associated with SCIs. In concordance with the ndings of Smits at al. SCI were associated with cervical fractures in most cases. Den Ouden reported on a SCI rate of 8.5% in their population of traumatic vertebral fracture patients in 2016 17 38 . These estimations are in range of the average cost for the respective G-DRG on which our analysis is based on.
The increase in fracture incidence poses a challenge for stakeholders in the health care systems. Fragility fractures were projected to 928,000 in 2025 with an estimated socioeconomical burden of €11,261 million [20]. Vertebral fractures, even treated conservatively, incur substantial healthcare utilization and costs. However, the costs of work absenteeism and disability add to the overall costs, associated with vertebral fractures, and could not be accessed in this study. Further studies on the cost-effectiveness, cost utility and long-term costs of vertebral fractures are necessary to comprehend the total economic burden. The current analysis summarizes the up-to-date costs of vertebral fractures of the German population in 2019 and therefore provides important information for the planning of preventive and treatment strategies.

Strengths and limitations
The main strength of the current study is the nationwide analysis over one decade. To provide comprehensive information on different aspects of the inpatient treatment of vertebral fractures two data source were employed: Data based on the Destatis analysis (2009-2019) and data based on the InEK data and report browser (2019). The study is limited by the fact, that, although ICD-10 codes as a function of age and gender were available, it was not possible to determine the trauma mechanisms. Whereas correct coding of diagnosis can be assumed since DRG lump sum payment relies on it, which is strictly controlled by the Medical Service of Health Funds. Further, only inpatient data was available. Even though hospitalization is required in many cases, the reported fracture incidences may be underestimated as fractures treated in an outpatient setting are not included in the analysis. Moreover, a detailed analysis was conducted only for primary diagnoses. We showed, that in this population in 58.4% of cases an additional vertebral fracture was documented as a secondary diagnosis. The overall number of vertebral fractures might be underestimated, because secondary diagnoses were not screened for all relevant primary diseases.
The evaluation of 955,091 vertebral fractures between 2009 and 2019 showed a relevant increase in the incidence rate to 151 cases per 100,000 inhabitants, highlighting the challenge for stakeholders in the health care systems. This is underlined by the costs incurred of approximately €590 million in 2019. The age and gender distribution suggests an increase in the diagnosis of insu ciency vertebral fractures. Considering the relevant share of complex cases with the need of surgical and ICU treatment the expanding nationwide implementation of vertebral care centres appears reasonable and necessary.

Federal Statistical O ce of Germany (Destatis)
Data consisting of annual ICD-10 diagnosis codes from German medical institutions between 2009 through 2019 was provided by the Federal Statistical O ce of Germany (Destatis). Total number of vertebral fractures were quanti ed using the ICD-10 Codes "S12.0", "S12.1", "S12.2", "S22.0", "S32.0", "S32.1" and "S32.2" (Table 8)   In accordance with Section 17b of the German Hospital Financing Act (KHG), a universal, performance-based, and at-rate remuneration system has been introduced for general hospital services. The basis for this is the G-DRG system (German Diagnosis Related Groups system), whereby each inpatient case of treatment is remunerated by means of a corresponding DRG lump sum payment.
The InEK GmbH provides detailed data on the main diagnoses (based on ICD-10 codes), secondary diagnoses (based on ICD-10 codes), procedures (based on OPS-codes), and the G-DRG distribution. Data is accessible via the InEK Data Browser 40 . The Browser enables analysis back to the year 2019. The following comprehensive analysis was made only for the year 2019: Based on the ICD-10 codes for vertebral fractures, displayed in Table 8 data for total case numbers, numbers of in-hospital deaths, and numbers of cases treated on an ICU were extracted. Further, the data on secondary diagnoses, OPS-codes, time of hospital stay (mean ± standard deviation) and G-DRG codes were analyzed. A speci c G-DRG code corresponds to each one case that was determined based on the ICD code. Cases that were treated for the same diagnosis within the same year were grouped together. Thus, no duplicates were recorded.
To estimate the cost for inpatient treatment of vertebral fractures the G-DRG Report browser was employed 41 . The G-DRG Report Browser 2019 displays the calculation results for the G-DRG system 2019 on DRG level. The share of cases according to the Patient Clinical Complexity Level (PCCL) was adopted from the InEK Data Browser. The PCCL value is calculated in a complex procedure from the secondary diagnosis values (complication or comorbidity level values -CCL) and indicates the severity of the complication or comorbidity based on results between 0 (no CC) and 6 (most severe CC). Costs according to the distribution of applied G-DRG codes, that were used in at least 0.05% of cases were added up proportionally and calculated as mean value per case. Figure 1 The  Supplementary Files