Introduction

The proportion of Germanys population older than 65 years of age is expected to rise to 34% by the year 2034 and similar developments are reported for other developed countries [1, 2]. In addition, older patient’s lifestyle changed to higher activity levels, which also results in a higher number of accidents and more elderly trauma patients [3, 4]. In this context, it has been demonstrated in recent years that the principles of care for younger patients cannot be readily applied in the treatment of severely injured elderly [5]. In addition to differences in accident mechanisms and the influence of co-morbidities as well as long-term medication, physiological changes, such as the response to blood loss, are also difficult to compare with a younger collective. For these reasons, changes have already been made in the treatment of this subgroup. For example, it is recommended that higher age alone should be considered as a criterion for trauma team activation and that special teams should be established for the treatment of geriatric trauma [3, 6]. It is well-known that increasing age is an independent risk factor for mortality regardless of injury severity, where there is an increased risk of mortality, especially if the patient is older than 74 years of age [7,8,9]. Furthermore, the Revised Injury Severity Classification (RISC) II score demonstrated that age greater than 55 years makes an independent contribution to outcome that increases with age [10].

Length of hospital stay, along with mortality, is used as a quality criterion for the care of severely injured patients. Here, the Injury Severity Score (ISS), the injury pattern, the occurrence of complications, and the presence of pre-existing conditions is assumed to increase resource use and length of stay on Intensive Care Units (ICU) as well as the overall in-patient stay [11, 12]. Due to the numerous concomitant circumstances complicating treatment in the elderly, one might assume that especially severely injured geriatric patients are more likely to undergo a complicated long-term stay in intensive care and prolonged overall hospitalization [13]. Especially in the complex treatment of severely injured elderly patients, consequently longer courses seem likely to occur and cause excessive resource consumption and high costs.

Therefore, in this study, we focused on the identification of independent risk factors that might prolong the ICU-LOS and overall duration of hospitalization in severely injured elderly.

Methods

Data

The TraumaRegister DGU® (TR-DGU) of the German Trauma Society (Deutsche Gesellschaft für Unfallchirurgie, DGU) was founded in 1993. The aim of this multi-center database is a pseudonymized and standardized documentation of severely injured patients.

Data are collected prospectively in four consecutive time phases from the site of the accident until discharge from hospital: A) Pre-hospital phase, B) Emergency room and initial surgery, C) Intensive care unit and D) Discharge. The documentation includes detailed information on demographics, injury pattern, comorbidities, pre- and in-hospital management, course on intensive care unit, relevant laboratory findings including data on transfusion and outcome of each individual. The inclusion criterion is admission to hospital via emergency room with subsequent ICU/IMC care or reach the hospital with vital signs and die before admission to ICU. The infrastructure for documentation, data management, and data analysis is provided by AUC–Academy for Trauma Surgery (AUC—Akademie der Unfallchirurgie GmbH), a company affiliated to the German Trauma Society. The scientific leadership is provided by the Committee on Emergency Medicine, Intensive Care and Trauma Management (Sektion NIS) of the German Trauma Society. The participating hospitals submit their data pseudonymized into a central database via a web-based application. Scientific data analysis is approved according to a peer review procedure laid down in the publication guideline of TraumaRegister DGU®.

The participating hospitals are primarily located in Germany (90%), but a rising number of hospitals of other countries contribute data as well (at the moment from Austria, Belgium, China, Finland, Luxembourg, Slovenia, Switzerland, The Netherlands, and the United Arab Emirates). Currently, more than 28,000 cases from almost 700 hospitals are entered into the database per year. Participation in TR-DGU is voluntary. For hospitals associated with TraumaNetzwerk DGU®, however, the entry of at least a basic data set is obligatory for reasons of quality assurance. Given the background of quality assurance and anonymized data, informed consent is not necessary, which is confirmed by the review board of the official trauma registry of the german society of trauma surgery. Initial data and privacy agreements are obtained from the participating hospitals. The study was performed in accordance with the Declaration of Helsinki and following STROBE guidelines and the RECORD guidelines for observational studies (Reporting of studies Conducted using Observational Routinely Collected Data).

Inclusion and exclusion criteria

This retrospective cohort study included patients from the TR-DGU who were treated in participating German speaking trauma centers (Countries: Germany, Austria, Switzerland; GAS-Regions) between 01/2016 and 12/2020. We included hospitals from the GAS-Regions due to their homogenous medical care quality. The study period was chosen due to the fact, that some parameters were introduced in the questionnaire of the database since 2016. We wanted to examine a cohort with geriatric, severely injured patients. Therefore, this study only included patients aged 65 years or older with an ISS of at least 16 points. We excluded patients who were early transferred out (< 48 h) or transferred in with a delay of more than 2 days due to incomplete patient records and patients without ICU treatment. Patients who died or were transferred within the regular stay were excluded since there were no data available afterwards and could not be evaluated with respect to a prolonged stay.

The present study is in line with the publication guidelines of the TraumaRegister DGU® and registered as TR-DGU project ID 2021–030.

We defined our threshold for a prolonged stay (intensive care unit and hospital) on the distribution of the length of stay of the overall collective of severely injured elderly. Patients with a length of stay (LOS) above the 80% percentile were defined as the long-term group. This resulted in more than 13 days for the ICU LOS, and more than 25 days for hospital LOS. For patients who did not meet the criteria for an extended stay, deceased and transferred patients were subsequently excluded.

Coagulopathy was considered present if at least one of the following conditions were found on admission: Quick’s value ≤ 60%, or PTT ≥ 40 s, or International Normalized Ratio (INR) ≥ 1.4.

Statistical analysis

Continuous variables are presented with mean and standard deviation (SD), or as median with interquartile range (IQR, 1st and 3rd quartile) in case of a rather skewed distribution. Categorical variables are presented with number and percentage. Due to the large number of cases, formal statistical comparison of the two study groups was avoided since even minor differences (± 1%, or ± 0.05 SD) would turn out to be significant (p < 0.05).

Risk factors for a prolonged stay (hospital and/or ICU) were investigated with a multivariate logistic regression model. The independent predictors are listed in Table 5. Results are presented as odds ratios (OR) with a 95% confidence interval (CI).

Statistical analysis was performed using SPSS statistical software (version 25, IBM Inc., Armonk, NY, USA).

Results

A total of 202,817 patients from the TR-DGU were checked for eligibility. The median hospital length of stay was 18 [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28] days. A total of 6,335 deceased (median length of stay 2 [1,2,3,4,5,6] days) were excluded from further analysis just as 1,441 were transferred (LOS: 9 [6,7,8,9,10,11,12,13] days). After those exclusions, n = 15,255 geriatric patients aged 65 years or more qualified for analysis (Fig. 1). Of these patients, n = 8808 had a regular length of stay while n = 6447 (42.2%) met the definition for a prolonged stay (Tab. 1).

Fig. 1
figure 1

Definition of the study population (D/A/CH Germany/Austria/Switzerland, yrs. years, ISS injury severity score, ICU intensive care unit, LOS length of stay)

Full size image
Table 1 Basic demographic patient data of patients with and without prolonged LOS
Full size table

Mean age was 76.6 years with 75% (n = 11,961) of the patients being older than 70 years of age, where age was comparable between groups (regular LOS 76.9 yrs. vs prolonged stay 76.3 yrs.). 62.3% (n = 9.498) of the patients were male, with a higher proportion of male patients in the group with prolonged stay (regular stay 59.2% vs. prolonged stay 66.4%). The mean Injury Severity Score (ISS) was 24.0 ± 8.2 points, and 98.5% (n = 14,362) suffered a blunt trauma. 34.7% of the patients had multiple injuries according to the Berlin Definition of polytrauma (regular stay: 27.0% vs. prolonged stay: 45.2%) with a mean of 5.16 trauma-specific diagnoses. Higher ISS and more patients fulfilling the Berlin Definition of polytrauma were recorded for the prolonged LOS group, additionally the rate of lower extremity fractures in these patients was almost twice as high. Basic patient data is presented in Table 1 (Tab. 1).

Most common mechanism of injury were falls from lower heights (44%), while 37% were involved in traffic accidents. Attempted suicide was registered in 1.9% of all patients and was more frequently seen in patients with prolonged LOS (prolonged LOS 3.5% vs. 0.8% regular LOS). Details regarding the mechanism of trauma are listed in Table 2.

Table 2 Mechanism of trauma. (LOS: length of stay)
Full size table

A higher rate of patients that presented with longer stays were unconscious on scene (Glasgow Coma Scale, GCS ≤ 8 points): 6.8% regular LOS vs. 23.8% prolonged LOS (Table 3). Also shock on scene (regular LOS 4.3% vs. 11.4%) and shock on admission (regular LOS 3.8% vs. 12.9%) were observed more frequently in the prolonged LOS group. About half of the patients were on permanent platelet inhibition (46.7%), 23.4% used DOAC and 20.4% were on Vitamin-K-antagonists. While non-survivors were excluded from the regular LOS group, hospital mortality was 15.2% (n = 977) in patients with a prolonged stay.

Table 3 Prehospital data
Full size table

Mean/median length of ICU stay was 10.0/6 days and mean/median length of hospital stay was 22.5/18 days. The necessity of mechanical ventilation on ICU was given in 43.8% of all patients, where the rate of mechanical ventilation (regular LOS: 20.7% vs. prolonged LOS 75.4%) was higher in the group with prolonged stay. If mechanical ventilation was required, duration of ventilation (regular LOS: 0.5 ± 1.7 d vs. prolonged mechanical ventilation 11.7 ± 13.3 days) higher for patients with longer stays. Multiple organ failure occurred in 44.5% of patients with a prolonged stay (vs. 6.6% with regular LOS) (Tab. 4).

Table 4 Outcome parameter
Full size table

We identified 12 significant risk factors for a prolonged hospital or ICU stay using logistic regression analysis (Table 5). The need for mechanical ventilation (OR 2.55), and especially the prolonged need for ventilation (OR 37.8), is most important for an prolonged stay. Complications like sepsis (OR 4.12), multi organ failure (OR 1.85), or the occurrence of thromboembolic events (OR 2.72) predicted a longer stay. Extremity fractures also increased the risk for a prolonged stay significantly (OR for legs 2.34; OR for arms 1.37). Injuries to the head, the thorax, or the abdomen were less important, after adjustment.

Table 5 Adjusted risk factors for prolonged ICU or hospital LOS, identified with multivariate logistic regression analysis. Risk factors were sorted according to their decreasing importance (odds ratio)
Full size table

Discussion

In the present study, we were able to demonstrate that numerous independent risk factors in the polytraumatized elderly patient exist that contribute to a prolonged inpatient stay. In addition to extended ventilation time, the occurrence of organ failure and thromboembolic events, lower extremity injuries and suicide attempts were identified as risk factors that might possibly be influenced by further improvements in terms of prophylactic and therapeutic interventions, but also organizational and structural adjustments in the care of multiply injured elderly.

An internationally accepted distinct age threshold for the definition of a geriatric trauma patient remains outstanding which makes it difficult to clearly define a geriatric collective of trauma patients. Nevertheless, it is indisputable that elderly patients should be considered as a special group with multiple peculiarities. In this context, numerous studies have demonstrated a relationship between increasing age and mortality following trauma [14]. An in-house mortality of over 60% has been reported for multiply injured patients older than 85 years of age which is significantly higher than for younger patients [15]. Once surviving the initial trauma elderly patients are threatened with ongoing inactivity and long-term changes in their life conditions as well as long hospital stays and extended periods of rehabilitation [16, 17][18]. Based on observable physiological changes and effects on outcome, Advanced Trauma Life Support (ATLS®) recommends treatment in specialized trauma centers from age > 55 years [3]. These expectations can be supported by studies comparing the outcome before and after the implementations of specialized geriatric trauma centers [19]. Multiple studies defined the threshold for a geriatric population by an age of older than 65 years. Regardless of the incongruent age definition in multiple studies, the data presented here clearly meets the criteria of a geriatric collective with a mean age of > 75 years of age.

Length of hospital stay in trauma is identified as one of the main cost factors in the treatment of severely injured patients [20]. In addition, prolonged ICU-LOS is also known to be independently associated with increased mortality and the risk to sustain an adverse event increases by 6% per hospital day [21,22,23]. For these reasons, we focused on identifying factors that may condition a prolonged stay. Some studies describe disproportionate early patient death as a bias to predict those factors, due to the fact, that patients decease before they even reach the threshold for a prolonged stay. For this reason, we deliberately excluded patients who died prior to our threshold for a prolonged stay from the present study [24]. Once the threshold for prolonged stay was exceeded, mortality in our collective was 15.2%, which is slightly higher than described in age-independent collectives and again emphasizes the higher mortality of older patients [7]. However, no final statement can be made about the functional outcome from the registry data. Apart from mortality, due to the structure of this registry study, we are unable to make an informed statement about the long-term outcome of patients.

Comparable to threshold in regard of the definition of a geriatric population, there is disagreement about the definition for a prolonged inpatient stay or a prolonged intensive care stay, respectively. This complicates the definition and especially the comparability with the few existing studies analyzing prolonged ICU and hospitals stays in trauma. Recent studies use definitions that are arbitrarily set at, for example, 10 days as well as standard lengths of stay that are defined depending on the diagnosis [25]. We based our threshold for a prolonged stay on the average length of stay of the overall collective and defined it belonging to the top 80 percentage. In an age-independent study by Hwarjibe, for example, the threshold was 21 days [25]. Nonetheless, the rate of long-layers in this study was only 5%, it should be noted the ISS was markedly lower than in our patients (ISS mean 18 (12) in the prolonged group). Despite different definitions of prolonged in-hospital stays, this may indicate that prolonged LOS is more common in geriatric patients.

It is generally known that trauma patients make up a relatively small proportion of intensive care patients but consume a high level of resource commitment and have comparatively long intensive care courses [11]. Nevertheless, good outcomes have been described for trauma patients despite longer hospital stays [24]. Previously, high ISS, TBI, co-morbidities, and the occurrence of complications during the inpatient stay were identified as predictors of long hospital stays following trauma [12, 26]. Apart from that, a matched pair study detected that geriatric patients sustain more complications than their younger counterparts [27]. In one of the few studies looking at prolonged hospitalization of trauma patients, Trottier et al. identified higher ISS as a risk factor in addition to advanced age [28]. Our patients with prolonged stays also showed higher ISS, but this was not confirmed as an independent risk factor going along with a comparable mortality during intensive care hospitalization irrespective of the underlying fracture as shown in another study [29]. Mechanical ventilation, particularly prolonged ventilation times for more than 10 days and the development of complications such as sepsis and multiorgan failure were strongly associated with prolonged stay. These findings are supported by a study that has described prolonged ventilation was as a factor for extended ICU-stays in trauma [30]. At the same time, higher age is also known to be a risk factor for prolonged mechanical ventilation [31]. Although higher age was also hypothesized to be a risk, multivariate analysis in the mentioned study showed that age alone could not be considered as sole risk factor for delayed discharge. In this study, the development of urinary tract infections, sepsis and deep vein thrombosis were identified as potentially influenceable risk factors [24]. In a recent study, Spering et al. demonstrated that multiorgan failure–which was a risk factor for prolonged stay in our study—is more common in elderly trauma patients, reaffirming the necessity to take care of this complication in the elderly [1].

In addition to a serious injury of especially the lower extremities, attempted suicide was also identified as a cause for late discharge in our work. In clinical practice, transfer to psychiatric hospitals after trauma as well as transfer to rehabilitation clinics is often complicated when mobility of the patient is still insufficient. The present results confirm these clinical impressions and can thus be identified as a structural problem at the transition between acute and subacute treatment. Interestingly, this observation with problems to organize a subacute stay which then leads to a medically unnecessary elongation of the inpatient stay has been confirmed in an age-independent analysis for trauma patients. Here, 46% of extended stays were attributed to this factor, equaling an odd’s ratio of about 5% [25]. Consequently, there appears to be a fundamental potential for improvement here, as discharge delays to post-acute care facilities have been described as a risk factor for prolonged hospitalization in non-trauma patients as well [32] [33, 34].

We demonstrated that the need for RBC administration during the resuscitation in the emergency department was associated with an increased risk of prolonged stay. This is consistent with the results of several studies in surgical patients in which an association between the administration of blood products and prolonged stay was also found [35]. However, the present results must be interpreted with caution; a definite statement as to whether the administration of blood products is causative or associative with a prolonged stay cannot be made with certainty from the data available here.

Limitations

Certain limitations of this investigation must be acknowledged. As the present study is a retrospective registry analysis, all findings represent associations and do not claim any causality. There may be some predictors for length of stay, which we cannot correct for since they were not documented in the registry. Although the data quality in the TR-DGU is high based on multiple checks on data entry, registry data are general less valid than data from prospective randomized trials. Long-term data beyond discharge were not available, and the structure of the registry did not allow to perform follow-up examinations. Additionally, the mortality in the prolonged stay group may be conditioned in part by the existence of and compliance with an advance directive [36]. To make results more comparable, an overall definition for prolonged stay would be necessary.

Conclusion

In the present study, independent risk factors associated with prolonged hospitalization were identified for severely injured elderly patients. In addition to mechanistic criteria (e.g. suicide attempt) and specific injury patterns such as severe lower extremity injuries, logistic influences such as the interface with further rehabilitation could be identified as risk factors. Therefore, in addition to attention to an improvement in clinical treatment, structural changes in patient care should be included in further planning, as well.