Strategies for comparative analyses of registry data

Patient-Reported Outcome Measures in High-Risk Medical Device Registries: A Scoping Review

Little is known about the methods and outcomes of patient-reported outcome measure (PROM) use among high-risk medical device registries. The objective of this scoping review was to assess the utility and predictive ability of PROMs in high-risk medical device registries. We searched Ovid Medline, Embase, APA PsychINFO, Cochrane Library, and Scopus databases for published literature. After searching, 4323 titles and abstracts were screened, and 262 full texts were assessed for their eligibility. Seventy-six papers from across orthopedic (n = 64), cardiac (n = 10), penile (n = 1), and hernia mesh (n = 1) device registries were identified. Studies predominantly used PROMs as an outcome measure when comparing cohorts or surgical approaches (n = 45) or to compare time points (n = 13) including pre- and postintervention. Fifteen papers considered the predictive ability of PROMs. Of these, 8 treated PROMs as an outcome, 5 treated PROMs as a risk factor through regression analysis, and 2 papers treated PROMs as both a risk factor and as an outcome. One paper described PROMs to study implant survival. To advance methods of PROM integration into clinical decision-making for medical devices, an understanding of their use in high-risk device registries is needed. This scoping review found that there is a paucity of studies using PROMs to predict long-term patient and clinical outcomes in high-risk medical device registries. Determination as to why PROMs are rarely used for predictive purposes in long-term data collection is needed if PROM data are to be considered suitable as real-world evidence for high-risk device regulatory purposes, as well as to support clinical decision-making.

Level of Evidence 4

Early total care or damage control orthopaedics for major fractures ? Results of propensity score matching for early definitive versus early temporary fixation based on data from the trauma registry of the German Trauma Society (TraumaRegister DGU®)

PURPOSE

Damage control orthopaedics (DCO) und early total care (ETC) are well-established strategies for managing severely injured patients. There is no definitive evidence of the superiority of DCO over ETC in polytrauma patients. We conducted this study to assess the probability of a polytraumatised patient undergoing DCO. In addition, the effect of DCO on complications and mortality was investigated.

METHODS

We analysed data from 12,569 patients with severe trauma (Injury Severity Score ≥ 16) who were enrolled in the trauma registry of the German Trauma Society (TraumaRegister DGU®) from 2009 to 2016 and had undergone surgery for extremity or pelvic fractures. These patients were allocated to a DCO or an ETC group. We used the propensity score to identify factors supporting the use of DCO. For a comparison of mortality rates, the groups were stratified and matched on the propensity score.

RESULTS

We identified relevant differences between DCO and ETC. DCO was considerably more often associated with packed red blood cell (pRBC) transfusions (33.9% vs. 13.4%), catecholamine therapy (14.1% vs. 6.8%), lower extremity injuries (72.4% vs. 53.5%), unstable pelvic fractures (41.0% vs. 25.9%), penetrating injuries (2.8% vs. 1.5%), and shock (20.5% vs. 10.8%) and unconsciousness (23.7% vs. 16.3%) on admission. Based on the propensity score, patients with penetrating trauma, pRBC transfusions, unstable pelvic fractures, and lower extremity injuries were more likely to undergo DCO. A benefit of DCO such as reduced complications or reduced mortality was not detected.

CONCLUSION

We could identify some parameters of polytrauma patients used in the trauma registry (Traumaregister DGU®), which led more likely to a DCO therapy. The propensity score did not demonstrate the superiority of DCO over ETC in terms of outcome or complications. It did not appear to adequately adjust for the variables used here. Definitive evidence for or against the use of DCO remains unavailable.

The VACStent trial: combined treatment of esophageal leaks by covered stent and endoscopic vacuum therapy

BACKGROUND

Endoscopic treatment of esophageal leaks, mostly by covered stents or endoscopic vacuum therapy (EVT), has largely improved the clinical outcome in the last decade. However, both techniques suffer from significant limitations. Covered stents are hampered by a high rate of migration and missing functional drainage, whereas endoluminal EVT devices are limited by obstruction of the GI tract. The new design of the VACStent makes it a fully covered stent within a polyurethane sponge cylinder, allowing EVT while stent passage is still open. Initial clinical applications have demonstrated the fundamental concept of the VACStent.

METHOD

A prospective multicenter open-label study was performed with the primary endpoint safe practicality, complete leak coverage, and effective suction-treatment of esophageal leaks. Secondary endpoints were prevention of septic conditions, successful leak healing, and complications, in particular stent-migration, local erosions and bleeding.

RESULTS

Fifteen patients with different, mostly postoperative anastomotic leaks were enrolled in three centers. A total of 41 VACStents were implanted. The mean number of VACStents per patient was 2.7, with a mean duration of VACStent treatment of 15 days. The primary endpoint was met in all VACStent applications (41/41 implants), resulting in a leak healing rate of 80% (12/15 patients). Septic episodes were prevented in 93% (14/15 patients) and there was no mortality. There were no severe device-related adverse events (SADE) nor significant local bleeding or erosion. Minor stent-dislocation and migration, respectively, was observed in 7%. Oral intake of liquids or food was documented in 87% (13/15 patients). One anastomotic stenosis was seen during follow-up.

CONCLUSIONS

VACStent treatment is a safe and effective treatment in esophageal leaks which can be covered by the sponge cylinder. Its application was described as easy and resembling that of conventional GI stents, with an impressive clinical success rate comparable to EVT outcomes. The VACStent offers a new option for clinical treatment of critical situations in esophageal perforations and anastomotic sutureline failures.

Endotracheal intubation in trauma patients with isolated shock: universally recommended but rarely performed

PURPOSE

The indication for pre-hospital endotracheal intubation (ETI) must be well considered as it is associated with several risks and complications. The current guidelines recommend, among other things, ETI in case of shock (systolic blood pressure < 90 mmHg). This study aims to investigate whether isolated hypotension without loss of consciousness is a useful criterion for ETI.

METHODS

The data of 37,369 patients taken from the TraumaRegister DGU® were evaluated in a retrospective study with regard to pre-hospital ETI and the underlying indications. Inclusion criteria were the presence of any relevant injuries (Abbreviated Injury Scale [AIS] ≥ 3) and complete pre-hospital management information.

RESULTS

In our cohort, 29.6% of the patients were intubated. The rate of pre-hospital ETI increased with the number of indications. If only one criterion according to current guidelines was present, ETI was often omitted. In 582 patients with shock as the only indication for pre-hospital ETI, only 114 patients (19.6%) were intubated. Comparing these subgroups, the intervention was associated with longer time on scene (25.3 min vs. 41.6 min; p < 0.001), higher rate of coagulopathy (31.8% vs. 17.2%), an increased mortality (8.2% vs. 11.5%) and higher standard mortality ratio (1.17 vs. 1.35). If another intubation criterion was present in addition to shock, intubation was performed more frequently.

CONCLUSION

Decision making for pre-hospital intubation in trauma patients is challenging in front of a variety of factors. Despite the presence of a guideline recommendation, ETI is not always executed. Patients presenting with shock as remaining indication and subsequent intubation showed a decreased outcome. Thus, isolated shock does not appear to be an appropriate indication for pre-hospital ETI, but clearly remains an important surrogate of trauma severity and the need for trauma team activation.

Does a prior hysterectomy complicate transvaginal/transumbilical hybrid NOTES cholecystectomy?-a comparative analysis of prospectively collected data

PURPOSE

Hysterectomy alters the anatomy of the posterior vaginal vault used as access for transvaginal/transumbilical hybrid NOTES cholecystectomy (NC), creating potential consequences for the feasibility and complication rate of the procedure. Therefore, the aim of our retrospective analysis of prospectively collected data was to analyze the postoperative course after NC in previously hysterectomized (PH) patients compared with patients who had not undergone hysterectomy (NH).

METHODS

A total of 126 NH patients and 50 PH patients aged over 42 who had an NC from 12/2008 to 04/2021 were compared regarding age, body mass index (BMI), ASA classification, number of percutaneous trocars, need for intraoperative urinary bladder catheterization, length of procedure, conversion rate, and intraoperative and postoperative complication rate according to the Clavien/Dindo classification, Comprehensive Complication Index (CCI), mortality, and hospital length of stay.

RESULTS

PH patients were older than NH patients (63.0 vs 51.5 years; P < 0.001) but did not differ significantly in ASA classification (P = 0.595) and BMI (26.8 vs 27.9 kg/m²; P = 0.480). They required more percutaneous trocars (P = 0.047) and longer procedure time (66.0 vs. 58.5 min; P = 0.039). Out of all 287 scheduled NC only one had to be "converted" to traditional laparoscopic cholecystectomy. Intraoperative and postoperative complication rates, Clavien/Dindo classification, CCI, need for intraoperative urinary bladder catheterization, and length of stay did not differ significantly.

CONCLUSION

Our results indicate an increased degree of difficulty of NC in PH patients, although there is no major impact on intraoperative and postoperative complication rates. Urinary bladder perforation is a specific access-related complication in PH patients.

Simultaneous Casualty Admissions-Do they Affect Treatment in the Receiving Trauma Center?

Background

Simultaneous trauma admissions expose medical professionals to increased workload. The impact of simultaneous trauma admissions on hospital allocation, therapy, and outcome is currently unclear. We hypothesized that multiple admission-scenarios impact the diagnostic pathway and outcome.

Methods

The TraumaRegister DGU® was utilized. Patients admitted between 2002–2015 with an ISS ≥ 9, treated with ATLS®- algorithms were included. Group ´IND´ included individual admissions, two individuals that were admitted within 60 min of each other were selected for group ´MULT´. Patients admitted within 10 min were considered as simultaneous (´SIM´) admissions. We compared patient and trauma characteristics, treatment, and outcomes between both groups.

Results

132,382 admissions were included, and 4,462/3.4% MULTiple admissions were found. The SIM-group contained 1,686/1.3% patients. The overall median injury severity score was 17 and a mean age of 48 years was found. MULT patients were more frequently admitted to level-one trauma centers (68%) than individual trauma admissions were (58%, p < 0.001). Mean time to CT-scanning (24 vs. 26/28 min) was longer in MULT / SIM patients compared to individual admissions. No differences in utilization of damage control principles were seen. Moreover, mortality rates did not differ between the groups (13.1% in regular admissions and 11.4%/10,6% in MULT/SIM patients).

Conclusion

This study demonstrates that simultaneous treatment of injured patients is rare. Individuals treated in parallel with other patients were more often admitted to level-one trauma centers compared with individual patients. Although diagnostics take longer, treatment principles and mortality are equal in individual admissions and simultaneously admitted patients. More studies are required to optimize health care under these conditions.

Does in-hospital trauma mortality in urban Indian academic centres differ between "office-hours" and "after-hours"?

INTRODUCTION

Trauma services within hospitals may vary considerably at different times across a 24 h period. The variable services may negatively affect the outcome of trauma victims. The current investigation aims to study the effect of arrival time of major trauma patients on mortality and morbidity.

METHOD

Retrospective analysis of the Australia-India Trauma Systems Collaboration (AITSC) registry established in four public university teaching centres in India Based on hospital arrival time, patients were grouped into "Office-hours" and "After-hours". Outcome parameters were compared between the above groups.

RESULTS

5536 (68.4%) patients presented "after-hours" (AO) and 2561 (31.6%) during "office-hours" (OH). The in-hospital mortality for "after-hours" and "office-hours" presentations were 12.1% and 11.6% respectively. On unadjusted analysis, there was no statistical difference in the odds of survival for OH versus AH presentations. (OR,1.05, 95% CI 0.9-1.2). Adjusting for potential prognostic factors (injury severity, presence of shock on arrival, referral status, sex, or extremes of age), there was no statistically significant odds of survival for OH versus AH presentations (OR,1.02, 95%CI 0.9-1.2).ICU length of stay and duration of mechanical ventilation was longer in the AH group.

CONCLUSION

The in-hospital mortality did not differ between trauma patients who arrived during "after-hours" compared to '"office-hours".

Do elderly trauma patients receive the required treatment? Epidemiology and outcome of geriatric trauma patients treated at different levels of trauma care

PURPOSE

In an ageing society, geriatric trauma displays an increasing challenge in trauma care. Due to comorbidities and reduced physiologic reserves, these patients might benefit from an immediate specialised care. The current study aims to clarify the prevalence and outcome of geriatric trauma depending on the level of the primary trauma centre.

METHODS

Data sets of 124,641 patients entered in the TR-DGU between 2009 and 2016 were included. Geriatric trauma was defined above 65 years and ISS ≥ 9. Analysing the prevalence, the age structure of all trauma cases registered in 2014 was compared to demographic data of the German Federal Statistical Office. Differences in injury pattern, in-hospital care and outcome between the primary levels of care were analysed.

RESULTS

In comparison to their share of population, geriatric patients are highly overrepresented in the TR-DGU. Despite minor injury mechanisms, severe head injuries are common. A tendency to under-triage can be observed, as level II and III trauma centres receive a higher percentage of older patients. Nevertheless, there is no effect on the mortality. 10% of these patients require an early transfer to a higher levelled trauma centres mainly due to severe head and spine injuries. Surprisingly, pre-clinical available signs such as GCS or blood pressure were not altered in these patients.

CONCLUSION

Patients above the age of 65 years represent a second group with high risk for traumatic injuries besides younger adults. Despite low-energy trauma mechanisms, these patients are prone to suffer from severe injuries, which require specialised care. Current admission practice appears adequate, as pre-clinical available symptoms did not correlate with injuries that demanded an early inter-hospital transfer. Specialised geriatric triage scores might further improve admission practice.

Evolution of high-energy pelvic trauma in southern Finland: a 12-year experience from a tertiary trauma centre

Purpose

We compared incidence, demographics, and injury mechanisms in severely injured patients with and without a pelvic ring fracture treated at a tertiary trauma centre. We also analyzed the changes in injury mechanisms that lead to high-energy pelvic trauma.

Methods

Data on severely injured adult patients (New Injury Severity Score [NISS] ≥16) from Helsinki Trauma Registry over the years 2006–2017 were reviewed. Patients with a pelvic ring fracture (PRF) and those without (N-PRF) were analyzed. Further subgrouping regarding time of the accident (2006–2009, 2010–2013, 2014–2017) was made. A comparison between groups was performed according to age, age > 60, gender, American Society of Anesthesiologists classification, injury scoring and mechanism, and 30-day in-hospital mortality.

Results

We included 545 PRF and 1048 N-PRF patients. Pelvic ring fracture patients were more likely to be female (39% vs 22%, p < 0.001), to be more severely injured (NISS 35.2 vs 30.4, p < 0.001), injured due to a high fall (41% vs 25%, p < 0.001), to have self-inflicted injuries (23% vs 8%, p < 0.001), and to have higher 30-day in-hospital mortality (13% vs 9%, p = 0.005). During the study period, we noted increasing mean age and proportion of patients aged > 60, improvement in outcome (shown by decreasing 30-day in-hospital mortality rate) in both groups, and a decrease in motor vehicle accidents (MVAs) leading to pelvic trauma (30–16%).

Conclusions

High-energy pelvic trauma can no longer be characterized as traffic accident injuries among young men. MVAs leading to pelvic trauma are decreasing and the most common injury mechanism is high fall. The patients are older and often female. Every fourth high-energy pelvic trauma was due to attempted suicide.

Changes in anaesthetic use for trauma patients in German HEMS - a retrospective study over a ten-year period

Background

Airway management and use of intravenous anaesthetics to facilitate tracheal intubation after major trauma remains controversial. Numerous agents are available and used for pre-hospital rapid-sequence induction (RSI). The aim was to investigate usage and potential changes in administration of intravenous anaesthetics for pre-hospital RSI in trauma patients over a ten-year period.

Methods

Based on a large helicopter emergency medical service (HEMS) database in Germany between 2006 and 2015, a total of 9720 HEMS missions after major trauma leading to RSI on scene were analysed. Administration practice of sedatives and opioids were investigated, while neuromuscular blocking agents were not documented in the database.

Results

With respect to administration of sedatives, independent from trauma mechanism and specific injury patterns the use of Etomidate decreased dramatically (52 to 6%) in favour of a more frequent use of Propofol (3 to 32%) and Ketamine (9 to 24%; all p < 0.001) from 2006 to 2015. The use of Benzodiazepines increased slightly, while the utilization rate of Barbiturates remained constant. In patients with Shock Index > 1 at initial contact, the administration rate of Etomidate dropped significantly as well. This decline was mainly substituted by Ketamine and particularly Propofol. In patients with GCS ≤ 8 upon initial contact, a similar distribution compared to the general trauma population could be observed.

With respect to opioids, mainly Fentanyl has been administered for RSI in trauma patients (2006: 69,6% to 2015: 60.2%; p < 0.001), while the use of sufentanyl showed a significant increase (0.2 to 8.8%; p < 0.001).

Conclusions

This large study analysed prehospital administration of anaesthetics in trauma patients, showing a substantial change from 2006 to 2015 despite the lack of any high-level evidence. Etomidate has shifted from the main sedative substance to virtual absence, indicating that the recommendation of an established national guideline was transferred into clinical practice, although based on weak evidence as well. The pre-hospital use of Propofol showed a particular increase. Fentanyl has been the main opioid drug for RSI in trauma, however Sufentanyl has become increasingly popular. The mechanisms and advantages of the different substances still have to be elucidated, especially in head injury and bleeding trauma.

Surgical treatment strategies in pediatric trauma patients: ETC vs. DCO-an analysis of 316 pediatric trauma patients from the TraumaRegister DGU®

PURPOSE

External fixation within the damage control concept in unstable multiple trauma patients is widely accepted. Literature about its usage in the pediatric trauma population, however, is rare. The aim of the present study was to elucidate the factors associated with the application of external fixation in the severely injured child.

METHODS

Patients with severe trauma aged 0-54 years documented in the TraumaRegister DGU^® were included in this study. Demographic data, pattern of injury, injury severity, use of the damage control orthopedics (DCO) or early total care (ETC) concept, duration of mechanical ventilation, intensive care stay, and total hospital stay as well as the occurrence of complications and mortality were evaluated. Statistical evaluation was performed using SPSS (Version 21.0.0) using Chi square tests and linear regression models.

RESULTS

While injury severity was comparable between children and adults, type of accident and injury patterns showed significant differences, Overall, the majority of surgical fracture stabilization in AIS_Extremity ≥ 3 injuries followed the DCO concept in adults (60.3%) and the ETC protocol in children (49.4%). Conservative treatment was chosen for only 11.6% of all children and 9.6% of all adults. An increasing injury severity, AIS_Extremity ≥ 3 and AIS_Extremity ≥ 3 in ≥ 2 body regions, and a more advanced age were found to be independent factors in the use of the DCO concept in children.

CONCLUSION

Use of external fixation increases with age and plays a minor role in the very young trauma population. However, this does not produce a difference in outcome between children and adults.

Trauma-induced coagulopathy upon emergency room arrival: still a significant problem despite increased awareness and management?

PURPOSE

Over the last decade, the pivotal role of trauma-induced coagulopathy has been described and principal drivers have been identified. We hypothesized that the increased knowledge on coagulopathy of trauma would translate into a more cautious treatment, and therefore, into a reduced overall incidence rate of coagulopathy upon ER admission.

PATIENTS AND METHODS

Between 2002 and 2013, 61,212 trauma patients derived from the TraumaRegister DGU® had a full record of coagulation parameters and were assessed for the presence of coagulopathy. Coagulopathy was defined by a Quick's value < 70% and/or platelet counts < 100,000/µl upon ER admission. For each year, the incidence of coagulopathy, the amount of pre-hospital administered i.v.-fluids and transfusion requirements were assessed.

RESULTS

Coagulopathy upon ER admission was present in 24.5% of all trauma patients. Within the years 2002-2013, the annual incidence of coagulopathy decreased from 35 to 20%. Even in most severely injured patients (ISS > 50), the incidence of coagulopathy was reduced by 7%. Regardless of the injury severity, the amount of pre-hospital i.v.-fluids declined during the observed period by 51%. Simultaneously, morbidity and mortality of severely injured patients were on the decrease.

CONCLUSION

During the 12 years observed, a substantial decline of coagulopathy has been observed. This was paralleled by a significant decrease of i.v.-fluids administered in the pre-hospital treatment. The reduced presence of coagulopathy translated into decreased transfusion requirements and mortality. Nevertheless, especially in the most severely injured patients, posttraumatic coagulopathy remains a frequent and life-threatening syndrome.

Is the shock index based classification of hypovolemic shock applicable in multiple injured patients with severe traumatic brain injury?-an analysis of the TraumaRegister DGU®

BACKGROUND

A new classification of hypovolemic shock based on the shock index (SI) was proposed in 2013. This classification contains four classes of shock and shows good correlation with acidosis, blood product need and mortality. Since their applicability was questioned, the aim of this study was to verify the validity of the new classification in multiple injured patients with traumatic brain injury.

METHODS

Between 2002 and 2013, data from 40 888 patients from the TraumaRegister DGU^® were analysed. Patients were classified according to their initial SI at hospital admission (Class I: SI < 0.6, class II: SI ≥0.6 to <1.0, class III SI ≥1.0 to <1.4, class IV: SI ≥1.4). Patients with an additional severe TBI (AIS ≥ 3) were compared to patients without severe TBI.

RESULTS

16,760 multiple injured patients with TBI (AIS_head ≥3) were compared to 24,128 patients without severe TBI. With worsening of SI class, mortality rate increased from 20 to 53% in TBI patients. Worsening SI classes were associated with decreased haemoglobin, platelet counts and Quick's values. The number of blood units transfused correlated with worsening of SI. Massive transfusion rates increased from 3% in class I to 46% in class IV. The accuracy for predicting transfusion requirements did not differ between TBI and Non TBI patients.

DISCUSSION

The use of the SI based classification enables a quick assessment of patients in hypovolemic shock based on universally available parameters. Although the pathophysiology in TBI and Non TBI patients and early treatment methods such as the use of vasopressors differ, both groups showed an identical probability of recieving blood products within the respective SI class.

CONCLUSION

Regardless of the presence of TBI, the classification of hypovolemic shock based on the SI enables a fast and reliable assessment of hypovolemic shock in the emergency department. Therefore, the presented study supports the SI as a feasible tool to assess patients at risk for blood product transfusions, even in the presence of severe TBI.

[Potential for the survey of quality indicators based on a national emergency department registry : A systematic literature search]

BACKGROUND

Emergency department processes are often key for successful treatment. Therefore, collection of quality indicators is demanded. A basis for the collection is systematic, electronic documentation. The development of paper-based documentation into an electronic and interoperable national emergency registry is-besides the establishment of quality management for emergency departments-a target of the AKTIN project. The objective of this research is identification of internationally applied quality indicators.

METHODS

For the investigation of the current status of quality management in emergency departments based on quality indicators, a systematic literature search of the database PubMed, the Cochrane Library and the internet was performed.

RESULTS

Of the 170 internationally applied quality indicators, 25 with at least two references are identified. A total of 10 quality indicators are ascertainable by the data set. An enlargement of the data set will enable the collection of seven further quality indicators. The implementation of data of care behind the emergency processes will provide eight additional quality indicators.

CONCLUSION

This work was able to show that the potential of a national emergency registry for the establishment of quality indicators corresponds with the international systems taken into consideration and could provide a comparable collection of quality indicators.

[Using data from registries like the TraumaRegister DGU(®) for effectiveness evaluations]

In some situations like a rare disease or an emergency intervention, a randomized controlled trial is rather difficult or even impossible to conduct. In these cases, the use of observational studies or registries for the assessment of effectiveness is discussed. Results from registries, however, are subject to the same methodological limitations as any non-randomized comparison. Specifically, the comparability of selected patient groups is not ensured. Using three examples from the TraumaRegister DGU(®), a nationwide registry for severely injured patients in Germany, the advantages as well as the risks of outcome evaluation with registry data will be discussed: 1) whole-body computed tomography in the early in-hospital phase, 2) pre-hospital volume resuscitation, and 3) surgical treatment according to the principle of damage control. The most important prerequisite for any successful outcome evaluation using registry data is the documentation and availability of so-called confounding variables (prognostic factors), the absence of an accepted standard treatment, i.e., both interventions are routinely applied in similar situations, and, finally, the use of sophisticated statistical methods for multivariate adjustment of results.

Change of initial and ICU treatment over time in trauma patients. An analysis from the TraumaRegister DGU®

BACKGROUND

Clinical guidelines have been standardized for pre- and in-hospital trauma management in the last decades. Therefore, it is known that prehospital management has changed significantly. Furthermore, in-hospital course may be altered to reduce complications and length of stay (LOS). However, the development of trauma patient in-hospital management as well as LOS in the intensive care unit (ICU) has not been investigated systematically over a long-term period in Germany. Aim of our study is to examine the changes in in-hospital management and LOS in the ICU in moderately and severely injured patients.

METHODS

Patients documented in the TraumaRegister DGU® (TR-DGU) of the German Trauma Society from 2000 to 2011 and admitted to ICU were included in this study. Demographic data, the pattern of injury, injury severity, duration of mechanical ventilation, LOS in the ICU, hospital LOS, and discharge destination were evaluated. The mean values and the standard deviations are shown. The constant variables were calculated with changes over time analyzed by linear regression analysis, and categorical variables were calculated with the chi-square test.

RESULTS

A total of 18,048 patients were analyzed. The rate of patients being intubated at the time of ICU admission decreased from 86.8 % in 2000 to 60.0 % in 2011 (p < 0.001). The time of mechanical ventilation decreased from 7.5 ± 10.5 to 4.7 ± 8.7 days. The intensive care unit LOS was reduced from 11.7 ± 12.8 to 9.0 ± 11.3 days and the length of hospital stay from 27.9 ± 28.7 to 21.1 ± 20.4 days (both p < 0.01). The ICU LOS remained stable in the subgroup of mechanically ventilated patients (12.7 ± 13.2 day in 2000, 12,6 ± 12.9 in 2011, p = 0.6), whereas it was reduced in non-mechanically ventilated patients (5.5 ± 6.8 days in 2000, 3.6 ± 4.5 days in 2011; p < 0.001).

CONCLUSIONS

The reduction LOS in the analyzed dataset is mainly explained by the relevantly reduced rate of patients being intubated at the time of ICU admission. Our data demonstrate that trauma patients' in-hospital course is influenced by reduced intubation rate at the time of ICU admission.

Association between volume of severely injured patients and mortality in German trauma hospitals

BACKGROUND

The issue of patient volume related to trauma outcomes is still under debate. This study aimed to investigate the relationship between number of severely injured patients treated and mortality in German trauma hospitals.

METHODS

This was a retrospective analysis of the TraumaRegister DGU® (2009-2013). The inclusion criteria were patients in Germany with a severe trauma injury (defined as Injury Severity Score (ISS) of at least 16), and with data available for calculation of Revised Injury Severity Classification (RISC) II score. Patients transferred early were excluded. Outcome analysis (observed versus expected mortality obtained by RISC-II score) was performed by logistic regression.

RESULTS

A total of 39,289 patients were included. Mean(s.d.) age was 49.9(21.8) years, 27,824 (71.3 per cent) were male, mean(s.d.) ISS was 27.2(11.6) and 10,826 (29.2 per cent) had a Glasgow Coma Scale score below 8. Of 587 hospitals, 98 were level I, 235 level II and 254 level III trauma centres. There was no significant difference between observed and expected mortality in volume subgroups with 40-59, 60-79 or 80-99 patients treated per year. In the subgroups with 1-19 and 20-39 patients per year, the observed mortality was significantly greater than the predicted mortality (P < 0.050). High-volume hospitals had an absolute difference between observed and predicted mortality, suggesting a survival benefit of about 1 per cent compared with low-volume hospitals. Adjusted logistic regression analysis (including hospital level) identified patient volume as an independent positive predictor of survival (odds ratio 1.001 per patient per year; P = 0.038).

CONCLUSION

The hospital volume of severely injured patients was identified as an independent predictor of survival. A clear cut-off value for volume could not be established, but at least 40 patients per year per hospital appeared beneficial for survival.