Outcome after polytrauma in a certified trauma network: comparing standard vs. maximum care facilities concept of the study and study protocol (POLYQUALY)

The German Trauma Society (Deutsche Gesellschaft für Unfallchirurgie, DGU) started a nationwide multicentre study based on voluntary participation of trauma centres in 1993. This registry data collection currently contains some 160.000 cases with over 130 single items per patient until December 2013. The aim of this trauma registry was to analyse the performance of individual trauma centres, including the pre-hospital management, the clinical care setting, and outcome data. The main results of the Trauma Register DGU® [4] were published and served as the empirical and conceptual basis for the “White Book of the Severely Injured Patient” (2nd ed, 2012, [5]).

The White Book delineated the creation of a nationwide system of regional trauma networks. Similar to the American trauma system, a ranking of trauma facilities was introduced where all participating hospitals were classified into basic, (“Lokales Traumazentrum”, comparable to US level III), standard (“Regionales Traumazentrum”, comparable to US level II), and maximum (“Überregionales Traumazentrum”, comparable to US level I) care facilities. Strict rules for patient transfer were introduced with the aim to "get the right patient at the right time to the right hospital". According to the White Book the maximum transportation time for a major trauma patient from the scene should not exceed 30 minutes. The patient has to be transported to the next available standard or maximum care facility. If this time interval cannot be adhered to (either by ground or air transportation), the nearest basic trauma facility has to initially stabilize the patient and then transfer him/her as soon as possible to the next higher level of care institution. All hospitals within a region were audited and the network was not certified until an agreement on cooperation and uniform patient documentation was signed and the German S3 guideline on polytrauma care [6] has been implemented.

The effects of this major innovation in the German health care system cannot be predicted. Quite generally, it remains unclear whether trauma networks indeed improve the quality of care. More recently, trauma networks or trauma systems have been established in a number of different countries. Thus far, the reports on the results on an international level have been controversial. The Major Trauma Outcome Study in the USA, a trauma registry with more than 150.000 patients [7] has demonstrated the superiority of treatment of multiple injured patients in dedicated trauma centres [8]. In a Dutch study [9], the introduction of a regional trauma system reduced mortality and changed admission rates for all types of trauma. Other reports from the United Kingdom and from Switzerland could not show any improvement for mortality [10, 11]. The authors reached different conclusions in their respective health care settings due to the difference in the health care systems.

Based on these various philosophies the design of trauma centres differs greatly between countries with respect to size and catchment area (e.g. North America: centralization into a few maximum care facilities, Germany: decentralized health care system with many small to medium-sized facilities). Therefore it is difficult to apply the results of foreign studies to the German situation. Improving the quality of care also includes a profound understanding of care processes [12], which is especially true for the treatment of polytrauma patients in the acute period, due to the critical timeframe and the multidisciplinary co-operation required [13]. The “blueprint” is a quality management tool derived from the service industry, which structures and visualizes complex, customer-centred service processes and helps identify critical points (“moments of truth”) along the service process and within the interaction of different stakeholders [14, 15]. The blueprint method has been successfully transferred to patient-centred health care processes by some authors, e.g. with regard to mammography performance or outpatient care of hypertension [16, 17]. The method has not yet been applied to the acute care situation.

The aim of the present project is to evaluate the performance of the first certified regional trauma network in Germany, the Trauma Network Eastern Bavaria (TNO). In order to assess quality of care for polytrauma (POLYQUALY) the following specific research questions will be addressed:

We hypothesize that the quality of care and therefore the outcome will be equal in hospitals with standard care in comparison to hospitals with maximum care facilities in a region if all hospitals are organized into a network structure that allows mutual support including the use of latest technologies such as telemedicine consultation. Furthermore, we hypothesize that TNO outperforms German Trauma Register DGU^® 2008 reference data and that the quality of care in the standard and maximum care facilities of TNO complies with guideline-derived benchmarks. In case our hypotheses hold true the qualitative barriers and facilitators analysis will be able to explain what the driving forces were that made this favorable results possible. In case the study results contradict our hypotheses the barriers and facilitators analysis can explain the hindering forces and may show ways how to improve the health care situation in the region.

The present project will also consider gender aspects. The majority of trauma patients (about 70 %) are male. In this study males and females will be enrolled according to the actual incidence of trauma events. After major trauma, quality of life differs between genders. Women have a 2.5 fold greater risk than men for developing psychological problems following major trauma and intensive care treatment (post-traumatic stress disorder, PTSD) [18, 19]. In order to address these gender-specific differences, quality of life outcomes will be compared between males and females.

The study design includes a prospective quantitative analysis and a related explorative qualitative analysis (see Fig. 1).

The prospective multicentre study will (a) compare two cohorts of major trauma patients, those treated in maximum care facilities vs. those treated in standard care facilities, (b) compare results of the two cohorts with historical controls derived from the German Trauma Register DGU^® 2008 reference data and (c) compare results of the two cohorts with benchmarks including 5 outcome standards derived from the S3 guideline: risk adjusted morality (<15 %), door to CT-time (<20 min), ventilator free days (>22 days), ICU free days (>20 days), GOS (>80 % excellent and good).

Patients will be recruited within the 25 hospitals that comprise TNO (Table 1). Inclusion criteria for the prospective cohort study are: major trauma ISS ≥ 16 (Injury Severity Score), treatment in any of the care facilities within the trauma network TNO, no age limitations. Exclusion criteria are ISS < 16, care outside TNO. Based on current register data it is expected that n = 1.100 patients will be assessed for eligibility and n = 800 be included and analyzed. For the qualitative analysis of barriers and facilitators, focus groups will be conducted with representatives of different professional groups involved in the TNO including emergency physicians, paramedics, dispatch centre staff, emergency room staff, trauma surgeons, rehabilitation staff and social workers. The focus group methodology can be transferred to other trauma networks and can be spread throughout Germany.

Clinical information will be collected using the TraumaRegister^QM form. This quality management (QM) form underwent a rigorous development process and has been issued by the German Trauma Society. TraumaRegister^QM is a one-page, 40-item short form that includes among other things the following contents: date and site of referral, type of trauma, pre-hospital management, Glasgow Coma Scale (GCS), in-hospital management, ICU management and outcome at discharge (Glasgow Outcome Scale, GOS). A number of validated scores can be delineated from this documentation, such as AIS, ISS, TRISS, RISC, [20] for evaluation of adjusted outcome. TraumaRegister^QM will be completed by the participating hospitals via a web-based system. Data will be transferred to AUC (Akademie der Unfallchirurgie) and from there to the Centre for Clinical Studies Regensburg.

Another data source which will be used as a historical control is the German Trauma Register DGU^® reference data base. Quality of life will be assessed using the EQ-5D form which is a short form widely used in large scale cohort studies. This instrument underwent rigorous psychometric testing, is part of the trauma-specific QL measurement system POLO-Chart [21] and generates indices that can be used for further socio-economic analyses. A German version and population-based German reference data are available. EQ-5D assessment will be performed 6, 12 and 24 months after trauma (Fig. 2). Assignment to the respective clinical information documented in TraumaRegister^QM will be accomplished via an index number.

In the course of the qualitative part of this project the semi-standardized interviews and focus groups will be conducted face-to-face, audio-taped and transcribed verbatim. For the analysis of barriers and facilitators, the developed blueprint will be used as a basis for focus group discussions.

The use of TraumaRegister^QM is mandatory for participating hospitals via the TNO contract and hospital staff will be trained to complete the form via a web-based system. Non-compliant hospitals are easily detectable. Assistance with data collection and checks on plausibility will be provided by a data manager hired for this study and affiliated with the Centre for Clinical Studies at the University Hospital Regensburg. Thus, a near complete (>95 %) documentation of all major trauma patients with a minimum of missing values for the two-year recruitment period can be expected. The quality management of the qualitative data collection and analysis will be performed according to the standards set by [22]. Among others, it is planned to utilize “respondent validation” by presenting the results and discussing them with interviewees and focus group members in interactive workshops.

Prospective cohort study: Statistical analyses of the prospective data will primarily dwell on descriptive statistics using counts, percentages, means/standard deviations, medians/interquartiles/ranges and confidence intervals. In order to interpret descriptive results they will be compared with secondary data from the Trauma Register DGU^® reference data bank 2008 and the benchmarks derived from the S3 guideline on major trauma. A logistic regression will be performed with 30 day mortality as the dependent variable and the ISS and TRISS scores as covariates. A p value of < 0.05 will be considered statistically significant. All other statistical tests will be chosen according to the scale level of the variable and the number of variables used in the model (t-test, Mann–Whitney U-Test, chi-squared test, risk-adjusted multivariable regression models) and will be performed in an explanatory manner, leaving alpha at < 0.05.

Qualitative study: Interview and focus group transcripts will be critically examined using thematic content analysis [23]. Themes will be identified using a grounded conceptualisation process [24]. Transcripts will be repeatedly read before and after coding to ensure proper categorisation of data. For the blueprint, the information on the health care process will be visualized as a two-dimensional picture: The horizontal axis represents the chronology of actions by the patient and the health care professional. The vertical axis distinguishes between different areas of actions, i.e. direct interactions between patient and professional, professional actions visible and invisible to the patient, support processes, and planning, managing and controlling activities [15, 17].

Only authorized persons in the participating hospitals will have access to the data base, identifying themselves with user codes and passwords. Data transfer from peripheral data capture institutions to the data analysis centre complies with security and safety standards analogue to SSL coding. Data of the prospective study will be anonymised. Nevertheless, the combination of a set of variables documented in each TraumaRegister^QM form (data of referral and reception, referring and receiving hospital, age, gender, diagnosis) will allow for matching of two or more forms in case a patient is being referred from one hospital to another. TraumaRegister^QM data will be matched with respective EQ-5D forms using a unique index number. The completeness and plausibility of forms and documentation will be cross-checked by the data manager.

Due to the observational nature of the prospective study no additional invasive diagnostic or therapeutic regimens are required. Data will be collected anonymously. The documentation and use of data generated with the TraumaRegister^QM form has been contracted with the German Trauma Society. Patients will be asked to give informed consent to take part in EQ-5D outcome evaluation.

The POLYQUALY study has been approved by the Ethics Committee of the University of Regensburg (reference number 10-101-0077), is funded by a grant from the German Federal Ministry of Education and Research (reference number 01GY1153), and is registered in the data base of the German Network of Health Services Research (www.​versorgungsforsc​hung-deutschland.​de/​) reference number VfD_Polyqualy_12_001978, date 10.Jan.2013, and in the German Clinical Trials Register DRKS00010039, 18.Feb.2016.