Scientific Papers

Effect of urban vs. remote settings on prehospital time and mortality in trauma patients in Norway: a national population-based study | Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine

Design and settings

We performed a register-based study of trauma cases included in the NTR between 1 and 2015 and 31 December 2020. Norway has a population of 5.5 million people, with approximately 43% of the population living in urban areas, 43% in suburban areas and 14% living in remote areas [10, 11]. It is a high-income country with a publicly funded healthcare system. A national trauma plan has been developed and implemented, and includes all stages of the chain of survival, from accident site to rehabilitation [12, 13]. Furthermore, the Norwegian health and hospital plan has been developed to ensure a coherent system of emergency services in and outside hospitals throughout the country [14]. Thirty-four trauma units (TU) and four major trauma centers (MTC) receive and treat trauma patients, and report data to the NTR. All TU and MTC have 24/7 trauma team availability led by an advanced trauma life support-educated experienced resident or a surgical consultant. Calls made to the national medical emergency number (113) are evaluated by specially trained emergency medical communication center (EMCC) personnel using the ‘Norwegian Index for Medical Emergencies’ (Index), a criteria-based dispatch system for prehospital resources [15].

Data sources and study cohort

The NTR is a national clinical quality registry containing information about injured patients in Norway from accident to rehabilitation (according to the Utstein template [16]). The NTR received formal status as a national medical quality register in 2006 [17]. All patients are registered with a waiver of consent. Injuries are coded according to the Abbreviated Injury Scale (AIS) manual (2005 version, updated in 2008 [18]) by certified nurse registrars. The NTR holds information about patients who meet the following inclusion criteria: admitted through trauma team activation (TTA), admitted without TTA but found to have penetrating injuries to head, neck, torso or extremities proximal to knee or elbow, head injury with AIS ≥ 3 or New Injury Severity Score (NISS) > 12, or patients who die at the scene of injury or during transport [17]. We excluded patients with injuries from drowning, inhalation, hypothermia and asphyxia without concomitant trauma, patients who presented to hospital via private vehicle, police vehicle or other/unknown, patients who were not registered with the EMCC, patients missing centrality index score and patients with prehospital time intervals we considered to be outliers (response time > 120 min, on-scene time < 5 min or > 120 min, or transport time < 5 min or > 360 min). Multiple registrations on the same patient (i.e., transfers) were counted only once. We chose to include patients with NISS = 0 for the descriptive statistics analysis as we wanted to investigate the trauma system and the prehospital phase where the patients were believed to be seriously injured. In the regression model we have excluded patients with NISS = 0.

Data collection and management

Data collected from the NTR include time points from which we have calculated the prehospital time intervals for further analyses, illustrated in Fig. 1. Time cut-offs were applied to all prehospital time intervals (see exclusion criteria above).

Fig. 1
figure 1

Time points and time intervals

Other variables collected from the NTR include transport type (ground ambulance, rotor-wing, fixed-wing), mortality (measured by 30-day mortality) and accident municipality for urban–remote classification. Rotor- and fixed-wing transport modes were merged into ‘air ambulance’, of which rotor wing constituted 97%. Patient characteristics included age, gender, injury mechanism, dominant injury, NISS, prehospital advanced airway management, whether or not the patient was trapped on the accident site, whether the patient was transported to a TU or MTC, and prehospital treatment level among the prehospital crew. The latter variable is only an indicator of the crews’ qualifications, and we do not know the amount of experience among prehospital crew. The injury mechanism variable was re-categorized from the original NTR definitions, where four traffic-related injuries (motor vehicle, motorcycle, pedestrian and other) were merged into ‘transport-related’, and shot by firearm, stabbed by sharp object, explosion injury and other were merged into ‘other’. Patients without a Norwegian ID number were registered as ‘missing age’. The variable coverage was high (see appendix).

Measure of centrality: The centrality index of Norway

Statistics Norway’s centrality index (CI) provides a measure of the municipality’s centrality based on criteria such as travel time to workplaces and service functions [11]. Municipalities are categorized into six groups, where the proportion of inhabitants in each group is an important criterion for the classification [11]. Furthermore, the six CI groups are merged into urban (CI 1 and 2), suburban (CI 3 and 4) and remote (CI 5 and 6) areas. In 2020, a national municipality structure reform was accomplished, but we have used the original municipal division in this study.

Statistical analysis

Registry data were analyzed using descriptive statistical methods including number, frequency (percentage) and median. Data were tested for normality with Kolmogorov–Smirnov tests. Differences in prehospital time intervals between the three centrality index groups were analyzed using Mann–Whitney U tests. Effect size was calculated to Mann–Whitney U with Cohen’s classification of effect sizes, where < 0.3 = small effect, between 0.3 and 0.5 = moderate effect and > 0.5 = large effect. A simple logistic regression model was performed to assess the effect of centrality index groups on mortality. Further, a forward stepwise logistic regression modeling strategy was applied to investigate the effects of prehospital time and centrality index groups on 30-day mortality, where we adjusted for control variables we believed would affect the results, including NISS, age, gender, injury mechanism and prehospital treatment level among the prehospital crew. All independent variables were tested for multicollinearity. Nagelkerke R2 was used to evaluate model improvement. A p-value of < 0.05 was considered to be statistically significant. All analyses were performed using SPSS v. 27.0 (IBM Company, Chicago, IL, USA).

Source link