This retrospective study was conducted in three steps. First, we developed the simplified scale (sAIS) using the 100 most frequent injuries collected in the Northern French Alps Trauma Registry (TRENAU). Second, for the internal validation, we examined the sAIS classification performance by randomly selecting 100 injured patients in the TRENAU Registry, which were rated by 10 French physicians. Third, we externally validated the sAIS by randomly selecting 100 injured patients included from a different dataset, the Trauma Registry of Acute Care (TRAC) of Lausanne University Hospital (Lausanne, Switzerland), which were rated by eight Swiss physicians and two research nurses. The study was conducted and reported in accordance with the Standards for Reporting Diagnostic Accuracy Studies (STARD) 2015 guidelines [6, 7].

Two trauma registries from 14 trauma centres were used to randomly select study participants. We used data collected by the French TRENAU Registry between 1 January, 2013 and 31 December, 2014 [8] for the development and internal validation of the sAIS. The Registry includes two level I, one level II, and 10 level III trauma centres in an inclusive trauma system. External validation was completed using data from the TRAC collected from 1 June, 2019 to 1 June 1, 2021. The TRAC includes one level I trauma centre (Lausanne University Hospital) in an exclusive trauma system of the state of Vaud (Switzerland), regrouping seven general hospitals and one university hospital [9].

The two registries collected data following the Utstein template for the uniform reporting of data following major trauma [10]. The certified coder scoring the reference AIS (rAIS) and reference ISS (rISS) in the TRAC had six years of coding experience with 5,028 cases rated throughout her career. Another certified coder scored the rAIS in the TRENAU Registry. The AIS 2008 classification was used until 31 December, 2019 and the AIS 2015 since 1 January, 2020. Inclusion criteria were any suspected major trauma based on physiological, anatomical and anamnestic criteria. Exclusion criteria were patients with isolated burns (including electric injury), out-of-hospital traumatic cardiac arrest, asphyxia or hanging without other injuries, and drowning. The following data were extracted: rAIS; rISS; age; gender; type of trauma; mechanism of injury; heart rate; systolic blood pressure; Glasgow Coma Scale; and survival status at hospital discharge (alive or dead). Coders included in the study to score the sAIS and calculate the sISS were randomly selected among all physicians involved in trauma care in the emergency department (ED) or intensive care unit (ICU) in one trauma centre of each trauma system. Clinicians did not receive any previous training or certification in AIS coding.

We extracted the 100 traumatic injuries most frequently reported in the TRENAU Registry between 2013 and 2014 (Additional File 1), which represent 90% (in proportion of reporting) of all AIS diagnoses described in the registry. We classified the 100 diagnoses into six anatomical regions (head and neck, face, chest, abdomen and pelvis, extremities, external) and by severity from 1 (minor injury) to 6 (maximal injury) to develop the sAIS (Table 1). We checked if all organs and all types of injury (skeletal, vascular, neurological, internal organs) were represented in the classification. We ensured that every category of severity was represented for each organ. If not, we added a generic injury in the chart for the missing organ or missing type of injury (e.g., retina detachment) in order to cover all possible diagnoses. We grouped diagnostics in generic categories by severity to reduce the number of items of the condensed chart. The sAIS was designed to be used by non-trained healthcare professionals.

We internally validated the sAIS using data from the TRENAU Registry from 1 January, 2013 to 31 December, 2014. Eight physicians from the ED and two from the ICU of a level 1 trauma centre (Annecy-Genevois Hospital) were randomly chosen among the ED (n = 29) and ICU (n = 16) teams, without any previous experience in AIS coding. They were asked to independently calculate the simplified ISS (sISS) of 10 cases each using the sAIS (Table 1), reported by body region using a data collection sheet (Additionnal File 2), and blinded to the rISS reported in the trauma registry. The 100 patients were selected by stratified randomization according to the ISS severity. The physicians used the ED medical records and radiological reports (radiography, computed tomography, magnetic resonance imaging, ultrasound) to rate their 10 cases, if available. No cases were rated by more than one physician.

We externally validated the sAIS using patient cases from the TRAC Registry. A similar process was used as previously detailed for the internal validation. Six registrar physicians, two senior consultants and two clinical research nurses from the ED were chosen among the team (n = 39) by randomization to calculate the sISS of 10 cases per participant, i.e., 100 patients in total.

The reference diagnostic test for major trauma identification was the rISS calculated using the rAIS and scored by a specifically trained and accredited coder. The index diagnostic test under evaluation was the sISS, calculated using the sAIS and scored by non-trained healthcare professionals.

The primary outcome was the accuracy of the sISS calculated using the sAIS compared with the rISS calculated using the rAIS ©2008 and ©2015.

We present continuous data as means and standard deviation (SD) when normally distributed or medians and interquartile range (IQRs) when not normally distributed. We report categorical data as numbers and percentages. We used Student's t-test to compare continuous and normally distributed data and the Mann–Whitney test for continuous and non-normally distributed data. We defined a two-tailed p-value of < 0.05 as statistically significant.

First, we assessed the difference between the two methods at the whole trauma population level. We estimated the mean difference between the rISS and the sISS, which represents a measure of the accuracy. We considered a clinically relevant limit of equivalence of ± 4 ISS points for the bias. For a SD of the ISS of 9.5 and a limit of equivalence of 4 ISS points, 97 patients were required to ensure a power of 80% with a significance level of 5%. We estimated the Pearson correlation coefficient (r) as another measure of accuracy. The relationship between the sISS and the rISS was described by using scatterplots and local polynomial regression in a calibration plot.

Second, as a measure of the precision at an individual patient level and to examine the agreement between the sISS and the rISS, we used the Bland–Altman method to plot the bias and the limits of agreement (LoA). Assuming a normal distribution, the LoA represent the mean of the difference ± 2 SD of the difference. We considered a relevant LoA range of ± 9 ISS points. We used two different limits of ISS variation. For the precision at an individual level, we used a LoA range of ± 9 ISS points, corresponding to an increase in the severity of an injury from an AIS 4 to 5, as described by Ringdal et al. [11]. For the accuracy at the population level, we chose a narrower limit of equivalence of ± 4 ISS points as clinically relevant. In addition, as the ISS is used to classify major trauma (ISS ≥ 16), we assessed the agreement of major trauma classification by using the Cohen’s kappa statistic between the two methods [12]. We performed a complete case analysis as no missing values were reported.

As we suspected an imperfect gold standard bias, an independent trained coder reviewed the rISS of the patient cases of the external validation dataset when the difference between rISS and sISS was outside the calculated LoA limit. We performed a sensitivity analysis of the sISS compared with the corrected rISS. Analyses were performed with Stata version 16 (Stata Corporation, College Station, TX, USA).