Spinal cord injury in the context of major motor vehicle collision trauma: a retrospective ecological analysis of global estimates across income groups

Abstract

Introduction

Road traffic injuries (RTIs) are a leading cause of death globally, especially in low- (LICs) and middle-income countries (LMICs). Despite this burden, post-crash care remains underdeveloped. Many clinical principles of post-crash care focus on spinal cord injury (SCI), yet its incidence is poorly understood. The aim of this study was to describe the incidence of death and non-fatal RTI with a specific focus on SCIs using Global Burden of Disease (GBD) 2019 data.

Methods

A retrospective ecological analysis was conducted using GBD 2019 data for 204 countries and territories (2012–2019). We examined the MVC-related mortality, SCIs, and major non-fatal injuries stratified by income group and sex. Incidence rate ratios (IRRs) were compared across income groups using regression models. SCIs were analysed as a proportion of all non-fatal injuries across income groups.

Results

MVC-related mortality incidence rates were significantly lower in high-income (HICs) [IRR 0.79 (95% CI: 0.69–0.91); p = 0.001], UMICs [IRR 0.72 (95% CI 0.63–0.82); p < 0.001], and LMICs [IRR 0.67 (95% CI: 0.59–0.76); p < 0.001] compared with LICs. SCIs accounted for ~ 0.5% of non-fatal injuries, with a significantly higher proportion in LICs (p < 0.001). Males consistently showed higher injury and mortality incidence rates than females.

Conclusions

SCIs from MVCs are relatively rare, but disproportionately affect LICs. Strengthening bystander response, developing context-specific post-crash protocol, and improving prehospital systems may reduce disparities and improve outcomes, especially in resource-limited settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13049-025-01536-7.

Introduction

Road traffic injury (RTI) is the leading cause of death in children and young adults aged 5–29 years, with 93% of these fatalities occurring in low- and middle-income countries (L/LMICs) [1, 2]. In addition to the 1.2 million road deaths per year globally, an additional 20–50 million people incur significant injury and often long-term disability from road traffic injury [1]. Road traffic trauma contributes to individual and societal economic hardship costing L/LMICs approximately 5% of their gross domestic product (GDP) [2, 3]. The interaction between poverty, injury, economic status and recovery has been well reported; with an urgent need for intervention recognised by the United Nations, the World Health Organisation (WHO) and non-governmental organisations (NGOs) [1].

The ‘safe system’ describes five pillars for action to reduce road trauma deaths [1], and includes post-crash care as a core component. However, while prevention measures such as safer roads and vehicles have advanced in many settings, the post-crash phase has received comparatively less development. Many trauma deaths remain potentially avoidable with timely and effective intervention after the collision. A recent systematic review identified that between 4.9–11.3% of trauma deaths were definitely preventable and between 25.8–42.7% potentially preventable [4]. Post-crash care, the fifth pillar, is therefore a critical but often under-developed component of road safety. Barriers to improvement globally include variation in post-crash response and challenges in investment, training, resource availability and essential infrastructure.

The road injury chain of survival outlines a sequence of key steps that can improve survival following a collision when delivered promptly and effectively [5]. The earliest stages (early recognition, bystander assistance and activation of emergency services) are particularly important, yet often lack consistent or accessible guidance [6]. Many lay bystanders and professional responders (including prehospital emergency medical, fire and rescue, and police services) report uncertainty about how to manage potential spinal injury at this stage, partly due to limited or conflicting recommendations [7]. These concerns may discourage early intervention and assessment, delaying extrication and access to time-critical care.

Recent work from the United Kingdom (UK) Trauma Audit Research Network (TARN) registry demonstrated that spinal cord injury is relatively uncommon among patients trapped following a motor vehicle collision (MVC), occurring in approximately 0.7% of such cases, whereas head, chest and abdominal injuries are much more frequent [8–10]. Building on these findings, subsequent UK biomechanical studies, qualitative exploration of patient preferences and consensus work have informed national guidance recommending that most patients should be encouraged to self-extricate or be gently assisted from their vehicle, with an emphasis on minimal assistance and gentle patient handling rather than prolonged immobilisation [11]. This shift aims to reduce delays to time-critical care and to enable bystanders and first responders to provide earlier support.

However, whether this approach is applicable in other settings is uncertain. The risk and pattern of injury following MVCs may differ across regions due to variation in vehicle design, restraint systems, road infrastructure and regulatory standards. Equivalent trauma registry data are not available in many countries, making it difficult to determine whether the low incidence of spinal cord injury observed in the UK is consistent elsewhere. Large-scale global datasets offer an opportunity to describe the international burden of MVC-related spinal cord injury and to assess whether injury patterns observed in the UK are reflected across different income and regional contexts. UK findings are used here as contextual evidence to illustrate the clinical relevance of proportional injury patterns, rather than as a formal comparator.

The aim of this study was to describe the global incidence of death and non-fatal injury following MVCs, with particular focus on spinal cord injuries, to help determine whether injury patterns observed in the UK are reflected across different income and regional contexts.

Methods

Data were sourced from the Global Burden of Disease, Injuries and Risk Factors (GBD) 2019 study. This study was conducted as a retrospective ecological analysis, using countries or income groups as the unit of observation, drawing on routinely modelled Global Burden of Disease estimates over time. This dataset contains estimated incidence, prevalence, mortality, and other metrics for 369 diseases and injuries for 204 countries and territories Within the GBD, mortality and morbidity estimates are based on a variety of data submitted directly from national registration systems, facility administrative records and other data sources, as well as data extracted from literature reviews of published evidence for each disease, injury and risk factors included in the estimates*. No additional primary data were extracted from the literature for this study; all analyses were conducted using estimates directly obtained from the GBD 2019 dataset. Data from the GBD study is publicly available from the Institute of Health Metrics and Health Evaluation (IHME) Global Health Data Exchange (GHDx, http://ghdx.healthdata.org/) [12].

Spinal Cord Injury (SCI) incidence, case numbers, and rates per 100,000 population were extracted for all ages for each country and territory. The same data for MVC-related mortality and major non-fatal injuries were also extracted for comparison to SCIs. To support comparisons to previous UK work, the same window of years 2012–2019 were selected [8–10]. ‘Motor vehicle road injuries’ (ICD10 V30-V79.9, V87.2-V87.3) was selected as the cause of death or injury, for all ages, and all genders; this GBD category excludes pedestrian (V01-04.9, V06-V09.9), motorcycle (V20-V29.9), cycle (V10-V19.9) and other road injury (V80-V80.9, V82-V82.9) types. For non-fatal data, all spinal cord injuries (listed as ‘Spinal Injuries’) were extracted alongside other non-fatal injury types, excluding minor injuries. Non-fatal injury types were categorised according to Abbreviated Injury Score (AIS) regions for meaningful comparison to previous work [8–10]. Data were extracted for both sexes combined, and separately for subgroup analysis. Full search criteria and categorisation of major non-fatal injuries according to AIS can be found in Appendix 1.

Median incidence rates per 100,000 persons with interquartile range (IQR) for SCIs, non-fatal injuries, and mortality were reported for each country and region for the whole period 2012–2019. Median values were selected due to substantial right-skewness and heterogeneity in incidence rates across countries. Regression analyses were undertaken as exploratory ecological analyses to describe broad temporal and income-group patterns, rather than to infer causal relationships at the individual or country level. Incidence rates were then aggregated by World Bank income group (low-income (LIC), lower-middle-income (LMIC), upper-middle-income (UMIC), and high-income (HIC)), with median (IQR) incidence rates reported by income group and stratified by sex for the whole study window [13].

A linear regression ecological analysis was performed to compare median incidence rates per 100,000 persons for all countries and regions and for each income group by year. An analysis of variance (ANOVA) was then performed to test the significance of year and income over time, and their interaction. A Quasi-Poisson regression was performed to compare incidence rates per 100,000 persons across income groups. Quasi-Poisson regression was selected over Poisson regression to account for overdispersion whereby variance exceeds the mean. Counts of cases were modelled with the log of the mid-year population from GBD estimates [14] as an offset to produce incidence rate ratios (IRRs) with Wald 95% confidence intervals (CIs) [15].

Proportions of non-fatal injuries, including SCI, were categorised by AIS region and reported globally and by income group by converting incidence rates to counts using mid-year population, summing them by AIS region, and dividing them by the total number of non-fatal injury cases. Median country-year proportions of SCIs amongst all non-fatal injuries were further calculated by income group. The Kruskal–Wallis test was used to compared the proportion of SCIs across all income groups, and the Mann–Whitney U test used for post-hoc pairwise comparison.

Statistical significance was deemed as p < 0.05. All data linkage, cleaning, and statistical analysis were conducted in R (version 4.2.2) [16]. There was no missing data. Allocation of income groups by country/territory is displayed in supplementary Fig. 1.

Data collection, pre-processing, analysis and reporting were completed in line with the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) statement [17] (Appendix 2).

Results

Incidence rates per 100,000 persons of mortality, non-fatal injury, and spinal cord injuries

The overall median incidence rate of MVC-related mortality was 5.8 per 100,000 population (IQR 3.8–9.4) (Table 1). Incidence rate ranged from 0.9 to 48.1 per 100,000 population (Fig. 1A). Higher incidence rate was observed in Africa, the Middle East, the Americas and Northern Asia (Fig. 1A). The median incidence rate of MVC-related non-fatal injuries (inclusive of SCIs) was 123 (88.7–182) per 100,000 persons and ranged from 28 to 844 per 100,000 population (Table 1). Higher incidence rates were noted in Eastern Europe, Northern Asia, and the Middle East (Fig. 1B). The median incidence rate per 100,000 persons of SCIs was 0.64 (0.40–1.0) per 100,000 persons, ranging from 0.14 to 3.2 per 100,000 persons globally (Table 1). The distribution of SCIs was generally in keeping with all non-fatal injuries, although the USA and Mexico had comparatively relatively higher rates (Fig. 1C). Males had higher incidence rates of MVC-related mortality, non-fatal injury, and SCI (Table 1).

Table 1.

Median (IQR) incidence rates of MVC-related mortality, major non-fatal injuries, and spinal cord injuries due to motor vehicle collisions per 100,000 population between 2012 and 2019, by sex and income group

Income Group	Mortality			Major Non-Fatal Injuries			Spinal Cord Injuries
	Total	Male	Female	Total	Male	Female	Total	Male	Female
Low Income	7.3 (5.2–10.2)	11.4 (9.0–14.8)	4.7 (3.3–7.7)	116 (89.5–135)	135 (103–205)	69.1 (53.7–119)	0.85 (0.51–1.1)	0.40 (0.31–0.61)	0.16 (0.12–0.26)
Lower-Middle Income	7.2 (4.6–10.6)	9.5 (7.6–13.7)	3.0 (2.1–5.1)	123 (98.7–182)	206 (116–327)	79.1 (49.6–162)	0.59 (0.40–1.0)	0.66 (0.39–1.0)	0.19 (0.13–0.37)
Upper-Middle Income	6.1 (4.1–10.4)	9.9 (6.0–14.7)	2.7 (1.8–4.6)	142 (93.1–205)	245 (157–316)	85.3 (57.1–126)	0.63 (0.40–1.1)	0.83 (0.55–1.0)	0.25 (0.15–0.33)
High Income	4.0 (2.6–5.8)	5.3 (3.2–8.5)	1.9 (1.2–2.7)	104 (81.6–185)	187 (124–275)	80.2 (59.0–118)	0.57 (0.37–0.90)	1.1 (0.77–1.88)	0.42 (0.25–0.66)
All Countries	5.8 (3.8–9.4)	8.8 (5.5–13.1)	2.7 (1.7–4.6)	123 (88.7–182)	196 (123–287)	80.1 (55.9–123)	0.64 (0.40–1.0)	0.80 (0.47–1.2)	0.27 (0.16–0.42)

Fig. 1.

Median incidence rate between 2012 and 2019 of MVC-related a mortality b major non-fatal injuries and c spinal cord injuries across each of the 204 countries and territories in the Global Burden of Disease dataset

Linear regression models showed that income group was a significant predictor of median incidence for MVC-related mortality, non-fatal injury, and SCIs (p < 0.001). There was no significant change in median incidence over time for MVC-related mortality (p = 0.61), non-fatal injury (p = 0.34), or SCI (p = 0.19). However, median injury incidence showed a significant year-income interaction (p < 0.001); median injury incidence rates increased approximately 2.9 cases per 100,000 persons per year (95% CI 1.0 to 4.7) in LMICs, whereas they decreased approximately 3.1 cases per 100,000 persons per year (95% CI −4.8 to −1.5) in UMICs and HICs. A similarly significant income-year interaction was observed for median SCI incidence rates (p = 0.02). In LICs, median incidence rates rose by 0.1 cases per 100,000 persons per year (95% CI 0.00 to 0.02), and in UMICs and HICs, case decreased by 0.02 (95% CI −0.04 to −0.00) and 0.01 (95% CI −4.8 to −1.5) cases per 100,000 persons per year, respectively.

Incidence rates of MVC-related mortality, non-fatal injuries, and spinal cord injuries by income group

The median incidence rate of MVC-related mortality between 2012–2019 was lowest in the high income group (4.0 per 100,000 population [IQR 2.6–5.8]). Upper-middle income countries had a median incidence rate of 6.1 deaths per 100,000 persons (IQR 4.1–10.4), lower-middle income countries 7.2 deaths per 100,000 persons (4.6–10.6 IQR), and low income countries 7.3 deaths per 100,000 persons (IQR 5.2–10.2) (Table 1). MVC-related mortality incidence rates were significantly lower in HICs (Incidence Rate Ratio, IRR 0.79 (95% CI: 0.69–0.91); p = 0.001), UMICs (IRR 0.72 (95% CI 0.63–0.82); p < 0.001), and LMICs (IRR 0.67 (95% CI: 0.59–0.76); p < 0.001) compared with LICs. The median incidence rate of MVC-related major non-fatal injuries was highest in upper-middle income countries (142 per 100,000 persons [IQR 93.1–205]), followed by lower-middle (123 per 100,000 persons [IQR 98.7–182]) and low income countries (116 per 100,000 persons [IQR 89.5–135]) (Table 1). Incidence rates for non-fatal injuries were not significantly different between income groups (Fig. 2). The median incidence rate of SCIs due to MVCs was highest in low income countries (0.85 per 100,000 persons [IQR 0.51–1.1]) (Table 1). Incidence rates of SCIs were not significantly different between income groups (Fig. 2). Supplementary Table 1 reports all IRR and 95% confidence intervals for each group. Incidence rates of mortality, non-fatal injury, and SCI were consistently higher in males across all income groups (Table 1).

Fig. 2.

Incidence rate ratios with 95% confidence intervals for MVC-related mortality, major non-fatal injuries, and spinal cord injuries between World Bank income groups; reference group = low income

Spinal cord injuries as a proportion of all major non-fatal injuries

Summed across all countries, SCIs accounted for around 0.5% of all cases of major non-fatal injury. Lower extremity injuries accounted for the vast majority of injuries (35.1%), followed by head injuries (19.3%), upper extremity injuries (8.4%), thorax injuries (7.69%), and abdomen and pelvic contents injuries (7.63%). The distribution of injuries by AIS region was comparable across income groups (Fig. 3). A full breakdown of proportions by AIS region and income group is available in Supplementary Table 2.

Fig. 3.

Distribution of major non-fatal injury burden by AIS body region by income groups and globally. Percentages are calculated by summing the total number of cases across all country-years to present a weighted mean by income group or globally. Spine denotes spinal cord injuries. Spine (other) denotes vertebral column fractures

The median country-year contribution of SCIs amongst all major non-fatal injuries due to MVCs was also low overall at 0.392% (IQR 0.377–0.832%) across all income groups (Fig. 4). The median percentage was the highest in low-income countries (0.833%, IQR 0.410–0.853%), and similar across lower-middle (0.382%, IQR 0.367–0.823%), upper-middle (0.385%, IQR 0.376–0.401%), and high income (0.394%, IQR 0.382–0.426%) countries. Proportions were significantly different between income groups (p < 0.001). Post-hoc testing revealed that the proportion of SCIs was significantly different between LICs and all other income countries, and between lower-middle income and high income countries.

Fig. 4.

Box and whisker plot of the percentage of spinal cord injuries between 2012–2019 by income group. The box represents the interquartile range (25th-75th percentile), the line within the box represents the median value. The Kruskal–Wallis test demonstrated that there were significant differences between proportion of spinal cord injuries by income group. Post-hoc testing with Mann Whitney U tests (horizontal brackets) demonstrated statistically significant differences between multiple income group proportions. **** = p < 0.0001, ns = non-significant

Discussion

Summary of key findings

This study demonstrates substantial global variation in mortality and major non-fatal injury rates following MVCs according to income group. Across all settings, spinal cord injury accounted for a small proportion of non-fatal MVC-related injuries (approximately 0.5–1%). This is consistent with UK registry analyses, where head, chest and abdominal injuries form the predominant burden of serious harm [8–10]. This finding is clinically relevant; in many settings current post-crash practices continue to prioritise spinal protection and movement minimisation; yet spinal cord injuries in survivors are relatively uncommon compared with other time-critical injuries. The emphasis placed on avoiding spinal movement may therefore not be proportionate to clinical need in all contexts.

In addition to describing incidence, this study applied the GBD dataset to allow comparison of injury distributions and proportional contribution of spinal cord injury across income settings. GBD is widely used for global burden estimation but has not previously been used to explore the relative pattern of injury among survivors, or across multiple years by income group. Demonstrating that the GBD dataset can be used in this way provides a foundation for further exploratory work.

Comparison with previous literature

Reported incidence rates of spinal cord injury in this study align with previous global estimates. WHO has estimated annual SCI incidence (traumatic and atraumatic) to be between 4–8 new cases per 100,000 population, with road traffic collisions a major contributing mechanism [18]. Estimates from this analysis suggest that MVCs account for approximately 13–25% of all SCIs globally, which is comparable to WHO figures. Findings from Kumar et al. and Jazayeri et al. also reported higher traumatic SCI incidence in L/LMICs compared with HICs, consistent with the patterns observed here [18, 19]. Previous studies exploring SCI in the context of MVCs have largely relied on regional or national trauma datasets, which limits the ability to compare across settings [18, 20–23]. By contrast, the present approach allows for cross-country comparison by income group and over time.

Interpretation of global differences

Non-significant annual changes in incidence rates for mortality, major non-fatal injury and SCI may reflect the absence of large-scale, system-wide improvements in post-crash care over the study period. However, the divergence seen between income groups, with increasing non-fatal injury and SCI incidence in L/LMICs and decreasing rates in UMICs and HICs, suggests evolving differences in injury patterns, road environments and post-crash response capacity. These patterns likely reflect a combination of road-use, infrastructure and care-system differences.

Higher injury incidence rates were consistently observed in males compared with females across all income groups, reflecting established global trends [24]. However, previous UK and US analyses suggest that sex-related differences in mortality and injury patterns are context-dependent and influenced by factors such as crash dynamics, restraint fit, injury recognition and triage [10, 25]. These inconsistencies highlight the value of sex-disaggregated reporting to inform equitable post-crash care and responder strategies.

The relatively low proportion of SCI among survivors across income groups is consistent with UK registry findings in trapped MVC patients, where head, chest and abdominal injuries predominate [8]. The higher proportional burden of SCI in some low-income settings, despite similar incidence rate ratios, suggests variation in injury distribution and survivorship. Potential contributing factors include differences in road-user behaviour, vehicle safety standards and enforcement of restraint use.

Implications for post-crash care and patient handling

The low absolute proportion of SCI relative to other severe injuries provides important context for the development of post-crash guidance. In many settings, rescue strategies continue to emphasise movement minimisation and prolonged extrication. However, extended entrapment may delay time-critical interventions for more prevalent life-threatening injuries. Recent work has suggested that self-extrication and gentle assistance may prioritise rapid access to care while avoiding unnecessary immobilisation [10, 26–30]. The suitability of these approaches will depend on context, available training and system capacity.

Limitations

This study relies exclusively on GBD data and excludes pedestrian, motorcycle, cycle and other road injury types. While the GBD provides broad global coverage, its estimates are influenced by the quality and completeness of underlying data. Variability in data availability, particularly in L/LMICs, requires model-based estimation, which may introduce uncertainty. These limitations should be considered when interpreting the findings, although in the absence of comprehensive trauma registry data for most countries, GBD estimates remain a useful resource for examining global patterns of disease and injury.

Future research and system priorities

A major gap remains in understanding the injury burden among those who do not survive to hospital. However, the absence of routine data on injury patterns in pre-hospital deaths limits the ability to determine which injuries are survivable with optimal early care. Strengthening pre-hospital trauma surveillance, including structured review of fatal MVC cases, is an important priority. For survivors, the lack of detail in GBD on SCI level, severity and mechanism limits interpretation of the nature and complexity of injuries across contexts; such information would further inform decisions on patient handling and extrication.

Finally, the study underscores system-level priorities for improving post-crash outcomes. These include improving seat-belt and road-safety adherence, strengthening early clinical-care pathways, supporting community-based first response programmes, and ensuring data-driven evaluation across trauma systems. Local adaptation and co-development of guidance will be essential to ensure relevance across diverse settings.

Conclusion

This study highlights substantial global variation in mortality and major injury rates following motor vehicle collisions, alongside a consistently low proportion of spinal cord injury among survivors. Although the proportion of spinal cord injury is higher in some low-income settings, head, chest and abdominal injuries remain far more common across all regions. Understanding this relative pattern is important, as current post-crash care in many settings continues to prioritise spinal protection and movement minimisation, which may delay access to time-critical interventions for more prevalent injuries.

These findings support the need to contextualise spinal precautions within a broader assessment of injury risk, and to develop responder and bystander guidance that enables early, safe and timely extrication, assessment and treatment. Further work should explore how such approaches can be adapted and applied across different resource settings, particularly where structured trauma registry data are limited, and pre-hospital fatality patterns remain insufficiently understood.

Authors’ contributions

TN,JC: initial concept, acquisition, analysis and initial drafting. All authors contributed to the interpretation of data, reviewing and final approval of the work.

Data availability

Data from the GBD study is publicly available via the Institute of Health Metrics and Evaluation (IHME) Global Health Data Exchange (GHDx, [](http:/ghdx.healthdata.org) [http://ghdx.healthdata.org/](http:/ghdx.healthdata.org)) [12] (https:/app.readcube.com/library/f2c917b6-1dcf-4786-ad4a-3ce6c7b3c6f9/all?uuid=9811896540587384&item_ids=f2c917b6-1dcf-4786-ad4a-3ce6c7b3c6f9:20c44967-d639-4117-bd43-14d948e26742).

Declarations

Competing interests

The authors declare no competing interests.