Life expectancy and long-term survival after traumatic spinal cord injury: a systematic review

Abstract

INTRODUCTION

Spinal cord injuries have a considerable impact on healthcare in terms of mortality and morbidity. To address the difficulties faced by people affected by this condition and to raise awareness among stakeholders and policymakers, it is crucial to understand factors impacting survival. The purpose of this study is to systematically review the literature on life expectancy in people with traumatic spinal cord injury (tSCI), identifying key factors influencing mortality and survival.

EVIDENCE ACQUISITION

We conducted a systematic review, searching the literature for articles published up to July 2023 in PubMed, Web of Science, Cochrane Library, Google Scholar, and PEDro. Study outcomes had to be one of survival rate, life expectancy, standardized mortality ratio, or mortality rate. Only original research articles published in English were included. The quality of evidence was evaluated with the MINORS scale. The level of evidence was categorized according to the OCEBM model.

EVIDENCE SYNTHESIS

A comprehensive literature search yielded 102 articles, after the selection process 20 studies were included in our review. The main factors negatively influencing survival and life expectancy included higher neurological level of injury (NLI), completeness of the lesion, need for mechanical ventilation, increasing age, and male gender. The development of SCI-related comorbidities also negatively impacted survival as well as the lack of specialized care, especially in low-income countries. Additionally, pre-injury health status and personal income may affect survival.

CONCLUSIONS

Current literature shows that people affected by tSCI have a shorter life expectancy compared to the general population, highlighting some factors as possible predictors. It is difficult to compare available evidence due to the methodological heterogeneity across studies, which makes it challenging to draw generalizable conclusions on life expectancy in people with tSCI. Further studies are required to address these issues and accurately estimate life expectancy accounting for gaps in the management of people affected by tSCI to improve their care.

European Journal of Physical and Rehabilitation Medicine 2024 October;60(5):822-31

DOI: 10.23736/S1973-9087.24.08462-4

© 2024 THE AUTHORS

SYSTEMATIC REVIEW

(Cite this article as: Zadra A, Bruni S, De Tanti A, Saviola D, Ciavarella M, Cannavò G, et al. Life expectancy and long-term survival after traumatic spinal cord injury: a systematic review. Eur J Phys Rehabil Med 2024;60:822-31. DOI: 10.23736/S1973-9087.24.08462-4)

Introduction

Spinal Cord Injury (SCI), whether traumatic or non-traumatic, elicits profound changes in the lives of affected individuals, impacting functionality and psychosocial well-being. The incidence of traumatic spinal cord injury (TSCI) worldwide, regardless of the cause, ranges from 5.1 to 150.48 cases per million/year according to national studies.¹ TSCI predisposes to various secondary medical conditions throughout life, which might interfere with health in its multiple dimensions: physical functioning, social activities, productive employment, and quality of life.^{2, 3} The economic burden of SCI on healthcare is very high, as is the risk of long-term complications and rehospitalizations.⁴ A review of the frequency of secondary health conditions in people with spinal cord injury found that a high prevalence rate (50% and higher) was chronic pain, bowel and bladder problems, muscle spasms, and fatigue.⁵ The mortality rate is extremely high during the first year after the injury occurs,^6-9 and it depends on numerous factors, such as the severity of the trauma, the management of early complications, the quality of medical assistance during rescue and the acute phase, admittance to specialized centers.^10-13 It has also been demonstrated that these variables can affect long-term survival and quality of life as well.^{11, 12, 14-16} The USA Model Systems represents the preeminent data source for gauging long-term trends in mortality rates and life expectancy.¹⁰ Several studies reported elevated mortality rates in SCI compared to the general population not only in the first year after injury but also in the following years. Over the last decades, acute survival rates have improved, but longer-term mortality rates have not decreased at the same pace.^{10, 17-19} It is undeniable that we have witnessed an extension of life expectancy following TSCI, even for the most severe lesions, which raises the need for long follow-up to better assist individuals aging with TSCI.²⁰ However, lately, life expectancy has not significantly increased for those who survive the first year after the injury.^{17, 21} Additionally, over the years we have assisted to medical advancements that have shifted the primary causes of death in the SCI population. For instance, before the mid-1970s, renal failure and other urinary tract complications were the leading cause of death in people with SCI, but at present we face a different scenario.^{22, 23}

To effectively support people with SCI, a multifaceted approach is mandatory: it is crucial to adapt healthcare systems, and social and insurance services, allocating available resources in a meaningful manner. For this reason, obtaining scientific evidence on life expectancy and the main factors detrimental to survival is essential to managing resources efficiently, identifying the main areas for surveillance and prevention, and targeting educational interventions for people with SCI and their caregivers.

This study aims to systematically review the literature on the life expectancy of people affected by tSCI, summarize the data available on the long-term survival outcome of individuals with SCI, address open questions on this topic, and identify some key factors influencing mortality/survival. The secondary objective is to evaluate predictors of life expectancy, highlight the main issues that warrant consideration, and raise awareness among stakeholders and policymakers of the challenges faced by people with SCI, with the final goal of improving the lives of people affected by this condition.

Evidence acquisition

The review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.^{24, 25}

Data sources and literature search strategy

A literature search for journal articles published up to July 2023 was carried out using these databases: PubMed, Web of Science, Cochrane Library, Google Scholar, and PEDro. Keywords and Mesh included: “spinal cord injur*/classification”[MeSH Terms] OR “spinal cord injur*/complications”[MeSH Terms] OR “spinal cord injur*/diagnosis”[MeSH Terms] OR “spinal cord injur*/epidemiology”[MeSH Terms] OR “spinal cord injur*/mortality”[MeSH Terms] OR “spinal cord injur*/prevention and control”[MeSH Terms] OR “spinal cord injur*/rehabilitation”[MeSH Terms] OR “spinal cord injur*/therapy”[MeSH Terms]) OR “spinal cord injur*”[tw] OR “spinal cord classification”[tw] AND “Life Expectancy”[Mesh] OR “Longevity”[Mesh] OR “Mortality, Premature”[Mesh] OR “Disability-Adjusted Life Years”[Mesh] OR “Survival Rate”[Mesh] OR “Cause of Death”[Mesh] OR “Vital Statistics”[Mesh] AND “Spinal Cord Injuries”[Mesh] OR “Central Cord Syndrome”[Mesh] OR “Spinal Cord Compression”[Mesh] OR “Spinal Cord Injuries”[Mesh] OR “life expectancy” [tw] OR “survival rate” [tw] OR “standardized mortality ratio” [tw] OR “health expectancy” [tw] OR “Long-term survival” [tw] AND “Italy/Epidemiology”[Mesh] OR “Germany/epidemiology”[Mesh] OR “United States/epidemiology”[Mesh] OR “Australia/epidemiology”[Mesh] AND “evidence-based medicine”[Mesh] OR “Biostatistics/methods”[Mesh] OR Evidence-Based Medicine[tw]. The identified articles were further analyzed for inclusion criteria by three reviewers (AZ, SB, JB); retrieved articles were also scanned for additional papers that may have been missed in the original database search. Results were limited to publication dates between 2000 and 2023 and published in the English language.

Finally, to be included in this systematic review, the published manuscript needed to contain all necessary data (see below in data analysis).

Study selection

We used the following inclusion criteria: 1) the study design included randomized controlled trials, quasi-experimental design, cohort or historical cohort study, case-control study, or cross-sectional study; systematic reviews were used to identify additional relevant studies; 2) patients had to be ≥16 years of age at the time they sustained the SCI; 3) the study had to consider a measure of survival or mortality in people with traumatic SCI; 4) the study had to include a control or comparison group from the general population; 5) the paper had to be published in the English language; 6) study had to provide mortality rates stratified by age, sex, or race group; 7) had to report a specified follow-up time ≥12 months when mortality was assessed; 8) one of the study outcomes had to be survival rate, life expectancy, standardized mortality ratio, or mortality rate.

We used the following exclusion criteria: 1) mortality rate data could not accurately be extracted from the article; 2) studies that included individuals with SCI who died on-scene of the accident or before hospital admission; 3) letters, editorials, and reviews.

Statistical analysis

For each study included in this review, the following data have been extracted: information on patient characteristics (age, injury etiology, etc.), primary outcomes, follow-up time, and study design (retrospective versus prospective).

Independently, three investigators (A.Z., S.B., J.B.) reviewed all potentially relevant titles and removed duplicates to determine which abstracts to review. Then, the same investigators reviewed the abstracts to select the final full manuscripts for review and data extraction. The abstract review was carried out independently, with arbitration performed by J.B. to resolve any disagreement between reviewers.

Data extraction and study rating process

A data extraction tool was employed to synthesize pertinent information from each research endeavor, encompassing study design, geographic location, sample size, principal endpoints, and duration of follow-up. Subsequently, the compiled data underwent rigorous scrutiny by all three researchers to ensure completeness and applicability. Independently, each author assessed the quality and evidence strength of all the studies. The quality entailed an examination of internal validity factors (such as study design, reporting quality, potential biases, and confounding variables) and external validity considerations (including generalizability) utilizing the Methodological Index for NOn-Randomized Studies (MINORS).²⁶ This instrument assigns a numerical score, ranging from 0 to 24, to gauge study quality. Additionally, the level of evidence conferred by each study was classified using the Oxford Centre for Evidence-Based Medicine (OCEBM).²⁷ Studies falling under exclusion criteria (specifically, systematic reviews, case series, and opinion-based papers) were systematically omitted. Any discrepancies in MINORS scoring or OCEBM categorization were resolved through deliberation among the three evaluators.

Data synthesis

Extracted data and quality and level of evidence were summarized for each study.

Evidence synthesis

Identification of studies

An overview of the study identification and selection process is provided in Figure 1. The initial search returned 97 manuscripts; an additional five references were identified through reference review, resulting in 102 articles. Of these articles, 42 duplicates were removed, leaving 60 potentially relevant articles. Thirty were removed based on examination of the title and/or on the review of the abstract. After the first step, the full texts were further narrowed to 30 full manuscripts included in the systematic review based on inclusion/exclusion criteria. Studies were removed mainly due to the inability to extract accurate survival/mortality data or a lack of data on the outcomes under study (N.=4). For example, the study by Imai et al.²⁸ did not report confidence intervals or standard errors and for this reason, it was not included in the review. Therefore, this review is based on a selection of 20 articles. A detailed breakdown of the studies’ characteristics is summarized in Supplementary Digital Material 1 (Supplementary Table I).^{13-15, 17-20, 29-40}

Figure 1.

—Flow diagram of the literature selection process.

The studies were published between 2000 and 2023. The examined population across the included studies was from the US (nine studies), Europe (seven studies), and the UK (two studies), Australia (two studies). Fourteen studies were retrospective, while three were prospective in design. A median sample size of 2492 individuals was examined, with studies having a sample size ranging from 100 to 18,988 people. Concerning the follow-up time, studies showed a median of 9.5 years, ranging from 0.5 to 70 years. The median methodological quality score for all the included studies, based on the MINORS criteria, was 10/14 (cut-off ≥10). Because the included studies were observational, five items (from 8 up to 12; totaling 10 points) on the MINORS were not applicable for all studies included. It was observed that none of the studies were able to achieve a full score due to several limitations that were common among most of the included studies. These limitations include an incomplete description of how the sample was representative of the population of interest, a limited description of the characteristics of those who dropped out, use of unreliable or invalid measures (or failure to report on reliability and/or validity), insufficient reporting of how participants were lost to follow-up, and how differing length of follow-up time was accounted for in statistical analyses, and inadequate sample size.

All studies were rated at level 2b evidence (defined as an individual cohort study by the OCEBM Levels of Evidence Working Group, 2011), given the exclusion criteria.

Life expectancy literature in SCI patients and investigated outcomes

There are multiple ways to estimate the long-term survival and life expectancy of individuals.⁴¹ Several methods have been utilized to study life expectancy and long-term survival among people with spinal cord injuries, each with its advantages and disadvantages.^{41, 42} Noteworthy, life expectancy estimates vary according to the chosen method of statistical analysis.⁴² For example, the use of age at injury or a general population cohort life table in SMRs can significantly increase estimated life expectancies, particularly for younger ages.³²

Generally, life expectancy is presumed to be shorter in individuals with spinal cord injury compared to the general population.^{14, 33, 43} Thietje et al.⁴⁰ investigated the differences in life expectancy between individuals with paraplegia and those with tetraplegia. The study revealed that life expectancy was significantly higher in individuals with paraplegia, with an average of 34 years compared to 25 years for those with tetraplegia. However, the authors found no significant difference in life expectancy between individuals with higher versus lower spinal cord lesions or complete versus incomplete injuries.⁴⁰ In a more recent study, the same authors found a correlation between post-injury life expectancy and the level of lesion and severity of injury. The study indicated that the overall post-injury life expectancy was 25.0 years, with a significant difference in life expectancy depending on the level of the lesion. Specifically, individuals with a high level of lesions had an average survival time of 17.3 years, whereas tetraplegics with lower-level lesions survived for an average of 22.3 years.³⁰ Paraplegic patients with high-level lesions had a mean survival time of 31.1 years, whereas all type AIS D patients had a mean survival time of 28.9 years. However, the difference between these two groups was not significant.³⁰ Another piece of evidence present in the literature regards the standard mortality ratios (SMRs). It is known that individuals with spinal cord injury (SCI) were two times more likely to die compared to the general population and crude SMRs revealed that overall mortality in traumatic SCI was between 1.47 and 2.8 times higher than in the general population.⁴⁴ There is evidence indicating that SMRs were higher among women compared to men.^{37, 43, 45} Additionally, the mortality rate and risk of death show a positive correlation with increasing age, completeness of lesion, and higher neurological level of injury.⁴⁶

Studies have shown that people who sustain a full spinal cord injury during their youth are likely to have a shorter lifespan. It is also known that those with spinal cord injuries during their youth often have a higher risk of developing secondary complications over the long term. These complications may include issues with their respiratory and cardiovascular systems, bowel and urinary functions, as well as pressure ulcers and chronic pain.⁴⁷ A key factor negatively associated with survival rate is the need for mechanical ventilation, in addition to age.^{16, 39, 48-50} Moreover, several other factors should be considered as plausible predictors of survival, including poor education, maladaptive behaviors (substance abuse, suicide attempts), and poor quality of care.⁵¹ Pre-injury health status, history of smoking, alcohol and drug abuse, and obesity are factors relevant to life expectancy for persons with SCI, as they are in the general population.^{52, 53} Finally, people with SCI have specific factors that can increase their risk of mortality. These factors include dysphagia, neurogenic bowel dysfunction, neurogenic bladder, bladder cancer, frequent infections, and psychiatric comorbidities, particularly depression, which can lead to suicide.^54-56

We now consider the most investigated outcomes included in the reviewed studies: survival rate, life expectancy, standardized mortality ratio, and mortality rate.

Survival

The range of median survival was 2.8³⁸ and 43⁵⁷ years after SCI.^{38, 57-59} One-year survival rates ranged from 79% to 100%, 5-year survival from 85% to 96%, and 10-year survival from 81 to 93%.^{14, 17, 37, 58} Studies comparing survival rates between males and females generally showed higher survival rates among females.⁴³ Altogether these results confirm the findings summarized in a previous review of this literature.⁴³

Only a few studies have reported on survival rates based on SCI characteristics. The studies reporting survival information based on the neurological level of injury (high versus low) or completeness (complete versus incomplete) have found that there is a negative relationship between survival rates and a higher neurological level of injury or complete lesion.⁴³ A study by Middleton et al.,¹⁴ which provided both long-term follow-up visits and detailed information related to SCI characteristics, showed consistent trends for 5-, 10-, and 20-year survival, and also highlighted that among people with AIS A, B, or C, those with a C1-C4 lesion had better survival rate compared to those with a C5-C8 lesion, meaning that the higher mortality rate following spinal cord injury in the first year is associated with higher lesions. To account for this, deaths occurring within the first year were excluded from survival calculations in subsequent years.¹⁴

Life expectancy

The life expectancy of individuals affected by SCI was reduced as compared to the general population.^{14, 33} Also, the difference in life expectancy increased as the severity of the lesion and the age of the individual increased.^{39, 40, 60} For example, as reported also by Chamberlain et al.,⁴³ Strauss et al.¹⁷ calculated a percent life expectancy as compared to the general population for a 25-year-old white male (with traumatic SCI with nonviolent etiology, 3 years since injury, and a grade A, complete lesion) of 68.2% for C6-C8; 58.9% for C5; 51.9% for C4; and 49.9% C1-C3 lesions.¹⁷ Middleton et al.¹⁴ estimated percent life expectancy for individuals aged 25 with a T1-S5 AIS A, B, or C level and severity of 88%; 74% for C5-C8 AIS A, B, or C; and 69% C1-C4 AIS A, B, or C lesion.⁴³ Similarly, even if it is not the focus of the present analysis, in the USA pediatric population (21 years and younger), the life expectancy of males with nonviolent SCI etiology and an attained age of 5 years, 2 or more years post-injury, was 72.5% for incomplete paraplegia (AIS grades B, C) and 60.2% for complete tetraplegia (C5).^{43, 60}

Mortality risk

Among studies that compared mortality rates across different age groups at the time of injury, the hazard ratio (HR) increased with age. The ratios ranged from 1.03 to 3.44,^{38, 61} and the odds ratio (ORs) ranged from 1.02 to 1.31.⁶⁰ Comparing the mortality of males and females across studies, HRs ranged from 0.94 to 1.40,^{38, 62} and ORs ranged from 1.15 to 1.60.^{33, 46} Overall, the literature indicates a consistently higher hazard ratio in males as compared to females, in line with the results reported in previous review.⁴³

Except for a few studies, it has been consistently found that the hazard ratio increases with higher neurological levels of injury and complete lesions. These measurements are typically taken through either AIS or Frankel grading. However, comparing individuals affected by tetraplegia and those affected by paraplegia can be difficult due to variations in grouping, such as age groups and the level of neurological injury, methodological differences, and various adjustment techniques applied to potential confounding variables.

Standardized mortality ratio

SMRs were described as stratified for several characteristics such as cause of death, paraplegia, tetraplegia, incomplete versus complete lesions, gender, age, and AIS grade.⁴³ Most studies reporting overall SMRs were from Europe,^{6, 31, 38, 58} and one study was from Australia.¹⁴ SMRs for people with SCI ranged between 1.47 and 5.00.³⁸ On average, SMRs were found to be higher in women compared to men,^{31, 37, 38} with increasing age,^{14, 38, 58} and with higher levels of neurological injury and completeness of the lesion.^{14, 31, 37, 58} SMRs were positively associated with cerebrovascular diseases, infectious diseases, suicide or accidental poisoning, and pneumonia/influenza.^{14, 15, 37, 58}

Causes of death after SCI

Several studies have examined the causes of death in the SCI population, and we reported the most frequent in Table I (N.=22).^{14, 15, 29, 31, 37, 40, 43, 58, 63, 64} Pneumonia was the most reported (N.=5), followed by heart disease (N.=3). A single study provided raw data on the number of deaths by cause and gender. It revealed that heart disease was the primary cause of death among males, while females had an equal number of deaths from heart disease and gastrointestinal disorders.³¹ In the research conducted by Hagen et al.,⁵⁸ SMRs were documented based on gender for specific causes of mortality; notably, a huge contrast in SMRs between males and females emerged in accidental poisoning and suicide (3.7 compared to 37.6, respectively).⁵⁸

Table I. —Leading causes of death.

Most reported leading cause of death	N. of studies
Pneumonia and influenza	5
Other respiratory diseases	1
Ischemic heart disease	3
Diseases of the pulmonary circulation	-
Non-ischaemic heart disease	-
Cancer	2
Diseases of the nervous system	-
Septicemia	2
Infectious and parasitic	-
Other bacterial diseases	-
Cerebrovascular disease	-
Diseases of the artery	-
Diseases of the digestive system	-
Diseases of the urinary system	-
Unintentional injuries	-
Mental disorder	-
Suicide	2
Accidental poisoning	-
Musculoskeletal and skin disease	-
All other external	-
Endocrine	-
Other and unknown	-

Survival and mortality trends

In conclusion, some studies have examined the trends in survival and mortality rates over time, but the results have been mixed. All these studies were conducted using a sample population from high-income countries where SCI management was adequate, and the follow-up periods were extended, ranging from 12 to 51 years.^{14, 17, 29, 37, 58, 62, 65} Middleton et al.¹⁴ found evidence of progressive improvement in life expectancy and survival rate, particularly among people affected by paraplegia. Saunders et al.⁶⁵ reported an overall decrease in the SCI mortality rate, while Hagen et al.⁵⁸ found no significant change in SMRs over time.^{14, 58, 65} Other studies identified improvements in mortality rate during the first year post-injury over time, although not since the 1980s.^{10, 66} Finally, a study with a follow-up of 40 years found higher SMRs in the period from 1976 to 1982 compared to 1961 to 1975.³⁷

Discussion

Life expectancy is the average expected years of life remaining for a person based on their age, sex, and other factors. It is often calculated using available data and literature.⁶⁷ In the debate on the life expectancy in people who suffer a spinal cord injury, “life expectancy” should not be confused with the actual survival time of an individual, which is difficult to predict with certainty, even in the general population. Nevertheless, these studies are crucial for case management and improving the quality of life of these patients. To date, a robust body of evidence highlights that life expectancy in the spinal cord injury population is reduced as compared to the general population.^{14, 33} A different piece of information available in the literature is represented by the standardized mortality ratios (SMRs). SMRs for individuals with SCI extended from 1.5 to 3.^{43, 44} Defined SMRs as a comparison between the considered SCI population and the healthy general population, SMRs were higher for complete versus incomplete lesions and for women versus men.^{38, 43, 44} This latter finding is probably due to the longer life expectancy of the female general population. Key factors to survival rate are age, severity of lesion (complete versus incomplete), neurological level of injury (high versus low), and the need for chronic mechanical ventilatory support.^{30, 39} Other variables potentially affecting the outcome include pre-injury health status and habits (smoking, athletic condition, substance abuse), level of education, employment status, and annual incoming.^{34-36, 68}

Studies have shown that the in-hospital mortality rate among individuals with SCI can vary significantly depending on the healthcare organization and level of clinical skills involved.⁴³ This suggests that the quality of clinical care is critical to the outcome of in-hospital mortality. It is also generally observed that countries with higher income levels tend to have better clinical outcomes than low-income countries.^{43, 69} Additionally, research indicates that admission to specialized spinal rehabilitation centers can lead to a lower incidence of secondary conditions and better outcomes.^{43, 69}

Subgroups within the individuals with SCI and comparison

Mortality and survival outcomes vary depending on lesion characteristics such as the level and severity. Generally, mortality rates increase with higher AIS grades and higher neurological injury levels.^{14, 31} This supports that individuals with spinal cord injuries require specialized care in centers that possess a high level of competence in SCI management. This care should be focused on the treatment of health conditions that are specific to SCI, such as neurogenic bladder and bowel, pulmonary complications, and skin ulcers. It is important to implement a specialized approach for individuals with SCI and high lesion levels. This is to minimize the gap in life expectancy and survival rates within the SCI population.⁷⁰ For example, after a cervical SCI, breathing difficulties can arise due to the paralysis of respiratory muscles. Care that is focused on respiratory management has proven to be effective in improving respiratory function.^{70, 71} Inconsistencies in mortality data and longevity outcomes across studies may also be influenced by non-clinical factors such as referral patterns, admission criteria, and rehabilitation management.⁷² These inconsistencies could be partly attributed to differences in study design. For instance, Cao et al.⁵⁹ excluded individuals who died during hospitalization, censored all observations at the end of the study in 2009 and had a follow-up time of 11 years. On the other hand, Strauss et al. excluded in their 2006 study patients who required ventilatory support, resulting in significant effects on the outcomes studied.¹⁷

Trends compared to the general population

People with SCI have a higher risk of death than the general population, especially those with tetraplegia compared to paraplegia.⁴³ The above-mentioned increased risk can be attributed to conditions and causes of death specific to SCI. Lidal et al.³⁷ highlighted that point reporting much higher mortality among both males and females due to urogenital disease, but not for ischemic heart disease. Indeed, the mortality due to this last condition was comparable to that of the general population.³⁷ Finally, this finding also emphasizes the role of comorbidities associated with SCI in mortality.

Altogether, the studies indicate that individuals with SCI have a lower life expectancy and a higher mortality rate compared to the general population. The severity and level of the injury play a significant role in determining life expectancy, with individuals with paraplegia having a higher life expectancy than those with tetraplegia. Healthcare professionals and policymakers may use these findings to inform treatment decisions and develop interventions to improve the health outcomes of individuals with SCI. Noteworthy, in the existing literature, there is some degree of uncertainty around the estimates of life expectancy/standard mortality ratios due to the differences in study characteristics such as inclusion/exclusion criteria, male-to-female ratio, and the proportion of tetraplegics. This uncertainty is also due to the scarcity of data and to different quality of care, which are related to the country of origin. For instance, a study carried out by Sabre et al.³⁸ in Estonia reported a disproportionately high SMR compared to other countries included in the study. However, Estonia has a lower socioeconomic status and 75% of the deaths that occurred within the study population were among people younger than 60 years. Since the SMRs were calculated using a life table based on the entire European Union, this could explain the observed inconsistency in SMRs.³⁸

Finally, the literature reflects the robust evidence of a higher risk to die for tetraplegic compared to paraplegic, supporting what was observed in the previous reviews conducted by van den Berg et al.⁴⁴ and Chamberlain et al.⁴³

Conclusions

According to recent research, individuals with SCI tend to have a shorter life expectancy compared to the general population. Noteworthy, it is fundamental to recognize that two factors play a crucial role in this reduction, namely the characteristics and localization of the lesion and the development of comorbidities. These two factors are related to the quality of the care provided during inpatient rehabilitation and discharge process, as it is well-known that specialized care and the availability of units specialized in the treatment of SCI are crucial to mitigate the risk of secondary conditions and premature mortality. The quality of care is, in turn, linked with the geographical location and different systems of health management. As a result, these variables and several other factors may contribute to a shorter life expectancy, as an outcome after SCI. Despite the importance of the topic, it remains challenging to achieve general conclusions due to the methodological differences across studies. This absence of methodological standardization makes it hard to compare the results of the examined studies.⁴³ Unfortunately, the current recommended guidelines for reporting on observational studies (STROBE) or guidelines suggested by the International Spinal Cord Society (ISCoS) are only followed by a few journals in the field and are rarely adhered to by authors themselves. This results in inconsistent quality of reporting.^{73, 74} Discrepancies or variations in study results can be due to different sample sizes, different inclusion/exclusion criteria, different lengths of follow-up, and sample population characteristics (e.g., the proportion of tetraplegic and paraplegic patients, mixed age population, the proportion of ventilator-dependent patients, etc.). Several studies have excluded participants based on age, but there were no consistent criteria for age exclusion across the studies. This led to difficulties in comparing the results as the studies did not all begin follow-up at the same time after spinal cord injury (SCI), nor did they all include deaths that occurred within the first year after SCI. As a result, the true standardized mortality ratio (SMR) for the outcomes being considered may have been underestimated.¹⁴ Moreover, different definitions of in-hospital mortality and/or different methods to document it may contribute to observed heterogeneity.

Most of the examined literature focuses on people affected by SCI residing in high-income countries such as the US and EU, therefore these results lack external validity. Several differences in healthcare structure, social welfare, and socioeconomic status of individuals depend on the considered population’s residing country and have been shown to influence life expectancy.

Therefore, further methodologically homogeneous studies are necessary to assess mortality and life expectancy, stressing once more the conclusion drawn by Chamberlain et al.⁴³ These studies should employ STROBE guidelines⁷⁵ and “Standardization of Data for Clinical Use and Research in Spinal Cord Injury.”⁷⁶ Also, it would be desirable to obtain more data from developing countries. Finally, a future meta-analysis is needed to statistically quantify the data summarized in this review and give new insight into life expectancy after spinal cord injury.

Supplementary Digital Material 1

Supplementary Table I

Characteristics of the studies included.^{13-15, 17-20, 29-40}