Cervical spine (c-spine) injuries are major health hazards that have catastrophic effects on the patients, with unpredictable outcomes ^[1]. Depending on the level and severity of spinal cord injury (SCI), the patient can have varied physical and functional limitations. The management and rehabilitation of these injuries not only affects the patient and his family but the socioeconomic balance of society as a whole. The grades of recovery of neuromotor deficit, as well as the associated morbidity and mortality, are unpredictable and several factors are responsible for it. Whether to operate or not, the timing of surgery and the nature of surgery are some of the controversial questions to which the answers are not clear. Factors such as early decompression have been reported to improve outcomes and diminish mortality by several reviews and articles [2,3]. However, the role of other factors such as demography, level of injury, initial grades of weakness, magnetic resonance imaging (MRI) changes, and vitals is not very clear in terms of their effect on neurological and functional improvement. The present systematic review of the literature was conducted to assess the role of such factors in the overall prognosis of patients having traumatic cervical spine injury.

Search methodology

The search was conducted based on PRISMA guidelines and a checklist ^[4]. Three databases, PubMed, EMBASE, and SCOPUS, were searched on April 12^th, 2021, using the keywords as depicted in Table 1. A total number of 2313 hits were obtained. We also performed a secondary search from the references from all the articles selected as per the predefined criteria.

Inclusion and exclusion criteria

Studies of any design from the last 15 years evaluating multiple factors related to outcomes in traumatic cervical spine injuries, including mortality, were all assessed. Cadaveric studies, conference abstracts, case reports, and any studies that included non-traumatic cases or other spinal levels were all excluded. We also excluded non-English articles and systematic reviews (Fig. 1).

Data collection and analysis

Two reviewers (P.K. and R.K.R) independently screened the studies. The title of the present study was utilized to assess the articles that seemed fit for inclusion, and their abstracts were read. In case of any discrepancies during screening, full texts were studied. The articles that pertained to the study question were identified, and finally, these short-listed articles were included in the review for systematic analysis. We did not do the statistical meta-analysis because of different and multiple variables assessed by the individual articles making the data pool and assessment improbable. Any conflicts between the two authors were resolved by discussion involving the other co-authors to arrive at a final consensus. Data extracted were collected and registered on a structured form that included names of the authors and the journal, year of publishing, demographic parameters such as age, sex, and the number of patients, and the other factors assessed by individual studies, such as lesion type (complete or incomplete), American Spinal Injury Association (ASIA) grade, level of cervical spine injury, MRI findings, use of steroids. This was summarized in a tabular form (Table 2).

Search and screening results

A total of 73 studies were identified by the word search on all three databases [5-11]. After initially assessing the titles for assessment of the role of multiple variables in the clinical outcome of cervical spine injuries, relevant studies were identified. Abstracts were read of all these studies, and finally, after excluding the repetitive or duplicate studies, 13 studies were further identified, and full texts were read, out of which seven studies were included in the review. These studies have been published between the years 2003 and 2019. Six out of the seven studies are from the last 5 years, 2013–2017 [6-11].

Study characteristics

The majority of the studies are retrospective in design, with two prospective studies done recently in 2016–2017. The minimum and maximum sample size in the included studies is 37 and 412, respectively [5, 6]. The role of age, sex, level of injury, use of intravenous (IV) steroids, MRI findings, pre-operative ASIA grading, treatment options, timing of surgery, blood pressure, Subaxial Cervical Spine Injury Classification (SLIC) score, and requirement of ventilator support are all evaluated by one or more individual studies. Daneshvar et al. evaluated the effect of these variables on the in-hospital mortality of elderly patients with a mean age of 75 years ^[6]. Liu et al. determined the 8-year mortality rates in their study population by Kaplan–Mier analysis and the effect of various factors on this outcome ^[11]. Burke et al. studied the effect of the timing of surgery on the outcome by dividing patients into three groups ^[10]. The remaining four studies evaluated relatively younger patients under 45 years of age with outcomes defined as improvement in ASIA grades and scores [5, 7-9].

Factors assessed (Table 3)

Sex

Sex as a factor in the outcome after cervical spine injuries was studied by three of the authors included in our review [6, 7, 11]. The majority of the patients included in these studies are males, most likely because males have higher involvement in road traffic accidents and workplace injuries ^[12]. Daneshvar et al. included 28 males and eight females in their study group but found no significant role of gender in terms of in-hospital mortality (odds ratio [OR] = 1.02, P = 0.98) ^[6]. There were 58 males and 15 females in the study by Park et al., who did not find any significance of the role of gender in the neurological recovery of the patients (OR = 1.22, P = 0.73) ^[7]. Similar to the other studies, Liu et al. also did not find any significant role of gender in 8-year mortality (38.8% in males vs. 40.4% in females, P = 0.667), which was the primary outcome evaluated in the study ^[11].

On overall review, we get a clear idea that the role of the sex of a patient is not significant in the ultimate recovery or mortality after these injuries.

Age

Theoretically, with advancing age, the prognosis should be inferior in view of decreased body reserves and comorbidities. Daneshvar et al. divided their patients into two groups of more than and <80 years of age ^[6]. The mean age was 75 years. Although the total in-hospital mortality was 38%, age was not a significant factor (OR = 1.82, P = 0.51). The mean age of the patient cohort in the study by Park et al. patients was 44.23 years (range, 14–83) ^[7]. At 3 months postoperatively there was no statistically significant difference in improvement in ASIA grades in traumatic cervical spine injury patients managed with decompression, with respect to the age of the patient (OR = 1.00, P = 0.87). The study by Srinivas et al. included patients with a mean age of 35. 13 years also failed to show any direct impact of the age of patients on their recovery ^[9]. In patients with age <30 years, 31.6% showed recovery, whereas 60% of patients above 30 years showed improvement (OR = 0.31, P = 0.11). Although this finding was not significant statistically, a higher number of patients in the older age group showed improvement, and this was attributed to the younger age group possibly sustaining higher energy traumas. Similarly, Cao et al. divided patients into functional restoration and non-functional restoration groups based on neurological recovery ^[8]. They found no significant difference in outcomes based on age (29.4 vs. 31.9, P = 0.14). On the other hand, the study by Pollard et al. showed a direct impact of age on neurological recovery in incomplete injuries ^[5]. The mean age of their patients was 36 years. They observed that patients of <18 years of age are more likely to show motor recovery than older groups (P = 0.002) but found no difference in sensory recovery. Likewise, in the study by Liu et al., a highly significant difference was noted (P = 0.002) in the 8-year mortality rates between different age groups, with higher age groups showing increased mortality ^[11]. This mortality rate for the older age group (>50 years) was 50.2%, whereas that in the younger age group (<50 years) was 32.4%. In our review, we found conflicting reports on the effect of age on the prognosis of cervical spine injury. Because of the heterogeneity in the mean age of different studies, definite conclusions cannot be drawn from this data.

Level of injury

It is expected that the higher the level of cervical spine injury, the worse would be the outcome because of the involvement of higher nerve roots that are responsible for respiration. However, Park et al. found no significant difference in neurological recovery in injuries above and below C5 levels (OR = 0.68, P = 0.42) ^[7]. Cao et al. also found no significant difference in injury segment between the two groups (P = 0.96) ^[8]. Srinivas et al. showed that in patients with incomplete injuries, 71% of those with injuries between C5–C7 improved compared to 60% of those between C2–C4 ^[9]. Although this was not statistically significant, the trend of more improvement in lower levels could be ascertained. Daneshvar et al. also found that injuries above C4 had 7 times the risk of in-hospital mortality when compared to injuries below C4 (OR = 7.1, P = 0.01). Similar findings were reported by Liu et al., who observed that injuries at lower cervical levels (C5–C8) had significantly lower mortality rates than injuries at higher cervical levels (Relative risk [RR] = 0.487, P < 0.05) ^[11]. Findings from these studies do appear to confirm the theory that injury to lower cervical levels might have improved outcomes and a better prognosis.

IV steroids

The role of IV steroids has been a big debate in the management of SCI among neurosurgeons and orthopedicians. NASCIS II guidelines suggest a favorable role, whereas the latter studies do not [13,14].

Similar findings were reported by the three studies in our review, which tried to assess the effect of steroids on overall neurological outcomes after cervical spine injuries [5, 7, 8]. Pollard et al. found no significant difference in motor recovery with or without the use of methylprednisolone ^[5]. They did show a 1-point increase in initial sensory recovery with steroids, but ultimate final scores were equivalent. Park et al. found no role of IV steroids on neurological recovery (OR = 0.46, P = 0.123) ^[7]. Cao et al. studied the effect of giving steroid therapy within 8 h of admission ^[8]. They found a significant difference in univariate analysis (P = 0.02), but the significance was not seen in multivariate analysis (P = 0.07). With this review, we can arrive at a consensus that the administration of IV steroids does not have any significant impact on outcomes in cervical spine trauma.

MRI findings

Three studies assessed the role of MRI findings in the overall neurological recovery of traumatic cervical spine injuries [7-9]. Park et al. found no significant role of MRI changes on the neurological recovery associated with cervical injuries, both with respect to high signal intensity (OR = 0.20, P = 0.053) and maximum cord compression (OR = 0.99, P = 0.967) ^[7]. On comparison between patients having edema and contusions on MRI, Srinivas et al. observed that patients with edema were more likely to improve, with 65.2% of such patients showing improvement compared to none in the contusion group (P = 0.001) ^[9]. In addition, when <3 segments were involved, 75% of patients showed improvement as compared to only 42% in patients with involvement of more than 3 segments (P = 0.003). Three segments consisted of adjacent vertebrae and intervening discs on mid-sagittal cuts. Cao et al. observed that the length of cord edema did not have any significant effect on functional improvement in patients with cervical SCI (P = 0.27) ^[8]. Maximum spinal cord compression (MSCC) had a highly significant impact, with lower compression observed in the functional restoration group (P = 0.00).

ASIA grade

Complete lesions (ASIA A) are known to have worse outcomes. Park et al. showed a significant role of ASIA grades in neurological recovery. ASIA A has the worst outcome, with grades B, C, and D having better outcomes (OR = 5.00, 7.31, 1.76, P = 0.027) ^[7]. Daneshvar et al. had 15 patients with ASIA A, two of B, six of C, and 14 of D. There was a trend of more mortality in grades A/B when compared to grades C/D. However, this did not reach statistical significance (OR = 3.3, P = 0.09). When complete injuries were compared with incomplete injuries, complete injuries were found to have significantly higher mortality (OR = 5.1, P = 0.03). Srinivas et al. had no patients with complete injuries showing any improvement, whereas Pollard et al. excluded such patients from their studies in the context that such injuries are prone to have the poorest outcomes [5, 9]. Furthermore, only one patient with ASIA B showed improvement, while the maximum improvements were with ASIA D followed by ASIA C, with 83% and 78% of patients improving. Even the Functional score (FIM) showed maximum improvement in ASIA C ^[9]. Lie et al. included only ASIA A and B in their study cohort and observed high 8-year mortality rates of 40.7% and 35.4%, respectively ^[11]. Overall, findings from multiple studies in our review confirm that higher ASIA grades and complete injuries have a significant association with worse outcomes.

Treatment

Four of the studies in the review included patients who were operated on and assessed the factors post-surgery [7-10]. The other three treated some with decompression and some conservatively [5,6,11]. Decompression enhances remaining neurological function and prevents secondary injuries [7, 15]. However, the efficacy of bony decompression as compared to additive durotomy and duraplasty is a matter of debate and could be a factor leading to no difference when compared to conservative treatment in the included studies [16, 17]. Daneshwar et al. operated on 30 out of 37 patients but could not find any role of the method of treatment on mortality in the elderly (OR = 1.3, P = 0.76). This is an important finding, especially in the elderly who are often associated with medical comorbidities where surgery could be complicated ^[6]. Similarly, incomplete injuries and cervical stenosis managed conservatively when compared with decompression showed similar outcomes in the study by Pollard et al. ^[5] Liu et al. found that patients who were treated conservatively had significantly higher 8-year mortality rates than those who were treated with surgery (58.3% vs. 28.3%, RR = 0.380, P = 003) ^[11]. Most authors in our review have operated on most patients or treated those patients conservatively who had significant comorbidities or complete injuries with poor prognoses. The available data seems to suggest that surgery is the preferred method of treatment for cervical spine injuries both for decompression and stabilization, which can help in early mobilization.

Timing of surgery

The timing of surgery has been studied by multiple researchers who suggest that the neuroprotective efficacy of decompression is enhanced by early intervention [2, 3, 18]. However, some studies disapprove of this theory ^[19]. Even the definition of early and late surgery is not clear, with different studies defining early as within 1 day, 2 days, 3 days, or even within 1 week [5, 20]. In the present review, Daneshvar et al. showed no difference in mortality between early and late operated groups, defined as within 2 days and beyond that (OR = 1.1, P = 0.88) ^[6]. Similarly, Park et al. did not find any difference between early (<24 h) and late decompression (OR = 1.31, P = 0.590) ^[7]. Liu et al. did not find any significant difference in mortality rates between patients received early (0–3 days) and midterm surgery (3–7 days). They found higher mortality rates in patients who received late surgery but statistical significance was not reached, most likely due to the small sample size ^[11]. Srinivas et al. defined early decompression within a week of injury and did not find any significant difference in the outcomes with 50 % of patients operated early and 44.8 % operated late showing neurological recovery ^[9]. Burke et al. divided the patients into three groups based on the timing of surgery – ultra-early (<12 h), early (12–24 h), and late (>24 h). The improvement in ASIA grade was 1.3 for the ultra-early group as compared to 0.5 for the early group (P = 0.02). There was no significant difference between the early and late groups (P = 0.76) and no difference when traditional early groups (0–24 h) were compared with the late group (P = 0.40). They recommended surgical decompression within 12 h as a way to improve neurological outcomes ^[10]. Cao et al. also found that the functional restoration group had significantly lower time from injury to operation than the non-functional restoration group (1.2 ± 0.5 vs. 4.4 ± 2.0) on both univariate and multivariate analysis (P = 0.00) ^[8]. Pollard et al., on the other hand, showed better outcomes with delayed anterior decompression for cervical spine incomplete injuries. This is in contrast to most evidence in the literature and could be because of the small sample size as per the authors, and thus, they gave no definite conclusion. The timing of surgery has been one of the most controversial debates with regard to the management of cervical spine injuries. Some surgeons believe that a patient needs to be stabilized and then taken up for surgery, whereas others recommend immediate decompression. It is also affected by the resources available, something which is more of an issue in developing countries. In our review, we feel that the aim of treating surgeons should be early decompression but the patient’s overall condition and resource availability has to be kept in mind. More studies are needed from developing nations to ascertain the extent to which the timing of surgery affects neurological outcomes in these patients and how surgery can be expedited for the same.

SLIC score

Cao et al. did a scoring of all patients by the SLIC and found that it had a highly significant effect on the functional improvement in their study; lower scores in the functional restoration group (P = 0.00) ^[8]. ROC curve was constructed, and Area Under Curve was 0.91 for a SLIC score of more than 7.5. Daneshvar et al. found no role in the presence/absence of dislocation on the in-hospital mortality in their patient cohort (OR = 2.4, P = 0.27) ^[6].

 Ventilator dependence

The study by Liu et al. on 8-year mortality studied the effect of the requirement of ventilator requirement. They found that in-hospital ventilator support had a significantly higher mortality risk than those who did not need it (RR = 2.397, P < 0.05). No other study in the review studied this factor.

Blood pressure on admission

Only Park et al. assessed these variables. The AANS/CNS guidelines recommend avoiding hypotension with systolic blood pressure (SBP) <90 mm Hg in the 1^st week of injury with maintenance of mean arterial pressure (MAP) of 85–90 mm Hg. The study showed that mean MAP during 1^st week is not a significant factor in the outcome, but SBP and MAP values at admission correlated significantly with the recovery (P = 0.004 and 0.003). However, ASIA A patients showed no improvement and relation to these values ^[7].

C-spine injuries with neurological deficits are often associated with a high number of complications, primarily because of respiratory muscle involvement and subsequent respiratory failure. In addition, immobilization due to neurological deficit can lead to decubitus ulcers, deep venous thrombosis, and pooling of respiratory secretions, which in turn lead to pulmonary infections [21, 22]. The primary aim of surgery and stabilization in these patients is early mobilization on a wheelchair which in turn helps in avoiding the above-stated complications. Adequate chest physiotherapy and rehabilitation are of utmost importance in the recovery of patients with cervical spine trauma. There has been extensive research to ascertain the factors that could aid in predicting the prognosis of these injuries. Previous work has focused mainly on mortality, treatment options, and functional outcomes; however, the literature on the role of one or more factors in predicting these rates is scarce [22, 23]. We specifically focused on cervical trauma as these tend to be the most catastrophic with the involvement of all four limbs and looked into the available evidence that could support the role of various demographic, injury, and treatment-related factors in the ultimate outcomes and mortality of these patients. Theoretically, a more severe and complete injury at a proximal cervical level should be associated with worse outcomes because proximal injuries would have a greater impact on respiratory function, thus leading to more mortality. Findings by Daneshvar et al. and Liu et al. confirm these presumptions regarding mortality associated with injury to higher cervical levels [6, 11]. With regards to the severity of the injury, the findings of most authors in the review showed poor functional improvement and high mortality with higher ASIA grades, thus giving weight to the assumption that the severity of injury directly correlated with outcomes [6, 7, 9, 11]. Despite multiple studies which have given evidence against the benefit of IV steroids in SCI, these continue to be used extensively and indiscriminately across the world. Unrestricted steroid use can lead to immunosuppression, which in turn can predispose the patient to nosocomial infections, most predominantly fungal. The studies included in our review also confirm that IV steroids have no significant role in the prognosis and recovery after cervical spine injuries [5, 7, 8]. Surgical decompression and stabilization is the treatment of choice for patients with cervical spine injury, and this has been echoed by the authors in the review. Conservative management should be reserved for patients who present with respiratory complications or those who are at high risk of post-surgery complications. Liu et al. also confirmed this bias when they observed that patients who were ventilator-dependent had significantly higher mortality after surgical management ^[11]. The best time for surgery is often debated because many studies show that when the surgery is done after injury, can greatly affect the patient’s final recovery and results. [2, 3, 10, 18, 19]. According to Pollard et al. theory behind early decompression is that the neuroprotective effect is enhanced by reducing the time between injury and surgery ^[5]. Burke et al. tried to differentiate between ultra-early surgery (<12 h), early surgery <12–24 h) and late surgery (>24 h) ^[10]. In literature, the definition of early decompression is variable with periods from 1 day to 1 week being considered early by different researchers ^[20]. In the present review, we observed conflicting evidence by the different authors on the role of the timing of surgery on neurological outcomes and mortality. However, because most studies are retrospective and with a small number of patients, further prospective large-scale evidence is required. This evidence also needs to be tailored according to the resources available at any given center to clearly define protocols and streamline management. Due to varied definitions of the early surgical category, we recommend a formalization of this category to <24 h, for further studies. Age and sex are the most common demographic factors that have been assessed for their impact on outcomes of cervical spine trauma. No role of sex has been ascertained by the various authors, and the evidence in our review is conclusive [6, 7, 11]. However, there is yet no consensus on the role of age in prognosis and results after surgery for these injuries. Older age group patients having medical comorbidities are prone to increased risk of complications in the perioperative period. Daneshvar et al. did not find any difference in mortality between patients above and below 80 years ^[6]. Pollard et al. showed that younger patients have better outcomes in incomplete cervical spine injuries, and Liu et al. showed lower mortality in younger age groups [5, 11]. A well-designed, prospective study is needed to provide a categorical role of age in cervical spine injuries. Another factor for which conflicting reports have been seen in the literature is the role of MRI findings on the recovery from these injuries. There is limited research regarding this, and the data available suggests that the amount of cord compression and the length of the cord involved plays a role in the severity of the injury and ultimate outcome [8, 9]. However, there are studies that have found no significance of MRI ^[7]. This is another factor which needs further research to ascertain its role. Maintenance of blood pressure during the early period post-injury has been shown to be a vital entity. Hypotension is not advisable by AANS/CNS guidelines on the management of cervical spine injuries ^[24]. Higher MAP values during the early period post-injury have been shown to have a better recovery ^[25]. Park et al., however, only found SBP and MAP at the admission of value in terms of prognosis of incomplete injuries ^[7]. None of the other included studies discussed this aspect, and this is a factor which needs further evaluation, especially because it can be modified to improve outcomes based on recommendations. Overall, this review evaluates the role of multiple and diverse factors on the outcomes of cervical spine injuries treated with decompression. Although the number of studies in the literature is limited, there is conclusive evidence on the role of level of injury, ASIA grade, IV steroids, and surgical decompression in the management of these injuries. However, there are still factors such as age, timing of surgery, and role of blood pressure and MRI findings that need further research. With this limited data from these seven studies, this cannot be considered concrete evidence and we recommend further prospective large multi-centric research to evaluate these factors further.

c-spine injuries require decompression and stabilization to aid in physiotherapy and rehabilitation. Complete injuries and proximal levels of injury above the C4 vertebrae are associated with the worst outcomes. There seems to be a limited role of IV steroids in recovery after cervical trauma. Advanced age and medical comorbidities should not be a hindrance to operate, and maintenance of high mean arterial pressure and systolic pressure initially could have favorable results. The role of the timing of surgery and MRI findings need to be evaluated further through prospective large-scale trials to provide a high level of evidence and to help reach a consensus.

This study emphasizes the need for timely and successful surgical intervention in cervical spine injuries, especially for patients with low ASIA-grade lesions below C5. It emphasizes the importance of conducting robust prospective studies to assess potentially modifiable parameters, including age, MRI results, and surgery scheduling to improve clinical outcomes.