The treatment of spinal injuries has evolved from a predominantly nonoperative field to a subspecialty within spinal care, offering a wide range of surgical and nonsurgical options. Nonsurgical care generally includes variable periods of recumbency and some form of external orthosis or cast. Surgical care has aimed to improve on nonoperative spine-specific outcomes by more reliably restoring physiologic spinal alignment, providing a more optimal environment for improvement of neurologic outcomes and facilitating early mobilization and reactivation. Perhaps most importantly, early mobilization is thought to probably decrease morbidity and mortality, especially in the multiply injured patient.
The advent of modern trauma care with its rationally derived management algorithms, such as the Advanced Trauma Life Support system, initially relegated surgical care of the injured spine to a more secondary role after management of cranial, torso, and musculoskeletal trauma. With the evolution of early mobilization as a means of potentially reducing morbidity and mortality, there has been increased interest in the role of early surgical stabilization of spine fractures. Early surgical intervention may, indeed, have a number of benefits: (1) improved pulmonary function in patients with concurrent chest trauma by allowing for more rapid mobilization without bracing; (2) reduced pain induced by spinal column instability and, thus, shortened intensive care unit (ICU) care and hospital length of stay (HLOS); and (3) decreased mortality by reduction of incidence and severity of sepsis and respiratory failure. Furthermore, the role of early surgical decompression in aiding neurologic recovery from spinal cord injury (SCI) continues to receive extensive attention in basic scientific and clinical research, without a clear answer to date.
Despite the many advantages of early surgery for spine trauma, potential downsides do exist. There is concern that a second, early physiologic injury may have a detrimental effect on various associated injuries, such as to the lungs and brain. This problem may be compounded by the potential for increased hemorrhage, with associated hypotension, which may also be deleterious to the recovery of associated SCIs. Other disadvantages include operating in a setting of potentially missed or underestimated associated injuries, operating without a full understanding of the intricacies of a specific injury, and operating under less-than-ideal conditions relative to the complexity of the surgery and resources required. Circumstances of this nature may lead to over- or underutilization of surgery.
In this systematic review, we concentrated on the timing of surgical intervention for spine trauma relative to its impact on morbidity and mortality of patients with thoracolumbar (TL) fractures, with and without associated injuries. Outcome variables included time to discharge, ICU length of stay (ICULOS), incidence of postoperative infection and other complications, and mortality. Unfortunately, there is a lack of consensus regarding the definition of acute or early intervention in spine trauma. Time to early intervention has been described as ranging from anywhere between 8 and 72 hours.1,6,9,10 The absence of consistent timelines and the need to control for comorbidities, overall injury burden, and other confounders present formidable challenges to precise scientific analysis. Despite these limitations, we felt it important to answer the following question: does early stabilization in TL spine trauma decrease morbidity and mortality? Although this question was addressed for both neurologically intact and neurologically compromised patients, the potential role of early surgery in neurologic recovery is a separate topic, which we did not specifically address.
Materials and Methods
Electronic Literature Database
The literature search is outlined in detail elsewhere.10a A systematic search was conducted in Medline, EMBASE, and the Cochrane Collaboration Library for literature published from January 1990 through December 2008 on the effect of fixation timing for TL spine fractures. We limited our results to humans and to articles published in the English language. Reference lists of key articles were also systematically checked. We included articles that compared early with late surgical stabilization in adults who sustained a blunt destabilizing TL injury. We excluded studies that had neurologic recovery as their main outcome, articles that contained 10 or fewer patients in each treatment group, and those articles that evaluated the timing of surgery in cervical fractures. Outcomes of interest included the number of ventilator days, HLOS and ICULOS, respiratory complications, and mortality and morbidity (Figure 1).
Each retrieved citation was reviewed by 2 independently working reviewers (D.C.N. and J.R.D.). Most articles were excluded on the basis of information provided by the title or abstract. Citations that seemed to be appropriate or those that could not be excluded unequivocally from the title and abstract were identified, and the corresponding full-text reports were reviewed by the 2 reviewers. Any disagreement between them was resolved by consensus. From the included articles, the following data were extracted: patient demographics, spine segment injured, injury severity, timing of stabilization, and patient safety outcomes.
Level of evidence ratings was assigned to each article independently by 2 reviewers using the criteria established by The Journal of Bone and Joint Surgery, American Volume (J Bone Joint Surg Am)16 for prognostic studies, and modified to delineate criteria associated with methodologic quality and described elsewhere (See Supplemental Digital Content 1, individual study ratings, tables, individual study ratings, available at: http://links.lww.com/BRS/A413).
Patient safety outcomes were reported as the proportion of patients experiencing death or respiratory complications or the mean number of days spent in the hospital, on ICU, and on a ventilator. Data were summarized in tables, and qualitative analysis13 was performed considering the following 3 domains: quality of studies (level of evidence), quantity of studies (the number of published studies similar in patient population, condition treated, and outcome assessed), and consistency of results across studies (whether the results of the different studies lead to a similar conclusion).15 We judged whether the body of literature represented a minimum standard for each of the 3 domains using the following criteria: for study quality, at least 80% of the studies reported needed to be rated as a level of evidence I or II; for study quantity, at least 3 published studies were needed, which were adequately powered to answer the study question; for study consistency, at least 70% of the studies had to have consistent results. The overall strength of the body of literature was expressed in terms of the impact that further research may have on the results. An overall strength of “high” means that further research is very unlikely to change our confidence in the estimate of effect. The overall strength of “moderate” suggests that further research is likely to have an important impact on our confidence in the estimate of effect and, may therefore, change the estimate. A grade of “low” means that further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate, whereas “very low” means that any estimate of effect is very uncertain.10a,13–17
We identified 68 articles from our literature search evaluating outcomes associated with the timing of stabilization of TL spine fractures. From these potential articles, we judged 19 to undergo full-text review. After full-text review, we excluded 10 of the articles for the following reasons: 3 articles contained data from cervical fractures only; 2 articles reported on neurologic recovery rather than complications; 3 articles included results from all spine levels, but did not stratify results by spine level; 1 article provided no data; and 1 article had <10 patients in the late treatment group (see Figure 2 for detailed results of the literature search). More information on excluded articles can be found in Table 1 (Supplemental Digital Content, available at: http://links.lww.com/BRS/A413). The remaining 9 articles provide a comparison of early versus later fracture stabilization on patient safety (Table 1; Supplemental Digital Content, available at: http://links.lww.com/BRS/A413).
One study was graded as level II, whereas the remaining 8 studies were graded level III (Table 2; Supplemental Digital Content, available at: http://links.lww.com/BRS/A413). One study was a quasirandomized control trial,1 1 was a prospective cohort study,10 and the remaining 7 were retrospective cohort studies, with data obtained from registries.2,3,5–7,10,12 The quasirandomized control trial did not have concealed allocation nor was it adequately powered to detect important differences in complications. It also excluded patients with multisystem injuries and its inclusion criteria included “patients who did not have an obstacle for urgent surgery,” limiting its usefulness in determining the role of early surgical intervention in the multiply injured patient. The prospective cohort study lacked independent assessment of complications, did not control for potential confounding factors, had an inadequate sample size to detect important differences, and had no description of whether cointerventions were applied equally in both early and late stabilization groups. Only 1 study controlled for an array of possible confounding factors.9
Hospital and Intensive Care Length of Stay
Of the 9 eligible articles, 71–3,5–7,12 report on the HLOS and ICULOS. Of these 7, 4 present data separately for thoracic and lumbar fractures, 1 includes only thoracic fractures, and 2 limit the population to TL junction fractures. The range of the mean HLOS in those with thoracic fractures range from 10 to 39 days in the early group compared with 14 to 43 days in the late group (Table 2; Supplemental Digital Content, available at: http://links.lww.com/BRS/A413). The range of the average HLOS for patients with TL junction or lumbar fractures was 8 to 13 days for the early group compared with 14 to 26 in the late group. In all studies, the HLOS was significantly lower in the early group compared with the late group. The mean ICULOS ranged from 2 to 13 days in the early group compared with 5 to 20 days in the late group in patients with thoracic fractures; P < 0.05 in most studies, reporting this outcome (Table 2; Supplemental Digital Content, available at: http://links.lww.com/BRS/A413). In patients with TL junction or lumbar fractures, the range of average ICULOS was 0 to 4 days in the early group and 0 to 10 days in the late group.
Respiratory Morbidity and Mortality
Morbidity, as measured by the mean number of days on a ventilator and the proportion of patients with respiratory complications, and mortality are displayed in Table 3. Four studies report the number of ventilator days for thoracic and lumbar fractures,3,5–7 and 1 study reports on thoracic fractures only.12 Two reports from Kerwin et al (1 in 20056 and 1 in 20075), using overlapping data, report no statistical difference overall in mean ventilator days between the early and late groups among patients with thoracic fractures (2 days compared with 4 days). However, Kerwin et al's 20056 study identified significantly lower mean ventilator and ICU days in patients with SCI whose fractures were stabilized within 72 hours, with the greatest effect of early surgery being seen in patients with SCI with thoracic fractures. Kerwin et al's 20075 study defined early stabilization as <48 hours rather than 72 hours, which may explain, in part, why this difference was not borne out in the latter study. Three other studies reported statistically significant differences between treatment groups with mean number of ventilator days ranging from 1 to 7 days in the early group compared with 5 to 13 days in the late group. There was no statistical difference in the mean number of ventilator days in patients with lumbar fracture in any of the 4 studies reporting this outcome in lumbar fractures with the mean number of days ranging from 1 to 2 days in the early group and 1 to 4 days in the late group.3,5–7
Morbidity resulting from respiratory complications was less in patients with thoracic fractures when stabilized early versus late. Pneumonia was reported less frequently in those with thoracic fractures stabilized early; 3% versus 37% by Croce et al3 (P < 0.001) and 7% versus 33% by Kerwin et al6 (P = 0.029), in patients with SCI.6 However, the patients in the early and late groups were not statistically matched relative to injury characteristics and, most notably, those in the late group had significantly higher chest abbreviated injury scores at the time of presentation. Lung failure occurred in 19% of the early stabilization patients compared with 31% in the late group in the study by Schinkel et al.12 A subanalysis of their data reveals that the differences between the early and late stabilization groups depend on how severely the patient is injured. The late stabilization group had significantly higher rates of lung failure when their injury severity score (ISS) was > 26 (17% vs. 41%, early vs. late in those with ISS from 26 to 38), whereas the late stabilization group that had a score over 38 had significantly higher ventilator days (median, 5 vs. 15 days, early vs. late). The implication of their study was that the sickest patients benefit most from early surgical intervention.
In a group of patients sustaining thoracic or lumbar fractures, early stabilization within 48 hours resulted in half the respiratory failure after controlling for age, lung injury, and Glasgow Coma Scale score.9 There were no statistical differences in respiratory complications in the 2 studies evaluating TL junction fractures,1,2 keeping in mind that 1 of these studies excluded multiply injured patients1 or in the 2 studies reporting on lumbar fractures.3,5–7
Mortality rates vary among studies ranging in thoracic fractures from 0% to 11% in the early group and 0% to 17% in the late group (Table 3). Schinkel et al12 reported a statistically significant lower mortality in patients treated within 72 hours compared with those treated after 72 hours, 6.2% versus 17%. When stratified by ISS, those treated after 72 hours continued to have higher mortality despite injury severity (ISS <26, 3% vs. 13%; ISS 26 to 38, 5% vs. 9%; ISS >38, 10% vs. 27%). However, Kerwin et al,5 in 2007, reported a slightly higher mortality in patients treated within 48 hours compared with those treated later, 5.6% versus 0%. In an earlier study in 2005 using overlapping data with the 2007 study, Kerwin et al6 compared patients using a 72-hour cutoff and stratified by the presence or absence of SCI. There were no deaths in those with SCI. Among those without SCI, 2 (11%) patients died in the early group and no patients died in the late group. The sample size for this analysis included only 18 and 26 patients in the early and late groups, respectively. Because of their findings and the concern that early surgery may be associated with higher mortality, Kerwin et al7 investigated the relationship between timing of surgery and mortality in greater detail by reviewing 871 patients from a national trauma database. They found no difference in mortality in 2 groups of ISS- and Glasgow Coma Scale score-matched patients whose spine fractures had been stabilized before and after 72 hours. There were no statistical differences in mortality between groups in the 4 studies reporting on lumbar fractures3,5–7 or the 1 study reporting mortality for all TL fractures.10
The overall strength of evidence to assess whether early stabilization in TL fractures reduces morbidity is “low”; that is, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate (Table 4). The overall strength of evidence to assess mortality is “very low” such that the estimate of effect is very uncertain.
Entering this systematic review, significant limitations were anticipated. However, it was imperative to work toward the possible answer to this critically important clinical question, not only to provide direction to clinicians, management and administrators, and payers for justification, but to illustrate what must be done from a research perspective going forward to answer this question with a higher degree of certainty. Most studies are retrospective analyses of prospectively collected registry data that limit the examination of potential factors that may affect the risk of complications and death, such as hypotension at the time of the initial presentation, metabolic acidosis, and history of blood transfusion. Even when multiple prognostic factors are available in the registry, these are often not considered in the analysis. In our review, among studies that had multiple prognostic data available, only the study by McHenry et al9 controlled for 2 or more potentially confounding factors. Data registries are also prone to selection bias if the database is regional and if different hospitals apply different standards for inclusion and exclusion. Furthermore, treatment given for trauma patients is often determined by injury severity, potentially confounding results of treatment by indication.11 Although registry data can be subject to bias, affecting its internal validity, it has its strengths, particularly when evaluating safety or complications. The data are usually collected prospectively by independent coordinators and done so in a standardized fashion. They usually provide large sample sizes, which are needed to provide the power necessary in identifying rare outcomes or overcoming large variability in the data and multiple prognostic factors to answer more common questions. Registries also provide strong external validity or generalizability by studying heterogeneous populations across different regions, with different care providers and care paths.4,14 One final difficulty is adequate follow-up in this patient population. Although elective surgical spine patients tend to be a captive, compliant population to study trauma patients are often multiply injured, transferred to level 1 centers from other regions, and generally a more transient younger population, making follow-up of these patients a challenge for even the most established of research programs.
Within the context of the above limitations and past research infrastructure and resources in which to study these patients, this review has made a significant contribution in answering the question of the use of early surgical treatment in the patient with TL trauma. The results of the systematic review representing the best available literature suggest that early fixation of TL fractures probably decreases complications, reduces hospital stay, and likely facilitates early rehabilitation. No comments can be made regarding mortality from the review. It would also seem that doing early surgery in these patients is safe and does not compromise safety.
Another significant confounder in the review of large multicenter patient registries is the potential effect of a specific institution's clinical volume on patient outcomes. This issue has been illustrated in a recent publication by Macias et al.8 In their review of 4121 patients with SCIs treated at 701 centers over 7 states, they found through a complex multivariate statistical analysis that the risk of long-term paralysis was one-third lower in patients treated at a designated trauma center, and that there was a positive correlation between patient outcomes and the treating institution's surgical volume (Table 5).
When the systematic review results are blended with consensus expert opinion and related literature from early fixation in other bodily systems, the case for early fixation becomes more compelling when one considers risks, benefits to the patients, and potential cost savings to the payers. There is certainly enough consensus based on the literature and clinical expertise to study this question more rigorously in a multicenter trial, with appropriate control for important confounding variables identified from this review.
Question: Does Early Spinal Stabilization in TL Spine Trauma Decrease Morbidity and Mortality?
On the basis of low quality evidence, the systematic review suggests that early fixation of unstable thoracic fractures will reduce morbidity. Current evidence does not suggest that early fixation of lumbar fractures provides a similar clinical benefit. On the basis of very low quality evidence, no comment can be made regarding the impact of early fixation on mortality.
Patients with TL trauma should undergo early (<72 hours) stabilization of their injury to reduce morbidity and possibly mortality.
A weak recommendation is defined as something the majority of patients would have done and the majority of clinicians would recommend, but a number of the patients or clinicians would not choose to do based on certain clinical circumstances or patient preferences.
- Stabilization of thoracic spine fractures within 72 hours is safe and results in decreased respiratory morbidity and length of ICU and hospital stay, though the strength of supporting evidence in the current literature is low.
- Other than decreased length of hospital stay, no such differences were observed with early stabilization of lumbar fractures.
- There is not enough evidence to determine the effect of the timing of stabilization on mortality in thoracolumbar fractures.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.spinejournal.com).
The authors thank Ms. Nancy Holmes, RN, for her administrative assistance and Mr. Jeff Hermsmeyer, BS, for his assistance in searching the literature, abstracting data, and proofing.
1.Cengiz SL, Kalkan E, Bayir A, et al. Timing of thoracolomber spine stabilization in trauma patients; impact on neurological outcome and clinical course. A real prospective (RCT) randomized controlled study. Arch Orthop Trauma Surg
2.Chipman JG, Deuser WE, Beilman GJ. Early surgery for thoracolumbar spine injuries decreases complications. J Trauma
3.Croce MA, Bee TK, Pritchard E, et al. Does optimal timing for spine fracture fixation exist? Ann Surg
4.Hunter D. First, gather the data. N Engl J Med
5.Kerwin AJ, Frykberg ER, Schinco MA, et al. The effect of early surgical treatment of traumatic spine injuries on patient mortality
. J Trauma
6.Kerwin AJ, Frykberg ER, Schinco MA, et al. The effect of early spine fixation on non-neurologic outcome. J Trauma
7.Kerwin AJ, Griffen MM, Tepas JJ 3rd, et al. Best practice determination of timing of spinal fracture fixation as defined by analysis of the National Trauma Data Bank. J Trauma
2008;65:824–30; discussion 830–1.
8.Macias CA, Rosengart MR, Puyana JC, et al. The effects of trauma center care, admission volume, and surgical volume on paralysis after traumatic spinal cord injury. Ann Surg
9.McHenry TP, Mirza SK, Wang J, et al. Risk factors for respiratory failure following operative stabilization of thoracic and lumbar spine fractures. J Bone Joint Surg Am
10.McLain RF, Benson DR. Urgent surgical stabilization of spinal fractures in polytrauma patients. Spine
10a.Dettori JR, Norvell DC, Dekutoski M, et al. Methods for systematic reviews on patient safety during spine surgery. Spine
11.Psaty BM, Koepsell TD, Lin D, et al. Assessment and control for confounding by indication in observational studies. J Am Geriatr Soc
12.Schinkel C, Frangen TM, Kmetic A, et al. Timing of thoracic spine stabilization in trauma patients: impact on clinical course and outcome. J Trauma
2006;61:156–60; discussion 160.
13.van Tulder M, Furlan A, Bombardier C, et al. Updated method guidelines for systematic reviews in the cochrane collaboration back review group. Spine
14.Weiss NS. Therapeutic safety. In: Clinical Epidemiology. The Study of the Outcome of Illness
. New York: Oxford University Press; 1986.
15.West S, King V, Carey TS, et al. Systems to rate the strength of scientific evidence. Evidence report/technology assessment no. 47. [Prepared by the Research Triangle Institute-University of North Carolina Evidence-Based Practice Center; contract no. 290-97-0011]. Rockville, MD: Agency for Healthcare Research and Quality; 2002.
16.Wright JG, Swiontkowski MF, Heckman JD. Introducing levels of evidence to the journal. J Bone Joint Surg Am A
17.Dettori JR, Norvell DC, Dekutoski M, et al. Methods for the systematic reviews on patient safety during spine surgery. Spine