Skip Navigation LinksHome > Blogs > The Spine Blog
The Spine Blog
Friday, March 27, 2015

Many studies have demonstrated high rates of “abnormal” MRI findings in asymptomatic patients. However, most of these have been lumbar spine studies. Cervical spine studies on the topic have generally included small numbers of patients across a limited age spectrum. Recently, some spine surgeons have advocated surgery in asymptomatic patients with cervical spinal cord compression with T2 cord signal changes. To shed more light on the topic, Dr. Nakashima and colleagues from Japan performed a study in which over 1,200 asymptomatic subjects equally distributed across the age spectrum and across both genders underwent cervical spine MRI. They evaluated disk bulging (defined as over 1 mm of bulging into the canal), spinal cord compression (SCC), and increased T2 cord signal intensity (ISI). Not surprisingly, the vast majority of subjects across all age ranges had some degree of disk bulging, with the frequency of disk bulging increasing with age. Spinal cord compression and ISI were very rare in patients under 50, though over 15% of men in their 70s had SCC and 8% had ISI. Rates of SCC and ISI were lower in women.


The authors should be congratulated on obtaining funding for and successfully completing this large scale study. Such large numbers are required in order to compare trends across the age spectrum, especially when rates of SCC and ISI are relatively low. For practicing spine practitioners, these findings come as no surprise. We frequently see patients with SCC and ISI who are not myelopathic and present only with neck pain. This cross sectional study represents a snapshot in time, so while it identifies asymptomatic patients with SCC and ISI, it does provide any prognostic information. The authors performed physical exams on all patients, but these results are not presented or correlated with the MRI findings. It would be interesting to know if those with SCC and ISI were more likely to have myelopathic signs on exam. Hopefully the authors will follow this cohort to see if those with SCC and ISI develop myelopathy. Those data would be very helpful for surgeons trying to decide what to do with asymptomatic patients or those with subtle myelopathic findings who have SCC and/or ISI on MRI. Currently, we do not know how they are best treated as we do not know the natural history of their condition.


Please read Dr. Nakashima’s article on this topic in the March 15 issue. Does this change how you view the asymptomatic patient with SCC or ISI on cervical spine MRI? Let us know by leaving a comment on The Spine Blog.

Adam Pearson, MD, MS

Associate Web Editor

Friday, March 20, 2015

Readmission rate is being used as a surrogate for healthcare quality, and it seems a likely metric to be used in pay for performance schemes in the future. Additionally, readmissions will not be reimbursed in bundled payment models, so hospital systems are trying to better understand the causes of readmission in order to avoid the costs associated with it. Like most measures of healthcare quality, readmission rate is a complex statistic that is strongly influenced by patient, disease, and treatment characteristics. While many administrative databases have studied this topic, fewer single institution studies have been performed. The latter offer the possibility of more detailed analyses that can involve chart review in addition to analysis of billing codes. Dr. Akamnonu and his colleagues from New York analyzed two years of consecutive lumbar spine surgery patients in order to determine the readmission rate for lumbar surgery at the Hospital for Joint Disease and identify risk factors for readmission. Over 1300 lumbar spine surgery patients were identified and classified according to their ICD-9 diagnosis and treatment codes. Patients with codes for deformities were excluded. The authors reported 43 unplanned readmissions within 90 days of index discharge, yielding an overall readmission rate of 3.3%, which is relatively low compared to the published literature on the topic. Univariate analyses revealed that readmitted patients had significantly longer index length of stay, higher comorbidity burden, higher rates of discharge to care facilities rather than home, a higher proportion of patients with spondylolisthesis, and a higher rate of posterior-only surgery. Fusion, age, gender, and insurance type were not associated with readmission. Seventy-five percent of readmissions were for infections and wound complications.


Some of the results of this study were expected, while some were not. Higher length of stay, increasing comorbidities, discharge to rehab or another hospital, and a spondylolisthesis diagnosis would be expected to be associated with readmission as they tend to be associated with older and sicker patients and more complex procedures. Surprisingly, increasing age, fusion, and an anterior approach—which in most cases probably represented AP surgery—did not increase readmission rate. These results reflect the univariate nature of the analysis, which was not able to account for confounders. For example, patients undergoing fusion and anterior surgery were significantly younger, and likely healthier, than those undergoing decompression and posterior-only surgery. The low number of readmitted patients precluded a multivariate analysis, which would have been necessary to control for confounders and identify independent risk factors for readmission. Additionally, the diagnostic classifications were based on ICD-9 codes rather than typical clinical diagnoses. It is unclear how spondylosis differs from disc displacement/DDD. The reliance on ICD-9 diagnostic codes was problematic as review of the readmitted patients revealed that 12 were actually degenerative scoliosis patients, and these were eliminated from the analysis. However, the authors did not perform a chart review of the entire cohort, so inappropriately coded degenerative scoliosis patients who were not readmitted were left in the denominator, likely resulting in an artificially low readmission rate. While this study offered an interesting snapshot into readmissions at a single institution, the methods really were not much different from a large administrative database study. As such, the same limitations were present without the benefit of large numbers.


Please read Dr. Abamnonu’s article on this topic in the March 15 issue. Does this change how you view readmissions following routine lumbar spine surgery? Let us know by leaving a comment on The Spine Blog.

Adam Pearson, MD, MS

Associate Web Editor

Friday, March 13, 2015

The Evidence-Based Recommendations for Spine Surgery group published another  biannual literature reviews this month, evaluating articles from across the spectrum of spine care. They started with two SPORT papers reporting 8 year results from the disk herniation arm of the study. The primary 8 year paper demonstrated a persistent advantage of surgery compared to non-operative care even in long-term follow-up. A subgroup analysis showed that while the absolute degree of improvement varied significantly across subgroups, the treatment effect of surgery—the relative advantage of surgery compared to non-operative treatment—tended to vary much less. In fact, some groups with worse surgical outcomes (i.e. those with high levels of back pain and patients with symptoms for greater than 6 months) actually had a greater treatment effect of surgery due to their even worse non-operative outcomes. The authors of the current review discuss the methodological limitations of SPORT introduced by patient crossover and the observational nature of the study, yet they still hold it up as an example of a pragmatic research design that represents some of the best spine research ever done. The next article they discussed is a double-blinded, placebo-controlled randomized trial comparing the effects of perioperative gabapentin and pregabalin to placebo for lumbar discectomy patients. The paper demonstrated that the gabapentinoids significantly decreased pain scores and post-operative tramadol use without major complications related to the medications. This research design represents the pinnacle of EBM and yields conclusions that are difficult to dispute. Whether or not the results of this Indian study performed in patients undergoing inpatient discectomy are generalizable to North American or European patients treated with outpatient diskectomy remains to be seen. Another RCT included in this review is the FDA IDE study comparing 7 year outcomes between ACDF and the Prestige cervical disk arthroplasty (CDA) device. The authors report similar improvements in the neck disability index, though the CDA group had better neurological outcomes and a lower re-operation rate. The review questions if different decompression techniques were responsible for the different neurological outcomes and point out that the decision to re-operate is a very subjective outcome. They also raise the question of the role of industry funding and possible bias by authors with financial ties to the makers of the device. A study comparing the cost-effectiveness of lumbar total disk arthroplasty to intensive rehabilitation for chronic low back pain is also included. It reports widely varying results depending on what utility measure is used, indicating cost-effectiveness analysis (CEA) can yield very different results depending on the techniques employed.


All of the papers included in this review represent the spine community’s strong efforts to create high quality evidence that could potentially be used to guide clinical decision making. This review highlighted RCTs comparing different treatments for spinal disorders and demonstrates that Level 1 studies can be performed in the spine world. At the same time, the challenges in designing and carrying out experimental studies were also made clear by the review. While SPORT was designed to include an RCT, the data from the intent-to-treat analysis was very difficult to interpret due to high levels of crossover. The spine research community learned that RCTs comparing surgical to non-operative care are unlikely to be successful in healthcare environments in which patients can easily obtain care outside of the study (i.e. the United States). The cervical disk arthroplasty study, while representing a Level 1 study design, makes it clear that the interpretation of results is subjective, and conflicts of interest are likely to affect that process. As the focus on value becomes sharper, we will continue to see more cost-effectiveness analyses. The CEA discussed in the current review demonstrates that this is still a science early in its development, and subtle technical differences in the analysis can yield markedly different results. While CEA is a useful tool, the spine community should realize that multiple subjective judgments must be made while performing such a study, and caution must be used when applying the results to policy and payment decisions. The spine literature still has a long way to go in order to get us to truly evidence-based care, but this latest EBM update certainly shows that we are making progress.


Please read the latest EBM review in the March 1 issue and the accompanying editorial by Dr. Fischer. Will any of these articles change how you practice? Let us know by leaving a comment on The Spine Blog.


Adam Pearson, MD, MS

Associate Web Editor

Friday, March 06, 2015

The best treatment for central cord syndrome and timing of surgical intervention, if elected, is poorly understood. There is limited data comparing surgical versus non-operative outcomes and early versus delayed surgery. A 2010 survey of the Spine Trauma Study Group concluded that early surgery was indicated in patients with a significant deficit (ASIA C), while patients with a mild deficit (ASIA D) should be initially treated non-operatively, with the decision to operate in a delayed fashion based on their degree of neurological improvement.1 A recent paper using the National Inpatient Sample showed that while only 25% of central cord syndrome patients underwent surgery on their initial admission for the injury in 2003, the rate of surgery was 87% in 2010.2 On this background, Dr. Grauer and his group at Yale evaluated the National Trauma Databank (NTDB) to determine risk factors for mortality and adverse events in patients admitted for central cord syndrome who underwent surgery on their initial admission. Their analysis included over 1,000 patients admitted in 2011 and 2012, with a mean age of 57 years. Over 75% were men, and 55% were injured in falls. The average time to surgery was 3.5 days. In multivariate analysis evaluating risk factors for mortality, less time to surgery and a higher Charlson Comorbidity Index were independent risk factors for mortality. Every day surgery was delayed after admission was associated with a 19% reduction in the odds of mortality. Patients taken to the OR on the day of admission had a 4% chance of mortality compared to a 1% mortality rate for those taken to surgery 7 days after admission. Conversely, increased time to surgery resulted in a small but significant increase risk of minor adverse events and a trend towards a higher rate of serious adverse events.


The finding that earlier surgery was associated with increased mortality in this group is an interesting finding given that there is an increasing trend towards operative treatment, and some experts have advocated early surgery in order to maximize the chance of neurological improvement.3 Contrary to this study, delayed surgery has been shown to increase mortality in the hip fracture population, another elderly group with similarities to the central cord syndrome population.4 Before drawing strong conclusions from an administrative database study, one must consider the limitations. While mortality is accurately captured in the NTDB, severity of neurologic injury is not, so it was not possible to adjust the outcomes for this important variable. It is possible that patients with more severe neurological deficits underwent earlier surgery, and this potential confounder could explain some of the differences in mortality. Mortality is an important outcome in this population, though residual neurological deficit is the other half of the equation not evaluated by this study. If early surgery leads to better neurological outcomes, as some have suggested, some patients might prefer to take on a higher risk of mortality in return for better long-term function.5 Unfortunately, the evidence on the effect of surgical timing on neurological outcomes in central cord syndrome is very weak, so we currently cannot inform our patients that earlier surgery will improve their neurological outcomes. The authors have certainly raised the possibility that early surgery in central cord syndrome might increase the risk of mortality. If replicated in prospective studies, this finding might temper enthusiasm for early surgery, especially if there is no major neurological advantage to early decompression.


Please read Dr. Grauer’s article on this topic in the March 1 issue. Does this change your opinion about early surgery for central cord syndrome? Let us know by leaving a comment on The Spine Blog.

Adam Pearson, MD, MS

Associate Web Editor




1.            Lenehan B, Fisher CG, Vaccaro A, Fehlings M, Aarabi B, Dvorak MF. The urgency of surgical decompression in acute central cord injuries with spondylosis and without instability. Spine 2010;35:S180-6.

2.            Brodell DW, Jain A, Elfar JC, Mesfin A. National trends in the management of central cord syndrome: an analysis of 16,134 patients. Spine J 2015;15:435-42.

3.            Riew KD, Kang DG. Central cord syndrome: is operative treatment the standard of care? Spine J 2015;15:443-5.

4.            Moja L, Piatti A, Pecoraro V, et al. Timing matters in hip fracture surgery: patients operated within 48 hours have better outcomes. A meta-analysis and meta-regression of over 190,000 patients. PLoS One 2012;7:e46175.

5.            Fehlings MG, Vaccaro A, Wilson JR, et al. Early versus delayed decompression for traumatic cervical spinal cord injury: results of the Surgical Timing in Acute Spinal Cord Injury Study (STASCIS). PLoS One 2012;7:e32037.


Friday, February 27, 2015

Extension-distraction injuries in the ankylosed spine have long been recognized as unstable injuries with high rates of spinal cord injury (SCI), surgical complications, and mortality. These poor outcomes result from both the unstable nature of these fractures as well as from the advanced age of the patients and their accompanying comorbidities. Additionally, surgical stabilization tends to require long posterior constructs that can be complicated by significant intra-operative blood loss and relatively high infection rates. The literature on this topic contains primarily case series including fractures from all regions of the spine treated with a variety of surgical and non-operative approaches. Dr. Robinson and his colleagues in Sweden elected to publish their series of 41 cervical extension-distraction injuries in the ankylosed spine, including both ankylosing spondylitis and DISH patients. All patients were treated with long, posterior constructs from the upper cervical spine (C2 in most patients, with 4 requiring fixation to C1 or the occiput) to the upper thoracic spine. The 27% of patients with SCI also underwent decompression. The average age was 71, though a 95 year old was included, and the vast majority of patients were men. Twenty four percent of patients died within 1 year, and the median survival was 52 months. Not surprisingly, older patients, males, smokers, and those with SCI were significantly more likely to die. The average blood loss was over 2L, with 17L of blood loss in one case. Twelve percent of patients developed surgical site infection, though most were treated with antibiotics without the need for debridement. All patients healed their fractures, and none suffered hardware failure.


This study makes a valuable contribution to the literature on these difficult fractures. The study population and treatment approach is more homogenous than most of the existing literature, and the authors have demonstrated that a long posterior construct is a highly effective approach to stabilizing these injuries and consistently results in healing. There is consensus in the spine community that these fractures are best treated surgically, and most spine surgeons likely agree with the authors of the current paper that a long posterior construct is best for most patients. The high blood loss and infection rate are consistent with prior reports, and anticipating these challenges should allow surgeons to deal with them more effectively. Additionally, these data should be shared with patients and their families so that they have realistic expectations and are not surprised when complications arise. The good news from this study is that if patients survive the initial injury and surgery, they will likely heal their fracture. Long-term outcomes are likely driven more by the patient’s age, comorbidities, and neurological status than by factors within the surgeon’s control. While this study is unlikely to change clinical practice, it adds more data to the literature on this topic and lends strong support to treating these fractures aggressively with long posterior constructs.


Please read Dr. Robinson’s article in the February 15 issue. Does this change how view extension-distraction injuries in the ankylosed spine? Let us know by leaving a comment on The Spine Blog.


Adam Pearson, MD, MS

Associate Web Editor

About the Blog

Spine Journal
This Blog provides a forum for discussion about high impact articles published in Spine, including the bi-annual publication of "Evidenced-Based Recommendations for Spine Surgery." Website users can use this forum to discuss how the articles have affected their practice and query the authors about their findings and recommendations.