Comparative effectiveness research (CER) has impending significance for the field of spine surgery. This article outlines the rationale for comparative effectiveness research and reviews recommended priorities of spinal surgery emphasis. It also examines recent key studies of CER in the spine surgery literature and associated cost-effectiveness studies. It concludes with a discussion of the direction of CER in the spine surgery community.
Kalil G. Abdullah, Edward C. Benzel and Thomas E. Mroz
Matthew J. Grosso, Roy Hwang, Thomas Mroz, Edward Benzel and Michael P. Steinmetz
Reversal of the normal cervical spine curvature, as seen in cervical kyphosis, can lead to mechanical pain, neurological dysfunction, and functional disabilities. Surgical intervention is warranted in patients with sufficiently symptomatic deformities in an attempt to correct the deformed cervical spine. In theory, improved outcomes should accompany a greater degree of correction toward lordosis, although there are few data available to test this relationship. The purpose of this study is to determine if the degree of deformity correction correlates with improvement in neurological symptoms following surgery for cervical kyphotic deformity.
A retrospective review of 36 patients with myelopathic symptoms who underwent cervical deformity correction surgery between 2001 and 2009 was performed. Preoperative and postoperative radiographic findings related to the degree of kyphosis were collected and compared with functional outcome measures. The minimum follow-up time was 2 years.
A significant relationship was observed between a greater degree of focal kyphosis correction and improved neurological outcomes according to the modified Japanese Orthopaedic Association (mJOA) score (r = −0.46, p = 0.032). For patients with severe neurological symptoms (mJOA score < 12) a trend toward improved outcomes with greater global kyphosis correction was observed (r = −0.56, p = 0.057). Patients with an mJOA score less than 16 who attained lordosis postoperatively had a significantly greater improvement in total mJOA score than patients who maintained a kyphotic position (achieved lordosis: 2.7 ± 2.0 vs maintained kyphosis: 1.1 ± 2.1, p = 0.044).
The authors' results suggest that the degree of correction of focal kyphosis deformity correlates with improved neurological outcomes. The authors also saw a positive relationship between attainment of global lordosis and improved mJOA scores. With consideration for the risks involved in correction surgery, this information can be used to help guide surgical strategy decision making.
Kene T. Ugokwe, Iain H. Kalfas, Thomas E. Mroz and Michael P. Steinmetz
Pseudarthrosis and construct failure following single-level anterior cervical discectomy, fusion, and plate placement (ACDFP) rarely occur. Routine postoperative anteroposterior and lateral radiographs may be an inconvenience to patients and expose them to additional and potentially unnecessary radiation. No standard exists to define when patients should obtain radiographs following an ACDFP. The authors hypothesize that routinely obtaining static anteroposterior and lateral radiographs in patients who recently underwent a single-level ACDFP without new axial neck pain or other neurological complaints or symptoms is unwarranted and does not alter the long-term treatment of the patient.
The authors retrospectively reviewed the charts and radiographs of patients who underwent a single-level ACDFP between January 1, 2000, and December 31, 2005. All patients underwent a single-level ACDFP and had routine cervical radiographs obtained at various intervals after surgery.
Twenty-one patients underwent ACDFP at C5–6, 14 patients underwent surgery at C6–7, 11 patients at C4–5, and 7 patients at C3–4. None of the intraoperative radiographs demonstrated malposition of the graft or instrumentation. Based on subjective reporting by the patients, the vast majority (49 of 53) showed improvement in neck and arm pain, and/or neurological dysfunction following surgery. Overall, 5 patients (9%) demonstrated abnormalities on their postoperative radiographs. No patients were returned to the operating room as a result of postoperative radiographic findings. The sensitivity of plain radiographs in this patient series or the percentage of patients with new symptoms that had an abnormality related to the construct on plain radiography was 50%. The specificity of plain radiographs or the percentage of patients who were asymptomatic and had normal radiographs was 94%. The positive predictive value was 25%; that is, there was a 25% chance that patients with symptoms would have a construct abnormality on postoperative radiographs. The negative predictive value was 98%; that is, 98% of patients without symptoms will have normal radiographs.
Pseudarthrosis and construct failure following single-level ACDFP occur rarely, and patients with new symptoms following surgery are as likely to have normal radiographic findings as they are to have abnormalities identified on their postoperative plain radiographs. Routinely obtaining postoperative radiographs at regular intervals in asymptomatic patients following single-level ACDFP does not appear to be warranted.
Matthew D. Alvin, Daniel Lubelski, Edward C. Benzel and Thomas E. Mroz
Cervical spondylotic myelopathy (CSM) often can be surgically treated by either ventral or dorsal decompression and fusion. However, there is a lack of high-level evidence on the relative advantages and disadvantages for these treatments of CSM. The authors' goal was to provide a comprehensive review of the relative benefits of ventral versus dorsal fusion in terms of quality of life (QOL) outcomes, complications, and costs. They reviewed 7 studies on CSM published between 2003 and 2013 and summarized the findings for each category. Both procedures have been shown to lead to statistically significant improvement in clinical outcomes for patients. Ventral fusion surgery has been shown to yield better QOL outcomes than dorsal fusion surgery. Complication rates for ventral fusion surgery range from 11% to 13.6%, whereas those for dorsal fusion surgery range from 16.4% to 19%. Larger randomized controlled trials are needed, with particular emphasis on QOL and minimum clinically important differences.
Andrew T. Healy, Prasath Mageswaran, Daniel Lubelski, Benjamin P. Rosenbaum, Virgilio Matheus, Edward C. Benzel and Thomas E. Mroz
The degenerative process of the spinal column results in instability followed by a progressive loss of segmental motion. Segmental degeneration is associated with intervertebral disc and facet changes, which can be quantified. Correlating this degeneration with clinical segmental motion has not been investigated in the thoracic spine. The authors sought to determine if imaging-determined degeneration would correlate with native range of motion (ROM) or the change in ROM after decompressive procedures, potentially guiding clinical decision making in the setting of spine trauma or following decompressive procedures in the thoracic spine.
Multidirectional flexibility tests with image analysis were performed on thoracic cadaveric spines with intact ib cage. Specimens consisted of 19 fresh frozen human cadaveric spines, spanning C-7 to L-1. ROM was obtained for each specimen in axial rotation (AR), flexion-extension (FE), and lateral bending (LB) in the intact state and following laminectomy, unilateral facetectomy, and unilateral costotransversectomy performed at either T4–5 (in 9 specimens) or T8–9 (in 10 specimens). Image grading of segmental degeneration was performed utilizing 3D CT reconstructions. Imaging scores were obtained for disc space degeneration, which quantified osteophytes, narrowing, and endplate sclerosis, all contributing to the Lane disc summary score. Facet degeneration was quantified using the Weishaupt facet summary score, which included the scoring of facet osteophytes, narrowing, hypertrophy, subchondral erosions, and cysts.
The native ROM of specimens from T-1 to T-12 (n = 19) negatively correlated with age in AR (Pearson’s r coefficient = -0.42, p = 0.070) and FE (r = -0.42, p = 0.076). When regional ROM (across 4 adjacent segments) was considered, the presence of disc osteophytes negatively correlated with FE (r = −0.69, p = 0.012), LB (r = −0.82, p = 0.001), and disc narrowing trended toward significance in AR (r = −0.49, p = 0.107). Facet characteristics, scored using multiple variables, showed minimal correlation to native ROM (r range from −0.45 to +0.19); however, facet degeneration scores at the surgical level revealed strong negative correlations with regional thoracic stability following decompressive procedures in AR and LB (Weishaupt facet summary score: r = −0.52 and r = −0.71; p = 0.084 and p = 0.010, respectively). Disc degeneration was not correlated (Lane disc summary score: r = −0.06, p = 0.861).
Advanced age was the most important determinant of decreasing native thoracic ROM, whereas imaging characteristics (T1–12) did not correlate with the native ROM of thoracic specimens with intact rib cages. Advanced facet degeneration at the surgical level did correlate to specimen stability following decompressive procedures, and is likely indicative of the terminal stages of segmental degeneration.
Syed K. Mehdi, Vincent J. Alentado, Bryan S. Lee, Thomas E. Mroz, Edward C. Benzel and Michael P. Steinmetz
Ossification of the posterior longitudinal ligament (OPLL) is a pathological calcification or ossification of the PLL, predominantly occurring in the cervical spine. Although surgery is often necessary for patients with symptomatic neurological deterioration, there remains controversy with regard to the optimal surgical treatment. In this systematic review and meta-analysis, the authors identified differences in complications and outcomes after anterior or posterior decompression and fusion versus after decompression alone for the treatment of cervical myelopathy due to OPLL.
A MEDLINE, SCOPUS, and Web of Science search was performed for studies reporting complications and outcomes after decompression and fusion or after decompression alone for patients with OPLL. A meta-analysis was performed to calculate effect summary mean values, 95% CIs, Q statistics, and I2 values. Forest plots were constructed for each analysis group.
Of the 2630 retrieved articles, 32 met the inclusion criteria. There was no statistically significant difference in the incidence of excellent and good outcomes and of fair and poor outcomes between the decompression and fusion and the decompression-only cohorts. However, the decompression and fusion cohort had a statistically significantly higher recovery rate (63.2% vs 53.9%; p < 0.0001), a higher final Japanese Orthopaedic Association score (14.0 vs 13.5; p < 0.0001), and a lower incidence of OPLL progression (< 1% vs 6.3%; p < 0.0001) compared with the decompression-only cohort. There was no statistically significant difference in the incidence of complications between the 2 cohorts.
This study represents the only comprehensive review of outcomes and complications after decompression and fusion or after decompression alone for OPLL across a heterogeneous group of surgeons and patients. Based on these results, decompression and fusion is a superior surgical technique compared with posterior decompression alone in patients with OPLL. These results indicate that surgical decompression and fusion lead to a faster recovery, improved postoperative neurological functioning, and a lower incidence of OPLL progression compared with posterior decompression only. Furthermore, decompression and fusion did not lead to a greater incidence of complications compared with posterior decompression only.
Michael F. Shriver, Valerie Zeer, Vincent J. Alentado, Thomas E. Mroz, Edward C. Benzel and Michael P. Steinmetz
There are a variety of surgical positions that provide optimal exposure of the dorsal lumbar spine. These include the prone, kneeling, knee-chest, knee-elbow, and lateral decubitus positions. All are positions that facilitate exposure of the spine. Each position, however, is associated with an array of unique complications that result from excessive pressure applied to the torso or extremities. The authors reviewed clinical studies reporting complications that arose from positioning of the patient during dorsal exposures of the lumbar spine.
MEDLINE, Scopus, and Web of Science database searches were performed to find clinical studies reporting complications associated with positioning during lumbar spine surgery. For articles meeting inclusion criteria, the following information was obtained: publication year, study design, sample size, age, operative time, type of surgery, surgical position, frame or table type, complications associated with positioning, time to first observed complication, long-term outcomes, and evidence-based recommendations for complication avoidance.
Of 3898 articles retrieved from MEDLINE, Scopus, and Web of Science, 34 met inclusion criteria. Twenty-four studies reported complications associated with use of the prone position, and 7 studies investigated complications after knee-chest positioning. Complications associated with the knee-elbow, lateral decubitus, and supine positions were each reported by a single study. Vision loss was the most commonly reported complication for both prone and knee-chest positioning. Several other complications were reported, including conjunctival swelling, Ischemic orbital compartment syndrome, nerve palsies, thromboembolic complications, pressure sores, lower extremity compartment syndrome, and shoulder dislocation, highlighting the assortment of possible complications following different surgical positions. For prone-position studies, there was a relationship between increased operation time and position complications. Only 3 prone-position studies reported complications following procedures of less than 120 minutes, 7 studies reported complications following mean operative times of 121–240 minutes, and 9 additional studies reported complications following mean operative times greater than 240 minutes. This relationship was not observed for knee-chest and other surgical positions.
This work presents a systematic review of positioning-related complications following prone, knee-chest, and other positions used for lumbar spine surgery. Numerous evidence-based recommendations for avoidance of these potentially severe complications associated with intraoperative positioning are discussed. This investigation may serve as a framework to educate the surgical team and decrease rates of intraoperative positioning complications.
Michael F. Shriver, Jack J. Xie, Erik Y. Tye, Benjamin P. Rosenbaum, Varun R. Kshettry, Edward C. Benzel and Thomas E. Mroz
Lumbar microdiscectomy and its various minimally invasive surgical techniques are seeing increasing popularity, but a systematic review of their associated complications has yet to be performed. The authors sought to identify all prospective clinical studies reporting complications associated with lumbar open microdiscectomy, microendoscopic discectomy (MED), and percutaneous microdiscectomy.
The authors conducted MEDLINE, Scopus, Web of Science, and Embase database searches for randomized controlled trials and prospective cohort studies reporting complications associated with open, microendoscopic, or percutaneous lumbar microdiscectomy. Studies with fewer than 10 patients and published before 1990 were excluded. Overall and interstudy median complication rates were calculated for each surgical technique. The authors also performed a meta-analysis of the reported complications to assess statistical significance across the various surgical techniques.
Of 9504 articles retrieved from the databases, 42 met inclusion criteria. Most studies screened were retrospective case series, limiting the number of studies that could be included. A total of 9 complication types were identified in the included studies, and these were analyzed across each of the surgical techniques. The rates of any complication across the included studies were 12.5%, 13.3%, and 10.8% for open, MED, and percutaneous microdiscectomy, respectively. New or worsening neurological deficit arose in 1.3%, 3.0%, and 1.6% of patients, while direct nerve root injury occurred at rates of 2.6%, 0.9%, and 1.1%, respectively. Hematoma was reported at rates of 0.5%, 1.2%, and 0.6%, respectively. Wound complications (infection, dehiscence, orseroma) occurred at rates of 2.1%, 1.2%, and 0.5%, respectively. The rates of recurrent disc complications were 4.4%, 3.1%, and 3.9%, while reoperation was indicated in 7.1%, 3.7%, and 10.2% of operations, respectively. Meta-analysis calculations revealed a statistically significant higher rate of intraoperative nerve root injury following percutaneous procedures relative to MED. No other significant differences were found.
This review highlights complication rates among various microdiscectomy techniques, which likely reflect real-world practice and conceptualization of complications among physicians. This investigation sets the framework for further discussions regarding microdiscectomy options and their associated complications during the informed consent process.
Haariss Ilyas, Joshua L. Golubovsky, Jingxiao Chen, Robert D. Winkelman, Thomas E. Mroz and Michael P. Steinmetz
The objective of this study was to evaluate the incidence and risk factors for 90-day readmission and reoperation after elective surgery for lumbar spinal stenosis (LSS).
The authors performed a retrospective consecutive cohort analysis of patients undergoing posterior lumbar decompression with or without fusion for LSS with claudication from January 2014 through December 2015.
Data were collected on 1592 consecutive patients. The mean age at surgery was 67.4 ± 10.1 years and 45% of patients were female. The 90-day reoperation rate was 4.7%, and 69.3% of the reoperations occurred within the first 30 days. The 90-day readmission rate was 7.2%. Multivariable analysis showed that postoperative development of a surgical site infection (SSI; odds ratio [OR] 14.09, 95% confidence interval [CI] 7.86–25.18), acute kidney injury (AKI; OR 6.76, 95% CI 2.39–19.57), and urinary tract infection (UTI; OR 3.96, 95% CI 2.43–6.37), as well as a history of congestive heart failure (CHF; OR 3.03, 95% CI 1.69–5.28), were significant risk factors for readmission within 90 days. Male sex (OR 0.60, 95% CI 0.38–0.92) was associated with decreased odds for readmission. With regards to reoperation, development of SSI (OR 25.06, 95% CI 13.54–46.51), sepsis (OR 7.63, 95% CI 1.52–40.59), UTI (OR 2.54, 95% CI 1.31–4.76), and increased length of stay (LOS; OR 1.25, 95% CI 1.17–1.33) were found to be significant risk factors. A subsequent analysis found that morbid obesity (OR 6.99), history of coronary artery disease (OR 2.263), increased duration of surgery (OR 1.004), and LOS (OR 1.07) were significant risk factors for developing an SSI.
Overall, this study found rates of 4.7% and 7.2% for reoperation and readmission, respectively, within 90 days: 30.7% of the reoperations and 44.7% of the readmissions occurred beyond the first 30 days. A diagnosis of SSI, AKI, UTI, and history of CHF were significant factors for readmission, while male sex was associated with decreased odds for readmission. A diagnosis of SSI, sepsis, UTI, and increased LOS were found to be significant predictors for reoperation. Understanding 90-day complication rates is imperative because there has been increased discussion and healthcare policy extending the global postoperative window to 90 days. Current literature supports a readmission rate of 3%–9% after spine surgery. However, this literature either is limited to a 30-day window or does not stratify between different types of spine surgeries.
Sina Pourtaheri, Akshay Sharma, Jason Savage, Iain Kalfas, Thomas E. Mroz, Edward Benzel and Michael P. Steinmetz
The flexed posture of the proximal (L1–3) or distal (L4–S1) lumbar spine increases the diameter of the spinal canal and neuroforamina and can relieve symptoms of neurogenic claudication. Distal lumbar flexion can result in pelvic retroversion; therefore, in cases of flexible sagittal imbalance, pelvic retroversion may be compensatory for lumbar stenosis and not solely compensatory for the sagittal imbalance as previously thought. The authors investigate underlying causes for pelvic retroversion in patients with flexible sagittal imbalance.
One hundred thirty-eight patients with sagittal imbalance who underwent a total of 148 fusion procedures of the thoracolumbar spine were identified from a prospective clinical database. Radiographic parameters were obtained from images preoperatively, intraoperatively, and at 6-month and 2-year follow-up. A cohort of 24 patients with flexible sagittal imbalance was identified and individually matched with a control cohort of 23 patients with fixed deformities. Flexible deformities were defined as a 10° change in lumbar lordosis between weight-bearing and non–weight-bearing images. Pelvic retroversion was quantified as the ratio of pelvic tilt (PT) to pelvic incidence (PI).
The average difference between lumbar lordosis on supine MR images and standing radiographs was 15° in the flexible cohort. Sixty-eight percent of the patients in the flexible cohort were diagnosed preoperatively with lumbar stenosis compared with only 22% in the fixed sagittal imbalance cohort (p = 0.0032). There was no difference between the flexible and fixed cohorts with regard to C-2 sagittal vertical axis (SVA) (p = 0.95) or C-7 SVA (p = 0.43). When assessing for postural compensation by pelvic retroversion in the stenotic patients and nonstenotic patients, the PT/PI ratio was found to be significantly greater in the patients with stenosis (p = 0.019).
For flexible sagittal imbalance, preoperative attention should be given to the root cause of the sagittal misalignment, which could be compensation for lumbar stenosis. Pelvic retroversion can be compensatory for both the lumbar stenosis as well as for sagittal imbalance.