Dennis R. Knapp

Rates of infection after spine surgery based on 108,419 procedures: a report from the Scoliosis Research Society Morbidity and Mortality Committee.

Study Design. Retrospective review of a prospectively collected database. Objective. Our objective was to assess the rates of postoperative wound infection associated with spine surgery. Summary of Background Data. Although wound infection after spine surgery remains a common source of morbidity, estimates of its rates of occurrence remain relatively limited. The Scoliosis Research Society prospectively collects morbidity and mortality data from its members, including the occurrence of wound infection. Methods. The Scoliosis Research Society morbidity and mortality database was queried for all reported spine surgery cases from 2004 to 2007. Cases were stratified based on factors including diagnosis, adult (≥21 years) versus pediatric (<21 years), primary versus revision, use of implants, and whether a minimally invasive approach was used. Superficial, deep, and total infection rates were calculated. Results. In total, 108,419 cases were identified, with an overall total infection rate of 2.1% (superficial = 0.8%, deep = 1.3%). Based on primary diagnosis, total postoperative wound infection rate for adults ranged from 1.4% for degenerative disease to 4.2% for kyphosis. Postoperative wound infection rates for pediatric patients ranged from 0.9% for degenerative disease to 5.4% for kyphosis. Rate of infection was further stratified based on subtype of degenerative disease, type of scoliosis, and type of kyphosis for both adult and pediatric patients. Factors associated with increased rate of infection included revision surgery (P < 0.001), performance of spinal fusion (P < 0.001), and use of implants (P < 0.001). Compared with a traditional open approach, use of a minimally invasive approach was associated with a lower rate of infection for lumbar discectomy (0.4% vs. 1.1%; P < 0.001) and for transforaminal lumbar interbody fusion (1.3% vs. 2.9%; P = 0.005). Conclusion. Our data suggest that postsurgical infection, even among skilled spine surgeons, is an inherent potential complication. These data provide general benchmarks of infection rates as a basis for ongoing efforts to improve safety of care.

Rates of new neurological deficit associated with spine surgery based on 108,419 procedures: A report of the scoliosis research society morbidity and mortality committee

Study Design. Retrospective review of a prospectively collected, multicenter database. Objective. To assess rates of new neurologic deficit (NND) associated with spine surgery. Summary of Background Data. NND is a potential complication of spine surgery, but previously reported rates are often limited by small sample size and single-surgeon experiences. Methods. The Scoliosis Research Society morbidity and mortality database was queried for spinal surgery cases complicated by NND from 2004 to 2007, including nerve root deficit (NRD), cauda equina deficit (CED), and spinal cord deficit (SCD). Use of neuromonitoring was assessed. Recovery was stratified as complete, partial, or none. Rates of NND were stratified based on diagnosis, age (pediatric < 21; adult ≥ 21), and surgical parameters. Results. Of the 108,419 cases reported, NND was documented for 1064 (1.0%), including 662 NRDs, 74 CEDs, and 293 SCDs (deficit not specified for 35 cases). Rates of NND were calculated on the basis of diagnosis. Revision cases had a 41% higher rate of NND (1.25%) compared with primary cases (0.89%; P < 0.001). Pediatric cases had a 59% higher rate of NND (1.32%) compared with adult cases (0.83%; P < 0.001). The rate of NND for cases with implants was more than twice that for cases without implants (1.15% vs. 0.52%, P < 0.001). Neuromonitoring was used for 65% of cases, and for cases with new NRD, CED, and SCD, changes in neuromonitoring were reported in 11%, 8%, and 40%, respectively. The respective percentages of no recovery, partial, and complete recovery for NRD were 4.7%, 46.8%, and 47.1%, respectively; for CED were 9.6%, 45.2%, and 45.2%, respectively; and for SCD were 10.6%, 43%, and 45.7%, respectively. Conclusion. Our data demonstrate that, even among skilled spinal deformity surgeons, new neurologic deficits are inherent potential complications of spine surgery. These data provide general benchmark rates for NND with spine surgery as a basis for patient counseling and for ongoing efforts to improve safety of care.

Complication rates of three common spine procedures and rates of thromboembolism following spine surgery based on 108,419 procedures: A report from the scoliosis research society morbidity and mortality committee

Study Design. Retrospective review of a prospectively collected database. Objective. The Scoliosis Research Society (SRS) collects morbidity and mortality (M and M) data from its members. Our objectives were to assess complication rates for 3 common spine procedures, compare these results with prior literature as a means of validating the database, and to assess rates of pulmonary embolism (PE) and deep venous thrombosis (DVT) in all cases reported to the SRS over 4 years. Summary of Background Data. Few modern series document complication rates of spinal surgery as routinely practiced across academic and community settings. Those available are typically based on relatively low numbers of procedures or confined to single-surgeon experiences. Methods. The SRS M and M database was queried for lumbar microdiscectomy (LD), anterior cervical discectomy and fusion (ACDF), and lumbar stenosis decompression (LSD) cases from 2004 to 2007. Revisions were excluded. The database was also queried for occurrence of clinically evident PE and DVT in all cases from 2004 to 2007. Results. A total of 9692 LDs, 6735 ACDFs, and 10,329 LSDs were identified, with overall complication rates of 3.6%, 2.4%, and 7.0%, respectively. These rates are comparable to previously published smaller series. For assessment of PE and DVT, 108,419 cases were identified and rates were calculated per 1000 cases based on diagnosis, age group, and implant use. Overall rates of PE, death due to PE, and DVT were 1.38, 0.34, and 1.18, respectively. Among 82,082 adults, the rate of PE ranged from 0.47 for LD to 12.4 for metastatic tumor. Similar variations were noted for DVT and deaths due to PE. Conclusion. Overall major complication rates for LD, ACDF, and LSD based on the SRS M and M database are comparable to those in previously reported smaller series, supporting the validity of this database for study of other less common spinal disorders. In addition, our data provide general benchmarks of clinically evident PE and DVT rates as a basis for ongoing efforts to improve care.

Does bone morphogenetic protein increasethe incidence of perioperative complicationsin spinal fusion?: A comparison of 55,862 cases of spinal fusion with and without bone morphogenetic protein

Study Design. Retrospective review of a multi-institutional, multisurgeon database. Objective. Assess for associations between bone morphogenetic protein (BMP) use and rate of complications in spinal fusion. Summary of Background Data. BMP is commonly used in spinal surgery to augment fusion; however, there is limited evidence demonstrating its associated complications. Methods. We performed a retrospective analysis of all fusion cases submitted by members of the Scoliosis Research Society from 2004 to 2007. We stratified on the basis of the use of BMP and evaluated for complications and associated characteristics. Results. A total of 55,862 cases of spinal fusion were identified with BMP used in 21% (11,933) of the cases. Excluding anterior cervical fusions, there were no significant differences between fusions with and without BMP with regard to overall complications (8.4% vs. 8.5%; P = 0.5), wound infections (2.4% vs. 2.4%; P = 0.8), or epidural hematomas/seromas (0.2% vs. 0.2%; P = 0.3). Anterior cervical fusions with BMP were associated with more overall complications (5.8% vs. 2.4%; P < 0.001) and more wound infections (2.1% vs. 0.4%; P < 0.001) than fusions without BMP. On multivariate analysis for thoracolumbar and posterior cervical fusions, BMP use was not a significant predictor of complications (P = 0.334; odds ratio = 1.039; 95% confidence interval = 0.961–1.124; covariates were BMP use, patient age, revision vs. primary surgery). Multivariate analysis for anterior cervical spinal fusion demonstrated that BMP use remained a significant predictor of complications (P < 0.001, odds ratio = 1.6; 95% confidence interval = 1.516–1.721), after adjusting for the effects of patient age and whether the surgery was a revision procedure. Conclusion. BMP use with anterior cervical fusion was associated with an increased incidence of complications. Use of BMP was not associated with more complications in thoracolumbar and posterior cervical fusions.

Morbidity and mortality in the surgical treatment of 10,242 adults with spondylolisthesis.

OBJECT This is a retrospective review of 10,242 adults with degenerative spondylolisthesis (DS) and isthmic spondylolisthesis (IS) from the morbidity and mortality (M&M) index of the Scoliosis Research Society (SRS). This database was reviewed to assess complication incidence, and to identify factors that were associated with increased complication rates. METHODS The SRS M&M database was queried to identify cases of DS and IS treated between 2004 and 2007. Complications were identified and analyzed based on age, surgical approach, spondylolisthesis type/grade, and history of previous surgery. Age was stratified into 2 categories: > 65 years and ≤ 65 years. Surgical approach was stratified into the following categories: decompression without fusion, anterior, anterior/posterior, posterior without instrumentation, posterior with instrumentation, and interbody fusion. Spondylolisthesis grades were divided into low-grade (Meyerding I and II) versus high-grade (Meyerding III, IV, and V) groups. Both univariate and multivariate analyses were performed. RESULTS In the 10,242 cases of DS and IS reported, there were 945 complications (9.2%) in 813 patients (7.9%). The most common complications were dural tears, wound infections, implant complications, and neurological complications (range 0.7%-2.1%). The mortality rate was 0.1%. Diagnosis of DS had a significantly higher complication rate (8.5%) when compared with IS (6.6%; p = 0.002). High-grade spondylolisthesis correlated strongly with a higher complication rate (22.9% vs 8.3%, p < 0.0001). Age > 65 years was associated with a significantly higher complication rate (p = 0.02). History of previous surgery and surgical approach were not significantly associated with higher complication rates. On multivariate analysis, only the grade of spondylolisthesis (low vs high) was in the final best-fit model of factors associated with the occurrence of complications (p < 0.0001). CONCLUSIONS The rate of total complications for treatment of DS and IS in this series was 9.2%. The total percentage of patients with complications was 7.9%. On univariate analysis, the complication rate was significantly higher in patients with high-grade spondylolisthesis, a diagnosis of DS, and in older patients. Surgical approach and history of previous surgery were not significantly correlated with increased complication rates. On multivariate analysis, only the grade of spondylolisthesis was significantly associated with the occurrence of complications.

Mortality and morbidity in early-onset scoliosis surgery.

Study Design. Retrospective chart review. Objective. To accurately determine complication rates, particularly mortality rates, in surgically treated early-onset scoliosis. Summary of Background Data. The advent of modern segmental instrumentation for spinal fusion surgery in adolescent scoliosis has allowed for application of similar nonsegmental unfused techniques aimed at controlling scoliosis in the very young child. The dismal prognosis for these children without repeated spinal lengthening procedures is unquestioned, although no controlled trials exist. Many, if not most, of these children need surgery; however, the surgical complication rate is very high. Methods. During the study period, all surgically treated children with early-onset scoliosis seen at our institution were identified. Any patient who presented to our clinic with early-onset scoliosis that was surgically managed was included. The total number of procedures, type of implants, number and type of complications, geographic origin of the cases, and final outcomes were all assessed. Results. A total of 165 surgical procedures on 28 patients accrued during the study time period, including index implantation of instrumentation, lengthening, and definitive fusion, as well as operations performed for complications such as wound debridement and revision of failed implants. Clinical diagnoses included congenital scoliosis, syndromic and chromosomal abnormalities, cerebral palsy, and spinal muscular atrophy. There was a complication rate of 84% overall with a mortality rate of almost 18%. The only patients with no complications were those whose entire surgical course had been at our institution only. The mortality rate was equal in patients whose treatment was performed elsewhere versus exclusively in our center. Conclusion. This study underlines the grave severity of these scolioses particularly in syndromic children. The high mortality rate is alarming, suggesting that further study is needed in this area.

Outcomes after salvage procedures for the painful dislocated hip in cerebral palsy.

Background: The painful dislocated hip in the setting of cerebral palsy is a challenging problem. Many surgical procedures have been reported to treat this condition with varying success rates. The purpose of this study is to retrospectively evaluate and compare the outcomes of 3 different surgical procedures performed at our institution for pain relief in patients with spastic quadriplegic cerebral palsy and painful dislocated hips. Methods: A retrospective chart review of the surgical procedures performed by 5 surgeons for spastic, painful dislocated hips from 1997 to 2010 was performed. The procedures identified were (1) proximal femoral resection arthroplasty (PFRA); (2) subtrochanteric valgus osteotomy (SVO) with femoral head resection; and (3) proximal femur prosthetic interposition arthroplasty (PFIA) using a humeral prosthesis. Outcomes based on pain and range of motion were determined to be excellent, good, fair, or poor by predetermined criteria. Results: Forty-four index surgeries and 14 revision surgeries in 33 patients with an average follow-up of 49 months met the inclusion criteria. Of the index surgeries, 12 hips were treated with a PFRA, 21 with a SVO, and 11 with a PFIA. An excellent or good result was noted in 67% of PFRAs, 67% of SVOs, and 73% of PFIAs. No statistical significance between these procedures was achieved. The 14 revisions were performed because of a poor result from previous surgery, demonstrating a 24% reoperation rate overall. No patients classified as having a fair result underwent revision surgery. All patients receiving revision surgery were eventually classified as having an excellent or good result. Conclusions: Surgical treatment for the painful, dislocated hip in the setting of spastic quadriplegic cerebral palsy remains unsettled. There continue to be a large percentage of failures despite the variety of surgical techniques designed to treat this problem. These failures can be managed, however, and eventually resulted in a good outcome. We demonstrated a trend toward better outcomes with a PFIA, but further study should be conducted to prove statistical significance. Level of Evidence: III.

Usefulness of electromyography compared to computed tomography scans in pedicle screw placement.

Study Design. This is a retrospective analysis of 30 pediatric deformity surgeries. Objective. The purpose of this study was to evaluate the accuracy of neuromonitoring in comparison to postoperative computed tomography scans for pedicle screw position. Summary of Background Data. Triggered electromyography potentials in aiding the placement of lumbar pedicle screws are considered useful; however, this method is less accepted in thoracic screw placement. Methods. Thirty pediatric deformity surgeries were reviewed. All screws were placed using fluoroscopic assistance. Electromyography data were obtained on all screws. Every patient underwent postoperative computed tomography scanning. Computed tomography scans were assessed by all authors, and each screw was classified. Sensitivity, specificity, negative predictive value, and likelihood ratios were determined for the cut-off value of an electromyography ≥6 mA. Results. A total of 329 screws were reviewed. No complications occurred. An overall accuracy of 93% was obtained. No retained screw had greater than 2 mm medial pedicle wall breach. Nine screws were removed intraoperatively due to medial breach. The mean electromyography potential for all classes of screws was not statistically different (P > 0.1). The negative predictive value of the test was 0.92 in the thoracic spine and 0.93 in the lumbar spine. The negative likelihood ratios were 0.96 and 0.35 for the thoracic and lumbar spines respectively, and the positive likelihood ratio was 1.4 for the thoracic spine and 12.5 for the lumbar spine. Conclusion. Thoracic and lumbar pedicle screws are safe surgical options in the treatment of pediatric scoliosis. Comparison of electromyography potentials and postoperative computed tomography scans showed no statistically significant difference for all classes of screws. The likelihood ratio for electromyography testing was more clinically significant in the lumbar spine. A triggered electromyography value greater than or equal to 6 mA has a high likelihood of that screw being in the “safe zone.” However, there is no true electromyography cut-off value that guarantees accurate placement and avoidance of neurologic injury.

Assessment of Morbidity and Mortality Collection Data 2009

STUDY DESIGN A retrospective analysis of the morbidity and mortality data collected by the SRS in 2009 with comparison to previous years. OBJECTIVES Objective of this study was to assess the new format of morbidity and mortality data collection by the SRS in 2009 and evaluate the data collected with comparison to previous years. SUMMARY OF BACKGROUND DATA 2009 morbidity and mortality reporting format was dramatically changed from previous years. This was done in an attempt to simplify the reporting process and to narrow the reporting to only three sentinel events: death, blindness, and neurologic injury. Only deformity cases including scoliosis, kyphosis, and Grade III or greater spondylolisthesis were included. METHODS Results were obtained from the SRS M & M reporting summary for 2009. These included detailed analysis for complications including death, blindness, and neurologic injury. These were compared to similar statistics obtained from the years dating back to 2001. Blindness has not previously been reported. RESULTS 80.1% of SRS members submitted 35,267 deformity cases. Both the percentage of members submitting data and the number of deformity cases far exceeded any previous year total. 57.6% of cases involved scoliosis, 10.5% kyphosis, and 31.9% spondylolisthesis. Neurologic complications were lower in each major category (scoliosis, kyphosis, and spondylolisthesis) compared to previous years. The death rate was similar to prior reporting 0.12%. Three cases of blindness were reported, two occurred in AIS. All three resolved. CONCLUSIONS The altered format and requirement for all members to participate has dramatically increased the total number of deformity cases reported and percent of membership responding. Neurologic injury rates are decreased from previous years. Three cases of blindness occurred, all of which resolved. Further collection of data is needed to elucidate mechanism and prevention.

The Recognition, Incidence, and Management of Spinal Cord Monitoring Alerts in Early-onset Scoliosis Surgery.

Background: The objective of the research was to study the relevance of intraoperative neuromonitoring throughout all stages of surgical management in patients with progressive early-onset scoliosis (EOS). The routine monitoring of spinal cord potentials has gradually become standard of practice among spinal surgeons. However, there is not a consensus that the added expense of this technique necessitates monitoring in all stages of surgical management. Methods: A retrospective review of 180 surgical cases of 30 patients with EOS from July 2003 to July 2012 was performed. All monitoring alerts as judged by the neuromonitoring team were identified. Both somatosensory-evoked potentials and transcranial electric motor-evoked potentials were studied and no limiting thresholds for reporting electrophysiological changes were deemed appropriate. Results: Of 150 monitored cases there were 18 (12%) monitoring alerts. This represented 40% of the patient cohort over the 9-year study period. Conclusions: Index versus routine lengthening rate of alerts showed no significant difference in incidence of monitoring alerts. Conversely, several patients whose primary implantation surgeries were uneventful had monitoring alerts later in their treatment course. Intraoperative neuromonitoring is warranted throughout all stages of surgical management of EOS. Level of Evidence: Level IV. This study is a retrospective review of surgical cases of 30 patients with EOS.