By: 4 September 2018
EUROSPINE’s best in show

Get acquainted with three of the six abstract submissions for EUROPSINE 2018, which have been elected as Best of Show and will be presented in the Best of Show and Award papers session.


In-hospital complication rate following microendoscopic versus open lumbar laminectomy: A propensity score-matched analysis

Presentation by K Soma, T Oichi and Y Oshima

Lumbar spinal stenosis is one of the most common orthopaedic disorders in increasingly elderly populations. It causes spinal claudication, back and leg pain, and disability. There has recently been an increase in the popularity of minimally invasive surgery (MIS) for lumbar spinal disorders. Forley and Smith first described minimally invasive lumbar laminectomy using microendoscopic procedures in 1997. Microendoscopic laminectomy (MEL) is a modification of microscopic bilateral decompression via a unilateral approach. This technique is popular and reported outcomes are good. However, it remains unclear whether MEL is safer than open laminectomy, as no large study has been performed to compare postoperative outcomes after the two procedures. It is imperative to know the risks of surgery to inform appropriate clinical decision-making. The purpose of this study was to compare the postoperative complications of patients undergoing MEL with those treated with open laminectomy, using a nationwide database in Japan.

Data of patients who underwent elective spinal surgery between July 2010 and March 2013 were extracted from the Diagnosis Procedure Combination database, a nationwide inpatient database in Japan. Clinical outcomes included length of hospital stay, occurrence of major complications (cardiac events, respiratory complications, pulmonary embolism, stroke and acute renal failure), surgical site infection, post-operative delirium and in-hospital death. Propensity score matching was performed to adjust for measured confounding factors, including patient age, sex, Charlson Comorbidity Index, BMI, smoking status, blood transfusion, duration of anaesthesia, number of operated disc levels, type of hospital and hospital volumes. The clinical outcomes of one-to-one propensity-matched pairs of the MEL and open laminectomy groups were compared.

Of 23,317 patients identified in the database, 1,536 underwent MEL (6.6 per cent). By one-to-one propensity score matching, 1,536 pairs were selected. The distributions of patient backgrounds were closely balanced between the MEL and open laminectomy groups. An analysis of 1,536 pairs revealed that there was a significantly lower incidence of major postoperative complications in those who underwent MEL (1.0 per cent versus 2.8 per cent for open laminectomy, risk difference 1.8 per cent, 95 per cent confidence interval [CI] 0.9 per cent–2.9 per cent), surgical site infection (0.5 per cent versus 1.6 per cent for open laminectomy, risk difference 1.1 per cent, 95 per cent CI 0.4 per cent–1.9 per cent) and postoperative delirium (1.1 per cent versus 2.3 per cent for open laminectomy, risk difference 1.2 per cent, 95 per cent CI 0.3 per cent–2.1 per cent). Length of hospital stay was significantly shorter in those treated with MEL (12 days versus 16 days for open laminectomy, p <0.001). There was no significant difference in in-hospital mortality between the groups (see table 1).








Our study yielded three important findings. First, patients treated with MEL had a significantly lower incidence of major postoperative complications (including cardiac events and stroke) than those treated with open laminectomy, even after adjustment for other risk factors using propensity score matching. Second, patients treated with MEL were less likely to develop SSI than those who underwent open laminectomy. Finally, patients treated with MEL were less likely to develop postoperative delirium than those who underwent open laminectomy. The less extensive surgical stress and reduction in tissue trauma in MIS probably accounts for the lower incidence of these postoperative complications and shorter length of hospital stay. A better understanding of the factors associated with postoperative complications will help inform clinical decision-making and provide appropriate information for patients with lumbar spinal canal stenosis scheduled for surgery.


Quality of life in males and females with idiopathic scoliosis

Presentation by Elias Diarbakerli, Anna Grauers, Anna Danielsson, Allan Abbott, Paul Gerdhem

In this hitherto largest comprehensive study comparing quality of life in males and females, we found that when compared with national norm scales, both males and females show significant and similar reductions in their quality of life, irrespective if treated or not. From a quality of life perspective, clinicians do not need to consider sex when deciding on treatment.

The prevalence of idiopathic scoliosis among males and females is similar in smaller curves, but males are heavily underrepresented among those with larger curves.

Females tend to develop more aggressive curves requiring treatment. Hence, females are most often targeted in studies on idiopathic scoliosis. In this study, we compared the quality of life in females (n=1308) and males (n=211). Using the validated and disease-specific SRS-22r, and the generic EQ-5D, we compared quality of life in 528 surgically treated individuals (78 males), 535 brace treated individuals (50 males) and 456 untreated individuals (83 males). Any treatment was instituted during adolescence, i.e. before skeletal maturity. For the whole population, mean (SD) age at the survey was 35.3 (14.9) years.

The results indicated that males with idiopathic scoliosis tended to have slightly higher scores in the scoliosis specific SRS-22r but not in the generic quality of life measurement EQ-5D when compared to females. The differences were overall small and did not reach the previously established thresholds for minimal clinical important differences. Male sex has been shown to be associated with better quality of life in the general population and previously published normative data on the SRS-22r show the same tendencies, in line with our results on idiopathic scoliosis patients. Comparisons were also made with previously published national norm values, showing significantly reduced quality of life scores for the individuals with idiopathic scoliosis.

Subgroup analyses based on treatment and age were performed in males and females.

Comparing males and females stratified on age showed that males tended to have higher scores in the group aged ≤19 years and in the group aged ≥40 years. In the individuals aged 20-39 years there was no difference at all between males and females. Treated males (braced or operated) scored higher than their female peers.

One obvious strength of the current study is the number of individuals included, giving good precision of the estimates. Furthermore, another strength of this study is that the quality of life results were compared to an untreated groups of individuals, i.e. the natural history, and also compared towards the national norms. These results highly contribute to the knowledge of how much impact the condition and its treatment exerts on the quality of life. This study also has limitations; the cross-sectional design itself limits the findings to descriptive results. Furthermore, we did not obtain any radiographs in conjunction with the survey.

This is the hitherto largest comprehensive study comparing quality of life with validated instruments in males and females with idiopathic scoliosis. In conclusion, quality of life in males and females with idiopathic scoliosis is overall similar with slightly higher scores in males. The same trend is seen within treatment groups where small differences do exist between males and females. Based on our findings, and out of a quality of life perspective, clinicians do not need to consider sex when deciding on treatment. When comparing with national norm scales, both male and female groups did show significant and similar reductions of their quality of life. Such reduction was also noted for the untreated groups.


What factors characterise an indication for surgery in lumbar spinal stenosis?

Presentation by Mannion AF, Serra-Burriel M, Becker HJ, and others, on behalf of the LSOS-Group

The outcome of surgery for central spinal stenosis can be very variable, which suggests that improvements in patient selection may be warranted. There are currently no established decision-making pathways to indicate when surgery is appropriate or not in these patients. Some studies have tried to tackle this problem by identifying the specific characteristics of groups of stenosis patients receiving either surgery or non-operative care, in order to reveal factors of potential importance in decision-making [1,2]. However, if one simply compares groups in relation to the treatment received, then patients who declined surgery or were unable to undergo surgery for whatever reason are analysed in a group together with patients for whom no such surgical indication ever existed, and this confounds the analyses. No similar comparisons have been done in relation to groups of patients for whom there was considered to be an indication for surgery (regardless of the subsequent treatment received) versus those for whom it was deemed there was no indication. The present study sought to explore precisely this issue, by carrying out a cross-sectional analysis of the presenting/baseline characteristics (imaging, symptoms, clinical history, demographics, etc.) of patients with and without an indication for surgery who were participating in the “Lumbar Spinal Stenosis Outcome Study” (LSOS) [3]. The LSOS is a large-scale, multicentre (seven sites), prospective observational study being carried out in Switzerland on patients undergoing surgical or non-operative treatment for neurogenic claudication due to central spinal stenosis.








The main aim of the present analysis was to evaluate, in a comprehensive multivariable model, the baseline factors that distinguish between patients with and without an indication for surgery. Ultimately, with a follow-on study of the 12-month patient-reported outcomes, we would hope to evaluate whether the result of surgery was better in those patients who fulfilled the criteria for a “surgical indication”, as identified by our multivariable model. If successful, these criteria could then be used as a basis for developing decision aids to establish clearer thresholds/indications for surgery.

For the present analysis, we included patients with sufficiently complete data for all variables of predictive interest registered in the database of the LSOS. A total of 557 patients with full observations were identified for inclusion: 280 (50.3 per cent) females and 277 males (49.7 per cent), with a mean age 72.8 (8.5) years. A logistic forward stepwise regression model was built with this sample to predict “having an indication for surgery, yes/no” as documented by the clinician referring the patient into the study. We used 66 variables in the model (demographic, clinical history, neurological, radiological, patient self-ratings for pain, disability, HRQL, psychological status) and controlled for study site fixed-effects.

Overall, 415 (73 per cent) patients had an indication for surgery and 142 (27 per cent) did not. The final model had a successful fit with an area under the receiver-operating characteristic (ROC) curve of 90.74 per cent (95 per cent CI, 87.84-93.63 per cent). Significant predictors of “indication for surgery” in the multivariable model (*p<0.05, ** p<0.01) included: Swiss nationality*, numbness of the foot*, no physical therapy in the last six months*, an increase in symptoms in the last three months*, worse score on the “back trouble thermometer”**, lower psychological distress (SCL-K-9)**, higher Roland Morris disability score**, grade C** or grade D** morphological spinal stenosis grade (rootlet to cerebrospinal fluid ratio on axial MRI, or Schizas grade), and less disc degeneration** (Pfirrmann grade). In other words, the most important variables shown to be associated with having an indication for surgery were a combination of severe/worsening symptoms and disability, less psychological distress, and radiological evidence of severe stenosis without accompanying disc degeneration. Classification tree analysis (see Figure 1) revealed two hierarchical steps to be of greatest statistical significance in classifying those with an indication for surgery: the first step was whether symptoms had worsened in the last three months, and the second was the presence of the two most severe grades (C or D) of radiological stenosis according to the Schizas grading system.

Limitations of the study include the relatively low proportion of patients with no indication for surgery, which could have biased our estimates of both relevant variables and goodness of fit; and the potential individual differences between referring clinicians in their thresholds for surgery.

In conclusion, we suggest that the variables identified as having the greatest predictive power should be combined into an algorithm for further investigation as potential “appropriate use criteria” to assist with decision-making in lumbar spinal stenosis. Our further studies will examine whether patients with a higher probability of having a surgical indication based on these criteria actually go on to have a significantly better outcome after surgery.


Azimi P et al J Neurosurg Sci. 2017;61(4). 

Kim H-J et al  Eur Spine J. 2015;24(2):339-347. 

Steurer J et al. BMC Musculoskelet Disord. 2010;11(2008):254.