The institutional ethical committee approved all aspects of this study that involved human subjects. All subjects signed an informed consent form to participate in the study under the Declaration of Helsinki.
Study population
The study population consisted of the same 40 subjects that participated in the ESLR interrater reliability study [14], 20 in the sciatic symptomatic group and 20 in the control group including nonspecific LBP patients and hip pain patients. These subjects were gathered from consecutive institutional spine center patients willing to participate in the study in order of appearance to the center. First, the study controller took a complete patient history and performed an in-depth clinical examination including neurological examination of the lower leg and ESLR. The subject allocation to groups was made by the study controller based on the combination patient history, symptoms and clinical findings which have been shown to be reliable in detecting sciatic patients. [11, 15] A complete list of inclusion and exclusion criteria is shown in Table 1. After group allocation, the interrater reliability test part for ESLR was performed by 2 independent blinded examiners (physiatry residents). [14] After the ESLR, each subject’s MRI of the lumbar spine was evaluated. The results of the traditional SLR (negative/positive) were retrieved from the patient’s medical records performed by the treating physician.
ESLR procedure and interpretation
The ESLR started similar to the traditional SLR: the patient lies supine on the examination table and the examiner passively lifts the subject’s leg with knee straight (fully extended), hip in neutral rotational position and ankle hanging free. The leg raise was continued until the first symptoms were evoked or the subject’s ongoing symptoms in the lower extremity are aggravated by 30%. If no responses are evoked by the hip flexion angle reaches 90 degrees, the leg raise is ceased, and the test is considered negative. The ESLR differs from the traditional SLR as the symptoms emerge: i) the evoked responses do not need to reach below the knee, ii) these responses do not need to emerge before 70 degrees but can happen anywhere from 0-90 degrees of hip flexion, and iii) a structural differentiation maneuver is added to the SLR. To evaluate whether the evoked responses are neural or musculoskeletal in origin, a location-specific structural differentiation movement is added to the test at the hip flexion angle where these responses have been evoked/provoked. The added movement is selected based on the location of the evoked responses: If the symptoms are located distally, below the knee, a hip internal rotation is performed at the same hip flexion angle of evoked responses. In case the symptoms are provoked proximally in the buttock and/or hamstring area, ankle dorsiflexion is the differentiating movement similar to the Bragard test. An integral part in performing these structural differentiation movements is that, given the structural continuum of the nervous system, a movement known to move the sciatic neural structures and its contiguous nerve roots [16-18] at an asymptomatic musculoskeletal location may evoke referred responses in the area of interest.
The ESLR was deemed positive if: i) at least part of the patient's symptoms were evoked (or present symptoms increased by 30%) with the traditional SLR, and ii) a differentiating movement that moves the nerve roots - but not the spine - increased the evoked symptoms, e.g. ankle dorsiflexion or hip internal rotation.
MRI procedure and classification
Each subject’s MRI was performed with a 1.5 T MRI device (Siemens Magnetom Aera, Erlangen, Germany). The imaging area was from L2 level to S3 level and centered to L4-L5 disc. T1- (repetition time (TR) 500ms, echo time (TE) 11ms, 20 slices, slice thickness (ST) 3mm, field of view (FOV) 300mm, in-plane resolution 0.8x0.8mm, flip angle (FA) 150°), T2-weighted sagittal sequences (TR 4000ms, TE 95ms, 20 slices, ST 3mm, FOV 300mm, in-plane resolution 0.8x0.8mm, FA 150°) and T2-axial images (TR 1500, TE 119ms, 80 slices, ST 2mm, FOV 200mm, in-plane resolution 0.8x0.8mm, FA 150°) were taken to detect the presence of nerve root compression (NC) and/or LDH in L4/5 and L5/S1 discs. The images were viewed and analysed with Sectra PACS workstation (Sectra Workstation IDS7, version 21.2.5.6173 – 2019 – Sectra AB, Sweden).
The MRI results were classified following the recommendations by Li et al. [19] For NC we used a 5-point scaling by van Rijn et al. [20], where groups 4 and 5 (i.e., possible root compression and definitely root compression, respectively) were judged as NC-positive. We used the Combined Task Force classification as described by Fardon et al. [21] for the definition of LDH where extrusion, herniation or sequestration was required to be visible in the L4/L5 or L5/S1 discs on the subjects’ MRI scans. The MRI images were analyzed independently and blinded from the patient’s clinical findings by 2 senior radiologists and a spine specialist clinician. From these results, a consensus for the LDH and NC results was formed. If there was a difference in the initial outcome (LDH or NC), the scans were analyzed together to form a conclusion for the outcome.
Outcome measures and statistical analysis
The size of the study population was determined following the recommendations for reliability studies in clinical sciences where the sample size of 40 was required for the Kappa statistic to be significantly greater than 0.40 (assuming 80% power and .05 significance), [22, 23] where 0.40 represents the value of the null hypothesis. The data were analyzed using IBM SPSS Statistics version 26 and Microsoft Excel for Office365 programs.
Demographic variables were expressed as means with standard deviations. The prevalence of LDH and NC was calculated within the study groups. The outcomes of ESLR and SLR were compared with the MRI results and cross-tabulated from which the according sensitivity and specificity values were calculated. Cohen’s Kappa statistic was used to assess the agreement between the MRI and both the ESLR and SLR results. 95% confidence intervals (95%CI) were calculated. To obtain the odds ratio (OR) with positive ESLR or SLR results for LDH or NC, a binary logistic regression analysis with subjects’ age, gender, height and weight was performed. P-values <0.05 were set to indicate statistically significant results. ESLR test’s validity was assessed for its positive result by combination of interrater agreement (Cohen’s Kappa) and percentage prevalence of both LDH and NC.