Prospective observational cohort study.
To evaluate the comparative responsiveness of Oswestry Disability Index (ODI, version 2.0), 24-item Roland-Morris Disability Questionnaire (RMDQ), 12-item Short-Form Health Survey (SF-12) physical component score (PCS), and SF-12 mental component score (MCS) in patients that underwent micro-discectomy for lumbar disc herniation.
Responsiveness is a context specific term and no study has reported the responsiveness of ODI version 2.0, RMDQ, SF-12 PCS, or SF-12 MCS in discectomy patients.
Responsiveness was assessed using distribution-based methods (effect size [ES], standardized response means [SRM ], SRM difference between patients who improved and those who did not show improvement [SRM difference]), and the anchor-based method (area under the curve [AUC] of receiver operating characteristic curves). Values of ES and SRM higher than 0.8, and AUC value higher than 0.7 suggest adequate responsiveness. Outcome instrument demonstrating the highest value of SRM difference and AUC was considered the most responsive.
Responsiveness was assessed in 98 participants at a mean follow-up time of 12 weeks postoperatively. The overall ES of RMDQ, ODI, SF-12 PCS, and SF-12 MCS was 2.15, 2.11, 2.08, and 0.86, respectively. The overall SRM of ODI, RMDQ, SF-12 PCS, and SF-12 MCS was 1.36, 1.43, 1.24, and 0.65, respectively. The SRM difference in RMDQ, ODI, SF-12 PCS, and SF-12 MCS was 2.64, 2.26, 1.32, and 1.29, respectively. The AUC of ODI, RMDQ, SF-12 PCS, and SF-12 MCS was 0.96, 0.96, 0.83, and 0.83, respectively.
ODI, RMDQ, SF-12 PCS, and SF-12 MCS demonstrated adequate responsiveness in a homogenous cohort of patients who underwent lumbar micro-discectomy. ODI and RMDQ are equally responsive and, thus, are interchangeable instruments for region specific outcomes. Both the SF-12 PCS and SF-12 MCS can assess the quality of life following lumbar micro-discectomy.
The Oswestry Disability Index (ODI) and the Roland-Morris Disability Questionnaire (RMDQ) are ideal outcome instruments to assess functional outcomes; additionally, 36-item Short-Form Health Survey (SF-36) and its shorter versions such as 12-item Short-Form Health Survey (SF-12) are useful instruments to assess quality of life in patients with chronic low backache due to their established validity, reliability, and responsiveness [
To date, no studies have been published on the comparative responsiveness of ODI, RMDQ, SF12 PCS, and SF12 MCS in patients who underwent discectomy for lumbar disc herniation. Therefore, the purpose of the study was to evaluate and report the comparative responsiveness of these instruments. Region specific outcome instruments are more responsive than generic outcome instruments [
This was a quantitative, single site, prospective, observational, and cohort study comparing the responsiveness of validated outcome instruments in patients who underwent lumbar microdiscectomy. Our hospital research and clinical audit department approved this study protocol. Written informed consent was given by patients prior to surgical procedures. A prospective survey linked to service evaluation was undertaken. The study was an audit survey linked to service evaluation and approved by the research and clinical audit department of Countess of Chester Hospital, UK (approval no., 2520). As this study was an audit linked to service evaluation and improvement, institutional review board approval was waived off by the research and clinical audit department.
We recruited consecutive patients who underwent lumbar discectomy from 12th January 2009 to 4th October 2010 to the study. All the cases were either performed or supervised by one spinal surgeon. All the patients underwent non-instrumented surgical intervention (microdiscectomy) for lumbar disc prolapse and had primary complaints of leg pain which was unresponsive to nonoperative methods of treatment, such as analgesics and physiotherapy. The patients had a preoperative magnetic resonance imaging (MRI) scan to confirm the diagnosis. For a patient to be considered for surgical intervention, clinical examination had to correlate with the MRI scan findings. The exclusion criteria included patients undergoing instrumented spinal surgery, decompression for tumors, trauma, and infection. Written informed consent was given by patients prior to surgical procedures.
Data related to age, sex, primary indication of surgical intervention, level of pathology and type of surgical procedure were collected on the day of the surgery. Additionally, patients completed ODI version 2.0, RMDQ, and SF-12 preoperatively in pen and paper. Six to 12 weeks after the surgery, patients returned to clinic and repeated the ODI version 2.0, RMDQ, and SF-12. Patients were asked to evaluate the change in their clinical condition and rate the success of the operative intervention.
A 6-point transition scale (external anchor) was administered to patients to evaluate clinical condition following the spinal operation. Possible options on this scale were ‘cured,’ ‘much better,’ ‘a bit better,’ ‘the same,’ ‘a bit worse,’ and ‘much worse.’ This 6-point scale was converted into binary scale for the standardized response mean (SRM) and receiver operating characteristic (ROC) analysis. Patients who responded ‘cured’ and ‘much better’ were classed as ‘responders,’ and the remaining answers were classed as ‘non-responders.’
ODI version 2.0 has been shown to be valid, reliable, and responsive in chronic backache patients [
The RMDQ is a valid, reliable, and responsive instrument in chronic backache [
The SF-12 is a generic instrument shown to be valid, reliable, and responsive in patients with low backache [
IBM SPSS Statistics ver. 19.0 (IBM Corp., Armonk, NY, USA) was used to analyze data. The mean change in score was calculated by subtracting the mean postoperative score from the mean preoperative score. For the ODI and RMDQ, a positive value of the change score indicates improvement in clinical condition. For the SF-12, a negative value of the change score indicates improvement. Paired
The effect size (ES) is the ratio between the mean change in score of an outcome instrument and the SD of the score of the outcome instrument at the baseline. The SRM is the ratio between the mean change in score of an outcome instrument and the SD of change in score of the outcome instrument. Both ES and SRM were interpreted using Cohen’s criteria, where values of 0.2, 0.5, and 0.8 are considered as small, moderate, and large effects, respectively [
The SRM was calculated separately for patients who responded to the surgical intervention (SRM responder) and those who did not respond to the surgical intervention (SRM non-responder). The difference between SRM responder and SRM non-responder was called ‘SRM difference.’ The outcome instrument that has the highest value of ‘SRM difference’ is considered to be the most responsive [
ROC curve was used to discriminate responders from non-responders. The area under the curve (AUC) is calculated from the ROC curve, and an AUC value of 0.5 (no discriminatory accuracy) indicates poor ability to discriminate responders from non-responders; an AUC value of 1.0 (perfect discriminatory accuracy) indicates perfect discriminatory ability of the instrument. The AUC value of <0.70, 0.70–0.79, and ≥0.80 are considered as indicators of suboptimal responsiveness, optimal responsiveness, and excellent responsiveness along with accurate discriminatory ability of an outcome instrument, respectively [
The patient response rate regarding clinical condition subsequent to operative intervention was 93 cases (94.9%).
The overall ES of ODI, RMDQ, SF-12 PCS, and SF-12 MCS was 2.11, 2.15, 2.08, and 0.86, respectively. All were large effect as per Cohen’s criteria. The overall SRM of ODI, RMDQ, SF-12 PCS, and SF-12 MCS was 1.36, 1.43, 1.24, and 0.65, respectively. All were large effects as per Cohen’s criteria, except for SF-12 MCS which had a moderate effect. Setting self-reported change in clinical condition as an external anchor (
All outcome instruments demonstrated an adequate level of external responsiveness. It was difficult to conclude if ODI or RMDQ is better as 95% confidence interval (CI) of AUC of both outcome instruments tended to overlap. Setting self-reported change in clinical condition as an external anchor (
Our study shows that the commonly used region-specific patient-reported functional outcome instruments and generic quality of life outcome instruments accurately demonstrate improvement patients that underwent lumbar micro discectomy.
The SRM difference suggests that RMDQ had the highest responsiveness. As expected, RMDQ and ODI showed higher SRM difference compared to SF-12 PCS, which had a greater SRM difference than SF-12 MCS. Only one previous study has compared the SRM difference in ODI and RMDQ in a heterogenous cohort of patients with various surgically treated lumbar spine conditions such as lumbar canal stenosis, lumbar disc herniation, spondylolisthesis, and failed back syndrome [
We determined that both ODI and RMDQ have near perfect discriminatory ability. SF-12 PCS and SF-12 MCS had an identical AUC and as expected, the AUC of ODI and RMDQ was higher than that of SF-12, consistent with previous studies [
The large ES ODI, RMDQ, SF-12 PCS, and SF-12 MCS suggest optimal responsiveness for all patient-reported outcome instruments in our study. Three studies showed that the ES of ODI was higher than SF-12 PCS, which was greater than SF-12 MCS, which is consistent with our findings [
We showed large ESs for overall SRM values of ODI, RMDQ, and SF-12 PCS, and moderate effect for SF12 MCS. The overall SRM of ODI was higher than the SRM of SF-12 PCS, which was higher than that of SF-12 MCS. Four similar studies [
Studies comparing responsiveness of ODI and RMDQ after surgical intervention are scarce. One previous study reported higher SRM difference and higher AUC of ODI suggesting that the ODI was the most responsive instrument in their study [
The responsiveness of ODI and RMDQ was higher than that of SF-12 (PCS and MCS) based on greater ES, SRM, SRM difference, and the area under the curve. This is consistent with previous observations wherein region and condition-specific outcome instruments demonstrated higher responsiveness compared to generic outcome instruments [
In our study, responsiveness metric (SRM and ES) showed that SF-12 PCS had optimal responsiveness as SRM, and ES were above the threshold value of 0.8; also the AUC of SF-12 was above the threshold value of 0.8 and, thus, was excellent. Additionally, we observed higher responsiveness of SF-12 PCS compared to SF-12 MCS because SF-12 PCS exhibited higher values of ES, SRM, SRM difference, and the area under the curve compared to SF-12 MCS. The SRM of SF-12 MCS was suboptimal (SRM <0.8) but the remainder of parameters met the threshold of 0.8. This was on expected line especially as SF-12 MCS is a generic quality of life instrument. Therefore, we recommend the use of SF-12 PCS and SF-12 MCS to assess the quality of life after lumbar discectomy.
The present study is the first study to compare responsiveness of ODI, RMDQ, and SF-12 in homogenous group of patients that underwent micro-discectomy for lumbar disc herniation. The study was methodologically robust as we used both distribution- and anchor-based methods to evaluate responsiveness with a large sample size, based on COSMIN (consensus-based standards for the selection of health measurement instruments) checklist [
Several limitations of the study warrant discussion. External validity of the present study could be affected because this was a single center study and operated or supervised by a single surgeon performing a single specific surgical procedure. Thus, these results are applicable only to lumbar disc herniation treated with lumbar discectomy, and cannot be applied to patients having backache and sciatica secondary to other causes such as spondylolisthesis or spinal canal stenosis. The study was performed as part of service evaluation audit of spinal surgery services at our district general hospital. Patients were discharged to the care of the general practitioner at either 6 weeks or 3 months depending on the clinical progress of the patient. Hence, long-term follow-up was not undertaken in the study. Finally, the assignment of subjects into responders and non-responders was arbitrary and it is possible that the significance of the difference of SRM between the two groups could be due to the simplification of division.
ODI, RMDQ, SF-12 PCS, and SF-12 MCS demonstrate adequate responsiveness in a homogenous cohort of patients who underwent lumbar micro-discectomy for intervertebral disc herniation. The ODI and RMDQ are equally responsive and either instrument can be used to determine region specific outcome. SF-12 PCS and SF-12 MCS can be used to assess the quality of life after lumbar micro-discectomy.
No potential conflict of interest relevant to this article was reported.
Conception and design: KV, IB; data collection and assembly of data: KV, IB; provision of patients: IB; administrative support: IB; data analysis and interpretation: KV, IB; manuscript drafting: KV; and critical revision of manuscript and supervision: IB
Mean and 95% Cls of PRO. There was significant inprovement in values of all PROs as the 95% CIs of preoperative and final follow-up scores did not overlap. CI, confidence interval; PRO, patientreported outcome instruments; ODI, Oswestry Disability Index (version 2.0); RMDQ, 24-item Roland-Morris Disability Questionnaire; SF-12, 12-item Short-Form Health Survey; PCS, physical component score; MCS, mental component score.
Area under the curve of patient-reported outcome instruments using ROC curve using change in clinical condition from the patient’s perspective as an external anchor. ROC, receiver operating characteristic; ODI, Oswestry Disability Index (version 2.0); RMDQ, 24-item Roland-Morris Disability Questionnaire; SF-12, 12-item Short-Form Health Survey; PCS, physical component score; MCS, mental compo� nent score.
Demographic and clinical features at baseline for lumbar discectomy cohort
Characteristic | Value |
---|---|
Total no. of cases | 98 |
Age (yr, n=98) | 44±11.1 |
Gender (n=98) | |
Male | 51 (52.0) |
Female | 47 (48.0) |
Location of pain (n=97) | |
Leg pain | 17 (17.5) |
Backache and leg pain | 80 (82.5) |
Affected extremity (n=97) | |
Single leg pain | 89 (91.8) |
Pain in both legs | 8 (8.2) |
Sensory symptoms (n=93) | |
Presence of paresthesias | 65 (69.9) |
Presence of numbness | 65 (69.9) |
Side of operation (n=97) | |
Right | 49 (50.5) |
Left | 45 (46.4) |
Bilateral | 3 (3.1) |
No. of affected levels (n=98) | |
Single level | 96 (98.0) |
Two levels | 2 (2.0) |
Affected levels (n=98) | |
L3/L4 | 2 (2.2) |
L4/L5 | 45 (45.9) |
L5/S1 | 49 (50.0) |
L4/L5+L5/S1 | 2 (2.0) |
Type of surgery (n=98) | |
Primary discectomy | 85 (86.7) |
Revision discectomy | 13 (13.3) |
Values are presented as mean±standard deviation or number of cases (%).
Patient response and global assessment of change after operative intervention (N=93)
Outcome response | Frequency (%) |
---|---|
Change in clinical condition from patient perspective (external anchor) | |
Patient response | |
Cured | 29 (31.2) |
Much better | 53 (57.0) |
Bit better | 5 (5.4) |
The same | 2 (2.2) |
Bit worse | 2 (2.2) |
Much worse | 2 (2.2) |
Created sub-groups (for SRM difference and AUC) | |
Improved (responder) | 82 (88.2) |
Unimproved (non-responder) | 11 (11.8) |
SRM, standardised response mean; AUC, area under the curve.
Overall pre- and postoperative values, difference between preoperative and postoperative values, and statistical significance of the change in the value of outcome instruments
Outcome instrument | Preoperative | Postoperative | Change in score | 95% Confidence interval of change in score | |
---|---|---|---|---|---|
ODI | 51.9+15.02 | 20.9+20.32 | 31.6+23.21 | 26.8 to 36.5 | <0.0001 |
RMDQ | 14.9+4.68 | 5.0+5.94 | 10.1+7.00 | 8.6 to 11.5 | <0.0001 |
SF-12 PCS | 29.5+6.41 | 42.7+10.33 | -13.3+10.73 | -15.6 to -11.0 | <0.0001 |
SF-12 MCS | 42.2+11.03 | 51.4+10.88 | -9.5+14.56 | -12.6 to -6.4 | <0.0001 |
Values are presented as mean±standard deviation, unless otherwise stated.
ODI, Oswestry Disability Index (version 2.0); RMDQ, 24-item Roland-Morris Disability Questionnaire; SF-12, 12-item Short-Form Health Survey; PCS, physical component score; MCS, mental component score.
SRM and AUC in different sub-groups of patient that underwent lumbar micro-discectomy
Responsiveness parameter | ODI | RMDQ | SF-12 PCS | SF-12 MCS |
---|---|---|---|---|
SRM responder | 1.91 | 2.09 | 1.53 | 0.83 |
SRM non-responder | -0.35 | -0.55 | 0.21 | -0.46 |
SRM difference | 2.26 | 2.64 | 1.32 | 1.29 |
Receiver operating curve | ||||
AUC | 0.96 | 0.96 | 0.83 | 0.83 |
95% Confidence interval of AUC | 0.92 to 1.00 | 0.92 to 1.00 | 0.73 to 0.94 | 0.72 to 0.93 |
<0.0001 | <0.0001 | 0.001 | 0.001 |
The outcome instrument having the highest value of SRM difference is considered to be the most responsive instrument. The outcome instrument having the highest value of AUC is considered to be the most responsive instrument.
SRM, standardised response mean; AUC, area under the curve; ODI, Oswestry Disability Index (version 2.0); RMDQ, 24-item Roland-Morris Disability Questionnaire; SF-12, 12-item Short-Form Health Survey; PCS, physical component score; MCS, mental component score.