Our results demonstrate that conventional lumbar laminectomy without fusion is a safe treatment for spinal stenosis in patients 70 years and older. There were no differences in outcomes between patients who experienced complications and those who did not. Unselected patients recruited from general clinical practice can expect statistically significant improvement in health related quality of life, functional status and pain. Patients' statements about their benefit of the operation and the calculated effect size on quality of life indicate that these improvements are clinically meaningful. Our results support results from Fredman et al and Ragab et al that within an elderly population increased age is not a predictor of worse outcome [7, 8].
We used the EQ-5 D score to assess health related quality of life. It has shown good reliability and proved useful for monitoring outcome of patients undergoing low-back surgery . At baseline mean EQ-5 D index score was 0.32 which is lower compared to a similar age-group of the general Swedish population (0.79) , which is expected to be comparable to a Norwegian population due to similarities between the countries with respect to genetics, demography, socio-economic and health care system. Disease specific mean EQ-5 D index score in patients between 20-88 years suffering low back pain in Sweden is 0.66 , which is slightly better than the mean score after surgery in our material (0.60). We believe this, at least partly, can be explained by the fact that our patients had longstanding chronic low back problems (median duration of symptoms of 100 weeks) and 29 (28.7%) of our patients had ASA score of greater than 2, indicating severe co- morbidity. Baseline EQ-5 D and improvement after surgery is comparable to what Jansson et al demonstrated recently in a Swedish cohort undergoing decompressive surgery for spinal stenosis, with or without fusion, in a slightly younger population .
The ODI at baseline was comparable to what others have reported [15, 20, 21] Mean change in ODI in the present study was 5.1 points less at one year than in surgically treated patients in the SPORT trial from Weinstein et al, although comparable to results from previous trials [4, 20, 22]
In this population 70 years and older there does not seem to be an additional age-effect with respect to outcome or complications. Here we report complications occurring at comparable rates with other trials including elderly patients, although different study design and methods for registration makes direct comparison difficult [7–9, 23]. In our study, patients above median age did not have worse outcomes or more complications compared to patients below the median age.
Whether type of surgery is a predictor for outcome remains controversial [1, 21, 24–27]. A recent study looking at trends in elderly undergoing spinal surgery for lumbar spinal stenosis reported increased use of complex fusions . Compared to Rosen et al, using minimal invasive technique, we report higher major complication rates (6% vs 0%) but less overall complications (18% vs 39%). Also, we report slightly (4.7 points) less improvement in ODI, although their follow-up time was somewhat shorter . Glassman et al demonstrated a good clinical outcome, but more complications, in patients 65 years or older treated with single-level lumbar arthrodesis . Other studies have demonstrated that fusion procedures were associated with increased risk of major complications and death compared to decompression alone [28, 29]. These findings could indicate that minimal invasive decompression might be a treamtnet as effective, and probably safer, compared to conventional laminectomy with or without fusion.
The only predictor found was that longer duration of radiating pain in the leg(s) predicted less improvement in functional status measured with changes in ODI. Although, it is not known for how long patients should try conservative treatment for lumbar spinal stenosis or when the prognosis for a favorable outcome starts to decline. Studies in patients with sciatica indicate unfavorable outcome if pain has lasted longer than 6-8 months, but we did not reproduce this finding in an elderly population with lumbar spinal stenosis [17, 30]. Future studies might help us to identify the best timing of surgery.
Since our patients represent a consecutive sample from regular practice at a clinic where laminectomies are performed by several surgeons with different level of experience, we expect the external validity of our study to be good, and the data would be suitable for risk factor analyses. Although, the surgeons selected patients for surgery as part of regular practice at our institution, and this selection at doctors' discretion could lead to selection bias. Complications were prospectively assessed at 3 and 12 months - and not at discharge from the hospital or in retrospect by chart reviews. This may reduce sensitivity for minor complications, but is sensitive for delayed and major complications. The study was not designed to assess the efficacy of the surgical treatment, but it indicates that the effectiveness is acceptable. Weaknesses in our study is a relative short follow-up with for evaluating HRQL, functional status and pain - but long enough to assess safety. We used outcomes after re-operations in the 7 patients that were registered more than once in the study period, this could lead to biased results in favor of surgical treatment. Due to a small sample size it is difficult to make any meaningful conclusion regarding mortality.