Background & Aims
Lumbar spinal stenosis with spondylolisthesis is a common, disabling condition. Decompression surgery, with or without additional fusion is indicated for patients who experience minimal improvement with non-surgical care[1]. Additional fusion surgery is expensive, potentially unsafe, and ineffective beyond decompression alone on average[2,3]. However, fusion surgeries are increasing[4], and experts suggest that certain patients benefit from additional fusion[1].
A previous trial aiming to investigate whether expert identification of appropriate surgery is superior to random assignment (i.e., whether specific groups benefit) was unable to recruit enough participants, and it is unlikely future studies will[5]. Observational data must inform the precise indications for surgery[6]. Target trial emulation is a framework that improves causal inference from observational data[7]. We aim to estimate whether there are subgroups for whom fusion is superior compared to decompression alone.
Methods
Using observational data from the Norwegian Spine Surgery Register (NORSpine) emulated the NORDSTEN trial – the index trial[2]. As in the trial, patients were aged 18-80 with a diagnosis of spondylolisthesis and pain in the back or legs for ?3 months. Participants were excluded if they had previous surgery at the level of the spondylolisthesis. We compared decompression alone, to decompression with additional fusion. Our primary outcome was the between-group difference on the Oswestry Disability Index at 1-year follow-up. Spinal surgeons decided subgroups based on plausible effect-modifying variables, these were: back pain intensity, body mass index, sex, age, ASA score, smoking status, and difficulty walking. We estimated treatment effects and controlled for confounding using standardisation, with a sensitivity analysis using inverse probability weighting. We checked how well we controlled for confounding by comparing effect estimates between the NORDSTEN trial, and our emulation.
Results
In the NORSpine registry between 1 Jan 2007 to 17 June 2021, 3817 patients underwent surgery for lumbar spinal stenosis with spondylolisthesis. 2179 patients were included in our primary analysis, 770 received fusion surgery, 1409 received decompression alone. Our analysis of the primary outcome in the observational emulation of the NORDSTEN trial had ‘estimate agreement’ with the index trial (trial finding = 0.7 / 100 95% CI -2.8 to 4.3; emulation finding = 2.4 / 100, 95% CI 0.7 to 4.1), in favour of fusion surgery. Differences in subgroup effect estimates ranged from -1.4 to 2.3 (with negative results favouring decompression, positive favouring fusion), none of which were statistically significant differences between subgroups. There were no meaningful differences to our results when estimated using inverse probability weighting, except lower precision.
Conclusions
Our results indicate that we were able to successfully reduce confounding in our primary analysis emulating the NORDSTEN trial [2], indicating our results may be viewed as causal. With the larger sample attained from the NORSpine Registry, we extended the analysis of the index trial to include estimation of subgroup effects. Our results indicate there are unlikely to be any meaningful differences in benefits from decompression with additional fusion, or decompression surgery alone.
References
1. Katz JN, Zimmerman ZE, Mass H, et al. Diagnosis and Management of Lumbar Spinal Stenosis: A Review. JAMA 2022;327(17):1688-99. doi: 10.1001/jama.2022.5921
2. Austevoll IM, Hermansen E, Fagerland MW, et al. Decompression with or without Fusion in Degenerative Lumbar Spondylolisthesis. New England Journal of Medicine 2021;385(6):526-38. doi: 10.1056/NEJMoa2100990
3. Försth P, Ólafsson G, Carlsson T, et al. A Randomized, Controlled Trial of Fusion Surgery for Lumbar Spinal Stenosis. New England Journal of Medicine 2016;374(15):1413-23. doi: 10.1056/NEJMoa1513721
4. Sastry RA, Chen J-S, Shao B, et al. Patterns in Decompression and Fusion Procedures for Patients With Lumbar Stenosis After Major Clinical Trial Results, 2016 to 2019. JAMA Network Open 2023;6(7):e2326357-e57. doi: 10.1001/jamanetworkopen.2023.26357
5. Ghogawala Z. SLIP II Registry: Spinal Laminectomy Versus Instrumented Pedicle Screw Fusion (SLIP II) – NCT03570801. Clinicaltrials.gov, 2016.
6. Hernán MA, Wang W, Leaf DE. Target Trial Emulation: A Framework for Causal Inference From Observational Data. JAMA 2022;328(24):2446-47. doi: 10.1001/jama.2022.21383
7. Hernán MA. Methods of Public Health Research — Strengthening Causal Inference from Observational Data. New England Journal of Medicine 2021;385(15):1345-48. doi: 10.1056/NEJMp2113319
Presenting Author
Harrison Hansford
Poster Authors
Harrison Hansford
BSc(Hons)
UNSW
Lead Author
Aidan Cashin
UNSW/Neuroscience Research Australia
Lead Author
Margreth Grotle
PhD
Oslo Metropolitan University
Lead Author
Bjørnar Berg
PhD
Oslo Metropolitan University
Lead Author
James McAuley
PhD
Neuroscience Research Australia
Lead Author
Tor Ingebrigsten
MD
The Arctic University of Norway
Lead Author
Christian Hellum
MD
University of Oslo
Lead Author
Matthew Jones
PhD
University of New South Wales
Lead Author
Hopin Lee
PhD
University of Exeter
Lead Author
Tore Solberg
MD
The Arctic University of Norway
Lead Author
Ivar Austevoll
Haukeland University Hospital
Lead Author
Issa Dahabreh
Harvard University
Lead Author
Topics
- Epidemiology