Neurosurgery. 2026;99(3):643-653
The outcome of cervical surgery encompasses not only the patient’s condition at one year, but also the disability experienced while recovering. Asada and colleagues examine this distinction by comparing cervical disc replacement with anterior cervical discectomy and fusion using a cumulative, model-based outcome measure. Their findings associate arthroplasty with a lower burden of neck-related disability during the first postoperative year, without a corresponding advantage in cumulative arm pain. This approach adds a useful dimension to preoperative counseling, although its observational design and the limited validation of the new measure require a cautious interpretation.
Objective
To develop and internally evaluate a modified integrated health state (mIHS) measure and compare cumulative postoperative disability after one- or two-level cervical disc replacement (CDR) and anterior cervical discectomy and fusion (ACDF) for degenerative cervical radiculopathy.
Methods
The authors retrospectively analyzed a prospectively collected, multisurgeon institutional registry covering February 2016 to September 2024. Of 1,273 screened patients, 327 were included: 159 underwent ACDF and 168 underwent CDR. Hybrid constructs, procedures involving three or more levels, missing baseline Neck Disability Index (NDI) scores and baseline NDI below 20 were excluded. The cohort flowchart also excluded myelopathy and trauma.
NDI and neck and arm pain scores were collected before surgery and at intervals through 12 months. Postoperative scores were normalized to baseline, and nonlinear mixed-effects models estimated recovery trajectories. The mIHS summarized the area under these modeled trajectories; it is not a conventional NDI point score. Propensity-score overlap weighting balanced measured baseline characteristics. Patients with shorter follow-up remained eligible, with missing postoperative observations handled under a missing-at-random assumption.
Main results
The one-year NDI-based mIHS was lower after CDR than ACDF (2.68 versus 3.71; difference 1.03, 95% CI 0.80-1.25; p < 0.001), corresponding to a 27.7% relative reduction in this modeled measure. Cumulative neck pain also favored CDR; cumulative arm pain did not differ significantly. Two-level CDR had a similar mIHS to one-level CDR, whereas two-level ACDF carried greater cumulative disability than one-level ACDF. These results describe recovery burden, not a 27.7% reduction in the conventional NDI score.
Interpretation – operative relevance
For patients in whom both procedures are appropriate, the findings support discussing the course of recovery alongside the expected endpoint. Preserved segmental motion after CDR offers a plausible explanation, but this study does not establish it as the cause of better recovery. Nor does it justify extending arthroplasty indications, choosing a particular prosthesis or altering decompression technique. The practical contribution is to procedure selection and counseling rather than a new operative maneuver or proof of long-term superiority.
Limitations
Treatment was not randomized. Before weighting, CDR patients were younger and differed in comorbidity, pathology and number of treated levels. Statistical adjustment cannot remove unmeasured confounding, including the reasons a surgeon selected one procedure over another. Postoperative analgesic consumption was not reliably captured.
Excluding low baseline NDI scores limits applicability to patients with milder disability. Model-based estimates depend on assumptions about recovery trajectories and missing data. The mIHS underwent internal validation only, and its minimum clinically important difference remains undefined; statistical separation therefore cannot be equated with an established patient-perceived benefit threshold.
Follow-up was limited to one year. Imaging was incomplete and nonstandardized, making fusion, motion and heterotopic ossification analyses vulnerable to selection bias. The study does not establish comparative long-term revision rates or device-specific safety.
Clinical takeaway
When counseling suitable candidates for cervical arthroplasty or ACDF, discuss how recovery may unfold, not simply the expected one-year result. This study associates CDR with less cumulative neck-related disability, but its novel outcome measure and observational design do not support a universal preference for arthroplasty.