The Choice of Spinal Cord Stimulation Versus Targeted Drug Delivery in the Management of Chronic Pain: Validation of an Outcomes Predictive Formula.

OBJECTIVE

In 2020, Mekhail et al published a formula that predicted the likelihood of a successful outcome for those who undergo spinal cord stimulation (SCS) for long-term pain management, based on retrospectively collected clinical and demographic data from one major medical center. The aim of this study is to validate such a predictive formula, prospectively, in a cohort of patients from multiple medical practices that are more representative of real-life clinical practice.

MATERIALS AND

METHODS

For the study, 939 patients who underwent successful SCS or targeted drug delivery (TDD) trials at multiple independent medical centers in the USA were enrolled into the Medtronic product surveillance registry data base before they underwent SCS or TDD device implantation, from 2018 to 2020. The registry data were collected prospectively but not specifically for this study. The data examined included demographic information, pain diagnosis, pain scores (visual analog scale [VAS]), Oswestry Disability Index scores, and quality-of-life scores at baseline and six months after implant. Because our goal is to validate the previously published predictive formula, in addition to the outcomes data previously mentioned, we collected the variables necessary for such a task: sex, age, depression, the presence of neuropathic pain, spine-related pain diagnosis, and persistent spinal pain syndrome "post laminectomy syndrome." Spine-related pain diagnosis included subjects with chronic spine pain who never had back surgery and whose pain was not radicular nor neuropathic.

RESULTS

Of 619 patients with SCS, 138 (22%) achieved ≥ 50% reductions of the baseline VAS at six months. The logistic model predicts SCS success with an area under the receiver operating characteristic curve (AUC) of 80% in the current validation data set. Of 320 patients with TDD, 147 (46%) achieved ≥ 50% reduction of the baseline VAS at six months. The logistic model predicts TDD success with an AUC of 78% in the current validation data set.

CONCLUSION

The study provides real life validation of the previously published predictive formula(4).