Rationing by assumption: De-implementation of spinal injections requires rigorous comparative evidence.

Ballengee et al. propose an implementation science framework for de-implementing epidural steroid injections (ESIs) for chronic low back pain (CLBP). This commentary identifies several critical limitations of their proposal. First, the target population is inadequately defined; the term "non-specific" CLBP is never operationalized, inviting misapplication to patients with specific pathoanatomical diagnoses. Second, the evidentiary anchor, a recent BMJ clinical practice guideline, is subject to a formal retraction request from 34 professional organizations citing inappropriate pooling of heterogeneous procedures, data extraction errors, and violations of GRADE methodology. Third, the proposed replacement therapies (including cognitive behavioral therapy, physical therapy modalities) demonstrate comparably modest effect sizes, yet are held to a lower evidentiary standard. Fourth, indirect cost-utility comparisons do not favor de-implementation: published analyses consistently place ESIs within accepted willingness-to-pay thresholds, while behavioral alternatives carry higher cost-per-QALY ratios. Fifth, the structural incentive analysis applied to proceduralists must be applied symmetrically to the professions that would absorb redirected clinical volume. De-implementation should be held to the same evidentiary rigor as implementation, requiring head-to-head comparative effectiveness and cost-effectiveness data before removing a therapeutic modality from the clinical armamentarium. PERSPECTIVE: This commentary argues that de-implementing epidural steroid injections without head-to-head comparative effectiveness and cost-effectiveness evidence risks restricting access to a therapy that remains cost-effective and that may facilitate participation in multimodal care.