Impact of diabetes mellitus on spinal injection two-week outcomes in patients with predominant lower extremity pain.

Abstract

To evaluate the effect of diabetes mellitus (DM) on the effectiveness of spinal injections in patients with predominant lower extremity pain. This retrospective study included 218 patients (109 with DM and 109 without DM, matched by propensity scores; mean age 68.4 ± 9.6 years; 111 men) who underwent lumbar spinal injections in 2017. Treatment effectiveness was assessed by calculating the relative reduction in numerical rating scale (NRS) scores 2 weeks post-injection. A generalized estimating equation analysis was conducted to determine whether DM influenced overall treatment outcomes and to perform subgroup analyses based on age (< 50 years, ≥ 50 years), sex, pain onset (acute/subacute [≤ 6 months] vs. chronic [> 6 months]), preprocedural pain severity (mild [NRS 0‒3], moderate [NRS 4‒6], severe [NRS 7‒10]), and pain symmetry (symmetric vs. asymmetric). Fisher's exact test was used to compare adverse event rates between patients with and without DM. DM did not significantly impact the overall effectiveness of spinal injections 2 weeks after the injection (relative NRS reduction difference [95% confidence interval], -5.5% [-13.3% to 2.2%]; P = 0.163). However, patients with DM and chronic pain (-16.7% [-25.6% to -7.9%], P < 0.001) or mild preprocedural pain (-20.2% [-37.1% to -3.2%], P = 0.019) exhibited reduced treatment effects 2 weeks after the injection. Adverse event rates did not significantly differ between groups (P > 0.99). There was no difference in the spinal injection effectiveness between patients with and without DM 2 weeks after the injection. However, their therapeutic effect may be diminished in patients with chronic or mild preprocedural pain, warranting consideration of a potential diagnostic overlap with diabetic neuropathy.