Impact of Dialysis on 30-Day Outcomes After Spinal Fusion Surgery for Pathologic Fractures: Insights from a National Quality Registry.

Background: Patients with chronic renal failure undergoing hemodialysis have been shown to have poor overall health, osteoporosis, and altered bone metabolism. However, the impact of hemodialysis on patient outcomes after spinal fusion remains unknown. We sought to assess the effect of dialysis on 30-day perioperative and postoperative outcomes after cervical and lumbar fusion for pathologic compression fractures.

Methods: We queried the National Surgical Quality Improvement Program from 2009 to 2016 for patients undergoing cervical or lumbar fusion for compression fractures. Three-to-one propensity score matching using sex, age, body mass index, and number of operated levels was used to match patients not undergoing dialysis with those undergoing dialysis. Multivariable conditional regression was used to identify the association between dialysis and 30-day clinical outcomes, after adjusting for confounders.

Results: A total of 48,492 patients undergoing cervical fusion were identified; 156 (0.32%) of these were on dialysis. On multivariable regression, dialysis dependency was associated with increased operative time (regression coefficient [coef.], 15.93; 95% CI, 0.4-31.5; P = 0.045), length of stay (coef. 6.06; 95% CI, 4.64-7.48; P < 0.001), 30-day readmissions (odds ratio [OR], 1.07; 95% CI, 1.02-1.12; P = 0.009), any complications (OR 1.08; 95% CI, 1.03-1.13; P = 0.002), and serious complications (OR, 1.08; 95% CI, 1.02-1.14; P = 0.012). A total of 25,417 patients undergoing lumbar fusion were identified; 51 of these (0.2%) were on dialysis. On multivariable regression, dialysis dependency was associated with significantly higher length of stay (coef. 2.98; 95% CI, 1.28-4.68; P < 0.001).

Conclusions: Our analyses indicated that dialysis dependency is associated with poor perioperative and postoperative outcomes after cervical/lumbar fusion for pathologic compression fractures.