Robotic-assisted pedicle screw placement fails to reduce overall postoperative complications in fusion surgery

Abstract

Background context: Surgeons have increasingly adopted robotic-assisted lumbar spinal fusion due to indications that robotic-assisted surgery can reduce pedicle screw misplacement. However, the impact of robotic-assisted spinal fusion on patient outcomes is less clear.

Purpose: This study aimed to compare rates of perioperative complications between robotic-assisted and conventional lumbar spinal fusion.

Study design/setting: Retrospective cohort study.

Patient sample: A total of 520 patients undergoing lumbar fusion were analyzed. The average ages of patients in the robotic-assisted versus conventional groups were 60.33 and 60.31, respectively (p=.987). Patients with a diagnosis of fracture, traumatic spinal cord injury, spina bifida, neoplasia, or infection were excluded.

Outcome measures: This study compared the rates perioperative major and minor complications for elective lumbar fusion between each cohort.

Methods: This study screened hospital discharges in the United States from 2010 to 2014 using the National Inpatient Sample and the Nationwide Inpatient Sample (NIS). The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes were used to identify 209,073 patients who underwent conventional lumbar fusion (ICD 81.04-8) and 279 patients who underwent robotic-assisted lumbar fusion (ICD 81.04-8 and ICD 17.41, 17.49). Major and minor complications were identified using ICD-9-CM diagnosis codes. The robotic-assisted and conventional fusion groups were statistically matched on age, year, sex, indication, race, hospital type, and comorbidities. Univariate and multivariate logistic regression were used to compare risks of major and minor complications.

Results: We matched 257 (92.11%) robotic-assisted patients with an equal number of patients undergoing conventional lumbar fusion. Minor complications occurred in 16.73% of cases in the conventional group and 31.91% of cases in the robotic-assisted group (p<.001). Major complications occurred in 6.61% of the conventional cases compared to 8.17% of robotic-assisted cases (p=.533). For robotic-assisted fusion, multivariate analysis revealed that there was no difference in the likelihood of major complications (OR=0.834, 95% CI=0.214-3.251) or minor complications (OR = 1.450, 95% CI=0.653-3.220).

Conclusions: In a statistically matched cohort, patients who underwent robotic-assisted lumbar fusion had similar rates of major and minor complications compared to patients who underwent conventional lumbar fusion.

Keywords: Complication; Cost; Fusion; Inpatient outcomes; Length of stay; Lumbar spine; National Inpatient Sample; Robotic-assisted surgery.