P123. Have robotic procedures improved perioperative and long-term outcomes after overcoming the learning curve? A 2-year analysis

• Published in: The spine journal Vol. 22; no. 9; p. S186
• Main Authors: Passias, Peter G, Krol, Oscar, Imbo, Bailey, Joujon-Roche, Rachel, Tretiakov, Peter, Williamson, Tyler, Owusu-Sarpong, Stephane, Diebo, Bassel G., Vira, Shaleen N.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.09.2022
• ISSN: 1529-9430; 1878-1632
• DOI: 10.1016/j.spinee.2022.06.380

Robot-assisted surgical techniques are being increasingly implemented in spine surgery to increase accuracy and mitigate surgeon fatigue. There are known variations in the learning curve required for effective use; however, literature on the effect of robot-assisted techniques on perioperative outcomes, after overcoming the learning curve, remains scarce. To identify differences in perioperative outcomes and complication rates between robot-assisted and unassisted lumbar interbody fusions (LIFs), and the learning curve based on surgeon caseload. Retrospective review of a single-center stereographic database. A total of 230 robot-assisted lumbar spine fusion surgery patients. Perioperative outcomes [EBL, LOS], postoperative complication rates, radiographic alignment, HRQLs. Robot-assisted spinal fusion cases with baseline and up to 2-year HRQL and radiographic followup were isolated from a single-surgeon database. Cases were ranked by the date of surgery into 2 cohorts representing the first and last 50% of cases or early vs late, respectively. Univariate analysis was used to assess baseline, surgical and radiographic profiles of the 2 cohorts. A total of 230 patients met inclusion criteria (age: 56±12.5, BMI: 30±6, 42% female). Average levels fused was 2.3±1.5, mean operative time was 281±110min, EBL 298±274ml and LOS was 3.9±2.5. Average UIV was L3 and average LIV L5. Two percent of patients had an ALIF, 36% XLIF/LLIF, 57% TLIF, 6% PLIF and 19% had an osteotomy. The late group was older (58 vs 51 years) and had a higher CCI (1.3 vs.7). The late group had a greater number of levels fused (2.5 vs 1.9, p=.005), greater usage of interbody devices (1.5 vs 1.3, p=.033), more laminectomies (31% vs 10%), less decompressions (51% vs 98%) and more osteotomies (25% vs 13%). Late group had a lower usage of ALIF (5% vs 1%), higher usage of XLIF/LLIF (20% vs 48%), lower usage of TLIF (42% vs 84%). The late group had a lower EBL (272ml vs 331ml, p=.16), shorter op-time (261min vs 302min, p=.03), and comparable LOS (4.3 vs 4.3 p=.74). The late group had a lower rate of discharge to rehab (8% vs 50%, p<.001). Rate of reoperations, readmissions, and overall complications trended lower in the late group. At BL, late group had a comparable ODI (70 vs 64, p=.27) and BL NRS back pain (8.5 vs 8, p=.3). At 3M, late group had a lower ODI (48 vs 59 ,p=.114) and comparable NRS back (5 vs 5.8, p=.3). At 2-year followup, patients in the late group had a lower ODI (34 vs 42), lower NRS back pain (3.9 vs 5.2), and lower NRS leg pain (3.14 vs 4.2). Patients operated on in the second half of surgeons' full robotic case load demonstrated an improved perioperative course in the form of a shorter LOS, lower EBL, shorter operative time and a substantial reduction in proportions of patients discharged to rehabilitation facilities. Outcomes at 2 years show superior recovery as evidenced by lower pain and ODI scores. These findings suggest that once the robotic learning curve is overcome, outcomes may improve substantially. This abstract does not discuss or include any applicable devices or drugs.