Lightning Podiums: Value and Outcomes in Spinal Surgery - Room 801B

Presented by: R. Rampersaud


R. Rampersaud(1,2),(3), K. Sundararajan(1), A.V. Perruccio(1,4), R. Gandhi(1,2), J.R. Davey(1,2), K. Syed(1,2), C. Veillette(1,2), N.N. Mahomed(1,2)

(1) Krembil Research Institute, University Health Network, Arthritis Program, Toronto, ON, Canada
(2) University Health Network, Division of Orthopaedic Surgery, Toronto, ON, Canada
(3) Krembil Neuroscience Centre, Toronto, ON, Canada
(4) University of Toronto, Dalla Lana School of Public Health, Toronto, ON, Canada


Purpose: Surgeries for joint/spine disorders are among the most frequently performed elective procedures; they continue to demonstrate an upward trend in global annual volumes. Over 1.6 million joint replacement and spinal fusion surgeries were performed in the US in 2012 and 120,000 were performed in Canada in 2014, with corresponding hospital costs of US$32 billion and CAN$1 billion respectively. Despite these substantial costs, 20-25% of these procedures result in poor patient outcomes. This suggests that at a minimum, approximately US$6.4 billion and CAN$200 million are spent each year on elective joint/spine procedures for people who do not benefit from them. The purpose of this study was to evaluate the one year cost-utility of elective joint replacement and spinal decompression with or without fusion with respect to patient-reported surgical outcome.

Methods: Incremental cost-utility analysis from the hospital perspective, based on a single-centre prospective longitudinal cohort study. Hospital case cost data were obtained for a cohort of 403 elective surgical cases in 2011-2012, comprising inpatient spine (n = 48), outpatient spine (n = 39), inpatient hip (n = 150), and inpatient knee (n = 166) procedures. Costs were adjusted to 2016 Canadian dollars. Patients completed the SF-12/36 health status survey before and one year after surgery. Clinically important improvement was defined as a 4.6-point improvement in SF-12/36 Physical Component Summary score one-year post surgery. This criterion was used to categorize patients as "responders" and "non-responders". One year incremental cost-utility ratios (ICURs) were calculated for each group as cost per Quality Adjusted Life Year (QALY) gained following surgery. QALYs were determined using SF-6D utility scores before and one year after surgery.

Results: The cohort was 47% female and had mean age 64 years. Overall, 71% of cases were categorized as responders one year after

Surgery: 62% of spine cases, 85% of hip cases, and 63% of knee cases. In the combined group, mean per procedure hospital cost was $10,937 (SD=$6,657) and patients gained 0.12 QALY (SD=0.14) in the year after surgery, corresponding to $89,464 per QALY gained in the year after surgery. Patients categorized as non-responders had higher mean case cost: $12,447 versus $10,319 for responders (p=0.02). Non-responders also had substantially smaller QALY gains at one year, gaining 0.03 QALY versus 0.13 QALY in responders (p< 0.01). The resulting one year ICURs varied greatly with responder status: $64,015/QALY for responder cases compared to $460,399/QALY for non-responder cases. Estimates also varied by operative site: responder / non-responder ICURs were $115,694 / $1,574,451 for inpatient spine cases; $23,599 / $139,362 for outpatient spine cases; $57,975 / $118,068 for hip cases; and $77,761 / $1,121,714 for knee cases.

Conclusions: Considering the ICUR for all patients, elective joint replacement and spine surgery meet an acceptable threshold for cost-effectiveness. However, the cost-utility estimate for the 29% of patients who do not achieve clinically important change in physical status demonstrates that surgery in this subset of patients is grossly cost-ineffective. This is particularly the case for knee replacement and spinal fusion. These findings provide a strong economic argument for further research in surgical prognostication, and development of real-world predictive analytics and prognostic tools for joint replacement and spinal surgery.