Risk factors for reoperation after lumbar total disc replacement at short-, mid-, and long-term follow-up

Abstract

BACKGROUND CONTEXTThe reoperation rate following TDR (Total Disc replacement) has been established at short- and mid-term time points through the Food and Drug Administration Investigational Device Exemption (FDA IDE) trials. However, these trials include highly selected centers and surgeons with strict governance of indications. The utilization of TDR throughout the community needs further analysis. PURPOSETo identify the risk factors for lumbar spine reoperation in patients undergoing lumbar total disc replacement (TDR) at short-, mid-, and long-term follow-up. STUDY DESIGN/SETTINGThis study is a multi-center retrospective cohort study utilizing the New York Statewide Planning and Research Cooperative System database. PATIENT SAMPLEWe identified 1,368 patients who underwent an elective primary lumbar TDR in New York State between January 1, 2005 and September 30, 2013. OUTCOME MEASURESThe primary functional outcome of interest was lumbar reoperation, specifically the evaluation of independent risk factors for lumbar reoperation at a minimum of 2 years, with sub-analyses performed at 5 and ten years. METHODSInternational Classification of Diseases, Ninth revision codes were utilized to identify patients undergoing a primary lumbar TDR. We excluded patients with primary/revision lumbar fusion procedures and revision disc replacement procedures. Hospital academic status was determined by the Accreditation Council for Graduate Medical Education. Unique encrypted patient identifiers allowed for longitudinal follow-up for reoperation. Logistic regression models compared reoperation and no-reoperation cohorts, and were performed on sub-analyses for significant univariate predictors of reoperation. RESULTSBetween January 2005 and September 2013, 1368 patients underwent a primary lumbar TDR. Reoperation occurred in 8.8% by 2 years, 15.8% by 5 years, and 19.5% by ten years. Diabetics were more likely to have reoperations (7.5% vs 3.8%, p=.013). Teaching hospitals experienced a decreased reoperation rate compared to nonteaching hospitals at 2-year (5.0% vs 10.5%, p=.002), 5-year (10.7% vs 17.9%, p=.002) and 10-year (11.7% vs 21.9%, p=.045) follow-up. Lumbar fusion was the most common reoperation (14.2%). CONCLUSIONWe identified an 8.8% reoperation rate after inpatient lumbar TDR at 2-years, 15.8% at 5-years, and 19.5% at 10-years. When stratifying by teaching status, reoperation rates at teaching centers align with those reported in FDA IDE studies. Diabetes was the only patient factor influencing reoperation rate. There is a growing consensus that lumbar TDR is a durable and appropriate surgical option for lumbar degenerative disc disease. Proper indications are crucial to obtaining good outcomes with lumbar TDR.