Solving the Problem of Dropout in Long Term Outcome Studies: Assessing the Predictive Relationship Between Short-term (2 year) and Long-term (7 year) Outcomes in a Seven-year Randomized Controlled IDE Trial for Lumbar Artificial Discs. (16th November 2020)

• Record Type: Journal Article
• Authors: Braxton, Ernest E; Ferko, Nicole; Situ, Aaron
• Is Part Of: Neurosurgery. Volume 67(2010)Supplement 1
• Publication Date: 2020-11-16
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1093/neuros/nyaa447_706
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract: INTRODUCTION: Dropout is a problem for long-term studies. While the activL ® FDA IDE study had a reasonably low dropout of 27% (77/283) over 7 years, it was nevertheless valuable data lost to inform long-term efficacy of lumbar total disc replacement (TDR) in treating symptomatic single-level lumbar degenerative disc disease. Methods can be used to impute missing data; however, this can have limitations. METHODS: The activL ® and ProDisc-L ® groups who were available for follow-up at 2 and 7 years were analysed. Function, pain, and quality of life outcomes were captured as to whether there was clinically significant improvement (CSI) relative to baseline after these time points. To assess the predictive relationship between early and late outcomes, logistic regression models were fit. These models were used to predict 7-year outcomes for patients lost to follow-up (LTFU) after 2 years. RESULTS: Across outcomes analyzed, the percentage of patients who achieved CSI at 7 years was significantly higher in those who also achieved CSI at 2 years vs. those who did not. Using the outcome of ODI score, 93% of patients with CSI at 2 years also had CSI at 7 years; while only 43% without CSI at 2 years achieved CSI at 7 years. Baseline characteristics of patients who remained at 7 years were similar to patients who were available at 2 years but were LTFU thereafter. For ODI, 82% (75/91) of LTFU patients achieved CSI at 2 years. Of these patients, 93% (70/75) were predicted to achieve CSI at 7 years had they remained in the study. Other outcomes involved similar predictions. CONCLUSION: Our study showed that lumbar TDR patients were more likely to experience improvements at 7 years if they experienced improvement at earlier time points. These results suggest that short-term benefits of TDR are predictive of the long-term efficacy and that patients with CSI at 2 years lost to follow-up were highly likely to maintain improvements to 7 years.