General Session: Cervical Degenerative - Hall F

Presented by: I. Aleem

Author(s):

M.T. Shamaa(1), I. Aleem(2), B. Currier(3), M.J. Yaszemski(3), H. Poppendeck(3), P.M. Huddleston(3), J. Eck(4), J. Rhee(5), M. Bydon(6), B. Freedman(3), A. Nassr(7)

(1) Henry Ford Hospital, Surgery, Detroit, MI, United States
(2) University of Michigan, Orthopaedic Surgery, Spine Division, Ann Arbor, MI, United States
(3) Mayo Clinic, Orthopedics Spine, Rochester, MN, United States
(4) Center for Sports Medicine and Orthopedics, Chattanooga, TN, United States
(5) Emory University, Department of Orthopedic Surgery, Emory Spine Center, Atlanta, GA, United States
(6) Mayo Clinic, Neurosurgery, Rochester, MN, United States
(7) Mayo Clinic, Orthopedics, Rochester, MN, United States

Abstract

Introduction: Surgery for cervical degenerative myelopathy or radiculopathy focuses on addressing pain and disability while improving the patients' quality of life. Although Patient-Reported Outcomes (PROs) are being widely adopted, their interpretability may be limited by the accuracy of a patient's ability to recall preintervention impairment. Recall bias has been previously investigated in multiple orthopedic and lumbar spine studies, but recall accuracy in cervical spine patients remains unknown. We sought to characterize the accuracy of patient recall as a function of time on validated outcomes after cervical spine surgery.

Materials and Methods: We analyzed a consecutive series of patients undergoing cervical spine surgery for degenerative myelopathy or radiculopathy at a single institution. Using standardized questionnaires, we recorded preoperative neck and arm Numeric Pain Scores (NPS), Neck Disability Indices (NDI) and 36-Item Short Form Health Survey (SF-36). Patients were asked to recall their preoperative status through a standardized phone-call script and were subsequently stratified based on the timing of their recall into short-term (< 1 year) and long-term (> 1 year) follow-up sub-groups. Actual and recalled scores were compared using McNemar's or paired t tests, and relations were quantified using Pearson correlation coefficients. Characteristics between the subgroups were compared using Wilcoxon rank sum tests, t-tests, chi-square tests, or Fisher's exact tests as appropriate.

Results: Seventy-three patients with a mean age of 58.2 years (range 22 to 83 years) were included, with 34 and 39 patients in the short-term and long-term follow-up subgroups patients respectively. The mean period of recall from surgery was 4.6 months and 22.2 months for the short-group and long-term follow-up subgroups respectively. Compared to the preoperative scores, patients showed significant improvement in neck NPS (mean difference [MD] = -2.9, 95% CI -3.5 to -2.3), arm NPS (MD -3.4, 95% CI -4.0 to -2.8), and NDI (MD -12.4%, 95% CI -16.9 to -7.9). Patient recollection of preoperative status was more severe than actual for neck NPS (MD +1.5, p< 0.001), arm NPS (MD +2.3, p< 0.001), and NDI (MD +5.8%, p< 0.001) and this was maintained across the sub-groups. No difference in recall accuracy was noted in SF-36 scores signifying that patients could accurately recall their pre-operative physical and mental health status. Moderate correlation between actual and recalled scores for neck NPS (r= 0.41), arm NPS (r= 0.50), NDI (r= 0.67), and a strong correlation in SF-36 scores (r= 0.74) was observed. Further, the predominant symptom was switched from neck pain to arm pain or arm pain to neck pain on recall for 31.5% of patients.

Conclusion: Patient recollection of preoperative status after cervical spine surgery was significantly more severe than their actual preoperative status for neck pain, arm pain, and disability. Relying on retrospectively recalled data for outcome assessment does not provide an accurate measure of preoperative status. Prospective collection of PROs remains the gold standard to measure outcomes following cervical spine surgery.