Related Experiment Video
Updated: May 11, 2025
A Spine Robotic-Assisted Navigation System for Pedicle Screw Placement
Published on: May 11, 2020
Does PROMIS Identify Recall Accuracy and Bias in Elective Spine Surgery Patients?
Nicholas C Arpey1, Jacob R Staub, Bejan A Alvandi
1Department of Orthopaedic Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL.
Study Design:
Prospective study (level II evidence).
Objective:
The objective of this study was to assess recall accuracy and bias through PROMIS-PF (physical function) and PI (pain interference) scores in patients undergoing cervical or lumbar spine surgery.
Summary Of Background Data:
Patient-reported outcomes (PROs) have improved quantitative data availability; however, the interpretation of results may remain susceptible to confounding factors including recall accuracy. No studies to date have reported the accuracy of patient recall using PROMIS outcomes in spine patients.
Methods:
Patients who underwent elective lumbar or cervical spine surgery were identified at a single tertiary, academic institution. PROMIS-PF and PI CAT (computer adaptive tests) were administered preoperatively. After a minimum 2-year follow-up, patients were administered these questionnaires and asked to answer with their recalled preoperative status. Recall accuracy was assessed by comparing recalled and actual baseline PROMIS scores. Regression analyses were conducted to evaluate the agreement between actual and recalled scores. Multivariate logistic regression was performed to determine potential demographic and temporal sources of bias.
Results:
Lumbar surgery patients recalled significantly worse preoperative function (Δ -1.5, 95% CI (-2.8 to -0.3), P <0.05) and severe pain [Δ 2.1, 95% CI (0.5-3.6), P <0.01] than reported before surgery. Patients in the cervical cohort, in contrast, were more accurate in recall for both domains [PF Δ 1.8, 95% (CI -1.4 to 5.0), P >0.05 and PI Δ 0.0, 95% CI (-3.0 to 3.0), P >0.05]. The correlation between recalled and actual scores was moderate in both cohorts. Demographic and temporal variables did not significantly influence recall accuracy. Lumbar cohort patients who met PI MCID were more likely to accurately recall baseline pain scores.
Conclusion:
Recall inaccuracy is present in patients who undergo spine surgery; however, patients on average recall significantly worse preoperative status as measured by PROMIS PF and PI scores. Accurate recollection of preoperative status may influence patient perception of care.
