Poster Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

INTRODUCTION: PROMIS measures have gained considerable support for use as reliable and valid outcome tools across a variety of clinical conditions. However, empirical evidence is needed to establish minimum clinically important difference (MCID) values for PROMIS Physical Function (PF) and Pain Interference (PI) after elective spine surgery. The objective of this study was to examine the amount of change in PROMIS PF and PROMIS PI T-scores that best discriminated patient satisfaction and establish MCIDs for the Oswestry Disability Index (ODI) and back and leg pain (NRS) after lumbar spine surgery. 

METHODS: Data from 950 patients undergoing elective spine surgery who had preoperative and 12-month postoperative patient-reported outcomes (PROs) from a single center were analyzed. PROs included disability (ODI), NRS back pain, NRS leg pain, satisfaction with surgery, PROMIS PF, and PROMIS PI. For satisfaction, the anchor was 1+2 on the NASS satisfaction item. Anchors evaluated were 12.8-, 14-, and 20-point reduction in ODI and a 30% decrease from preop at 12-months on the ODI, NRS back pain, and NRS leg pain. These reflect published and commonly used MCIDs for disability and pain. ROC curve analysis with Youden’s index, sensitivity, and specificity were used to determine change score cut points. Potential alternative cut-points were evaluated by plotting the Youden index against the PROMIS change scores. Discrimination was visualized by plotting PRO scores over time grouped by whether or not the proposed PROMIS MCID scores were achieved.

RESULTS: For PROMIS PF, a change of >7.8 points best discriminated ODI MCIDs of 12.8-points, 14-points, and 30% decrease in the preoperative score (Figure 1a). A change of >8.9 best discriminated the ODI anchor of 20-points. The PF change scores that best discriminated back and leg pain were >5.8 and >4.6 points, respectively. Satisfaction was best discriminated by a PF change score of >4.2. For PROMIS PI, a change of ≤-9.5 points best discriminated ODI MCIDs of 12.8-points, 14-points, and 30% decrease in the preoperative score (Figure 1b). The cut point was ≤-10.5 for the ODI anchor of 20-points. The PI change scores that best discriminated back and leg pain were ≤-9.1 and ≤-9.6 points, respectively. Satisfaction was best discriminated by a PI change score of ≤-5.3. All ROC curves had p<0.001 and most ROC values were > 0.80. Evaluation of the Youden index against the change scores suggested that lower PF change scores may be appropriate to discriminate ODI MCIDs.

DISCUSSION: Spine surgeons and researchers are familiar with MCIDs for ODI, back pain, and leg pain. Understanding the point changes in PROMIS T-scores for PF and PI that best discriminate these MCID values are helpful as medical centers make the transition to PROMIS PROs. Future work will focus on evaluating these MCID values for specific procedures and varying levels of preoperative disability. An evaluation of these T-score point changes in an external sample is planned.