PROMIS Physical Function Score Strongly Correlates With Legacy Outcome Measures in Minimally Invasive Lumbar Microdiscectomy

Benjamin Khechen, BA; Brittany E. Haws, MD; Dil V. Patel, BS; Mundeep S. Bawa, BA; Islam M. Elboghdady, MD; Eric H. Lamoutte, BS; Sailee S. Karmarkar, BS; Kern Singh, MD


Spine. 2019;44(6):442-446. 

In This Article


The accurate evaluation of patient outcomes following spine surgery is necessary to ensure appropriate postoperative management and patient satisfaction. Although legacy measures have demonstrated efficacy in the measurement of PROs, the administration and collection of these instruments can be cumbersome. In addition to accurately tracking progress, PROMIS PF can offer patients convenience by minimizing the time required for survey completion. However, as this is a recently developed instrument, it has yet to be validated in many procedure-specific populations. To this end, our study sought to determine the validity of PROMIS PF when compared to legacy outcome measures in the MIS LD population.

The results of the current study demonstrate PROMIS PF score has utility as a postoperative outcome measure following MIS LD. Patients undergoing MIS LD exhibited significant postoperative improvements in PROMIS PF score, with greatest improvement being demonstrated in the first 12 weeks postoperatively. Significant postoperative improvements in ODI similarly demonstrated greatest improvement in the first 12 postoperative weeks. VAS Back and Leg pain scores also demonstrated significant improvements, with greatest improvement in the first 6-weeks postoperatively. Additionally, PROMIS PF score improvements strongly correlated with legacy measure improvements at all postoperative time points. These findings suggest PROMIS PF score may be used to estimate functional disability and back and leg pain following MIS LD. As such, potential exists for PROMIS PF to supplant these legacy measures as a stand-alone assessment of disability and pain following MIS LD.

In the present study, ODI scores exhibited strong correlations with PROMIS PF both preoperatively and at each postoperative time point. These findings have been previously demonstrated in the spine literature. In a population of patients surgically treated for lumbar spinal stenosis, Patel et al[1] demonstrated preoperative ODI to strongly correlate with PROMIS PF (|r| = 0.58, P < 0.01). Brodke et al[7] also identified a strong correlation between preoperative PROMIS PF and ODI score (|r| = 0.810, P < 0.05) in patients presenting with a spine-related complaint. Papuga et al[12] investigated the correlation between PROMIS PF and ODI in patients presenting to an orthopedic spine surgeon with back pain or disability complaint. PROMIS PF score was determined to strongly correlate with ODI, at both preoperative (r = 0.8264, P < 0.001) and postoperative time points (r = 0.7604, P < 0.05). These findings lend more support to the use of PROMIS PF as an assessment tool for functional disability. As the existing spine literature validating PROMIS PF as a functional disability assessment continues to grow, implementation of this instrument will allow for more effective survey compliance and tracking of functional outcomes.

VAS back and leg pain scores were observed to correlate strongly with PROMIS PF at each postoperative time point. However, only a moderate correlation was observed between preoperative VAS leg pain and PROMIS PF, and no correlation was identified with preoperative VAS back pain scores. Shahgholi et al[6] reported similar findings in patients who suffered a vertebral fracture secondary to osteoporosis or multiple myeloma. Using the Numeric Rating Scale (NRS), the authors identified moderate correlations between improvement in PROMIS and NRS score from preoperative to 30-day postoperative measures (P < 0.05), but failed to demonstrate correlations between preoperative PROMIS and NRS scores. In another study by Shahgholi et al,[13] the authors evaluated the validity of PROMIS in patients undergoing transforaminal epidural steroid injections for radicular pain. NRS pain scores significantly correlated with PROMIS PF scores at all timepoints (P < 0.05). Additionally, the authors demonstrated moderate-high correlations between PROMIS PF and NRS scores improvements from preoperative to 3-month scores, but did not observe correlations in improvement between 3- and 6-months.

A potential explanation for inconsistencies in preoperative pain correlations between PROMIS PF and pain scores may be due temporal differences in pain assessment between instruments. PROMIS asks patients to evaluate pain on the day of survey administration, while legacy measures such as NRS ask patients to evaluate their pain over the past 7 days. These findings, in combination with the results of the current study, suggest PROMIS PF score can offer utility in the assessment of postoperative improvement in pain, but may not accurately capture preoperative pain scores. Further work is necessary to more accurately characterize the relationship between these PRO instruments.

To the authors' knowledge, this is the first validity assessment of PROMIS PF in the MIS LD population. However, this study is not without limitations. As this study was conducted at a single institution, our findings may have reduced generalizability. Secondly, the retrospective nature of this study may have introduced an element of selection bias that could not be accounted for. Thirdly, correlations between PROMIS PF and Short-Form (SF)-12 scores were not included due to low SF-12 survey compliance. Additionally, low PROMIS PF survey compliance at 1- and 2-years postoperatively limited our analysis up to 6-months postoperatively. Lastly, determination of clinically meaningful change in PROMIS PF was not possible due to our small sample size. Additional, large-scale studies are necessary to identify clinically significant improvements in PROMIS score.