241. Towards a cervical deformity-specific outcome instrument: use of the patient-generated index to capture the disability of cervical deformity

Abstract

BACKGROUND CONTEXT Existing health outcome (HRQL) metrics fail to correlate with malalignment of the cervical spine and do not adequately capture disability from cervical deformity (CD). The purpose of this study is to introduce the cervical deformity patient generated index (PGI), where patients report their greatest difficulties related to their CD. These responses can then be used to identify the disabling features of CD that are not captured by existing HRQLs and determine items to be included in a CD specific HRQL. PURPOSE To introduce the Cervical Deformity Patient Generated Index (PGI) to determine disabling features of CD that are appropriate for a CD specific HRQL. STUDY DESIGN/SETTING Retrospective review of a prospective operative CD database and prospective development of a novel CD-PGI metric. PATIENT SAMPLE A total of 139 CD patients and 12 PGI patients. OUTCOME MEASURES Existing outcome metrics: NDI, NRS Back/Neck, mJOA, EQ5D, EQ5DVAS, and SWAL-QOL. Cervical Deformity Patient Generated Index (CD-PGI). METHODS CD patients completed the CD-PGI by describing the most important aspects of their disability. Patients wrote five things that bothered them most (Stage 1). Next patients rated the severity for each answer on a scale of 0 (max disability) to 10 (no disability) (Stage 2). Finally, patients “spent” up to 60 points on whichever responses they would like to improve (Stage 3). The PGI score was obtained by summing the product of Stage 2 and Stage 3 values for each response. PGI responses were categorized to find which aspects were most important to CD patients. A CD database was analyzed to assess whether legacy HRQLs correlate with cervical malalignment (CL, TSCL, cSVA, T1S, and C2S). Next PGI responses were analyzed to find how commonly responses were captured by existing cervical HRQL metrics: NDI, NRS Back/Neck, mJOA, EQ5D, EQ5DVAS and SWAL-QOL. Finally NDI and PGI were compared to elucidate major drivers of HRQL scores. RESULTS A total of 139 CD patients (mean cSVA: 46.2mm) and 12 PGI patients (mean cSVA: 62mm) were included. PGI responses were grouped into 6 categories: Pain, Sagittal Discomfort/ROM, Horizontal Gaze/Walking Safety, Activities of Daily Living (ADL), Social Life and Hobbies and Neurologic. mJOA correlated with CL (r=0.21, p=0.01), TS-CL (r=-0.20, p=0.02), and C2S (r=-0.18, p=0.03). PGI scores did not correlate with any HRQLs. 34/60 PGI responses (57%) were found to be captured by existing HRQLs. The EQ5D addressed 53% of PGI responses compared to 43% for NDI, 3% for mJOA address and 0% for SWALQOL. PGI-Pain, -neurologic, -social life, and -ADL responses were addressed by existing HRQLs. However, only 40% horizontal gaze and 0% sagittal discomfort responses were addressed. The main drivers of NDI score were reading, pain, and recreation questions, explaining 80% of variability (r2=0.80). The main drivers of PGI were ADL, sagittal discomfort and social life, explaining 75% of variability (r2=0.75). NDI-Concentration NDI-Reading, NDI-Driving and NDI-Sleep correlated with multiple individual PGI items including PGI-ADL, -sagittal discomfort, and -social life (all r>.75, p CONCLUSIONS Existing HRQL do not adequately capture CD disability and do not correlate with cervical malalignment. PGI items not addressed in existing HRQLs include Sagittal Discomfort/ROM and Horizontal Gaze/Walking Safety. In addition, the most important categories driving PGI scores were found to be ADLs, Sagittal Discomfort/ROM, and Social Life/Hobbies. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.