Assessment of spine surgery outcomes: inconsistency of change amongst outcome measurements

Abstract

Background context: Outcomes of spinal treatments are evaluated by clinical relevance: the proportion of patients who reach a minimum clinically important outcome change. Outcomes are evaluated through multiple measurements, and the inconsistency of outcome change across measurements is not known.

Purpose: The primary purpose of this study was to illustrate outcome inconsistencies after spinal surgery. Secondary goals of this study were to develop an index of overall change that incorporates outcome inconsistencies, to relate the index of overall change to patients' global assessment and satisfaction with treatment, to relate the index of global change to an intuitively understandable outcome: the level of tolerable pain.

Study design: This study is a review of prospectively collected patient-reported outcomes data.

Patient sample: Four hundred sixty patients from a large multicenter database were chosen. Those patients were included in the sample because they had undergone lumbar surgery and had baseline and 1-year follow-up scores. Baseline and 1-year follow-up scores for Oswestry Disability Index (ODI), physical component summary (PCS) of the Medical Outcome Study Short Form-36 (SF-36), numerical back and leg pain scales, and 1-year scores for satisfaction with results were included in the study.

Outcome measures: The outcome measures of the study were preoperative and 1-year postoperative scores for ODI, PCS, back pain scale, leg pain scale, health transition item of the SF-36, and satisfaction with results scales.

Methods: Oswestry Disability Index, SF-36, and pain scales were administered before and 1 year after spinal surgery. Satisfaction with results questionnaires were administered 1 year after surgery. The following threshold values were previously established and were used to evaluate outcome changes: minimum clinically important difference (MCID), substantial clinical benefit (SCB), and standard error of the mean. The following proportions of patients were determined according to outcome changes: "deteriorated," "no change," "below MCID," "above MCID," and "above SCB." The consistency of outcome change was determined amongst the four outcome measures. An index of overall change was developed and related to patients' answers to the health transition item of the SF-36 and to the satisfaction with results scale. The overall change index was also compared with the tolerable pain level.

Results: Only 40.5% of patients report consistent outcome changes on all four measures. The overall change index was significantly correlated to the global change and satisfaction scale (rho=.67, p less than .001). The overall change index was clearly associated with the tolerable pain level.

Conclusions: Efforts should be made to take into account the inconsistency of outcomes and to make clinical relevance more readily understandable by patients and clinicians.