Introduction
In Western societies, low back pain (LBP) has the largest disease burden [1]. It is associated with a substantial amount of morbidity, and complaints are multidimensional. Functional status is an important patient-related outcome when evaluating surgical and non-surgical interventions for LBP. One important feature of outcome instruments measuring functional status is their ability to detect meaningful change from the patient's perspective. In the absence of appropriate objective clinical outcome measures, the use of patient-reported outcome measures (PROM) to assess treatment outcome is commonly accepted [2]. The Oswestry Disability Index (ODI) [3], and the ODI version 2.1a [4], [5] in particular, is widely accepted and recommended as a condition-specific PROM in interventional studies [6]. As such, medical decision-making increasingly relies on this measure. Although most clinicians and researchers agree that the success of any intervention should be judged from the patient's perspective, to date no consensus exists for criteria indicating “success.”
In health services research, it is important to define clear criteria for treatment “success.” Success can be conceptualized in two ways: (1) relevant change or improvement, and (2) achievement of an acceptable state. With the first concept, the emphasis is on whether or not an individual has improved after an intervention [7], whereas with the second the emphasis is on whether or not the achieved outcome is acceptable from the patient's perspective [7]. The concept of change (minimum clinically important difference or change) is frequently used in spine research to assess treatment success. In relation to this, it is important to specify whether the observed change in an individual's scores is merely the result of measurement error or whether it constitutes a real change, and whether that change is also clinically relevant to the patient [8]. However, it is difficult to measure what is clinically relevant to patients [9], and methodological issues such as population dependency and baseline dependency [10] are encountered. Moreover, assessment of change does not indicate whether a “normal” or “healthy” symptom state is reached. For these reasons, we have previously used a more stringent definition of success based on achievement of values seen in “normal,” healthy populations [11]. The threshold used was the achievement of an ODI value, derived from “normal” subjects with little or no back pain, of ≤22 [4], [11]. The use of “normal, healthy population” values as the reference might, however, be criticized as the ODI is a condition-specific instrument.
An alternative approach to measuring success is to identify the value beyond which patients consider themselves well or consider their health state to be acceptable, ie, the concept of the patient acceptable symptom state (PASS) [12], [13]. Determination of the absolute cutoff value (threshold) at follow-up, equivalent to achievement of a PASS, would assist in interpreting scores at the individual level and would allow determination of the proportion of patients within a group who achieve this level, when evaluating the effectiveness or success of interventions. Achievement of this threshold might be more important than the achievement of a given change value, and it probably reflects the ultimate goal of treatment from the patient's perspective [12], [14]. The concepts of “feeling better” and “feeling good” are complementary but distinctly different; a patient's condition can be markedly improved by the intervention but can still be suboptimal [7], [12], [13].
The purpose of the present study was to estimate the score on the ODI (version 2.1a) corresponding to a “patient acceptable symptom state” in patients undergoing surgery for degenerative disorders of the lumbar spine. To assess the robustness of the findings, we performed sensitivity analyses with different definitions for “acceptable state” and in various subgroups of patients.