Evidence & Methods
For many low back pain patients without serious destructive disease or deformity, current diagnostic methods are of uncertain utility. Non-operative treatment of low back pain is often guided by a presumed underlying pathologic or structural disease entity, but can also be guided by an empirically classified presenting clinical syndrome. This article aims to compare these two approaches.
In this large, observational cohort, the authors have found that specific treatments guided by clinical syndrome correlated with apparent better outcomes.
Observation cohort studies such as this can suggest clinic superiority, but usually cannot provide that conclusion with much certainty. Caution is indicated in interpreting this study's findings as inclusion, exclusion, recruitment, diagnostic criteria and data sets, practitioner skills, and baseline patient characteristics may vary impact outcomes. Nonetheless, the classic work by Drs. Ian Macnab and John McCulloch stressed the importance of considering the empirically described clinical syndrome when the pathologic diagnosis is uncertain. This study indicates this approach remains worthy of consideration and further study today.
—The Editors
The usual search for anatomical sites and structural causes of LBP can interfere with treatment [1]. Abnormal images do not always indicate symptoms [6], [7], [8], [9], [10], [11], [12]. Overutilization of diagnostic procedures is not cost-effective, can adversely affect outcomes, may increase the likelihood of iatrogenic complications [13], and can lead to potentially inappropriate treatment [14], [15], [16].
Identifying LBP subgroups based on patient characteristics—clinical presentations established by the history and physical examination—may be a reasonable alternative [17]. The Quebec Task Force [18], an international LBP consensus group [19], the Cochrane Back Review Group [20], and others [2], [21], [22], [23] have all emphasized the need to classify LBP into meaningful clinically relevant subgroups.
The generic term “nonspecific low back pain” (NSLBP) has gained acceptance as a category for patients lacking a recognized anatomical or pathological cause [3]. LBP classifications frequently include a single NSLBP category despite the fact that LBP is heterogeneous [19], [24], [25], [26], [27]. Although it may serve as a diagnosis of exclusion for surgical indications, a nonspecific category offers no positive direction for nonsurgical management. One possibility for the failure to predict effective treatment is an inability to subgroup or classify patients with nonspecific LBP in a way that directs management decisions [28].
Waddell states that ordinary backache or NSLBP is mechanical pain of musculoskeletal origin in which symptoms vary with physical activities [1]. Because NSLBP includes a variety of mechanical conditions with varying responses to movement and posture, nonspecific seems to be an inappropriate label.
Effective LBP assessment should identify patient categories that lead clinicians to the most effective management. Previous research suggests that treating patients based on a classification approach results in better clinical outcomes than nonclassification-based treatment strategies [29], [30], [31]. However, these studies identify narrow putatively anatomic subsets applicable to some but not to most mechanical LBP patients [1]. A more robust and inclusive classification system is required. An appropriate starting point is recognition of the symptom clusters of back pain syndromes [1].
In the traditional medical paradigm, groups of signs and symptoms that appear together in a regular and predictable fashion but lack a definitive etiology are labeled syndromes. A carefully constructed history suggests a syndrome or pattern that is then supported by a concordant structured physical examination. Final verification of correct syndrome selection comes from a predictable positive treatment response.
The classification system in the present study is based solely and deliberately on the recognition of syndromes or patterns of pain with no direct reference to anatomic site or pathological process. In this system, the essential elements are determined by the location of the dominant pain, whether that pain is constant or intermittent and the particular movements or postures that exacerbate or alleviate the symptoms [32]. These recognized similarities allow categorization and the generation of a theoretical prototype against which subsequent clinical patterns can be compared and tested [33]. The intertester reliability of this subgroup identification has been previously documented (agreement=79%, kappa=0.61) [34].
The purpose of this study was to assess the effectiveness of this LBP classification system in a clinical setting. It addresses the research question: do patients assessed and treated in accordance with this classification system have superior outcomes to patients receiving more homogeneous generic care?
