LAUSR

Gordon Waddell described in 2004 that “back pain was a 20th-century medical disaster and the legacy reverberates into the newmillennium.” The Global Burden of Diseases Study quantifies that legacy, showing that low back pain (LBP) has been the most burdensome nonfatal condition globally during the past 25 years. Although the global point prevalence of LBP did not change significantly from 1990 to 2010, the disability burden increased by 43% in that time, causing tremendous socioeconomic inequalities and suffering. This suggests that the current management of LBP is not only inadequate, but also that outcomes are worsening. Traditionally, LBP has been conceptualized as a biomedical problem. However, limitations of this model have been evident for some time. For example, treatments targeting lumbar pathology such as spinal manual therapy or radiofrequency denervation are no more effective than inert interventions, and signs of spinal degeneration shown on lumbar imaging correlate poorly with symptoms. Furthermore, although the current diagnostic triage approach for LBP uses a biomedical approach to identify pathoanatomical causes of LBP, this framework is limited by the fact that the majority of LBP cases are classified as “nonspecific,” a label which is neither helpful to clinicians nor patients. The failings of the biomedical model to improve outcomes elicited the biopsychosocial model of LBP. In 2017, a revised biopsychosocial model of pain was published, incorporating evidence from recent pain science to show how interactions between biological, psychological, and social domains contribute to each individual’s unique pain experience. This reformulation suggests that if we are to provide effective treatments for LBP sufferers, it is this complex constellation of biopsychosocial factors that needs to be understood and addressed. Indeed, the U.S. Institute ofMedicine report on Relieving Pain in America urges recognition of the complex, multidimensional nature of pain. One framework to address the multidimensional nature of LBP is complexity science. Its application to health care overcomes a linear, reductionist perspective of the human body in which illness is considered to be caused by malfunction of its parts. Complexity science provides an alternative model: that illness (and health) result from complex, dynamic, and unique interactions between different components of the systems that humans operate within, and are composed of (eg, biochemical, physiological, psychological, and social) a framework that aligns with the reformulated biopsychosocial model of pain. Complexity science offers potential solutions to the challenges of implementing a biopsychosocial model of LBP in care, and clinical research and practice can be more effective if these complex relationships are concurrently addressed, rather than the focus being on individual parts of the problem. This article focuses on the application of complexity science to the development and evaluation of LBP interventions. In the design of treatments tested in clinical trials, the multidimensional nature of LBP has been largely neglected. If we genuinely embrace a formulation of LBP as an experience that results from a constellation of biopsychosocial drivers, then why would treatments directed at a single domain be expected to produce worthwhile outcomes? Yet, clinical trials commonly neglect this complexity and test unimodal treatments, targeting single domains of LBP, for example, analgesics or manual therapies, to address biological contributors. It is therefore unsurprising that LBP treatments frequently fail to show clinically important benefits in trials. Indeed, such treatments rarely reach clinically worthwhile effects for primary outcomes such as pain or disability, even if they are statistically significant. A review of nonspecific LBP in The Lancet showed that no studies of common treatments for acute LBP achieve reductions in pain that patients consider worthwhile ($30 points on a 0-100 pain scale). Notably, these treatments (eg, heat, massage, exercise, muscle relaxants, and acupuncture) were all unimodal, directed at biological (pathoanatomical or pathophysiological) targets. By contrast, there is Level 1 evidence that multimodal interdisciplinary treatment doubles the likelihood that people with chronic LBP return to work compared with unimodal treatments aimed at physical factors. A literature search of systematic reviews of treatments for LBP in the 30-year period since the introduction of the biopsychosocial model of LBP revealed that of 166 reviews, 80% described unimodal treatments targeting biological factors, such as injections, motor control exercises, surgery, and pharmacotherapies; only 5% of reviews focused on multimodal treatments such as multidisciplinary pain management or biopsychosocial approaches in primary care (Appendix 1, available at http://links.lww.com/PAIN/B52). A second way in which clinical research has failed to embrace complexity in addressing the multidimensional nature of LBP is the common practice of testing “one-size-fits-all” treatments in trials, assuming that nonspecific LBP is a homogenous condition. In addition to any differences in spinal pathologies that people with nonspecific LBP may have, patients are typically complex and heterogeneous regarding multiple biological, psychological, Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.