Physiotherapy movement based classification approaches to low back pain: comparison of subgroups through review and developer/expert survey

Mechanical Diagnosis and Treatment (MDT)

The primary objective of the MDT scheme is to determine if LBP symptoms can be abolished or reduced through application of direction-specific, repeated lumbar spine movements or sustained postures. Internal intervertebral disc displacement is a key component in the conceptual explanation of this model, particularly for the primary category, "derangement". Less common categories are termed "dysfunction" and "postural". With "dysfunction syndromes" it is thought that tissue has undergone "contraction, scarring, adherence, adaptive shortening, or imperfect repair" [37]. Intervention is based on the tissue remodelling theory. "Postural syndrome" is assumed to arise from joint capsule and ligament ischemia due to prolonged spinal end range positioning. Lifestyle factors believed to predispose individuals to each syndrome include high frequency of sitting, flexion biased posture and activities.

A principal assessment is the use of direction-specific "loading strategies" [79] to elicit a phenomenon termed "centralisation", i.e., movement of peripherally located pain to a more central location. Loading strategies involve repeated end range lumbar extension, flexion or side-gliding movements and are thought to relocate symptomatic displaced tissues in a "derangement syndrome", or stretch adhered or shortened tissue in a "dysfunction syndrome".

Treatment Based Classification (TBC)

The primary purpose of the TBC approach is to identify features at baseline that predict responsiveness to four different treatment strategies. Three levels of classification are implemented [11]. The first level determines whether the patient can be managed by physical therapy alone, will require multidisciplinary management, or require referral to another health care practitioner. The second level stages the patient based on the severity of symptoms and degree of disability. Stage I is the acute phase where the goal is symptom relief. Stage II is the subacute phase when symptom relief and quick return to normal function are encouraged. Stage III is for those who must return to high physical demands but demonstrate poor physical conditioning. The third level of classification consists of assignment of the patient to one of four treatment syndromes [11]: "manipulation", "stabilization", "specific exercise" or "traction". Interpretation of a collection of signs, symptoms and observations derives this decision. Signs from four primary orthopaedic tests include: (1) pain centralisation with two or more movements in the same direction, or pain peripheralisation in the opposite direction of centralisation, (2) straight leg raise range of motion, (3) "prone instability test" and (4) lumbar posterior-to-anterior mobility testing. Signs of nerve root compression are identified by a neurologic examination. Symptoms of relevance are: (1) pain duration, (2) pain location, (3) episode frequency and (4) fear avoidance behaviour [11]. Observation is used to detect presence or absence of aberrant motion or a lateral shift deformity. Age is considered a predictor for allocation to the "stabilisation" subgroup.

The "manipulation" subgroup treatment includes high velocity thrust manipulation directed towards the lumbo-pelvic region [11]. The "stabilisation" subgroup receives exercises aimed at "promoting stability by producing motor patterns of co-contraction among all spinal stabilising muscles" [11]. The "specific exercise" subgroup is treated primarily with repeated end range spinal movement in the direction found to elicit a centralisation effect. The "traction" subgroup receives static mechanical traction in prone along with exercises to centralise symptoms [11]. Although initiated as an 'expert panel consensus driven' approach, the TBC scheme has developed into a 'data driven' model, which incorporates evidence from 'clinical prediction rules' [80].

Pathoanatomic Based Classification (PBC)

The PBC scheme attempts to connect a symptomatic response to key orthopaedic tests and an assumed pathologic structure to direct treatment [12]. Nine of 11 categories consist of all possible structures that could cause LBP, based on discography and single injection diagnostic studies. The exceptions are "inconclusive" and "abnormal pain syndrome" categories, the latter defined by "abnormal illness behaviour" [81]. Syndromes are defined by symptom location and effect of mechanical loading.

This scheme employs a hierarchy approach in which structures that more commonly cause LBP are considered first [82, 83], followed by systematic inclusion or exclusion of less prevalent structures and conditions. Other than response to loading strategies, additional criteria used for diagnosis include: (1) symptom duration; (2) standing, walking, sitting, and lying tolerance; (3) age > 65 years; (4) mechanical aggravating factors; (5) symptom provocation with examination; and (6) "non-organic" signs. "Adherent nerve root" and "nerve root entrapment" categories have been excluded from the system due to low intertester reliability, and the criteria for "spinal stenosis syndrome" and "zygopophyseal joint problem" have been updated (Petersen, survey response).

Movement System Impairment Syndromes (MSI)

The major objective in the MSI scheme is to identify the direction of alignment, stress or spinal movement that elicits or increases symptoms [13]. The development of these subgroups is theorized to occur due to alteration in the precision of joint movement as the result of repeated movements and prolonged postures associated with daily activities. This is enhanced by age-related factors such as degenerative changes, habitual postures and repetitive direction-specific activities (Sahrmann, survey response). Prolonged postures and repeated movements are proposed to cause tissue adaptations that eventually result in a joint developing a susceptibility to movement in a specific direction. Stress related to the alteration in precise motion is assumed to cause tissue micro-trauma and eventually, macrotrauma. Repeated movements and associated tissue adaptations are reasoned to contribute to generalization of movement patterns including the imprecise joint motion that is direction-specific. Of particular interest is the concept of relative stiffness that is believed to contribute to and perpetuate a joint's directional movement susceptibility. The relative stiffness problem can pertain to intervertebral joints, as well as other joints such as the hip or shoulder. Studies of the MSI system have documented relative flexibility issues between the lumbar and hip region, where early or excess lumbar motion is observed with lower limb movement tests and with trunk sidebending [84, 85].

The MSI scheme focuses on modification of this altered alignment and motion to reduce spine symptoms and to redistribute movement to other joints [86]. This process involves correction and/or restriction of altered lumbar motion in the direction associated with symptoms and facilitation of movement of other joints. Subclassification is confirmed by "modified movement and alignment tests" that correct the impaired spine movement, redistribute motion, or change alignment to reduce or eliminate symptoms. Examination includes tests that "identify tissue adaptations contributing to the identified relative stiffness and associated movement patterns" (Sahrmann, survey response). Tissue adaptations contributing to the patient's altered movement and alignment impairments include motor control alterations in recruitment patterns and timing, de-recruitment, skeletal and muscular performance alterations, stiffness and length. Of critical importance is the biomechanical consideration of muscle performance and recruitment patterns.

O'Sullivan Classification System (OCS)

The objective of the OCS is to identify the underlying mechanisms that are considered to drive pain. One of these is the identification of maladaptive (pain provocative) spinal postures, movement patterns, and motor control behaviours associated with LBP, which are then used to target treatment. The development and persistence of these impairments is believed to occur due to a maladaptive response to pain ('movement' or 'control' alteration), resulting in a lack of variance in the way tasks are performed and further perpetuation of pain [14]. Impairments are characterized into either 'pain avoidance' (movement category) or 'pain provocation' (control category) behaviour. Pain response is further categorized as either 'adaptive' or 'maladaptive'. 'Adaptive' pain is considered a protective response secondary to an underlying pathological process, whereas 'maladaptive' pain behaviour is viewed as a process in which movement and cognitive behaviours drive the pain, causing a compromise to the neuromusculoskeletal system. Assessment involves subjective and physical examination of pain, cognitive and movement behaviours [14, 36]. The OCS involves five levels [87]: In level one, LBP disorders are separated into 'specific' vs. 'non-specific'. Specific LBP is based on radiological evidence matching the clinical presentation. In level two, the primary pain system is identified as either of a "peripheral" or "central" nervous system disorder. Central nervous system pain disorders (such as regional, neuropathic and fibromyalgia pain disorders) are defined as pain disorders that are constant and non-remitting in nature with no clear mechanical behaviour. Peripheral nervous system disorders refer to localized and anatomically defined pain influenced by mechanical factors, whereas central nervous system disorders refer to constant and widespread pain not clearly influenced by mechanical factors. If the disorder is peripheral, level three involves distinguishing LBP from pelvic girdle pain. Where pain is thought to arise from spinal structures, the fourth level identifies whether "control impairments" (no movement impairment in direction of pain but impaired control) or "movement impairments" (impairment of movement in the direction of pain provocation associated with fear avoidance behaviour) are drivers of the pain disorder. A directional or postural pattern is determined within each group [87]. The fifth level involves identification of psychosocial and/or lifestyle factors that may contribute to the disorder.

Management of non-specific chronic LBP disorders that are considered to be of the peripheral nervous system consist of changing movement and cognitive behaviours based on the disorder classification. "Control impairment" disorders involve enhancing control by training movement patterns in order to functionally unload pain sensitive structures. In contrast, the approach to management of "movement impairment" disorders is to facilitate movement in the direction of pain provocation via graded exposure in order to reduce fear avoidance behaviours. The OCS assesses maladaptive physical and cognitive 'drivers of pain' and implements a behavioural intervention that focuses on the cognitive and behavioural change (termed 'Cognitive Functional Therapy' [36]).

Philosophies

All schemes share the objective to identify directions, movements or control patterns that decrease or increase pain in order to direct treatment. In this respect, we considered that the five schemes could be clustered into two groups. One group (MDT, TBC, and PBC) uses the centralisation phenomenon to guide categorization, but schemes vary in loading dosage, assessment strategies, and intervention options. The other group (MSI and OCS) shares a parallel theme of 'modified' movement and alignment strategies and a common philosophy of promoting variety and precision of movement, although the proposed mechanism, movement observation focus, and emphasis on identification of concurrent psychosocial factors differ.

Assessment methods

The MSI and OCS schemes rely on identification of the symptomatic region or segment during functional lumbo-pelvic and thoraco-lumbar movement tests and are alike in their use of a variety of test positions. MSI tests focus on the influence of hip and pelvic movements on the lumbar spine in both weight bearing and non-weight bearing positions, whereas OCS tests focus on abnormal thoraco-lumbar and lumbo-pelvic movement primarily in weight bearing positions, determined by pain provocative activities reported by the patient.

Differences exist between the two clusters of schemes in frontal plane analysis, with an assessment preference for lumbar spine side gliding (i.e. lateral pelvic translation) (MDT, PBC, and TBC) versus side bending (MSI and OCS). Lumbar rotation is a key MSI assessment component and part of the OCS multi-directional movement categories, but not referred to in other schemes.

Patient appropriateness

The primary reason to implement a classification process for non-specific LBP is to facilitate the 'clustering' of signs and symptoms in order to help better direct treatment and improve efficiency of care. However, the application of the subgrouping systems is also likely to apply to specific LBP diagnoses. The application to patients with specific diagnoses has not been tested, nor is it clear whether having a specific 'pathoanatomic' LBP diagnosis facilitates treatment decisions better than an assessment based on the patient's presentation of movement, pain processing, and emotions, cognitions, and behaviours related to their LBP.

For development and reliability purposes, most studies of classification schemes exclude people with diagnoses such as severe kyphosis, scoliosis, ankylosing spondylitis, and rheumatoid arthritis. Spinal stenosis is integrated as a subgroup within some schemes (MDT and PBC) whereas it is considered a specific diagnosis in the OCS and requires further classification. Spinal stenosis is most often classified in the MSI "extension" or "extension-rotation" subgroup, or TBC "specific exercise--flexion" subgroup (Sahrmann and Fritz, survey responses). The OCS scheme considers disc prolapse with radicular pain, spondylolisthesis, degenerative disc disease with positive Modic changes [106] and positive neural provocation tests as specific LBP diagnoses and are assessed and classified within the scheme. The MSI and OCS scheme developers clarified that principles of the classification approach could be applied to the specific LBP population [39], but require a 'modified approach' (Sahrmann and O'Sullivan, survey responses), taking into account the known pathological influences on the condition. How this is operationalized is less defined.

Distinguishing 'mechanical' LBP and 'stage' of disorder

Incorporation of psychosocial factors

As psychosocial factors may contribute to or perpetuate LBP, it can be argued that they cannot be separated from a biomechanical view of LBP. All schemes consider psychological influences, particularly when psychological signs appear dominant in persistent LBP. However, there is disparity across schemes in identification of psychosocial aspects within movement-based categories.

MDT and PBC schemes identify subgroups with an 'abnormal or chronic pain state' by testing for Waddell's 'abnormal illness behaviour' signs [81] but prioritize treatment of the mechanical presentation first. "Chronic pain" is defined by the MDT and PBC schemes as symptoms that are persistent (i.e., > 12 weeks), widespread and increase with all activity. Patients demonstrate "exaggerated pain behaviour and mistaken beliefs and attitudes about pain and movement" [10]. The MDT rationale is that many psychological factors improve when the mechanical presentation and associated pain improves (Clare, survey response). PBC identifies depression as a key predictor of poor recovery and other psychosocial dimensions as poor prognostic indicators such as distress, job satisfaction, duration of sickness, unemployment and financial incentives (Petersen, survey response). How these variables are identified and integrated with the mechanical diagnosis categories is not specified.

Fear-avoidance is the key psychological feature addressed in the TBC scheme, as a low Fear Avoidance Beliefs Questionnaire (FABQ) score is a criterion for allocation to the "manipulation" subgroup. The classification approach can still be applied to those with a high FABQ, but with modification (Delitto, survey response). Likewise, the MSI scheme incorporates a measure for fear avoidance behaviour, but the developers/experts state this population subtype is atypical within their research and clinical setting (Sahrmann and Van Dillen, survey response). How the MSI scheme would address patients who present with high fear avoidance behaviour, or other psychological features, is not clear. The OCS advocates multidisciplinary management and functional rehabilitation in the presence of dominant psychosocial factors [87]. In the presence of non-dominant psychosocial traits, the OCS also integrates the assessment of psychosocial factors into decision-making and pursues a behavioural intervention.

Incorporation of neurophysiologic factors

The International Association for the Study of Pain has defined neuropathic pain as "pain initiated or caused by a primary lesion or dysfunction in the nervous system". In an effort to distinguish neuropathic pain from nociceptive system hyperexcitabiity (i.e., central sensitization) [116] a more precise definition of neuropathic pain has been proposed as "pain which arises as a direct consequence of a lesion or disease affecting the somatosensory system" [117]. Where the nociceptive system generates activity without adequate peripheral sensory ending activation, correlates of the neuroanatomic distribution, pain consistency, stimulus-response proportion, history of injury, and findings from a clinical neurologic examination may help to distinguish the likelihood of neuropathic versus nociceptive pain origin. Anatomic location of the lesion/disease can also help to divide neuropathic pain into peripheral and central nervous system subtypes. Recent studies suggest prevalence of neuropathic pain in patients with low back and leg pain is 37% [15] to 41% [118] and is characterised by certain signs, such as radicular pain, higher ratings of pain intensity, depression, panic, anxiety and sleep disorders [119]. LBP patients presenting with neuropathic features tend to have poorer outcomes, hence, consideration of LBP states beyond subcategories of nociceptive (mechanical) or inflammatory (chemical) pain in favour of considering LBP as a 'nervous system/neurodegenerative disease' [120] may provide more comprehensive pain assessment and treatment. It should be acknowledged that neuropathic pain could co-exist with other pain types (i.e., mechanical or chemical), and central sensitization (amplification of transmission pathways within the central nervous system) is one mechanism that can contribute to neuropathic pain. Deciding which pain subtype is more important or dominant is based largely on clinical judgement.

Although most classification schemes provide a biomechanical or presumed pathoanatomic explanation for the presence of pain, few attempt to distinguish between various pain mechanisms. The OCS differentiates pain systems into either a centrally mediated or peripherally mediated pain category [39]. These features are distinguished from one another within the classification process by clinical expert judgment based on a patient's pain behaviour. The "abnormal pain state" subgroup proposed by the MDT and PBC schemes potentially capture some neuropathic characteristics and the PBC scheme provides an "adverse neural tension" category which may further highlight a particular neuropathic subset. However, these categories imply the pain states are separate and do not co-exist or interrelate with more mechanically based subgroups.