Research Review By Dr. Michael Haneline©

Audio:

Download MP3

Date Posted:

October 2013

Review Title:

Mechanisms-based classifications of musculoskeletal pain:
  • Part 1: Symptoms and signs of central sensitization in patients with low back (± leg) pain
  • Part 2: Symptoms and signs of peripheral neuropathic pain in patients with low back (± leg) pain
  • Part 3: Symptoms and signs of nociceptive pain in patients with low back (± leg) pain

Authors:

Smart K, Blake C, Staines A et al.

Author's Affiliations:

Physiotherapy Department, St Vincent’s University Hospital, Elm Park, Dublin, Ireland.

Publication Information:

Manual Therapy 2012; 17: 336-44, 345-51 & 352-57.

Background Information:

Pain may be classified according to a mechanisms-based approach in which the classifications are derived from the major neurophysiological mechanisms that generate and/or maintain the pain. The three types of pain discussed in these 3 papers are as follows:
  1. ‘Central sensitization’ pain (CSP) has been defined as an amplification of neural signaling within the central nervous system (CNS) that brings about pain hypersensitivity. CSP occurs at the cellular level, affecting the spinal cord and/or supraspinal centers via enhanced synaptic excitability, lowered thresholds of activation and expansion of nociceptive input fields in the CNS. Please see the Related Reviews below – we have reviewed this concept in detail on RRS previously.
  2. ‘Peripheral neuropathic’ pain (PNP) refers to pain caused by a lesion or dysfunction in a peripheral nerve, dorsal root ganglion or dorsal root. These lesions may result from trauma, compression (i.e. entrapment neuropathies), inflammation or ischemia.
  3. ‘Nociceptive’ pain (NP) refers to pain that is caused by pathophysiological processes associated with activation of the peripheral receptive terminals of primary afferent neurons in response to noxious stimuli.
Further understanding of the mechanisms-based classifications of pain might be useful to clinicians in selecting treatments that are designed to target the dominant neurobiological mechanisms underlying patients’ pain.

A recent Delphi-type survey of pain consultants and musculoskeletal physiotherapists identified consensus-derived lists of symptoms and signs indicative of CSP, PNP and NP. There were thirteen symptoms and four signs suggestive of a dominance of CSP, nine symptoms and five signs suggestive of a dominance of PNP, and eight symptoms and four signs suggestive of a dominance of NP. The symptoms and signs (i.e. subjective and examination findings) for each classification are as follows.

Clinical indicators of CSP:

Subjective:
  • Disproportionate, non-mechanical, unpredictable pattern of pain provocation in response to multiple/non-specific aggravating/easing factors.
  • Pain persisting beyond expected tissue healing/pathology recovery times.
  • Pain disproportionate to the nature and extent of injury or pathology.
  • Widespread, non-anatomical distribution of pain.
  • History of failed interventions (medical/surgical/therapeutic).
  • Strong association with maladaptive psychosocial factors (e.g. negative emotions, poor self-efficacy, maladaptive beliefs and pain behaviors, altered family/work/social life, medical conflict).
  • Unresponsive to NSAIDS and/or more responsive to anti-epileptic (e.g. Lyrica) /antidepressant (e.g. Amitriptyline) medication.
  • Reports of spontaneous (i.e. stimulus-independent) pain and/or paroxysmal pain (i.e. sudden recurrences and intensification of pain).
  • Pain in association with high levels of functional disability.
  • More constant/unremitting pain.
  • Night pain/disturbed sleep.
  • Pain in association with other dysesthesias (e.g. burning, coldness, crawling).
  • Pain of high severity and irritability (i.e. easily provoked, taking a long time to settle).
Clinical examination:
  • Disproportionate, inconsistent, non-mechanical/non-anatomical pattern of pain provocation in response to movement/mechanical testing.
  • Positive findings of hyperalgesia (primary, secondary) and/or allodynia and/or hyperpathia within the distribution of pain.
  • Diffuse/non-anatomic areas of pain/tenderness on palpation.
  • Positive identification of various psychosocial factors (e.g. catastrophization, fear-avoidance behaviour, distress).
Clinical indicators of PNP:

Subjective:
  • Pain variously described as burning, shooting, sharp, aching or electric-shock-like.
  • History of nerve injury, pathology or mechanical compromise.
  • Pain in association with other neurological symptoms (e.g. pins and needles, numbness, weakness).
  • Pain referred in a dermatomal or cutaneous distribution.
  • Less responsive to simple analgesia/NSAIDS and/or more responsive to anti-epileptic (e.g. Neurontin, Lyrica)/anti-depression (e.g. Amitriptyline) medication.
  • Pain of high severity and irritability (i.e. easily provoked, taking longer to settle).
  • Mechanical pattern to aggravating and easing factors involving activities/postures associated with movement, loading or compression of neural tissue.
  • Pain in association with other dysesthesias (e.g. crawling, electrical, heaviness).
  • Reports of spontaneous (i.e. stimulus-independent) pain and/or paroxysmal pain (i.e. sudden recurrences and intensification of pain).
Clinical examination:
  • Pain/symptom provocation with mechanical/movement tests (e.g. active/passive, neurodynamic – such as SLR, brachial plexus tension test) that move/load/compress neural tissue.
  • Pain/symptom provocation on palpation of relevant neural tissues.
  • Positive neurological findings (including altered reflexes, sensation and muscle power in a dermatomal/myotomal or cutaneous nerve distribution).
  • Antalgic posturing of the affected limb/body part.
  • Positive findings of hyperalgesia (primary or secondary) and/or allodynia and/or hyperpathia within the distribution of pain.
Clinical indicators of NP:

Subjective:
  • Clear, proportionate mechanical/anatomical nature to aggravating and easing factors.
  • Pain associated with, and in proportion to, trauma or a pathological process or movement/postural dysfunction.
  • Pain localized to the area of injury/dysfunction (with/without some somatic referral).
  • Usually rapidly resolving or resolving in accordance with expected tissue healing/pathology recovery times.
  • Responsive to simple analgesia/NSAIDs.
  • Usually intermittent and sharp with movement/mechanical provocation; may be a more constant dull ache or throb at rest.
  • Pain in association with other symptoms of inflammation (i.e. swelling, redness, heat).
  • Pain of recent onset.
Clinical examination:
  • Clear, consistent and proportionate mechanical/anatomical pattern of pain reproduction on movement/mechanical testing of target tissues.
  • Localized pain on palpation.
  • Absence of or expected/proportionate findings of (primary and/or secondary) hyperalgesia and/or allodynia.
  • Antalgic (i.e. pain relieving) postures/movement patterns.
The symptoms and signs associated with the clinical classifications of CSP, PNP and NP have not been widely studied. Therefore, the purpose of this study (comprising 3 papers) was to identify clusters of symptoms and signs associated with the clinical classifications of CSP, PNP and NP in patients with low back (± leg) pain that presented for physiotherapy assessment.

Pertinent Results:

Four hundred sixty four patients were included in the study, with 256 patients in the NP group, 102 in the PNP group, and 106 in the CSP group.

Results for the Nociceptive Pain (NP) group:

Regression analysis revealed that a clinical classification of NP was predicted by the presence of three symptoms as well as the ‘absence’ of three symptoms and one sign.

The three symptoms that best predicted the NP classification when present were:
  1. Pain localized to the area of injury/dysfunction (with/without some somatic referral).
  2. Clear, proportionate mechanical/anatomical nature to aggravating and easing factors.
  3. Usually intermittent and sharp with movement/mechanical provocation; may be a more constant dull ache or throb at rest.
The three symptoms and one sign that had to be absent in order to predict the NP classification were:
  1. Pain variously described as burning, shooting, sharp or electric-shock-like.
  2. Pain in association with other dysesthesias (e.g. crawling, electrical, heaviness).
  3. Night pain/disturbed sleep.
  4. Antalgic (i.e. pain relieving) postures/movement patterns.
The strongest predictor of NP among the three symptoms that had to be ‘present’ was ‘Pain localized to the area of injury/dysfunction (with/without some somatic referral)’ which suggests that those patients were over 69 times more likely to be classified with a dominance of NP compared to those with non-NP. Patients with the other two symptoms that had to be present were respectively over 18 and 4 times more likely to be classified with a dominance of NP.

The group of symptoms and signs associated with a dominance of NP had a sensitivity of 90.9% and a specificity of 91.0%.

The positive predictive value was 92.7%, which means that a patient with these symptoms and signs had a 92.7% probability of being classified with NP. The negative predictive value was 88.9%, which means that a patient without these symptoms and signs had an 88.9% probability of being classified as Non-NP.

The positive likelihood ratio was 10.10 suggesting that the specified symptoms and signs are over 10 times more likely to be found in patients classified with NP than Non-NP. The negative likelihood ratio LR was 0.10, which indicates that the likelihood of the symptoms and signs being absent in patients classified with NP is very low (i.e. 0.10).

The diagnostic odds ratio was 100.7, which means that that the specified symptoms and signs are 100 times more likely to accurately than inaccurately predict a clinical classification of NP in patients classified with NP.

Results for the Peripheral Neuropathic Pain (PNP) group:

A clinical classification of PNP was predicted by the presence of two symptoms and one sign, as follows.
  1. History of nerve injury, pathology or mechanical compromise.
  2. Pain referred in a dermatomal or cutaneous distribution.
  3. Pain/symptom provocation with mechanical/movement tests (e.g. active/passive, neurodynamic, i.e. SLR) that move/load/compress neural tissue.
The strongest predictor of PNP was the symptom ‘pain referred in a dermatomal or cutaneous distribution’, and patients with it were over 24 times more likely to be classified with a dominance of PNP compared to those with non-PNP. Patients having the predictor ‘Pain/symptom provocation with mechanical/movement tests (e.g. Active/Passive, Neurodynamic, i.e. SLR) that move/load/compress neural tissue’ present were 14 times more likely to have been classified with a dominance of PNP compared to those with non-PNP and those having a ‘History of nerve injury, pathology or mechanical compromise’ were over 12 times more likely.

The cluster of symptoms and signs that lead to a classification of PNP had a sensitivity of 86.3% and a specificity of 96.0%.

The positive predictive value was 86.3%, which means that a patient with the specified symptoms and signs had an 86.3% likelihood of being classified with PNP. The negative predictive value was 96.0%, which means that a patient without the cluster of symptoms and signs had a 96.0% probability of being Non-PNP.

The positive likelihood ratio was 21.57, suggesting that the specified symptoms and signs are over 21 times more likely to be found in patients classified with PNP than Non-PNP. The negative likelihood ratio was 0.14, which means that that the likelihood of the cluster being absent in patients classified with PNP is 0.14. However, Jaeschke et al. (1) suggested that a negative likelihood ratio should be 0.1 or less in order to rule a condition out.

The diagnostic odds ratio was 150.86, which indicates that the presence of the cluster is 150 times more likely to accurately than inaccurately predict a clinical classification of PNP in patients classified with PNP.

Results for the Central Sensitization Pain (CSP) group:

A clinical classification of CSP was predicted by the presence of three symptoms and one sign, as follows.
  1. Pain disproportionate to the nature and extent of injury or pathology.
  2. Disproportionate, non-mechanical, unpredictable pattern of pain provocation in response to multiple/non-specific aggravating/easing factors.
  3. Strong association with maladaptive psychosocial factors (e.g. negative emotions, poor self-efficacy, maladaptive beliefs and pain behaviors, altered family/work/social life, medical conflict).
  4. Diffuse/non-anatomic areas of pain/tenderness on palpation.
The second symptom listed above was the strongest predictor of CSP and when present suggested that patients with ‘Disproportionate, non-mechanical, unpredictable pattern of pain provocation in response to multiple/non-specific aggravating/easing factors’ were over 30 times more likely to be classified as CSP when compared to non-CSP patients. Patients with ‘Diffuse/non-anatomic areas of pain/tenderness on palpation’ were 27 times more likely to be classified as CSP.

The group of symptoms and signs associated with a dominance of CSP had a sensitivity of 91.8% and a specificity of 97.7%.

The positive predictive value was 91.8%, which means that a patient with these symptoms and signs had a 91.8% probability of being classified with NP. The negative predictive value was 97.7%, which means that a patient without these symptoms and signs had a 97.7% probability of being classified as Non-NP.

The positive likelihood ratio was 40.64, which suggests that the specified symptoms and signs are over 40 times more likely to be found in patients classified with CSP than Non-CSP. The negative likelihood ratio LR was 0.08, which indicates that the likelihood of the symptoms and signs being absent in patients classified with CSP is very low.

The diagnostic odds ratio was 486.56, which means that that the specified symptoms and signs are about 480 times more likely to accurately than inaccurately predict a clinical classification of CSP in patients classified with CSP.

Clinical Application & Conclusions:

Clusters of symptoms and signs were identified that are associated with the clinical classifications of NP, PNP and CSP in patients with low back (± leg) pain. Each of the clusters were found to have good classification accuracy suggesting that the clusters might be useful for identifying pain arising from a dominance of these mechanisms. The clusters are presented in the Results section above, but are repeated here for clarity and emphasis.

The three symptoms that best predicted the Nociceptive Pain (NP) classification when present were:
  1. Pain localized to the area of injury/dysfunction (with/without some somatic referral).
  2. Clear, proportionate mechanical/anatomical nature to aggravating and easing factors.
  3. Usually intermittent and sharp with movement/mechanical provocation; may be a more constant dull ache or throb at rest.
The three symptoms and one sign that had to be absent in order to predict the NP classification were:
  1. Pain variously described as burning, shooting, sharp or electric-shock-like.
  2. Pain in association with other dysesthesias (e.g. crawling, electrical, heaviness).
  3. Night pain/disturbed sleep.
  4. Antalgic (i.e. pain relieving) postures/movement patterns.
A clinical classification of Peripheral Neuropathic Ppain (PNP) was predicted by the presence of two symptoms and one sign, as follows.
  1. History of nerve injury, pathology or mechanical compromise.
  2. Pain referred in a dermatomal or cutaneous distribution.
  3. Pain/symptom provocation with mechanical/movement tests (e.g. active/passive, neurodynamic, i.e. SLR) that move/load/compress neural tissue.
A clinical classification of Central Sensitization Pain (CSP) was predicted by the presence of three symptoms and one sign, as follows.
  1. Pain disproportionate to the nature and extent of injury or pathology.
  2. Disproportionate, non-mechanical, unpredictable pattern of pain provocation in response to multiple/non-specific aggravating/easing factors.
  3. Strong association with maladaptive psychosocial factors (e.g. negative emotions, poor self-efficacy, maladaptive beliefs and pain behaviors, altered family/work/social life, medical conflict).
  4. Diffuse/non-anatomic areas of pain/tenderness on palpation.

Study Methods:

This study utilized a cross-sectional, between-subjects design and was performed at four hospital sites and two private physiotherapy practices in Ireland. Fifteen experienced musculoskeletal physiotherapists participated in data collection.

Eligible patients were ? 18 years of age with low back (± leg) pain that were referred for physiotherapy assessment and/or treatment. Patients were excluded if they had a history of diabetes or central nervous system injury, were pregnant or had non-musculoskeletal low back pain (LBP).

All participants completed a standardized form that covered patient demographics. They were given a standardized clinical interview that covered details of their current low back pain, including aggravating and easing factors, diurnal variations and sensory disturbances, and their LBP history. Patients were then examined to assess their postural, movement and neurological status.

The clinicians completed a ‘Clinical Criteria Checklist’ that consisted of two parts.
  • Part 1 - Clinicians classified each patient’s pain based on clinical judgment as to the main mechanisms (i.e. NP, PNP, CSP) assumed to be causing the patient’s pain. Patients could also be assigned to a ‘mixed’ pain category (Mixed: NP/PNP, Mixed: NP/CSP, Mixed: PNP/CCP, Mixed: NP/PNP/CSP).
  • Part 2 - Clinicians completed a 38-item clinical criteria checklist that included 26 symptoms and 12 signs that were derived from a list of clinical criteria suggestive of a dominance of NP, PNP and CSP mechanisms. Clinicians could consider each criterion as ‘Present’, ‘Absent’, or ‘Don’t know’.

Study Strengths / Weaknesses:

This study employed a cross-sectional design, which is fairly low in the hierarchy of evidence (2). Cross-sectional studies merely take a slice of information out of a particular population. Further, there was no random assignment to groups and no control group was utilized.

Prior knowledge of the patients’ clinical classification (i.e. reference standard) may have influenced completion of the clinical criteria checklist because of their preconceived ideas about the nature of the clinical findings associated with the particular classification.

The study did not incorporate blinding, which would have strengthened its findings. In fact, the authors noted that blinded assignment of the reference standard and the completion of clinical criteria checklists by clinicians blinded to the patient’s clinical classification would have been desirable.

No diagnostic gold standard was used in this study, even though a gold standard is usually required when investigating the validity of tests. Instead, the best alternative ‘reference standard’ (i.e. the best available method for establishing the presence or absence of a condition of interest) was used. Nevertheless, estimates of classification accuracy found without a gold standard method to determine the true presence or absence of the target condition (i.e. CSP, PNP, NP) should be treated with caution.

The use of small to moderate sample sizes in studies that develop classification criteria tend to generate inflated odds ratios and estimates of classification accuracy. Therefore, cross-validation of the symptom and sign clusters in cohorts of patients in a variety of clinical settings is required in order to provide more accurate model estimates.

The authors suggested that subsequent validation studies of clinical criteria associated with NP, PNP and CSP should involve larger samples with a variety of neuromusculoskeletal disorders in order to obtain more robust estimates of classification accuracy.

Additional References:

  1. Jaeschke R, Guyatt GH, Sackett DL. Users’ guide to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients. JAMA 1994; 271: 703-7.
  2. Oxford Centre for Evidence-Based Medicine. Levels of evidence and grades of recommendation. Available from http://www.cebm.net/levels_of_evidence.asp.