Review of Neuropathic Pain Screening and Assessment Tools
- First Online:
- Cite this article as:
- Jones, R.C.W. & Backonja, M. Curr Pain Headache Rep (2013) 17: 363. doi:10.1007/s11916-013-0363-6
Chronic pain due to injury to or diseases of the nervous system, known as neuropathic pain (NP), is a common debilitating medication condition for which there are currently several symptomatically effective therapies. Therefore, early identification of NP in the primary and specialty care setting will avoid unnecessary delays in amelioration of symptoms. Given that it is associated with unique symptoms and physical exam signs, several assessment tools have been developed to aid medical practitioners in the identification of patients with NP. The majority of these tools have been developed to differentiate NP from nonNP and to quantify the severity of symptoms that define NP, and some have been used to aid in assessment of response to interventions. This focused review will describe the primary NP assessment tools that are currently available, and discuss their suitability for screening patients and for research applications. Wider use of NP assessment tools will facilitate the development of new therapies, further clarify the epidemiology of this condition, and improve the treatment of NP.
KeywordsNeuropathic painAssessmentScreeningLANSSNPSNPSIStEPDN-4NPQID painpainDETECTSF-MPQ-2
Pain due to injury of the nervous system, be it peripheral or central, has been recognized for centuries. With the introduction of animal models of neuropathic pain (NP), we have developed a better understanding of potential mechanisms of NP  and developed some effective treatments for it. Consequently, the identification of NP by physicians has become a focus of increasing interest and importance. Several clinical assessment tools have been developed to better characterize and diagnose NP for clinical and research purposes. All of these tools incorporate unique symptoms and signs of NP. Most were designed for use by general medical practitioners to distinguish NP from non-neuropathic pain (nonNP), and will be reviewed first. A few were developed for the experienced consultant to quantify the severity of NP and monitor treatment response, and will be reviewed second. In general, NP assessment tools should be differentiated from quantitative sensory testing (QST) of NP, as the latter is an approach that relies exclusively on physical exam signs and evoked phenomena produced by well-defined stimuli applied according to specific protocols.
There are many similarities among the various published NP assessment tools, with some important distinctions. In 1997, the Neuropathic Pain Scale (NPS) was published; a wave of publications soon followed, describing other NP tools developed almost simultaneously in different countries in Europe and the US. All were either initially developed to be self-administered, or quickly produced revised versions for this purpose. Most of the NP tools were constructed based on patients with multiple types of NP (a notable exception is the NPS that was developed in patients with postherpetic neuralgia (PHN). While most of the tools survey a similar set of symptoms, there are important differences that stem largely from the distinct compositions of NP patients upon which they were developed. The unique content of an assessment tool may lend additional validity of one test over another when applied to a specific NP patient that manifests those unique features. These differences will be detailed below.
Assessment Tools that Differentiate NP from nonNP
Leeds Assessment of Neuropathic Symptoms and Signs (LANSS)
Developed in 2001, the LANSS was the first assessment tool with the express purpose of discriminating NP from nonNP [2•]. It combines elements of patient symptom self-assessment and physical exam signs obtained by a trained medical practitioner. The S-LANSS is a modified version that relies on the patient to perform the physical exam components . The initial LANSS development study included subjects with NP primarily from peripheral nerve injury, including patients with Complex Regional Pain Syndrome type I (CRPS I). It relies on five “yes or no”-type questions intended to describe the essence of a group of related symptoms associated with NP: dysesthesia, dysautonomia, paroxysmal pattern, and allodynia. This method was chosen to putatively enhance the specificity of the descriptive portion of the assessment. The physical exam component of the LANSS tests for altered pinprick threshold and the presence of allodynia to brush with a cotton swab, while the S-LANSS has the patient rub the painful location with their index finger and press on the painful location with their fingertip to test for allodynia and hyperaglesia, respectively . Each item in the LANSS and S-LANSS is weighted differently based on its predictive importance (determined by taking the odds ratio for a given item and arbitrarily rounding down to the nearest whole value). A score of > 12 (out of a total possible 24) would be suggestive of NP with a sensitivity of 82 % and specificity and 80 %.
The LANSS is freely available, relatively simple to perform, and easy to score. Both the LANSS and S-LANSS have been validated in several languages, including Chinese  and Turkish . The reported specificity and sensitivity is relatively high compared to other tools, ranging from 82 to 91 % and 80 to 94 %, respectively. Of note, this tool includes an assessment of skin color changes; therefore, it may particularly well suited for identifying patients with CRPS.
The biggest shortcoming of the LANSS, and several other tools, is that it is not a quantitative assessment of NP severity. This shortcoming may partially explain why it does not correlate with the degree of injury present in cases of known neurologic dysfunction, e.g. electrodiagnostic findings in patients with carpal tunnel syndrome . One of the proposed advantages of the LANSS is its ability to discriminate NP from nonNP; however, patients with fibromyalgia, not traditionally thought to have a NP condition, reportedly scored high on the LANSS . This apparent lack of specificity has also been reported for other NP assessment tools, and may actually reflect the presence of NP-like characteristics in conditions like fibromyalgia, rather than a limitation of the tool.
It should also be noted that the use of different methods to assess hyperalgesia by the LANSS (pinprick) and S-LANSS (pressure) activate distinct neural pathways and could introduce variability between the two versions .
Neuropathic Pain Questionnaire (NPQ)
The NPQ is a self-report assessment that incorporates 12 questions pertaining to symptom quality (burning, sensitive to touch, shooting, numbness, electric, tingling, squeezing, freezing), exacerbating factors (allodynia, sensitivity to weather), and affective impact (unpleasantness, overwhelming nature). It was developed in a mixed patient population with either NP or nonNP based on chart review, although the exact etiologies were not specified except that patients with fibromyalgia, headache, and myofascial pain syndrome were specifically excluded. The patient scores each question on a scale from 0 to 100, then the scores are multiplied by a discriminant function coefficient, and totaled incorporating a set constant value. A positive score predicts the presence of NP, while a negative score predicts non-neuropathic pain with a sensitivity and specificity of 66.6 % and 71.4 %, respectively . A short-form of the NPQ (NPQ-SF), incorporating only three items from the original 12 (numbness, tingling, increased pain due to touch), has been demonstrated to have similar predictive properties [10••]. The NPQ has been translated and validated in Swedish, Korean and Chinese.
The NPQ is freely available, self-administered, and requires little time to complete. The inclusion of discriminant coefficients in the scoring of the NPQ lends a degree of face validity to its construction. Given that the NPQ was developed in a broad patient population with the express purpose of identifying NP from nonNP, it may be particularly useful for general practitioners seeking diagnostic information. It is also the only tool to incorporate sensitivity to changes in weather and, together with the NPS, affective dimensions of pain; therefore, it may be particularly useful in patients in whom these disease features are pronounced.
Of the discriminative NP tools available, the NPQ has one of the lowest reported accuracies . This is partly due to its intended purpose to be simple to administer, require a minimal amount of time, and require no input from a medical professional. The calculation of the NPQ is also somewhat more complicated than other tools, owing to the use of discriminant coefficient scores to weight the tool’s individual assessment items.
Douleur Neuropathique 4 Questions (DN4)
The DN4 was developed by the French Neuropathic Pain Group in a mixed patient population with either NP from predominantly peripheral nerve injury or nonNP from mainly osteoarthritis and inflammatory arthritis . It incorporates both symptoms (seven questions) and physical exam signs (three questions). It is one of the few tools to incorporate itch into the model. Based on the initial validation study, a score ≥ 4 (out of total possible of 10) achieved a specificity of 82.9 % and sensitivity of 89.9 % for identifying NP. Like the LANSS, a modified version of the DN4 without the physical exam component has been described with nearly equivalent accuracy. The DN4 has been translated and validated in Spanish, Yiddish, Chinese, German, and English.
The free availability and multiple translations of the DN4 has facilitated its use to the prevalence of NP in populations that have traditionally been unrepresented in the literature (for examples, see [12–14]). It is easy and rapid to employ, easy to score, and has reliability comparable to other tools. Interestingly, the DN4 has also been used to correlate specific sensory features of NP with abnormalities in peripheral nerve fiber function, e.g. A□ fiber dysfunction and paroxysmal pain and dysethesias, in contrast to A□ fiber dysfunction and spontaneous constant pain [15•]. It has also been used to characterize treatment effects .
The biggest shortcoming of this tool is that it is not quantitative. Some studies suggest that the DN4 may not retain its diagnostic accuracy in different patient populations, e.g. acute postoperative pain patients . This shortcoming is understandable, given that these patients were not included in its development.
The painDETECT is a self-assessment tool consisting of seven items relating to symptoms quality: burning, tingling/prickling, sensitivity to light touch, paroxysmal pattern, sensitivity to temperature, numbness, and sensitivity to slight pressure . It was constructed based on the authors’ analyses of patient descriptors, interviews with experts, and literature review. The questions are scored from 0 to 5, with 0 denoting a sensation never experienced by the patient and 5 denoting something the patient experiences very strongly. There is one question relating to pain pattern that is scored differently: persistent pain with some fluctuation (0), persistent pain with flairs (−1), flairs only (+1), and “pain attacks with pain between them” (+1). Finally, there is one question pertaining to radiation of the pain to other parts of the body, scored either 0 or +2. The painDETECT was validated in a prospective multicenter study in Germany, with the clinician diagnosis as the gold standard (similar to the validation studies of many of the NP assessment tools). A score of > 19 achieved sensitivity and specificity of 85 % and 80 %, respectively. The painDETECT has been translated and validated in Spanish, French and English.
The painDETECT is freely available, quick to administer, and easy to score. It has comparable accuracy compared to other neuropathic pain screening tools. Given the inclusion of radiation into the assessment, it may be particularly good for distinguishing NP in patients with radicular pain, such as in patients with neck and low back pain. Indeed, it has been shown to be predictive of response to epidural steroid injections .
As with other tools, the painDETECT is not as predictive of NP in patient populations with painful conditions, other than those upon which it was originally developed. For example, a study of cancer-related NP found that the painDETECT had a sensitivity and specificity of only 53 and 77 %, respectively. This difference may be due not only to differences in study population, but also clinician experience providing the “gold standard” diagnosis of NP.
The ID Pain is a self-assessment tool consisting of six “yes/no”-type questions: four questions query the presence of pins/needles, hot/burning, numbness, and electrical shocks symptoms; one assesses the presence of allodynia; and one describes the distribution to joints. The questions are scored 0 or 1 for a “no” or “yes” answer, respectively, while the question regarding limitation of pain to joints is scored a 0 or −1 if answered no or yes, respectively. Therefore, the maximum score is 5. In the initial validation study, a score > 2 had a sensitivity and specificity of 0.73 and 0.69, respectively, using a single clinician assessment as the gold standard. From this, cut-offs of scores ≥ 4, 2–3, and ≤ 1 on the ID Pain have been interpreted to reflect the presence of neuropathic pain as likely, possible, and unlikely, respectively. The ID Pain tool has been validated in several different languages, including Chinese, Spanish and Thai.
The ID Pain tool is brief, easy-to-use, and freely available. It was developed in a broad, but unspecified, population of chronic pain patients without headache; therefore, it may be particularly appropriate to apply as a screening tool in the primary care setting. It also incorporates an assessment of anatomical distribution of a subject’s pain; therefore it may be useful in distinguishing neuropathic from other types of localized pain, e.g. due to arthritis.
Like many other tools, the ID Pain is not quantitative. While relatively sensitive, the ID Pain tool may not possess the specificity of other tools, in certain patient populations, e.g. patients with pain related to breast cancer .
Standardized Evaluation of Pain (StEP)
The StEP is an assessment tool incorporating the most items, both symptoms and exam signs, of all the tools described in this review [21•]. From a large battery of items, hierarchical cluster analysis was used to identify six groups of symptom questions and ten physical exam signs based on their ability to distinguish NP from nonNP: location (superficial, deep), temporal characteristics (intermittent, constant), quality (burning, cramping, throbbing, sharp, pulling, stabbing, shooting, squeezing, dull, pins and needles, cold, stinging, electric shock, tender, spreading), exacerbation by activity and position, paresthesia and dysesthesia (tingling, itching, warm/cold), presence and intensity of pain, skin changes, decreased response to vibration, and sensitivity to blunt pressure, brush, pinprick, warmth and cold, temporal summation of sensations, and positive straight leg raise test. Interestingly, the StEP was developed in a cohort of patients with pain from diabetic peripheral neuropathy (DPN) and PHN, as well as both radicular and axial low back pain (LBP). However, the authors validated a restricted set of items, including three symptom questions (constant pain, pain quality described as burning and/or cold, and present of paresthesia and dysesthesia) and eight physical exam items (skin changes, blunt pressure, brush, vibration, pinprick, cold, temporal summation, and straight leg raise) for their ability to discriminate NP from nonNP specifically in patients presenting with LBP. Each item is weighted differently based on its relative discriminative ability, with positive straight leg raise scored at +7 and presence of skin changes −3. Neuropathic LBP is indicated by a score > 3 with a high sensitivity and specificity (92 and 97 %, respectively), surpassing both the DN4 and 1.5 T magnetic resonance imaging (MRI) of the lumbar spine in predicting neuropathic (i.e. radicular) LBP.
When applied to patients with low back pain, the StEP has diagnostic accuracy surpassing that of any tool described in this article, as well as MRI imaging of the spine. As the authors report in their study, the accuracy of the StEP is due in large part to the inclusion of sophisticated physical examination based signs. The authors also report that, with appropriate training, the exam can be completed in 10–15 min, much longer than that reported for other assessment tools with few, if any, physical examination tests.
The StEP is copyrighted, and therefore not freely available, but could be reproduced with permission. This tool is also not quantitative. Given its dependence upon extensive and specialized physical examination, it is not suitable for direct mailings and must be administered by well-trained personnel. Its clinical utility as a screening tool over other simpler ones in a busy community practice where 10 to 15 minutes can represent a significant expenditure of time raises serious questions about how realistic it would be to apply, in addition to its requirement for a highly skilled examination. Moreover, although it was developed in a more diverse group of NP patients, it has only been validated in patients presenting with LBP, and the high degree of accuracy is based predominantly upon the presence of a positive straight leg raise test.
Assessment Tools that Characterize Pain Symptoms in Patients with NP
Neuropathic Pain Scale
The NPS is the first assessment tool developed specifically for neuropathic pain [22••], and is not a screening tool per se. The sensory qualities it includes were selected based on the clinical experience of a physician involved in its development. It contains seven questions relating to symptom intensity and quality: intense, sharp, hot, dull, cold, sensitive to light touch, and itch. There is also one question on temporal pattern (constant plus flairs, constant with fluctuation, flair only), one question on pain unpleasantness, and two questions on intensity of deep and surface pain. Each question, except for that pertaining to temporal pattern, is rated by the subject on a 0–10 scale. No composite score is generated from these numbers, though some groups have generated one from NPS component subsets . The initial validation study demonstrated significant differences in description of pain by patients with PHN versus other NP conditions, as well as the differential effects of two different treatments (intravenous lidocaine or phentolamine) on the severity of NP symptoms in patients with CRPS-I. The NPS has subsequently been used in many studies examining the effectiveness of various pain treatments (for examples, see [24–26]). The NPS has been translated and validated in Spanish, French and German.
The NPS is freely available, widely employed, and has well-described utility for examining the ability of specific treatments to improve neuropathic pain symptoms. In addition, together with the DN4, it is one of the few described assessment tools that includes itch as a measured variable.
Because the NPS was developed by a single clinician, it is not surprising that it omits several qualities that are commonly described by NP patients, including “electric”, “tingling”, “radiating”, and “numbness”. This has been addressed by modification of the NPS into the Pain Quality Assessment Scale (PQAS) . The NPS has been employed by some as a screening tool, although it was not developed as such, and has only been fully validated for that purpose in patients with multiple sclerosis . The ease with which it is used to distinguish treatment effects on specific dimensions of a subject’s pain stands in stark contrast to how complicated it is to convert the NPS into a tool to characterize pain as neuropathic versus non-neuropathic . Therefore, it primarily represents a research tool for investigating treatment response in patients with NP, rather than a screening tool for use in the clinic.
Neuropathic Pain Symptom Inventory (NPSI)
The NPSI was developed in 2004 based on the clinical experience and analysis of the literature by a panel of seven French and Belgian experts . It includes four dimensions of NP: spontaneous ongoing pain, spontaneous paroxysmal pain, evoked pain, and paresthesia/dysesthesia. A pilot study with a preliminary list of questions was performed in an enriched cohort of NP patients with clinically defined lesions of the nervous system (e.g. CRPS-I patients were excluded). Based on these preliminary results, the final version of the NPSI was generated including a total of 12 items: four relating to spontaneous pain qualities (burning, squeezing, pressure, and duration within the last 24 hours), three relating to paroxysmal pain qualities (electric shock, stabbing, number of attacks in the last 24 hours), three relating to aggravating factors (brushing, pressure, cold), and two relating to abnormal sensations (pins/needles, tingling). Each question is rated by the patient from 0 to 10, and scoring the assessment involves summation of five subscores, derived from factor analysis of data from the pilot study as being independent and important identifiers of NP. The final version of the questionnaire was validated in a larger group of patients with well-defined NP, predominantly due to peripheral nerve injury, but also included patients with PHN, DPN, and multiple sclerosis. The NPSI demonstrated good test-retest reliability and correlation with patient and clinician assessment of improvement after initiation of treatments for NP. The NPSI has subsequently been used to assess the efficacy of treatments for NP [30, 31] and characterize symptoms of distinct NP conditions [32, 33]. The NPSI has been translated and validated in Spanish, German and English.
Like the NPS, the NPSI is a well-characterized method for quantifying distinct symptoms experienced by a diversity of NP patients as well as treatment effects. It is freely available, validated in multiple languages, and requires little time to perform (~7 min). In addition, it has been used to study the underlying basis for individual symptoms qualities [15•]. Consequently, it is a valuable research tool for many clinical studies on NP.
The NPSI was not designed as a screening tool, and although it generates a composite score, in contrast to other discriminative tools, there are no guidelines on cut-off scores indicating that NP is likely or unlikely. Therefore, it is not necessarily suitable for NP identification in a clinic setting. More recently, a derivation of the NPSI has been developed to address this question with diagnostic accuracy similar to other tools . Additional studies have not been performed to validate this approach. Interestingly, although the NPS and the NPSI are similar in that they are symptom assessment tools and not intended for screening purposes, they differ significantly in their content.
Short Form McGill Pain Questionnaire 2 (SF-MPQ-2)
The MPQ has been used for decades as the predominant method for characterization of chronic pain of all varieties. The desire to generate pathophysiology-specific assessment tools has led to a recent modification of the short form of the MPQ (SF-MPQ-2) to include just seven items that discriminate NP from nonNP, based on the study authors’ clinical experience and review of relevant research studies, as well as patient focus groups . The SF-MPQ-2 includes items divided into four descriptor subscales: continuous pain (throbbing, cramping, gnawing, aching, heavy, tender), intermittent pain (shooting, stabbing, sharp, splitting, electric-shock, piercing), NP (hot-burning, cold-freezing, sensitivity to light touch, itching, tingling/”pins and needles”, numbness), and affective (tiring/exhausting, sickening, fearful, punishing/cruel). Using a web survey of patients with a variety of chronic pain conditions, those patients who self-reported NP had significantly higher total scores as well as subscale scores on the neuropathic and intermittent descriptor domains. The authors went on to test the ability of the SF-MPQ-2 to detect a treatment effect of amitriptyline and ketamine in patients with DPN, and demonstrated a significant reduction in SF-MPQ-2 scores in the treatment groups .
The SF-MPQ-2 is freely available, self-administered, and simple to perform. Therefore, it meets several important criteria for a screening tool. However, like the NPS and NPSI, it does not generate a composite score that carries diagnostic value in discriminating NP from nonNP. It is more useful as a measure of treatment response (for example, see [36–38]). The SF-MPQ-2 has been translated into Persian.
The current gold standard in the diagnosis of NP is a detailed history and physical examination performed by an experienced clinician (although exactly what is entailed is usually not defined). Therefore, the use of patient reported diagnosis for the basis of the development of the SF-MPQ-2 is suspect. The SF-MPQ-2 has been subsequently validated in other studies that demonstrated scores are significantly higher in patients with clinician-diagnosed NP than those with nonNP , and it is also sensitive to treatments designed for NP . Therefore, its clinical utility does not appear to be significantly affected by the nature of its development.
The widespread interest in developing NP screening and assessment tools is reflected in the number of tools developed in multiple countries with different cultures and linguistic backgrounds. At first glance, all of these tools share a number of items [40, 41]; however, the form and structure of the questions is sufficiently different that they represent distinct assessment tools. This is apparent in studies comparing various NP tools, which, although limited in scope, represent a very encouraging development that needs to occur for these tools to receive wider acceptance. An example is a comparison of the accuracy of the NPQ and S-LANSS to identify NP in breast cancer patients that reported that the NPQ had less specificity but greater sensitivity than the S-LANSS . These results support the role of the NPQ as a screening rather than a confirmatory tool. However, another study done in Chinese pain patients reported a sensitivity and specificity of 52.9 and 91.4 %, respectively, for the NPQ, suggesting it is more appropriate as a confirmatory tool . These conflicting results highlight how accuracy changes depending upon the study and patient population. Other comparison studies have demonstrated equivalent accuracy of the LANSS and DN4 in a population of patients with peripheral neuropathy, and greater sensitivity of the ID Pain than the LANSS at identifying neuropathic pain in patients with breast cancer . The latter study confirms the utility of the ID Pain tool as a screening measure. More such studies are needed to delineate which tools are more appropriate for use in specific patient populations with NP pain.
Current State of Use of NP Assessment Tools
All of the NP tools have been translated in many languages, as described above, which would mean that the extent of their use globally is not certain. All of the publications about the use of NP assessment tools report use of these tools in various research settings. Though all of the published NP symptoms tools have been used in many studies, they have not been used in studies that would lead to registration of new therapies or medications, and surprisingly little is known about the extent of their use in clinical practice. One of the authors of this review (MB) has personal experience with the utility of one of these tools and how it contributes to decision-making in everyday practice. The ID Pain is used in the General Medicine clinic at the Veterans Administration Medical Center in Madison, WI, USA. As a standard of care, a numerical pain rating is obtained in all primary care patients and the ID Pain is administered to all patients who score ≥ 4 (out of total possible of 10). When those patients score ≥ 2 on the ID Pain, clinicians perform a sensory examination on the painful area and if NP is confirmed by the presence of sensory abnormalities (deficits, hypersensitivity or both), neuromodulating medications such as nortriptyline or gabapentin are prescribed, rather than opioid based analgesics, which are commonly used to treat moderate to severe pain.
NP assessment tools represent an important advancement in NP clinical research and are an important component in the assessment of NP in patients, capable of providing first-line information for building clinical evidence for the diagnosis of NP. The qualitative tools are simple and sufficiently accurate for use in clinical practice, and those that have quantitative properties are important research tools, in particular for developing new phenotypic profiles of NP patients based on symptomatology. Future studies will provide additional information about the performance characteristics of these tools in both research and clinical settings.
Compliance with Ethics Guidelines
Conflict of Interest
Dr. R. Carter W. Jones III and Dr. Miroslav “Misha” Backonja reported no potential conflicts of interest relevant to this article.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.