Expert performance in low-back disorder recognition using patient pain drawings

Data-driven identification of distinct pain drawing patterns and their association with clinical and psychological factors: a study of 21,123 patients with spinal pain

ABSTRACT

The variability in pain drawing styles and analysis methods has raised concerns about the reliability of pain drawings as a screening tool for nonpain symptoms. In this study, a data-driven approach to pain drawing analysis has been used to enhance the reliability. The aim was to identify distinct clusters of pain patterns by using latent class analysis (LCA) on 46 predefined anatomical areas of a freehand digital pain drawing. Clusters were described in the clinical domains of activity limitation, pain intensity, and psychological factors. A total of 21,123 individuals were included from 2 subgroups by primary pain complaint (low back pain (LBP) [n = 15,465]) or midback/neck pain (MBPNP) [n = 5658]). Five clusters were identified for the LBP subgroup: LBP and radiating pain (19.9%), radiating pain (25.8%), local LBP (24.8%), LBP and whole leg pain (18.7%), and widespread pain (10.8%). Four clusters were identified for the MBPNP subgroup: MBPNP bilateral posterior (19.9%), MBPNP unilateral posterior + anterior (23.6%), MBPNP unilateral posterior (45.4%), and widespread pain (11.1%). The clusters derived by LCA corresponded to common, specific, and recognizable clinical presentations. Statistically significant differences were found between these clusters in every self-reported health domain. Similarly, for both LBP and MBPNP, pain drawings involving more extensive pain areas were associated with higher activity limitation, more intense pain, and more psychological distress. This study presents a versatile data-driven approach for analyzing pain drawings to assist in managing spinal pain.

Colored Pain Drawing as a Clinical Tool in Differentiating Neuropathic Pain from Non-Neuropathic Pain

OBJECTIVES

This is a prospective, blinded, case-control study of patients with chronic pain using body diagrams and colored markers to show the distribution and quality of pain and sensory symptoms (aching, burning, tingling, numbness, and sensitivity to touch) experienced in affected body parts.

METHODS

Two pain physicians, blinded to patients' clinical diagnoses, independently reviewed and classified each colored pain drawing (CPD) for presence of neuropathic pain (NeuP) vs. non-neuropathic pain (NoP). A clinical diagnosis (gold standard) of NeuP was made in 151 of 213 (70.9%) enrolled patients.

RESULTS

CPD assessment at "first glance" by both examiners resulted in correctly categorizing 137 (64.3% by examiner 1) and 156 (73.2% by examiner 2) CPDs. Next, classification of CPDs by both physicians, using predefined criteria of spatial distribution and quality of pain-sensory symptoms, improved concordance to 212 of 213 CPDs (Kappa = 0.99). The diagnostic ability to correctly identify NeuP and NoP by both examiners increased to 171 (80.2%) CPDs, with 80.1% sensitivity and 80.6% specificity (Kappa = 0.56 [95% confidence interval: 0.44-0.68]). The severity scores for pain and sensory symptoms (burning, tingling, numbness, and sensitivity to touch) on the Neuropathic Pain Questionnaire were significantly elevated in NeuP vs. NoP (P < 0.001).

CONCLUSIONS

This study demonstrates good performance characteristics of CPDs in identifying patients with NeuP through the use of a simple and easy-to-apply classification scheme. We suggest use of CPDs as a bedside screening tool and as a method for phenotypic profiling of patients by the quality and distribution of pain and sensory symptoms.

Risk Factors for Postsurgical Foot Complaints One Year Following Degenerative Lumbar Spinal Surgery

MINI: Postsurgical foot complaints occurred frequently in 853 degenerative lumbar surgeries (prevalence, 20.6%; n = 176). Risk factor analysis showed that the incidence of postsurgical foot complaint was significantly higher in patients with preoperative foot symptoms (adjusted odds ratio, 5.532) and in those with preoperative sensory deficits on the leg (adjusted odds ratio, 1.904).

STUDY DESIGN

Retrospective.

OBJECTIVE

To investigate the prevalence and risk factors of postsurgical foot complaints (PFCs) following spinal surgery by using a modified pain drawing (PD) instrument.

SUMMARY OF BACKGROUND DATA

Although many patients report nonspecific foot symptoms with various clinical presentation, there is not a well defined diagnostic criterion. PDs are essential for measuring spinal surgery outcomes. We created a modified patient-physician communication-based PD instrument to overcome the limitations of the previous system.

METHODS

We included 853 consecutive patients who underwent decompression with or without fusion. PFCs were defined as sensory foot symptoms, including ambiguous sensations that were not clearly due to spinal pathology. Patients who complained of postoperative foot symptoms at more than two consecutive visits were assigned to the PFC group. The remaining patients were assigned to the asymptomatic group. We collected medical records using our PD instrument and compared variables between the two groups.

RESULTS

In total, 176 (20.6%) of the 853 patients had PFCs. The duration of preoperative leg pain was significantly longer in the PFC group than in the asymptomatic group (2.8 vs. 2.2 years; P = 0.048). The proportions of preoperative foot symptoms (82.9% vs. 43.3%) and sensory deficits on the leg (48.6% vs. 27%) were significantly greater in the PFC group than in the asymptomatic group (P < 0.001). Multivariable logistic regression analysis revealed two independent risk factors: the presence of preoperative foot symptoms (adjusted odds ratio, 5.532) and preoperative sensory deficits on the leg (adjusted odds ratio, 1.904).

CONCLUSION

PFCs occurred frequently after degenerative lumbar spinal surgery (prevalence, 20.6%). Based on our data using PD instrument, it can help reduce the incidence of PFCs if patients are informed and educated that preoperatively existing foot symptom and sensory deficits on the leg are significant risk factors for PFC development.

LEVEL OF EVIDENCE

4.

From Paper to Digital Applications of the Pain Drawing: Systematic Review of Methodological Milestones

Background

In a pain drawing (PD), the patient shades or marks painful areas on an illustration of the human body. This simple yet powerful tool captures essential aspects of the subjective pain experience, such as localization, intensity, and distribution of pain, and enables the extraction of meaningful information, such as pain area, widespreadness, and segmental pattern. Starting as a simple pen-on-paper tool, PDs are now sophisticated digital health applications paving the way for many new and exciting basic translational and clinical applications.

Objective

Grasping the full potential of digital PDs and laying the groundwork for future medical PD apps requires an understanding of the methodological developments that have shaped our current understanding of uses and design. This review presents methodological milestones in the development of both pen-on-paper and digital PDs, thereby offering insight into future possibilities created by the transition from paper to digital.

Methods

We conducted a systematic literature search covering PD acquisition, conception of PDs, PD analysis, and PD visualization.

Results

The literature search yielded 435 potentially relevant papers, from which 53 methodological milestones were identified. These milestones include, for example, the grid method to quantify pain area, the pain-frequency maps, and the use of artificial neural networks to facilitate diagnosis.

Conclusions

Digital technologies have had a significant influence on the evolution of PDs, whereas their versatility is leading to ever new applications in the field of medical apps and beyond. In this process, however, there is a clear need for better standardization and a re-evaluation of methodological and technical limitations that no longer apply today.

Digital Pain Drawings: Assessing Touch-Screen Technology and 3D Body Schemas

OBJECTIVES

To assess the consistency and level of agreement between pain drawings collected on (1) paper and a personal computer tablet; and (2) between a 2-dimensional (2D) line drawing and 3-dimensional (3D) body schema.

MATERIALS AND METHODS

Pain-free participants (N=24) recreated a premarked "pain" area from a 2D line drawing displayed on paper onto paper or tablet, and individuals with chronic neck pain (N=29) expressed their current pain on paper and tablet. A heterogeneous group (N=26) was recruited from cross-disciplinary pain clinic and expressed their pain on a 2D line drawing and a 3D body schema, as displayed on a tablet, and then completed an user-experience questionnaire.

RESULTS

Pain drawings showed moderate to high level of consistency and a high level of agreement for paper and tablet and between 2D line drawing and 3D body schema. A fixed bias (-1.0042, P<0.001) revealed that pain areas were drawn slightly smaller on paper than on tablet, and larger on the 2D than the 3D body schema (-0.6371, P=0.003), as recorded on a tablet. Over one-third of individuals with chronic pain preferred and/or believed that the 3D body schema enabled a more accurate record; 12 believed they were equal, and 3 preferred the 2D line drawing.

DISCUSSION

Pain drawings recorded with touch-screen technology provide equal reliability to paper but the size of the drawing slightly differs between the platforms. Although, 2D line drawings and 3D body schemas were similar in terms of consistency and reliability, it remains to be confirmed whether 3D body schemas increase the accuracy and precision of pain drawings.

Localizing value of pain distribution patterns in cervical spondylosis

Study Design

Prospective observational study.

Purpose

To investigate the value of pain distribution in localizing appropriate surgical levels in patients with cervical spondylosis.

Overview of Literature

Previous studies have investigated the value of pain drawings in its correlation with various features in degenerative spine diseases including surgical outcome, magnetic resonance imaging findings, discographic study, and psychogenic issues. However, there is no previous study on the value of pain drawings in identifying symptomatic levels for the surgery in cervical spondylosis.

Methods

The study collected data from patients with cervical spondylosis who underwent surgical treatment between August 2009 and July 2012. Pain diagrams drawn separately by each patient and physician were collected. Pain distribution patterns among various levels of surgery were analyzed by the chi-square test. Agreement between different pairs of data, including pain diagrams drawn by each patient and physician, intra-examiner agreement on interpretation of pain diagrams, inter-examiner agreement on interpretation of pain diagrams, interpretation of pain diagram by examiners and actual surgery, was analyzed by Kappa statistics.

Results

The study group consisted of 19 men and 28 women with an average age of 55.2 years. Average duration of symptoms was 16.8 months. There was no difference in the pain distribution pattern at any level of surgery. The agreement between pain diagram drawn by each patient and physician was moderate. Intra-examiner agreement was moderate. There was slight agreement of inter-examiners, examiners versus actual surgery.

Conclusions

Pain distribution pattern by itself has limited value in identifying surgical levels in patients with cervical spondylosis.

How fast pain, numbness, and paresthesia resolves after lumbar nerve root decompression: a retrospective study of patient's self-reported computerized pain drawing

STUDY DESIGN

A single-center retrospective study.

OBJECTIVE

To compare the speed of recovery of different sensory symptoms, pain, numbness, and paresthesia, after lumbar nerve root decompression.

SUMMARY OF BACKGROUND DATA

Lumbar radiculopathy is characterized by different sensory symptoms like pain, numbness, and paresthesia, which may resolve at different rates after surgical decompression.

METHODS

Eighty-five cases with predominant lumbar radiculopathy treated surgically were reviewed. Oswestry Disability Index score, 36-Item Short Form Health Survey scores (Physical Component Summary and Mental Component Summary), and pain drawing at preoperative and at 6 weeks, 3 months, 6 months, and 1-year follow-up were reviewed. Recovery rate between different sensory symptoms were compared in all patients, and between the short-term compression (<6 mo) and long-term compression groups.

RESULTS

At baseline, 73 (85.8%) patients had pain, 63 (74.1%) had numbness, and 38 (44.7%) had paresthesia; 28 (32.9%) had all these 3 component of sensory symptoms. Mean pain score improved fastest (55.3% at 6 wk); further resolution until 1 year was slow and not significant compared with each previous visit. Both numbness and paresthesia scores showed a trend of faster recovery during the initial 6-week period (20.5% and 24%, respectively); paresthesia recovery reached a plateau at 3 months postoperatively, but numbness continued a slow recovery until 1-year follow-up. Both Oswestry Disability Index score and Physical Component Summary scores (54.02 ± 1.87 and 26.29 ± 0.93, respectively, at baseline) improved significantly compared with each previous visits at 6 weeks and 3 months postoperatively, but further improvement was insignificant. Mental Component Summary showed a similar trend but smaller improvement. The short-term compression group had faster recovery of pain than the long-term compression group.

CONCLUSION

In lumbar radiculopathy patients after surgical decompression, pain recovers fastest, in the first 6 weeks postoperatively, followed by paresthesia recovery that plateaus at 3 months postoperatively. Numbness recovers at a slower pace but continues until 1 year.

LEVEL OF EVIDENCE

4.

Diagnosis of lumbar spinal stenosis: an updated systematic review of the accuracy of diagnostic tests

STUDY DESIGN

Systematic review of diagnostic studies.

OBJECTIVE

To update our previous systematic review on the diagnostic accuracy of tests used to diagnose lumbar spinal stenosis.

SUMMARY OF BACKGROUND DATA

A wide range of clinical, radiological, and electrodiagnostic tests are used to diagnose lumbar spinal stenosis. An accurate diagnosis is vital, because lumbar spinal stenosis may require specific medical advice and treatment. Therefore, it is important to know the accuracy of these diagnostic tests currently available.

METHODS

A comprehensive literature search was conducted for original diagnostic studies on lumbar spinal stenosis, in which one or more diagnostic tests were evaluated with a reference standard, and diagnostic accuracy was reported or could be calculated. Our previous systematic review included studies up to March 2004; this review is current up to March 2011. Included studies were assessed for their methodological quality using the QUADAS tool. Study characteristics and reported diagnostic accuracy were extracted.

RESULTS

Twenty-two additional articles in addition to the 24 included in the previous review met the inclusion criteria. Combined, this resulted in 20 articles concerning imaging tests, 11 articles evaluating electrodiagnostic tests, and 15 articles evaluating clinical tests. Estimates of the diagnostic accuracy of the tests differed considerably.

CONCLUSION

There is a need for a consensus on criteria to define and classify lumbar spinal stenosis. At present, the most promising imaging test for lumbar spinal stenosis is magnetic resonance imaging, avoiding myelography because of its invasiveness and lack of superior accuracy. Electrodiagnostic studies showed no superior accuracy for conventional electrodiagnostic testing compared with magnetic resonance imaging. These tests should be considered in the context of those presenting symptoms with the highest diagnostic value, including radiating leg pain that is exacerbated while standing up, the absence of pain when seated, the improvement of symptoms when bending forward, and a wide-based gait.

Reliability of a preliminary 3-D pain mapping program

OBJECTIVE

The objective of this preliminary investigation was to evaluate the test-retest reliability of a new pain assessment method referred to as 3-D pain mapping.

METHODS

In Study 1, 101 chronic noncancer pain patients from four sites reported their pain using the method on two occasions (separated by approximately 10 days). The patients marked intensity, surface location, and depth of pain on a 3-D computer display of a male or female body. The model body could be rotated in order to mark multiple pain locations. In Study 2, 25 patients from a single site were tested with a revised version of the mapping program used in Study 1. Each patient gave ratings on two occasions separated by approximately 1 week.

RESULTS

In Study 1, the intra-class correlations of the 3-D pain mapping measures were moderate to high for maximum pain intensity (0.73), vertical location of the point of maximum pain (0.94), and the number of pain marks (0.84). Correlations were low for the horizontal location of the point of maximum pain (0.56) and for the depth of pain (0.50). In Study 2, using the revised program, intra-class correlations were moderate for pain intensity (0.76), and high for the vertical (0.99) and horizontal (0.98) locations of the point of maximum pain, number of pain marks (0.89), and the depth of pain (0.84).

CONCLUSION

Three-dimensional pain mapping enables patients to report the location and intensity of their pain on all parts of the body, and such ratings are highly reliable. Future studies are needed to determine whether the clinical value of this method can improve the accuracy of pain diagnoses and the quality of pain management.

Technology Insight: imaging of low back pain

Chronic low back pain is a common condition that has significant economic consequences for affected patients and their communities. Despite the prevailing view that an anatomic diagnosis is often impossible, an origin for the pain can frequently be found if current diagnostic techniques are fully used. Such techniques include a mixture of noninvasive and invasive imaging. Prevalence data suggest that the intervertebral disc is one of the most common sources of back pain, accounting for around 40% of cases. The pathologic basis for discogenic low back pain might be full-thickness radial tears of the annulus fibrosus. Unfortunately, only MRI can image the internal morphology of the disc, and both CT and MRI lack the necessary specificity to validate this hypothesis. Many so-called radiographic abnormalities seen on CT and MRI are commonly encountered in asymptomatic individuals. Invasive techniques such as joint injections, nerve blocks and provocative discography can show the connection between an abnormal image and the source of low back pain, but do have notable related risks. The development of a noninvasive, low-risk technique that can show this connection is desirable.

Pain drawing scoring is not improved by inclusion of patient-reported pain sensation

STUDY DESIGN

This is a retrospective study of 250 patients who describe low back pain with pain drawings. A computer application using artificial neural networks was designed to analyze pain drawings and evaluate the contribution of pain sensation to drawing classification.

OBJECTIVE

The primary goal of this study was to assess the contribution of patient recorded pain sensation marks in classifying pain drawings into one of five broadly defined categories. The hypothesis was that including pain sensation would improve classification.

SUMMARY OF BACKGROUND DATA

With no perfect diagnostic test for patients with low back pain, many approaches have been proposed and are used. One common diagnostic tool is the pain drawing. Several quantitative methods have been proposed to score the drawings. Some methods use pain sensation in the scoring; however, the contribution of pain sensation has not been defined.

METHODS

A custom computer application classified the pain drawing. Data consisted of 250 pain drawings from patients with low back pain.

RESULTS

Patient recorded pain sensation is not necessary in computer-based scoring of pain drawings.

CONCLUSION

Patient-reported pain sensation does not improve classification when quantitatively scoring pain drawings.

Effects of subcutaneous administration of glutamate on pain, sensitization and vasomotor responses in healthy men and women

The present study aimed to investigate if (1) subcutaneous injection of glutamate induces pain, sensitization and vasomotor responses in humans and (2) if sex differences exist in these responses. Thirty healthy volunteers (men-15 and women-15) were included. Each subject received four subcutaneous injections (0.1ml; glutamate 100, 10, 1mM and isotonic saline 0.9%) into the forehead skin in two sessions separated by one week. Assessments of pain intensity (VAS), quality, distribution; area of pinprick hyperalgesia; pressure pain threshold (PPT) at the injection site; surface skin temperature and local blood flow were performed at predetermined time points. The highest concentration of glutamate evoked the highest pain intensity, the longest duration of pain and the largest pain area under the VAS-time curve (P<0.001) in both men and women, although responses in women were larger than in men (P<0.05). The face-chart pain area was the largest for the highest concentration of glutamate (P<0.001) and women drew a larger pain area than men (P=0.024). The area of pinprick hyperalgesia was the largest for glutamate 100mM (P<0.001) and women indicated a larger area than men (P<0.001). Concentration-dependent local vasomotor responses were found following the subcutaneous injection of glutamate but there was no sex difference in this effect. Glutamate 100mM significantly reduced the PPT values (P<0.001) without sex-related differences. The present study demonstrates for the first time that subcutaneous injection of glutamate evokes pain, vasomotor responses and pinprick hyperalgesia in human volunteers and that there are sex-related differences in some of these responses.

Diagnosis of lumbar spinal stenosis: a systematic review of the accuracy of diagnostic tests

STUDY DESIGN

Systematic review of diagnostic studies.

OBJECTIVE

To investigate the diagnostic performance of tests used to detect lumbar spinal stenosis.

SUMMARY OF BACKGROUND DATA

Little is known about the diagnostic accuracy of tests in detecting lumbar spinal stenosis. A systematic review will provide more insight in this topic.

METHODS

We performed a literature search in Medline (PubMed) and Embase for original diagnostic studies on lumbar spinal stenosis, in which one or more different tests were evaluated with a reference standard, and diagnostic values were reported or could be calculated. Two reviewers independently checked all abstracts and full text articles for inclusion criteria. Included articles were assessed for their quality using the Quadas tool. Study features and diagnostic values were extracted from the included studies.

RESULTS

Of the 24 articles included in this review, 15 concerned imaging tests, 7 evaluated "clinical tests," and 2 reported on other diagnostic tests. The overall quality was poor; only 5 studies scored positive on more than 50% of the quality items. Estimates of the diagnostic value of the tests differed considerably. The imaging studies showed no superior accuracy for myelography compared with CT or MRI. Overall, there is considerable variation in the clinical tests; some studies show high sensitivity, whereas others show high specificity.

CONCLUSIONS

Because of heterogeneity and overall poor quality, no firm conclusions about the diagnostic performance of the different tests can be drawn. Better-designed studies exploring the accuracy of diagnostic tests are needed to improve the diagnostic policy.

Body Chart Pain Location and Side-Specific Physical Impairment in Subclinical Neck Pain

OBJECTIVE

To test computer-using students to examine the relationship between location of neck pain as indicated on pain drawings and physical impairments compared with those subjects not reporting pain.

METHODS

This cross-sectional study enrolled 81 healthy student volunteers at the College of Rehabilitation Science, Daegu University, Korea, aged 18 to 30 years. Outcomes were endurance time of neck muscles and neck range of motion (ROM) sensitization or stretch effects on repeated range tests. Active neck ROM measures were taken twice, 10 minutes apart. Neck muscle endurance time was obtained using a horizontal head-holding test with a 10-minute goal. After all physical measurements were completed, information about any neck pain was collected and 4 groups were formed on the basis of the pain location noted on the body chart.

RESULTS

Sixty-seven subjects experienced recurrent neck pain. Nineteen had right-side pain, another 19 had left-side pain, 29 reported pain on both sides, and 14 did not experience neck pain. Neck muscle endurance time was significantly lower for all pain groups. For extension, left and right rotation movements at the second test, ROM decreased for subjects reporting subclinical pain and increased for those with no pain. Location of the pain to one side was related to the ROM decreased, in that the amount of reduction in the second-test rotation range was significantly greater on the side opposite to the pain.

CONCLUSIONS

The location of neck pain that occurs intermittently, but is not present during range testing, affects the second test when the rotation involves stretching of tissue on the side of pain.

Usefulness of pain drawings in identifying real or imagined pain: Accuracy of pain professionals, nonprofessionals, and a decision model

The aim of this study was to determine the accuracy of pain specialists, nonprofessionals, and a decision model in judging whether a pain diagram was marked by a patient with chronic pain or by a healthy volunteer. Two hundred twenty-eight pain drawings were shown in random order to 10 pain medicine physicians, 10 pain medicine fellows, 10 nonphysician specialists, and 10 nonprofessionals. One half of the drawings (n = 114) had been produced by patients treated at a pain center and the other half (n = 114) by healthy individuals who were instructed to mark the diagrams as if they had a pain problem. The nonprofessionals were found to be 51.5% accurate, pain medicine fellows 52.7%, nonphysician specialists 54.3%, and pain medicine physicians 55.2 % accurate at distinguishing drawings by actual pain patients from drawings from volunteers without pain. A decision model was able to achieve 68.9% accuracy in determining which drawings were made by pain patients and which drawings were made by healthy individuals. The results suggest that subjective assessment of pain drawings alone is not useful in determining whether someone has real or imagined pain. A decision model that makes decisions on the basis of the number of highlighted squares on the pain diagram can identify real pain drawings with greater accuracy than humans.

PERSPECTIVE

Pain drawings are clinically useful but have limitations. This study illustrates some of the benefits of computerized pain assessment and highlights the importance of not judging patients on the basis of one source of information.

Perceived pain and self-estimated activity limitations in women with back pain post-partum

BACKGROUND AND PURPOSE

In the general population many daily activities have an impact on low back pain. The aim of the present study was to describe pain intensity, localization, type of sensation and perceived activity limitation in women with different back pain patterns post-partum.

METHOD

In this cross-sectional survey 119 women with back pain persisting for two months after having given birth were interviewed and examined on average 7.2 months (range 6-10 months) post-partum. Based on pain provocation tests, four different back pain pattern groups were identified. Pain could be provoked in the area of the posterior pelvic/sacroiliac joints, in the lumbar spine, both in the posterior pelvic/sacroiliac joints and in the lumbar spine, and in none of the above areas. All women rated pain intensity on a visual analogue scale (VAS, 0-100 mm), and the pain localization and type of sensation were indicated on a pain drawing. They scored their activity limitations by use of the Disability Rating Index (DRI), which covers 12 daily activity items (VAS, 0-100 mm).

RESULTS

There was no significant difference (p = 0.12) in pain intensity (range of medians 19.5-10 mm) between the four groups. However, on average, most areas in the lower back (median 5 mm (range 2-14 mm)), were marked in the group with pain in both the posterior pelvic/sacroiliac joints and in the lumbar spine. The women in the three groups where pain was provoked in the lower area of the back had significantly (p < 0.01) more difficulties with movement-related daily activities than the group where no pain could be provoked.

CONCLUSIONS

The findings of this descriptive study suggest that back pain post-partum provoked by clinical tests considerably hampers movement-related activities. It seems important to pay special attention to the women where pain could be provoked in the lower back areas. The women should be identified early in the post-partum period to initiate adequate treatment.

Automated scoring of patient pain drawings using artificial neural networks: efforts toward a low back pain triage application

The goal of this research was to examine methods of automatically scoring patient pain drawings. Two hundred and fifty pain drawings were selected from the files of an orthopaedic surgeon who specializes in the treatment of low back pain patients. An artificial neural network was designed to score these drawings. The drawings were segmented into 85 regions following dermatomal mappings and from these regions the percent area in pain in each was computed and used as the neural network input variables. With five outcome categories (scores) we obtained a classification sensitivity of 49%, which is approximately as well as physician experts and discriminant analysis achieved using a subset of the same data. We conclude that an artificial neural network is well suited to automatically score patient pain drawings.

Repeatability of pain drawings in a low back pain population

STUDY DESIGN

This is a prospective test-retest repeatability study.

OBJECTIVE

The primary purpose of this study was to determine the test-retest repeatability, using a variety of scoring methods, of pain drawings in a group of chronic low back pain patients. The intraevaluator repeatability for the methods requiring subjective interpretation was also evaluated.

SUMMARY OF BACKGROUND DATA

Pain drawings have been used in a variety of applications, including documentation of symptom location, as a tool for diagnosis and as a psychological screening tool. Accordingly, there have been several methods described for interpreting the drawings and several groups have investigated evaluators' abilities to replicate their interpretation. However, there has been less investigation of patients' consistency in completing the drawings.

METHODS

The intraevaluator repeatability was determined for the two scoring methods requiring subjective interpretation by the evaluator rescoring the drawings 2 weeks after the initial scoring. To determine repeatability, drawings were completed on two occasions by 75 patients. Patients who indicated on a questionnaire that there had been no change in their pain location since their first clinic visit (the day the first drawing was completed) were included in the study. This subgroup consisted of 45 patients who completed pain drawings on two occasions separated by a mean of 244.2 days (range, 26-1197 days). Repeatability was assessed for several scoring methods described in the literature: penalty point system, overall visual inspection, body regions, and a grid method.

RESULTS

Patients were consistent in completing the drawings, assessed by the various scoring methods. The worst repeatability values were for some of the sensation types.

DISCUSSION

As reported by others, the intraevaluator repeatability was high. The results of this study, finding acceptable repeatability for most scoring methods for pain drawing completed on occasions separated by a relatively long period of time, support that the pain drawing is a stable instrument for use in chronic back pain patients.

Associations between patient report of symptoms and anatomic impairment visible on lumbar magnetic resonance imaging

STUDY DESIGN

A cross-sectional study comparing the relationship of symptoms with anatomic impairment visible on lumbar magnetic resonance imaging in 408 symptomatic subjects.

OBJECTIVE

To determine how various anatomic impairments, including the magnitude and location of nerve compression visible on lumbar magnetic resonance imaging, are associated with patient reports of pain, weakness, and dysesthesia.

SUMMARY AND BACKGROUND DATA

Anatomic impairments of the intervertebral disc, radicular canal, and associated soft tissues are prevalent in people with and those without low back pain or lower extremity radiculopathy. This has led to confusion in differentiating between symptom generators and benign variation visible on lumbar magnetic resonance imaging. Recent literature has suggested that the presence of nerve compression is an important finding in the prediction of symptoms. However, the threshold for meaningful nerve compression has not been described.

METHODS

In this study, 408 participants undergoing a diagnostic workup for low back pain, radiculopathy, and/ or completed a survey and pain drawing. Participants underwent standardized lumbar magnetic resonance imaging using a 1.5-T scanner. Two classification systems describing the spatial distribution of symptoms were developed. An additional system to quantify the magnitude of nerve and thecal sac compression was created. All systems were assessed for reliability, after which comparisons among variables were performed using Chi2 as well as simple and multiple logistic regression analysis.

RESULTS

The reliability coefficients for categorizing patients on the basis of pain drawing ranged from 0. 75 to 0.88. The S1-S2 segmental distribution was the most commonly reported location of symptoms, followed by L4-L5. The most common magnetic resonance imaging diagnosis was "unremarkable," followed by "disc impairment without nerve compression." Disc extrusion was present in 10.8% of participants. The reliability of classifying nerve compression visible on magnetic resonance imaging ranged from 0.27 to 1. Nerve compression was present in 37% of participants, and 18% had severe nerve compression. There were no significant associations between segmental distribution of symptoms and the presence of anatomic impairment. However, according to a collapsed classification scale, severe nerve compression and disc extrusion were predictive of pain distal to the knee (odds ratios, 2.72 and 3. 34). The self-report of weakness was associated mildly with severe nerve compression and disc extrusion, but not with other findings. Magnetic resonance imaging findings did not predict self-reports of dysesthesia.

CONCLUSIONS

The presence of disc extrusion and/or ipsilateral, severe nerve compression at one or multiple sites is strongly associated with distal leg pain. Mild to moderate nerve compression, disc degeneration or bulging, and central spinal stenosis are not significantly associated with specific pain patterns. Although segmental distributions of pain can be determined reliably from pain drawings, this finding alone is of little use in predicting lumbar impairment. The self-report of lower extremity weakness or dysesthesia is not significantly related to any specific lumbar impairments. [Key words: back pain, diagnosis, magnetic resonance imaging, nerve compression, pain drawing, pathology]

Relation between pain location and disc pathology: a study of pain drawings and CT/discography

OBJECTIVE

The purpose of this study was to determine whether pain location indicated in pain drawings was related to the specific lumbar disc level(s) that was abnormal in appearance and painful upon discographic injection.

DESIGN

Data were collected prospectively.

SETTING

This study was conducted in a spine specialty clinic.

PATIENTS

The study group consisted of 187 patients (118 men, 69 women; mean age = 37.2 years, range = 18-62 years) with low back pain with or without leg pain. All patients were undergoing computed tomography (CT)/discography at the three lowest lumbar levels for diagnostic purposes.

INTERVENTIONS

Pain drawings were completed the day of but prior to undergoing discography. Discographic pain responses were recorded with respect to the similarity to the patient's clinical symptoms. Pain drawings were classified based on the presence or absence of pain in five areas: low back and/or buttocks, posterior thigh, posterior leg, anterior thigh, and anterior leg. The drawings were scored with the system described by Ransford et al. (1976, Spine 1: 127-34), and those likely to be indicative of psychological problems were analyzed separately (n = 43).

OUTCOME MEASURES

Results were determined by analyzing the relation between the location of pain in the drawings and the specific lumbar disc level(s) found to be painful and disrupted by discography.

RESULTS

There was a significant relation between pain location indicated in the drawing and the lumbar disc level(s) identified as clinically painful and disrupted by CT/discography (p < 0.05, chi-square). Pain limited to the low back and buttocks was frequently associated with the absence of disc pathology (58.3%). When pain in the posterior thigh or leg was present but there was no pain in the anterior drawing, patients frequently had a positive L5-S 1 disc (> or =75%). In patients with anterior thigh pain, with or without posterior thigh or leg pain, the L4-5 disc was frequently symptomatic (>63%). The pattern of no posterior thigh or leg pain but with pain radiating into the leg anteriorly was most commonly associated with the L3-4 disc (71.4%).

CONCLUSIONS

The results of this study indicate that pain drawings may be helpful in identifying which specific discs are associated with pain complaints. As with any evaluation, the drawings should be considered in combination with findings from other assessments.

Pain drawings in the assessment of nerve root compression: a comparative study with lumbar spine magnetic resonance imaging

STUDY DESIGN

Prospective comparative study of pain drawings with findings on lumbar spine magnetic resonance imaging.

OBJECTIVES

To assess the ability of the pain drawing to predict the presence of nerve root compression.

SUMMARY OF BACKGROUND DATA

Most research work has concentrated on the ability of the pain drawing to act as a screening method for psychological distress with less work directed at the influence the anatomic abnormality has on the pain drawing.

METHODS

One hundred thirty-four consecutive outpatients attending for lumbar magnetic resonance imaging in the investigation of back and leg pain completed pain drawings and psychological testing immediately before the examination. The pain drawing was analyzed by previously reported criteria, and the magnetic resonance imaging was assessed independently for the presence of nerve compression by three radiologists. Multivariate stepwise discriminant analysis was used to identify patients with nerve compression on the basis of their pain drawing.

RESULTS

Nerve compression was predicted by numbness in the anterolateral aspect of the foot. There was considerable overlap in the appearances of the pain drawings between patients with and without nerve compression, and the pain drawing correctly classified only 58% of patients with nerve compression.

CONCLUSIONS

The pain drawing is not a good predictor of nerve compression on magnetic resonance imaging in a group of patients investigated for back and leg pain. It should be interpreted with caution and in light of the full clinical picture.

Pain maps from facial pain patients indicate a broad pain geography

Two hundred consecutive female patients, who were referred to a university-based facial pain clinic, were asked to mark all painful sites on sketches showing the contours of a human body in the frontal and rear views. The drawings were analyzed with transparent templates containing 1875 (frontal view) and 1929 (rear view) square cells of equal size. The average patient scored 71.8 cells in the frontal and 99.7 cells in the rear view (corresponding to 3.8% and 5.2% of the maximum possible scores). In individual patient drawings, however, up to 42.7% and 44.9% of all cells were marked. Only 37 cases (18.5%) exhibited pain that was limited to the trigeminal system. An analysis of the pain distribution according to the arrangements of dermatomes revealed three distinct clusters of patients: (1) pain restricted to the region innervated by the trigeminal nerves (n = 37); (2) pain in the trigeminal dermatomes and any combination involving the spinal dermatomes C2, C3, and C4, but no other dermatomes (n = 32); and (3) pain sites involving dermatomes in addition to the ones listed above (n = 131). Mean ages in the three clusters were 38.7, 35.5, and 37.5 years, respectively (p = 0.62, n.s.). Widespread pain existed for longer durations (median, 48 months) than conditions involving local and regional pain (median, 24 months) (p = 0.02, s.). Our findings showed that among a great percentage of persistent facial pain patients the pain distribution is more widespread than commonly assumed, and that the persistence of pain in the regional and widespread pain presentations is significantly greater than in cases with pain limited to the trigeminal system.

Test-retest reliability of patient reports of low back pain

Low back pain is, in large part, a subjective illness. Clinicians must use patient descriptions of the severity and location of low back pain and how it responds to various activities and positions to make diagnostic and treatment decisions. Therefore, it is important to understand how reliably patients describe these aspects of low back pain. The purpose of this study was to determine the test-retest reliability of a visual analogue scale measure of pain intensity, a pain drawing measure of pain location, and the pain response to activity and position questionnaire. Fifty-three subjects (28 men and 25 women) with a mean age of 54.2 years were recruited from an outpatient orthopaedic clinic. They completed the visual analogue scale, pain drawing, and pain response to activity and position questionnaire before and again immediately after seeing their physician. Thirty-three subjects also completed the visual analogue scale and pain drawing measure that evening and the next morning. Test-retest reliability of the visual analogue scale and pain drawing measure was examined using an intraclass correlation coefficient. Reliability of each item on the pain response to activity and position questionnaire was examined by calculating an unweighted Cohen's kappa. Overall, the three pain measures demonstrated fair to good test-retest reliability: 1) visual analogue scale = .66-.93, 2) pain drawing = .58-.94, and 3) pain response to activity and position questionnaire = .46-.89. The results of this study suggest that, although there is some variability in how consistently patients report various aspects of low back pain, the reliability of these pain measures is sufficient to permit their use in making clinical decisions and measuring treatment outcomes.

Degree of disc disruption and lower extremity pain

STUDY DESIGN

Data were collected prospectively from patient-completed pain drawings, lumbar discographic pain responses, and computed tomographic-discographic images.

OBJECTIVES

To determine if there were differences in pain location or the type of pain associated with the severity of symptomatic disc disruption.

SUMMARY OF BACKGROUND DATA

Lower extremity pain related to spinal pathology was for a long time attributed primarily to nerve root compression. However, this simple model could not explain all lower extremity pain. Other mechanisms such as biochemical agents have been implicated. Also, nerve endings have been found in the outer layers of the anulus. Such endings could be associated with pain referred from the disc into the lower extremities. Pain drawings have been used in several studies to investigate various back pain origins and provide an easily administered method to document pain location.

METHODS

Pain drawings were completed by 187 patients undergoing discography at the three lowest levels. The study group consisted of 118 men and 69 women with an average age of 37.2 years (range, 18-62 years). Computed tomographic discograms were scored using the Dallas discogram description, which assigns separate scores for discs with disruption of outer anular fibers (Grade 2) and those with disruption of the outermost anular layers associated with deformation or herniation of the outer anular well (Grade 3). The pain response provoked with each disc injection was recorded as pressure only or painless, pain dissimilar to clinical symptoms, similar to symptoms, or the exact reproduction of clinical pain, in this study, the similar and exact reproduction responses were combined and considered to be "symptomatic." The drawings were classified based on the presence or absence of pain in three regions: low back or buttocks, thigh, and leg. The drawings were also scored using the system described by Ransford, and those that were likely to be indicative of psychological problems were analyzed separately (N = 43).

RESULTS

There was no significant difference in the distal location of lower extremity pain among patients whose most severe symptomatic disc disruption was a Grade 2 compared with those with symptomatic Grade 3 disruption (62.2% vs. 61.7%; P > 0.75; chi-square). The figure was similar for patients with both symptomatic Grade 2 and 3 disruption (72.7%). However, patients with symptomatic Grade 2 disruption used significantly more symbols to describe their pain, and in particular aching pain, than did those with symptomatic Grade 3 disruption.

CONCLUSIONS

These results indicate that disc disruption passing into the outer layers of the anulus, but not resulting in deformation of the outer anular wall, was as frequently associated with lower extremity pain as were discs with more severe disruption deforming the outer anular wall; however, they were associated with a greater degree of aching pain. These findings support that lower extremity pain may be referred from the disc.