Classifying patients with lumbar spinal stenosis using painDETECT: a cross-sectional study

Background The pathological mechanisms of lumbar spinal stenosis are unclear. Family doctors in the primary care setting may perform medical examinations of patients with lumbar spinal stenosis. Our aim was to use the painDETECT questionnaire to quantify the pathological mechanisms of low back pain and/or leg pain caused by lumbar spinal stenosis. Methods We enrolled 102 patients (37 men, 65 women) who had been newly diagnosed with lumbar spinal stenosis at 2 facilities. The patients’ conditions were evaluated using the painDETECT questionnaire, Numerical Rating Scale, Roland-Morris Disability Questionnaire, Japanese Orthopaedic Association Back Pain Evaluation Questionnaire, and 36-Item Short-Form Health Survey. The pathological mechanisms of low back pain and/or leg pain caused by lumbar spinal stenosis were classified based on results of the painDETECT questionnaire as nociceptive pain, neuropathic pain, or unclear type of pain (mixed pain). Statistical analyses were performed using the Kruskal–Wallis test. A value of p < 0.05 was considered to indicate statistical significance. Results The mean age of all patients in this study was 70.3 ± 2 years. The male:female distribution was 37:65 (36.3:63.7 %). In all, 72 (70.6 %) patients had chronic pain (duration of ≥3 months), and 30 (29.4 %) had subacute or acute pain (duration of <3 months). The pain was classified as nociceptive in 59 patients (57.9 %), neuropathic in 18 (17.6 %), and unclear in 25 (24.5 %). The neuropathic pain group had a significantly lower quality of life (p < 0.05) than did the other groups. Conclusions Patients with neuropathic back and/or leg pain caused by lumbar spinal stenosis may have lower physical and/or psychological quality of life than patients with such pain caused by other mechanisms.


Background
Lumbar spinal stenosis (LSS) is the one of the most commonly encountered orthopedic disorders. Family doctors in the primary care setting may perform medical examinations of patients with LSS. Therefore, an understanding of the pathological mechanisms of LSS may be very helpful for these doctors.
LSS is defined as a reduction in the area of the spinal canal, lateral canal, and/or foramina. Symptoms of LSS may occur as a result of neurovascular mechanisms [1][2][3], such as reduced arterial flow in the cauda equina, venous congestion, increased epidural pressure, nerve root infiltration, and direct compression in the central canal and/or lateral recess [4]. Although the characteristic symptom of LSS is neurogenic intermittent claudication (NIC) [5,6], other symptoms may appear as well, including low back pain (LBP), radicular pain and/or numbness down the leg, and motor weakness in the legs [7][8][9][10]. It is therefore important to analyze the pathological mechanisms of LBP and/or leg pain caused by LSS to alleviate the symptoms.
We considered the pain mechanisms according to a classification that included neuropathic pain, nociceptive pain, and mixed pain [11,12]. Neuropathic pain is defined by the International Association for the Study of Pain as "pain initiated or caused by a primary lesion or dysfunction of the nervous system" [13]. It may be associated with abnormal sensations called dysesthesia or produced by normally nonpainful stimuli such as allodynia. Nociceptive pain may also be a result of the encoding and processing of harmful stimuli in the nervous system, reflecting the ability of the body to sense potential harm [14]. Because the developmental mechanisms responsible for neuropathic and nociceptive pain differ, treatment plans differ as well. Therefore, the pathological mechanisms of the pain should be precisely identified to arrive at an exact diagnosis of LSS-derived LBP and/or leg pain.
A recent study using the painDETECT questionnaire demonstrated that neuropathic pain was less common in patients with chronic LBP and that patients with neuropathic back and/or leg pain reported significantly more pain, disability, anxiety, depression, and reduced quality of life (QOL) than patients with nociceptive pain [15]. The pathological mechanisms of LSS, however, were unclear. Few studies have used the painDETECT questionnaire to quantify the pathological mechanisms of LSS [16,17]. The purpose of this study was to quantify the pathological mechanisms of LBP and/or leg pain caused by LSS using the painDETECT questionnaire, and to report quality of life data in patients with LSS. This study may help to fully elucidate the epidemiology of LSSderived LBP and/or leg pain in the primary care setting.

Methods
The ethics committees of the participating research institutions approved this study. Written informed consent was obtained from all patients prior to their participation.
This cross-sectional multicenter observational study was conducted nationwide at two hospitals with their attending spinal surgeons. The recruitment period was 1.5 years (1 April 2013 to 30 September 2014).

Patients
The study group included 102 consecutive patients (37 men, 65 women) who were newly diagnosed with LSS at two facilities: Aizu Medical Center Fukushima Medical University and the Fukushima Prefectural Minami-Aizu Hospital. Three spinal surgeons diagnosed the patients with LSS based on magnetic resonance imaging (MRI) findings of LSS, subjective symptoms, and neurological findings. All of the patients had subjective symptoms that included leg symptoms, LSS-related neurological findings, and MRI evidence of LSS. An independent radiologist assessed the MRI scans obtained for each patient at the time of diagnosis for evidence of LSS, including central, lateral recess, and foraminal stenosis. The MRI diagnosis of LSS was based on the following findings: (1) blockage of cerebrospinal fluid and/or the disappearance of fatty tissue surrounding the foramen on T1-weighted images in the sagittal view; (2) central stenosis, lateral recess stenosis, and/or foraminal stenosis caused by lumbar disc compression from the front, and/or increased size of the ligamentum flavum, and/or articular factors from the back in the axial view; and/or (3) compression of the nerve root surrounding the foramen in the coronal view. The ankle-brachial pressure index (ABI) was also checked in all patients to distinguish NIC from vascular intermittent claudication (ABI <0.9). The ABI is the systolic pressure at the ankle divided by the systolic pressure at the arm. It has been shown to be a specific and sensitive metric for the diagnosis of peripheral arterial disease [18].
The inclusion criteria for all patients were (1) a diagnosis of lumbar spondylosis, degenerative spondylolisthesis, or degenerative scoliosis with LSS; (2) pain and/or numbness in the lumbar dermatomal distribution; (3) motor or sensory neurological signs (hypoesthesia, hyperesthesia, allodynia, or dysesthesia) in the affected dermatomes; (4) cognitive capability to satisfy our inquiries; and (5) neurogenic NIC due to LSS. The exclusion criteria for all patients were (1) diagnosis of lumbar degenerative disease without LSS; (2) predominantly axial spinal pain; (3) rheumatoid arthritis; (4) known peripheral neuropathy; (5) history of workmen's compensation or disability issues; (6) chronic depression and use of antidepressant medication; and (7) patients with an ABI of <0.9.

Methods
The painDETECT neuropathic pain screening questionnaire scores [17] were used to identify the three pain subgroups of patients with LBP and/or leg pain caused by LSS: (1) those with nociceptive pain, (2) those with neuropathic pain, and (3) those in whom the type of pain was unclear [15,19]. We used the Japanese version of the painDETECT neuropathic pain screening questionnaire score [20]. The cut-off values for categorizing the type of pain based on the painDETECT neuropathic pain screening questionnaire scores were as follows: nociceptive pain, 0-12; unclear type of pain (mixed nociceptive and neuropathic pain), 13-18; and neuropathic pain, 19-38. The painDETECT neuropathic pain screening questionnaire is a reliable screening tool with high sensitivity, high specificity, and positive predictive accuracy. These parameters were each 84 % in a palmtop computerized version of the questionnaire and 85, 80, and 83 %, respectively, in a corresponding penciland-paper questionnaire [17]. Matsubayashi et al. [20] demonstrated that the Japanese version of the pain-DETECT neuropathic pain screening questionnaire has good validity and reliability, similar to that found with the original version.
Numerical rating scale (NRS) and Roland-Morris Disability Questionnaire (RDQ) scores [21] were used to compare the severity of subjective symptoms and the QOL associated with LBP at the patient's first medical examination. The patients used the NRS for selfevaluation of their leg pain and/or numbness. Both the NRS and RDQ scores were evaluated at the time of the first medical examination. We used the Japanese version of the RDQ score [22].
The Japanese Orthopaedic Association Back Pain Evaluation Questionnaire (JOABPEQ) [23,24] and the 36-Item Short-Form Health Survey (SF-36) [25] were used to compare the severity of subjective symptoms and the QOL associated with LBP and health at the patient's first medical examination. The JOABPEQ has five subscales, and the SF-36 has eight. For both questionnaires, higher scores indicate better QOL. We used the Japanese version of the SF-36 [26,27]. The SF-36 uses scores of 0 to 100, but the scores are not based on the norm.
The primary outcome was the type of pain-nociceptive, neuropathic, or unclear (mixed)-in patients with LBP and/or leg pain caused by LSS. At the first medical examination, we compared the patients' demographic and clinical characteristics; NRS (LBP); NRS (leg pain, leg numbness); and RDQ, JOABPEQ, and SF-36 scores among the three pain groups using the Kruskal-Wallis test. We also compared these outcomes according to the duration of symptoms-i.e., <3 months (acute and/or subacute phase) or ≥3 months (chronic phase) after the onset of LBP and/or leg pain caused by LSS. A p value of <0.05 was considered to indicate statistical significance. The statistical analyses were performed using StatView 5.0 statistical software (SAS Institute, Cary, NC, USA). The statistical power analysis of this study was performed using G*Power 3.1 (Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany) [28]. The power analysis utilized an effect size of 0.4, alpha level of 0.05, power (1 − beta error probability) of 0.95, and requirement of three groups. Finally, the total required sample size was calculated as 102.

Ethics
This study was approved by the ethics committees of the two participating research institutions: Aizu Medical Center Fukushima Medical University and Fukushima Prefectural Minami-Aizu Hospital. Informed consent was obtained from all patients.
The demographic and clinical characteristics of the patients with LBP and/or leg pain caused by LSS for each pain subgroup are shown in Table 1. The demographic and clinical characteristics of the patients with LBP and/ or leg pain caused by LSS were similar among the three pain subgroups.
Among all patients with LBP and/or leg pain caused by LSS, 59 (57.9 %) had nociceptive pain, 18 (17.6 %) had neuropathic pain, and 25 (24.5 %) had an unclear type of pain (mixed pain) at their first medical examination (Table 2). In all, 72 (70.6 %) patients had chronic pain (duration of ≥3 months), and 30 (29.4 %) had subacute or acute pain (duration of <3 months). Among the patients whose LBP and/or leg pain caused by LSS had been present for ≥3 months, 41 (56.9 %) had nociceptive pain, 14 (19.5 %) had neuropathic pain, and 17 (23.6 %) had pain of unclear (mixed pain) at their first medical examination ( Table 2). Among those whose LBP and/or leg pain caused by LSS had been present for <3 months, 18 (60.0 %) patients had nociceptive pain, 4 (13.3 %) had neuropathic pain, and 8 (26.7 %) had unclear pain (mixed pain) at their first medical examination ( Table 2). The populations of the three pain subgroups [nociceptive, neuropathic, and unclear (mixed)] with LBP and/or leg symptoms caused by LSS (regardless of pain duration) were thus similar.
NRS scores for LBP, leg pain, and leg numbness in patients with LBP and/or leg pain caused by LSS were not significantly different among the three pain groups (Table 3). However, the RDQ score in patients with LBP and/or leg pain caused by LSS was significantly lower in the neuropathic pain group than in the other groups (p < 0.05) ( Table 3). In contrast, there were no significant differences among the three pain groups regarding (1) the NRS scores for LBP, leg pain, or leg numbness or (2) the RDQ scores in patients with LBP and/or leg pain caused by LSS that had been present for ≥3 months or <3 months ( Table 3).
The five JOABPEQ subscales were LBP, lumbar function, walking ability, social life function, and mental health. The eight SF-36 subscales were physical functioning, physical role functioning, bodily pain, general health perceptions, vitality, emotional functioning, social role functioning, and mental health. The subscale scores of both questionnaires for each pain group are shown in Tables 4 and 5. There were statistically significant differences in three JOABPEQ subscales [lumbar function, social life function, and mental health (p < 0.05)] and in two SF-36 subscales [physical function and bodily pain (p < 0.05)] (Tables 4 and 5). In regard to pain duration, for patients whose pain had been present for ≥3 months after the onset of symptoms, there were statistically significant differences in two JOABPEQ subscales [lumbar function and mental health (p < 0.05)], but no statistically significant differences in any of the SF-36 subscales (Tables 4 and 5). For patients whose pain had been present for <3 months after the onset of symptoms, there were no statistically significant differences in any of the JOABPEQ subscales (p < 0.05), but there was a statistically significant difference in one SF-36 subscale [bodily pain (p < 0.05)] (Tables 4 and 5).

Discussion
The present study demonstrated four major points. (1) Overall, 58 % of the 102 patients with LBP and/or leg pain caused by LSS had nociceptive pain, 18 % had neuropathic pain, and 24 % had an unclear type of pain (mixed pain) at their first medical examination. (2) NRS scores for LBP, leg pain, and leg numbness in patients with LBP and/or leg pain caused by LSS were not significantly different among the three pain groups. The RDQ score in patients with LBP and/or leg pain caused by LSS, however, was significantly lower in the neuropathic pain group than in the other groups. (3) Three JOABPEQ subscales (lumbar function, social life function, and mental health) were significantly lower in the neuropathic pain group than in the other groups. (4) Two SF-36 subscales (physical function and bodily pain) were significantly lower in the neuropathic pain group than in the other groups.
The statistical power analysis performed in this study indicated a required total sample size of 102, and the power was 0.95. Therefore, we believe that the power of this study was adequate.
LSS may occur at different levels in the spinal canal. It may be caused by entrapment of nerve roots in the cauda equina due to hypertrophy of the osseus and soft tissue structures surrounding the lumbar spinal canal. Central canal stenosis may compress nerve roots in the cauda equina, whereas lateral recess stenosis and/or Table 5 SF-36 score for each pain type at the first medical examination   Parameter  Nociceptive pain  Type of pain unclear (mixed pain)  foraminal stenosis may compress nerve roots but spare the spine [29,30]. Although the lower limb symptoms associated with LSS are mainly attributed to mechanoreceptive compression of nerve rootlets and/or the cauda equina, they are also associated with inflammation, ischemia, malnutrition, nerve degeneration, and nerve injury. They consequently have a complicated pathophysiology. The pathological mechanisms of lower limb symptoms caused by LSS involve nociceptive, inflammatory, and/or neuropathic pain components, which may result from postural changes or persistent compression of the nerve roots and/or cauda equina while walking. The prevalence of neuropathic pain in the general population is unclear, although it has been reported at 3-9 % based on the results of various screening questionnaires in Europe and the United States [31][32][33][34]. Few studies, however, have used the painDETECT to analyze the pathological mechanisms of LBP and/or leg pain caused by LSS in primary care-referred patients. Beith et al. [15] studied patients from southeastern England who had LBP with or without leg pain and had been referred for physiotherapy. The authors reported that 59 % of the patients reported what was identified to be nociceptive pain, 16 % had neuropathic pain, and 25 % had an unclear type (mixed pain). This result is very similar to that obtained in the present study of the pathological mechanisms of LSS-derived LBP and/or leg pain (nociceptive pain, 58 %; neuropathic pain, 18 %; unclear pain (mixed pain), 24 %). Therefore, our findings may accurately elucidate the pathological mechanisms of LSSderived LBP and/or leg pain in the primary care setting.
The NRS scores for LBP, leg pain, and leg numbness in our patients with LBP and/or leg pain caused by LSS were not significantly different among the three pain groups. However, the RDQ score, three JOABPEQ subscale scores (lumbar function, social life function, and mental health), and two SF-36 subscale scores (physical function and bodily pain) were significantly lower in the neuropathic pain group than in the other groups. These data suggest that the neuropathic pain component produced significantly lower scores than its nonneuropathic pain components and reduced the patients' physical and/or psychological QOL. Therefore, patients with neuropathic back and/or leg pain caused by LSS should be diagnosed as soon as possible after referral. A previous study [35] compared JOABPEQ scores between patients with LBP who experienced either neuropathic pain or nociceptive pain as assessed by the Japanese version of the painDETECT. Their findings suggest that neuropathic pain affects the social and psychological well-being of patients with LBP and demonstrate that patients with neuropathic back and/or leg pain caused by LSS might have particularly low physical and/or psychological QOL.
Finally it is considered that the impact of this study on education, health services and research regarding primary care would suggest that it may be very important to elucidate the pathological mechanisms and the epidemiology of LSS-derived LBP and/or leg pain by classifying the patients with LSS using painDETECT in order to analyze either neuropathic pain or non-neuropathic pain, and this may be helpful to examine how to treatment for LSS-derived LBP and/or leg pain in primary care setting.
The present study has some limitations that require attention. First, we studied only a small population, although we believe that the power was adequate in this study. Future studies must plan to evaluate a larger population. Second, this study had a cross-sectional design-it was not a longitudinal study. We did not evaluate therapeutic efficacy for LSS in this study. Hence, future studies should evaluate the therapeutic efficacy for each pain group caused by LSS, including conservative versus surgical therapy. Third, it was considered the lack thorough clinical and laboratory investigation in this study, since our study utilizes only questionnaires to classify the pain introduced by LSS. A clinical study with post-licensure surveillance should be implemented, ideally by setting up a database that includes all patients seeking treatment for LBP and/or leg pain caused by LSS, minimizing losses to follow-up, and using validated methods to gather clinically relevant data including demographic information, clinical features, common co-morbidities, conservative and/or surgical treatments applied to each patient, experience and training standards of the care providers applying each treatment, and each patient's clinical evolution [36].

Conclusions
It may be important to analyze the pathological mechanisms of neuropathic pain and non-neuropathic pain in patients with LBP and/or leg pain caused by LSS. Patients with neuropathic back and/or leg pain caused by LSS may experience particularly low physical and/or psychological QOL.

Acknowledgments
No funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript. The submitted manuscript does not contain information about specific medical devices or drugs.

Availability of data and materials
Results from the dataset are presented in the paper. The raw data for this study is available from the first author upon request.