If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Centre of Excellence for Rehabilitation Medicine, UMC Utrecht Brain Centre, University Medical Centre Utrecht, UtrechtDe Hoogstraat Rehabilitation, Utrecht
Centre of Excellence for Rehabilitation Medicine, UMC Utrecht Brain Centre, University Medical Centre Utrecht, UtrechtDe Hoogstraat Rehabilitation, UtrechtUniversity of Groningen, University Medical Centre Groningen, Centre for Rehabilitation, Groningen
Centre of Excellence for Rehabilitation Medicine, UMC Utrecht Brain Centre, University Medical Centre Utrecht, UtrechtDe Hoogstraat Rehabilitation, UtrechtDepartment of Rehabilitation, Physical Therapy Science, and Sports, UMC Utrecht Brain Center, University Medical Center Utrecht, Utrecht
Centre of Excellence for Rehabilitation Medicine, UMC Utrecht Brain Centre, University Medical Centre Utrecht, UtrechtDe Hoogstraat Rehabilitation, UtrechtDepartment of Spinal Cord Injury and Orthopedics, De Hoogstraat Rehabilitation, Utrecht, the Netherlands
To describe the prevalence and characteristics of spinal cord injury (SCI)-related pain during initial inpatient rehabilitation and to investigate relationships with demographic and lesion characteristics.
Design
Cohort during inpatient rehabilitation.
Setting
Eight specialized SCI rehabilitation centers in the Netherlands.
Participants
Patients with newly acquired SCI admitted for inpatient rehabilitation between November 2013 and August 2019 (N=1432).
Interventions
Not applicable.
Main Outcome Measures
Presence of pain at admission and discharge. Logistic regression analyses were used to study the prevalence of pain related to sex, age, etiology, completeness, and level of injury.
Results
Data from 1432 patients were available. Of these patients 64.6% were male, mean age was 56.8 years, 59.9% had a nontraumatic SCI, 63.9% were classified as American Spinal Cord Injury Association Impairment Scale (AIS) D and 56.5% had paraplegia. Prevalence of pain was 61.2% at admission (40.6% nociceptive pain [NocP], 30.2% neuropathic pain [NeuP], 5.4% other pain) and 51.5% at discharge (26.0% NocP, 31.4% NeuP, 5.7% other pain). Having NocP at admission was associated with traumatic SCI. AIS B had a lower risk of NocP than AIS D at admission. Having NocP at discharge was associated with female sex and traumatic SCI. AIS C had a lower risk of NocP at discharge than AIS D. Having NeuP at admission was associated with female sex. Having NeuP at discharge was associated with female sex, age younger than 65 years vs age older than 75 years and tetraplegia.
Conclusions
SCI-related pain is highly prevalent during inpatient rehabilitation. Prevalence of NocP decreased during inpatient rehabilitation, and prevalence of NeuP stayed the same. Different patient and lesion characteristics were related to the presence of SCI-related pain. Healthcare professionals should be aware of these differences in screening patients on presence and development of pain during inpatient rehabilitation.
Pain is one of the secondary conditions that is frequently reported by patients with SCI. The prevalence of SCI-related pain is estimated at 61%, with a large amount of heterogeneity in studies. Pain is one of the most challenging symptoms to manage.
SCI-related pain is classified by the International Association of the Study of Pain into 2 main types: nociceptive pain (NocP) and neuropathic pain (NeuP).
Examples of musculoskeletal pain include pain resulting from joint arthritis, spinal fractures, muscle injury, rotator cuff tendinopathy, and muscle spasms.
Pharmacologic treatment strategy for NocP is based on the stepped care of paracetamol, nonsystemic anti-inflammatory drugs, and opioids in accordance with the World Health Organization Cancer Pain Relief.
For NeuP, pregabalin and gabapentin are the most evidence based treatment options and have shown a significant reduction in pain intensity in multiple studies but with limited evidence on long-term effects.
Besides pharmacologic interventions, frequently used nonpharmacologic treatments, such as acupuncture, exercise programs, cranial electrotherapy stimulation, and psychological therapies such as cognitive behavioral therapy are used.
It is important to gain a better understanding of SCI-related pain during inpatient rehabilitation to improve future treatment. Presence of SCI-related pain tends to change over time; therefore, studies on pain in community-dwelling patients with SCI might not relate to pain at inpatient rehabilitation.
Pain and its impact on inpatient rehabilitation for acute traumatic spinal cord injury: analysis of observational data collected in the SCIRehab Study.
Finding patient and lesion characteristics associated with SCI-related pain and type of pain can identify which patients are at risk for developing pain and facilitate patient screening during inpatient rehabilitation. Addressing pain at an early stage could potentially benefit treatment in the long-term. The objective of this study is to describe the prevalence of SCI-related pain during inpatient rehabilitation, describe the characteristics of such pain, and investigate possible associations to different patient and lesion characteristics.
Methods
Participants
This study uses data from the Dutch Spinal Cord Injury Database (DSCID).
The DSCID is a joint effort of the 8 SCI specialized Dutch rehabilitation centers. This study used data of the DSCID collected between November 2013 and August 2019. Patients were included in the current study if data on presence of pain (yes or no) were available at both admission and discharge, patients did not object to the sharing of deidentified data, and time between onset of SCI and admission to the rehabilitation center was within 9 months (to exclude readmissions).
Procedure
This study was conducted according the principles of the Declaration of Helsinki. Informed consent varies between centers. Some centers ask for written informed consent in advance, whereas other centers inform patients with oral and written information about the use of anonymous data for research purposes. Patients can decline the use of these data, including the data of the DSCID. Clinical data of patients with SCI were collected at admission and discharge. A designated member of the clinical staff is responsible for data collection at each center. Information is either retrieved from the electronic medical record retrospectively or added prospectively at admission of the patient to the rehabilitation center and entered into the DSCID anonymously. The notation in the electronic medical record or the admission to the rehabilitation center is always supervised by a SCI rehabilitation physician. In this way, entering NocP or NeuP is similar in all rehabilitation centers.
Instruments
The DSCID contains data on age, sex, etiology of SCI, various secondary conditions of SCI, functional status, and quality of life. American Spinal Cord Injury Association Impairment Scale (AIS) classification and neurologic level of injury were assessed according to the International Standards for Neurological Classification of Spinal Cord Injury, revised version of 2011.
Information on the presence of pain; the type of pain divided into NocP (which were subdivided into musculoskeletal and visceral pain), NeuP (which was divided into above level, at level and below level of injury, where at level pain extends to 3 levels below the level of injury), and other pain; intensity of pain on the 11-point numeric rating scale (NRS); treatment of pain; and type of treatment (oral medication, physiotherapy, psychology, other) was available from the DSCID.
Analysis
The statistical analyses were performed using SPSS Version 26.a Frequencies were reported on the sociodemographic characteristics of the included patients and the characteristics of pain. Independent samples t tests were used to compare means for continuous variables, and chi-square tests were used to compare categorical variables on baseline characteristics of patients included and excluded from the DSCID. An independent samples t test was used to compare mean time since onset between patients with pain and without pain. Logistic regression analyses were performed to examine whether a prediction on the prevalence of pain could be made using the variables age, sex, etiology, AIS classification, and level of injury for which unadjusted and adjusted odds ratios were calculated. Little's Missing Completely at Random tests were used to evaluate whether missing data were distributed randomly.
Results
From a total of 2700 patients registered in the DSCID, 1432 were included in the analysis. A flow diagram of inclusion of patients is shown in fig 1. Patient characteristics at admission of the included and excluded patients are shown in table 1. The included group contained less patients with AIS A and more patients with AIS D than the excluded group. Exclusion of patients readmitted to the rehabilitation center (time since onset >9 months) resulted in a significant difference in time since onset between included and excluded patients. Among the included patients, most common causes of traumatic SCI were falls (54.5%), traffic collisions (22.3%), and sports-related accidents (14.3%). Most common causes of nontraumatic SCI were vascular (22.4%), benign or malignant tumor growth (21.8%), degeneration (20.8%), and inflammation (15.9%). Most patients had an incomplete SCI in which AIS D occurred most frequently.
Prevalence of pain is reported in table 2. Pain was reported in 61.1% at admission and in 51.5% at discharge. NocP was reported in 40.4% of the patients at admission and in 26.1% at discharge. NeuP was reported in 30.5% of the patients at admission and in 31.7% at discharge. Change in pain is reported in table 3. Characteristics of NocP and NeuP are shown in tables 4 and 5, respectively. Above-level NeuP was described as other NeuP. This occurred in 3 patients, of which 2 had a combination with at-level and below-level NeuP.
Table 2Prevalence of pain at admission and discharge
% of included patients with nociceptive pain and with data on treatment. Multiple treatments or no treatment can be used in 1 patient. Therefore, total percentages can exceed or be lower than 100%.
Oral medication
475 (91.5)
270 (79.4)
Physiotherapy
77 (14.8)
96 (28.2)
Psychotherapy
1 (0.2)
4 (1.2)
Other
13 (2.5)
16 (4.7)
Abbreviations: IQR, interquartile range.
% of included patients with nociceptive pain and with data on treatment. Multiple treatments or no treatment can be used in 1 patient. Therefore, total percentages can exceed or be lower than 100%.
% of included patients with nociceptive pain and with data on treatment. Multiple treatments or no treatment can be used in 1 patient. Therefore total percentages can exceed or be lower than 100%
Oral medication
287 (71.6)
346 (80.8)
Physiotherapy
17 (4.2)
41 (9.6)
Psychotherapy
1 (0.2)
5 (1.2)
Other
5 (1.2)
14 (3.3)
% of included patients with nociceptive pain and with data on treatment. Multiple treatments or no treatment can be used in 1 patient. Therefore total percentages can exceed or be lower than 100%
Time since onset of SCI was not significantly different in the group with pain and without pain in the total sample (mean, 38.1 days and 38.2 days; P=.96 at admission and mean, 130 and 126 days; P=.39 at discharge, respectively) nor in the subgroups with nociceptive pain (mean, 44.8 and 47.6 days; P=.47 at admission and mean, 126 and 129 days; P=.56 at discharge, respectively) and neuropathic pain (mean, 38.7 and 37.3 days; P=.63 at admission and mean, 132 and 136 days; P=.25 at discharge, respectively).
Patient characteristics on pain
The odds ratios of NocP and NeuP in relationship to patient characteristics is shown in tables 6 and 7, respectively. At admission, patients with AIS B had a lower risk of having NocP than those with AIS D. Patients with traumatic SCI had a higher risk of having NocP. At discharge female patients and patients with traumatic SCI had a higher risk of having NocP, and patients with AIS C had a lower risk of having NocP than those with AIS D. Sex, age, etiology, and level of injury showed significant relations to the prevalence of NeuP. At admission, female patients had a higher risk of having NeuP. At discharge female patients, patients younger than 65 years compared with those older than 75 years, and patients with tetraplegia had a higher risk of having NeuP.
Table 6Prevalence of nociceptive pain in relationship to patient and lesion characteristics
Little's Missing Completely at Random tests showed that missing data were distributed randomly and did not occur more frequently in specific subgroups.
Discussion
In our population of 1432 patients with newly acquired SCI, pain is highly prevalent. The prevalence of NocP decreased during inpatient rehabilitation, and prevalence of NeuP stayed the same. Female patients and patients having traumatic SCI had a higher risk of having NocP. Female patients, younger patients, and patients with tetraplegia had a higher risk of having NeuP.
The prevalence of pain reported in this study is lower than reported in the current literature. One study performed in Switzerland on pain during inpatient rehabilitation reported a pain prevalence of 87% at admission and 74% at discharge.
Pain and its impact on inpatient rehabilitation for acute traumatic spinal cord injury: analysis of observational data collected in the SCIRehab Study.
This study did not make a distinction between different types of pain. Another study performed in Italy on pain at inpatient rehabilitation showed a prevalence of 72% during inpatient stay, with a prevalence of NocP and NeuP of 52% and 48%, respectively.
Most studies report the presence of pain in community-dwelling patients with SCI, which is different from pain prevalence at inpatient rehabilitation. In a recent meta-analysis, with large variation in time since onset, the pooled prevalence of musculoskeletal pain was 56%, and the pooled prevalence of visceral pain was 20%.
The differences in prevalence between the studies can also be explained by the different patient characteristics of the studies. Musculoskeletal pain could develop in long-term wheelchair users (range, 5-56 years) because of overuse of the upper extremity, spasticity, and muscle contractures.
During inpatient rehabilitation there seems to be a decrease in NocP. Table 3 shows that 296 patients ceased to have NocP during inpatient rehabilitation, while 104 developed NocP. This could be explained by pain resulting from the trauma or surgery disappearing over time. It is also possible that intensive physical treatment from a physical therapist during inpatient rehabilitation has a positive effect on managing NocP symptoms as has been described for SCI-related shoulder pain in those using wheelchairs for over 1 year.
Increased use of physical therapy as a treatment for NocP during inpatient rehabilitation is also described in table 4, where 28.2% of patients with NocP use physical therapy as a treatment compared with 14.8% at admission. Inactivity during hospitalization could also explain the higher pain levels of NocP at admission and decrease at discharge because training with a physical therapist increases during inpatient rehabilitation. The fact that inpatients start using wheelchairs, however, could explain the portion of patients that develop NocP during inpatient rehabilitation because rotator cuff pathology occurs more frequently in manual wheelchair users with SCI.
The prevalence of NeuP reported in this study is lower than other studies report. A wide range of percentages on the prevalence of NeuP exist in current literature, varying between 32% and 92%.
This study collected information during inpatient rehabilitation and therefore did not examine the prevalence of NeuP over a longer period of time. As shown in table 3, the number of patients that develop NeuP during inpatient rehabilitation is similar to the number of patients that cease to have NeuP (176 vs 164, respectively). NeuP mostly develops in a few years’ time, and prevalence increases after 1 year.
In the meta-analysis of Burke et al, pooled prevalence of at-level NeuP in patients with SCI was 19%, whereas pooled prevalence of below-level NeuP was 27%.
In this study, the prevalence of NeuP can only be generalized to patients with SCI during inpatient rehabilitation and not the entire population with SCI. Long-term follow-up at inpatient rehabilitation and after discharge would be useful to study the change in pain over time.
Demographic characteristics in this study were comparable with other studies on SCI. Mean age at admission was 57 years. The mean age in other studies on SCI range from 40-62 years.
There are differences on demographics in different parts of the world, for example, traumatic SCI is more prevalent in Northern America than Western Europe.
Pain and its impact on inpatient rehabilitation for acute traumatic spinal cord injury: analysis of observational data collected in the SCIRehab Study.
It is lower than in other studies reporting on pain in the 5-8 years post injury, where pain ranged from 5.7-5.8 on the NRS in musculoskeletal pain, 6.6 in visceral pain, and 5.7-6.4 in NeuP.
It is likely that assessment in other studies focus solely on pain, which could have led to a higher score on the NRS because of selection bias. Patients included in the current study were assessed on multiple questions regarding their SCI. This could have led to the large amount of missing data on NRS because the priority might not have been on pain during their inpatient rehabilitation.
A meta-analysis performed on determinants of pain in SCI did not find a relationship between pain and patient or lesion characteristics.
In this study women and patients having a traumatic SCI had a higher risk of having NocP, and female patients, patients younger than 65 years, and patients with tetraplegia had a higher risk of having NeuP. Other studies also found sex influenced SCI-related pain, where pain is related to female sex.
A study in the Swiss population found that chronic pain, with no distinction between types of pain, was more prevalent in the oldest age group (61 years or older) than in the youngest age group (16-30 years).
These studies did not differentiate between types of pain. The relationship between age and specifically having NeuP has not been described in previous research.
This study found a lower risk for NocP for patients with AIS B and AIS C. There was no relationship between the presence of NeuP and injury severity. Until now, only few studies have investigated the influence of injury severity on pain, mostly on the presence of NeuP. Different studies show conflicting results on the relationship between injury severity and pain. Norrbrink Budh et al reported that patients with AIS D had a significantly higher prevalence of SCI-related pain than patients with other AIS classifications.
A better understanding of the relationship between pain and injury severity can possibly be achieved if the pathophysiology of pain is further explained or a more in depth assessment of pain characteristics is performed.
Study limitations
Several limitations of this study deserve to be mentioned. First, this study only reports on pain during initial inpatient rehabilitation. Therefore the prevalence of pain can only be generalized to patients during inpatient rehabilitation, with a short time since onset of SCI. Second, this study describes presence on pain, type of pain, type of treatment, and NRS but does not report on further information such as type of medication or burden caused by pain. Third, the DSCID has a high percentage of missing data. Because data collection is either done prospectively at admission or by extracting information from the electronic medical record, the amount of missing data could have been caused by a lack of information in the electronical medical record. Also, in most cases there is a delay in entering data in the DSCID, making it difficult to retrieve missing data from the patient. Future funding could possibly create time in which data are collected routinely prospectively in the different centers. This could enhance data completeness and accuracy as has been described before.
Last, there is a significant difference in AIS classifications in the included and excluded group. Patients with complete SCI were less likely to have data on presence of pain at both admission and discharge. This could be explained by the fact that patients with complete SCI are admitted longer to the rehabilitation center and the information at discharge is not yet collected. In addition, there are more patients with AIS D in the analysis, which could have led to an overestimation of the prevalence of NocP because participants with AIS D were more likely to have NocP than those with AIS B and C.
Conclusions
This study showed that SCI-related pain is highly prevalent during initial inpatient rehabilitation. Prevalence of NocP decreased during inpatient rehabilitation, and prevalence of NeuP remained the same. Pain intensity was similar for different types of SCI-related pain, with a score on the 11-point NRS with means ranging from 4.19 (NocP at discharge) to 4.69 (NeuP at admission) . Female patients and patients with traumatic SCI have a higher risk of having NocP, and female patients, patients younger than 65 years, and patients with tetraplegia have a higher risk of having NeuP. Patients with AIS B and C have a lower risk of having NocP than those with AIS D. These differences in pain prevalence between different groups of patients with SCI are important in screening patients on having pain and developing pain during inpatient rehabilitation.
Suppliers
a.
SPSS Version 26; IBM, Armonk NY.
Acknowledgments
We thank these rehabilitation centers and the people involved in data collection.
Pain and its impact on inpatient rehabilitation for acute traumatic spinal cord injury: analysis of observational data collected in the SCIRehab Study.
The development of the Dutch Spinal Cord Injury Database (DSCID) was supported by Rehabilitation Netherlands (Revalidatie Nederland), Innovation in Rehabilitation program, grant IPR2011-12. All spinal cord injury rehabilitation centers in the Netherlands contributed to the DSCID.
Significant difference between groups.
00534-2&id=gr1.jpg){kind=link}