Spinal cord injury and polypharmacy: a scoping review

Purpose: The purpose of this scoping review was to map the scope of the literature on polypharmacy among individuals with spinal cord injury or dysfunction (SCI/D). Material and methods: Five electronic databases were searched for literature published between January 1990 and July 2018. The following keywords were searched using Boolean operators, wild cards, proximity operators and truncations: spinal cord injuries, multiple medications, polypharmacy. The initial search identified 1,459 articles; 1,098 remained after deduplication. Following the title and abstract screen, 81 full-texts were reviewed, and 18 met all of the eligibility criteria for inclusion in the review. Results: Of the 18 studies identified, less than half defined polypharmacy. Definitions varied in the types and number of medications. Older age, higher level of injury and greater severity of injury were factors related to polypharmacy. Negative clinical outcomes, such as drug-related problems and bowel complications were identified. Conclusions: This scoping review identified a paucity of research on polypharmacy post-SCI/D, highlighting a need for future research. To improve the state of knowledge, there is a need to better understand factors and clinical outcomes related to polypharmacy in persons with


Introduction
Damage to the spinal cord can be caused by a traumatic spinal cord injury or non-traumatic spinal cord dysfunction (hereafter referred to as SCI/D) [1,2].This damage to the spinal cord disrupts the transmission of motor, sensory and autonomic information from the brain to the body, resulting in impaired motor, sensory and autonomic function [3].Persons with SCI/D are at risk of developing secondary complications and multimorbidity, which are highly prevalent among persons with SCI/D [4,5].Due to the nature of secondary complications associated with SCI/D, there is a need for medications that act on both the central nervous system and the peripheral nervous system [6][7][8].As a result, many persons with SCI/D are on multiple medications [9].
Being on multiple medications, including prescription medications, over-the-counter medications and natural health products (e.g., vitamins, minerals, amino acids, probiotics, fish oils, herbal remedies, homeopathic preparations, non-medical marijuana), is often labeled polypharmacy [10].Polypharmacy is linked to an increase in the prevalence of falls [11], medication-related adverse events [12], medication errors [13] and poor adherence to medications among the elderly population [14].In a study examining drug-drug and drug-disease interactions, the likelihood of an adverse drug interaction with two different medications was 13%, while the likelihood of an adverse drug interaction when taking seven different medications was 82% [15].Similar results were found in studies of polypharmacy post-SCI/D, which highlighted several negative implications of multiple medication use in this population, including increased incidences of drug-related problems, higher medication and healthcare costs and an increased risk of fatigue and confused states [9,16,17].
Despite the high potential of adverse events related to polypharmacy among persons with SCI/D, there are a number of limitations of our knowledge on the topic for this population.First, there are only a few studies on the prevalence of polypharmacy post-SCI/D, with a range from 31% to 87% [9,[18][19][20].Second, there is a lack of understanding of how multiple medications persons with SCI/D might be taking interact with one another.Commonly prescribed classes of medications include: analgesicnarcotics, anticonvulsants, serotonergics, skeletal muscle relaxants, sedatives, tricyclic antidepressants and anxiolytics [9].These classes are identified as high-risk combinations because of the potential for overlapping effects, which can lead to adverse reactions [9].The frequent use of these medications increases the risk for adverse consequences, as the number of possible side effects and drug interactions increases drastically [21].Further, polypharmacy can have a negative impact on an individuals' health and their overall quality of life [19].Importantly, little is known about the appropriate use of medication among persons with SCI/D, the impact of polypharmacy and the importance of optimizing medication use in this population.
Given the potential risk for persons with SCI/D to experience polypharmacy, the purpose of this scoping review was to map the scope of the literature on polypharmacy among individuals with SCI/D.More specifically, we wanted to identify the types of studies that have been conducted on this topic, describe the sociodemographic and clinical characteristics of the participants in the studies (age, sex, type of injury, severity of injury, etc.), and highlight the components of polypharmacy under investigation (definition, factors, clinical outcomes, appropriateness, and patient, provider, caregiver experiences).Based on gaps identified in the review, suggestions for future research have been proposed.

Methods
This review was conducted using scoping review methodology outlined by Arksey and O'Malley and updated by Levac and colleagues [22,23].It also followed the reporting guidelines outlined by Tricco and colleagues [24].

Stage 1: Identifying the research question
The research question leading this scoping review was: What is reported in the literature about polypharmacy among individuals with SCI/D?Thus, the objectives of this review were to identify: (1) the types of research studies that have been conducted; (2) the sociodemographic and clinical characteristics of the individuals included in the studies; (3) how the authors defined polypharmacy; (4) factors related to polypharmacy; (5) clinical outcomes related to polypharmacy; and (6) the knowledge, attitudes, beliefs and experiences of caregivers, clinicians or individuals with SCI/D on polypharmacy.

Stage 2: Identifying relevant studies
The literature published between January 1, 1990 and July 31, 2018 was searched using five electronic databases (MEDLINE (Ovid Interface), EMBASE (Ovid Interface), PsycINFO (Ovid Interface), Cumulative Index to Nursing and Allied Health Literature (EBSCO Interface), and Cochrane Library).All database searches were conducted on June 29, 2018.A search for gray literature was conducted on TSpace, Spinal Cord Injury Research Evidence (SCIRE), the Rick Hansen Institute, Spinal Cord Injury Ontario, Paralyzed Veterans of American and the World Health Organization websites.
The search strategies (see Supplemental Material) were adapted for each database to include correct indexing and command line syntax.The following keywords were searched using Boolean operators, wild cards, proximity operators and truncations: spinal cord injuries, multiple medications, polypharmacy.The recommended gold standard search strategy for spinal cord damage not due to trauma was also incorporated into the search [25].

Stage 3: Study selection
Study selection was conducted using EndNote X8 (reference manager software system) and Bramer's method for deduplication [26].The purpose of the deduplication method is to remove duplicates from the exported articles.The initial search resulted in 1,459 articles and one article was identified from another source.The article identified from another source was found by continuously searching for new, relevant literature after the database search had taken place.No relevant articles were identified from the gray literature search.After removing the duplicates, 1,098 remained for title and abstract screening (Figure 1).
Included studies were required to meet the following criteria: (1) included individuals with SCI/D; (2) included individuals prescribed or taking multiple medications; and (3) published between January 1, 1990 and July 31, 2018.For studies to meet the second criteria (prescribed or taking multiple medications), the participants in the included studies had to be prescribed or taking two or more medications, which could include prescription medications, over-thecounter medications and natural health products.Studies were excluded if they met any one of the following criteria: (1) books, book chapters, opinion pieces, protocols, narratives or editorials; (2) studies on animals; or (3) poster or conference abstracts that did not have a published full-text article.Articles published prior to 1990 were excluded because the majority of research on polypharmacy among persons with SCI/D occurred after this date.
A Microsoft Excel (2016) spreadsheet was used to facilitate title and abstract screening.Two reviewers independently screened 200 articles and had a percent agreement of 0.93, representing good agreement, as a percent agreement over 0.80 is considered a high level of agreement [27].Any disagreements were discussed in person between the two reviewers.One reviewer (LC) screened the remaining 898 articles.After completing the title and abstract screening, 81 articles remained for full-text review.Two reviewers independently screened 16 articles and had a percent agreement of 1.00, so one reviewer (LC) screened the remaining 65 full-text articles.After screening the full-text articles, 18 articles were included in the scoping review (Figure 1).

Stage 4: Charting the data
Two reviewers used a Microsoft Excel (2016) spreadsheet to extract data from the included articles.The reviewers met in person to discuss any questions regarding the extraction of data; these meetings took place before extraction started and during the process.The following information was extracted from each article: general information about the article (title, authors, journal, publication year, publication country), study characteristics (objective, hypotheses, inclusion/exclusion criteria, method of data collection, study design, country study was conducted, frameworks guiding research and primary and secondary outcomes), study and population characteristics (sample size, age, sex, ethnicity, income level, education level, household composition, employment status, type of injury, years post-injury, geographical location, location of residence at time of study, other comorbidities, level of injury (the level of a SCI/D corresponds to the location of damage and is often classified as a cervical, thoracic, lumbar or sacral injury [3]) and severity of injury (depends on the level and the completeness; the severity of a SCI/D can be measured on the American Spinal Injury Association Impairment Scale [3], which classifies an injury from A to D: (A) complete, (B) sensory incomplete, (C) motor incomplete (more than half of key muscle groups below the level of injury have a muscle grade of 0-2) and (D) motor incomplete (at least half of key muscle groups below the level of injury have a muscle grade of 3 or greater)), polypharmacy (number of medications, types of medication, definition, factors and effects) and study outcomes and findings (results, key findings and conclusions).

Stage 5: Collating, summarizing and reporting results
Another Microsoft Excel (2016) spreadsheet facilitated the extraction and analysis of data from the included articles.This allowed for a thorough analysis and the ability to compare and contrast different variables of the included studies.Due to the nature of a scoping review, the details discussed in the results and discussion sections are based on how the authors presented the information in their article.In this scoping review, the terms medications, drugs and drug classes are not used interchangeably, as they each have their own distinct definition.

Results
Eighty-one full-text articles were reviewed, 63 were excluded, and 18 met all of the eligibility criteria to be included in this scoping review (Figure 1).The key study characteristics of the included articles are displayed in Table 1.Of the included articles, 16 were quantitative and the remaining two were a case report and a case discussion.The majority of studies were conducted in the United States (n ¼ 9) and Canada (n ¼ 2), with the remaining studies taking place in one of the following countries: Denmark, Sweden, Europe, Israel, Australia, United Kingdom and Norway.Most articles were published in the last ten years (n ¼ 13), with only five published before 2008.

Definition of polypharmacy
Seven of the included articles provided a definition of polypharmacy (Table 2) [9,[18][19][20]28,37,41]; definitions focused on the types of medications included and the number of medications.Some studies included only prescription medications, some included prescription medications, over-the-counter medications and natural health products and others did not report what types of medications were included in the definition of polypharmacy.The most common threshold for polypharmacy was five medications (n ¼ 4), but other definitions had thresholds of nine medications (n ¼ 1) and 10 medications (n ¼ 2).Among the studies that used five medications as a threshold for polypharmacy, there were differences in the types of medications that were included in this definition.For example, two definitions only included prescription medications [9,28], one included prescription medications, overthe-counter medications and natural health products [20], and one included prescription medication and over-the-counter medications but excluded vitamins [19].The method by which the information on medication use was obtained differed across the included studies (e.g., review of medical records or charts [20,29,30,33,34,36,37], survey data [19,32,35,[38][39][40] and administrative health data [9,18,28,31,41]).Only one of the included articles discussed the appropriateness of the medications [41].This intervention provided clinicians with a guide for tapering opioids and the guide stated that "opioids are not appropriate for chronic pain, only for cancer, postoperative pain, or a tooth abscess" (p.ix).The appropriateness of the rest of participants' medications was not discussed.
With respect to older age, Hope and colleagues conducted a retrospective chart review of 88 patients with SCI/D to identify patterns related to medication usage and the potential implications of these patterns [32].Older adults (aged 65þ) were more likely to use more medications than the younger group (average of 10.1 compared to 7.4).Similar results were seen in a cross-sectional survey that investigated medication use in persons with pediatric-onset SCI/D (age <19 years) [19].Among the 159 participants (mean age at time of interview was 35 years, range 26-59), polypharmacy was present in 49 (31%) participants and was significantly associated with older age.High rates of polypharmacy The case report and case discussion designs do not fit into a qualitative or quantitative method.
b Medicare Part D is a drug coverage plan that helps cover the cost of prescriptions drugs.
NR: not reported.among older adults (aged 66þ) with traumatic SCI were also identified in Guilcher and colleagues' retrospective population-based study using health administrative data on medication use post-SCI [18].They found that 87% of the participants were taking five or more drug classes and 56% were taking ten or more drug classes.Despite this finding, no statistical differences were observed for age among the older adults.
Level of injury was identified as a factor related to polypharmacy in two articles [19,30].A cross-sectional survey by Hwang and colleagues described the medications that were taken by adults with pediatric-onset SCI/D [19].Polypharmacy was significantly higher among participants with tetraplegia (n ¼ 92), compared to those with paraplegia (n ¼ 67), 40.2% compared to 17.9% (p ¼ 0.003).Similar results were identified in another study examining medication use in people with SCI/D living in a residential facility, as participants with cervical level injuries had the greatest number of medications per person [30].Despite a nonsignificant difference, participants with cervical injuries (tetraplegia) averaged 8.8 medications per person, compared to those with thoracic (7.9 medications per person), lumbar (8.0 medications per person) or another level (7.4 medications per person).
Severity of injury was also related to polypharmacy in two articles [19,29].A retrospective observational study that evaluated changes in medication patterns following a spinal cord lesion [29], found that those with more severe injuries (American Spinal Injury Association Impairment Scale A, B or C) had a higher increase in total medication consumption following their injury when compared to those with American Spinal Injury Association Impairment Scale D (average of 4.7 compared to 2.5).Hwang and colleagues' study demonstrated similar results in regard to severity of injury and medication use, as participants with more severe injuries had a greater use of medication [19].Significant differences in the frequency of polypharmacy were identified between the different American Spinal Injury Association Impairment Scale groups.
Other factors potentially related to polypharmacy, but mentioned in only one study, were: lower continuity of care between providers [18], type of injury (traumatic) [29], longer duration of injury [19], dual pharmacy users [28], gender (male) [29] and greater numbers of secondary health conditions [19].

Clinical outcomes related to polypharmacy
Eight articles reported outcomes related to polypharmacy and all found a negative relationship between polypharmacy and health outcomes.Negative outcomes included: drug-related problems (e.g., intoxication caused by drug interaction, adverse drug events), antibiotic-associated diarrhea and constipation.
The most commonly reported negative outcomes of polypharmacy were drug-(or medication) related, as seen in four of the eight included studies [9,20,31,33].Kitzman and colleagues examined the prevalence of polypharmacy, the level of polypharmacy specific to seven drug classes and risks for drug-related problems in the United States [9].The specific drug-related problems (e.g., nausea, drowsiness, etc.) were not reported, but they were serious enough for the healthcare provider to make note of them.When comparing the population that were on at least five medications, persons with SCI/D reported more drug-related problems than the controls and were 1.5 times more likely to experience a drug-related problem.When comparing persons with SCI/D to the controls, polypharmacy was 3.7 times more likely to be the reason for drug-related problems.Further, it was determined that persons with SCI/D were more likely to be prescribed five or more medications than the control group (56% compared to 27%) who were matched by age and sex, but without a SCI/D.Persons with SCI/D were also prescribed an average of 2.5 more high-risk drug classes than the controls, with 92% of persons with SCI/D being prescribed at least one high-risk medication compared to 44% of controls.
Patel and colleagues' retrospective chart review described the prevalence and use of pharmacotherapy, medication-related problems, prevalence of polypharmacy and participant demographics among patients with SCI/D [42].Of the 19 individuals whose charts were reviewed, 74% were on five or more prescription medications, over-the-counter medications, natural health products, vitamins and minerals, with a mean of eight products used concurrently per patient.Thirty-four medication-related problems such as: untreated condition, ineffective medications, adverse drug reactions, over or under dosage and noncompliance were reported by the patients.Of all the participants (n ¼ 19), 63% experienced at least one medication-related problem.
Vander and colleagues' case study described a woman who experienced drug toxicity following the addition of two new prescription medications [33].The patient was on a medication regimen of carbamazepine, baclofen, atenolol and fluvoxamine.Due to increased spasticity and hypertonic bladder without vesicoureteral reflux, both dantrolene and oxybutynin were added to her regimen.After the addition of these two new medications, the patient experienced drowsiness, confusion, unsteadiness and slurred speech.The administration of the new medications slowed the metabolism of carbamazepine, elevating the blood level concentration and resulting in carbamazepine intoxication.
Another type of drug-related problem was highlighted in Hand and colleagues' retrospective quasi-experimental study that found a high percent of adverse drug events in adults with SCI/D and polypharmacy compared to the incidence of the control group (able-bodied, matched controls without SCI/D) [31].Of the 2,779 participants with SCI/D, 31.5% experienced at least one adverse drug event, compared to 24.2% of matched controls.The most commonly reported adverse drug events included rash (25%), dermatitis (4%) and drug psychosis (0.5%).Among persons with SCI/D, approximately 14 participants reported experiencing drug psychosis, but the details of these experiences were not reported in the article.
In addition to drug-related problems, bowel complications such as antibiotic-associated diarrhea [39,40] and constipation [36] were identified in three of the eight articles that reported on outcomes.Wong and colleagues conducted two studies examining the prevalence of antibiotic-associated diarrhea and found similar results in both [39,40].In the international, multi-center study, of the 1,267 patients with SCI/D, 215 were on antibiotics and 32% of those were on more than one antibiotic [39].The top reasons for the antibiotic were urinary tract infections, infected pressure ulcers and skin infections.Fifteen percent of the patients on antibiotics developed antibiotic-associated diarrhea compared to 6% of those not on antibiotics (p < 0.01).Antibiotic-associated diarrhea was associated with: polypharmacy, multiple antibiotic use, use of high-risk antibiotics, older ages (65þ), tetraplegia, higher body mass index and hypoalbuminemia.
One study also reported the prevalence of constipation-related symptoms in persons with chronic SCI/D [36].In a cross-sectional interview study (n ¼ 161 persons with SCI/D) by Harari and colleagues, 69% of participants used four or more laxative, suppository or enema doses each month with an average of 27 doses.The use of six or more medications (polypharmacy) was associated with constipation, as were the use of benzodiazepines, tetraplegia, regular laxative use, Frankel grade A/B (SCI/D classification scale) and history of urinary outlet surgery.

Knowledge, attitudes, beliefs and experiences of polypharmacy
Only one of the included articles explored participants' beliefs about medications [38].Høgholen and colleagues examined whether adherence to medications was influenced by the participants' pain, spasms and/or beliefs about medications.Beliefs about medications were assessed using the Beliefs about Medicines Questionnaire, which consists of two sections: (1) necessity and concern for personal medicines and (2) beliefs about overuse and harm.Of the 105 participants with SCI/D, 97% reported strong beliefs about the necessity of their medications.Other common beliefs included: worries about the long-term effects of medications, worries about becoming dependent on medications, medications disrupt their life, medications cause unwanted side effects, medications make their life possible and without their medications they would be very ill.Additionally, the majority of participants believed that medications are generally overused, and just under half believed that medications are generally harmful.

Discussion
This scoping review mapped what is reported in the literature about polypharmacy in the SCI/D population, which only found a small number of studies.Our findings showed that under half of the included studies provided a definition for polypharmacy and of the studies that defined it, there were differing thresholds and types of medications included in the definitions.As well, the included studies identified several factors (older age, higher level of injury and greater severity of injury) and negative outcomes (drug-related problems and bowel complications) related to polypharmacy post-SCI/D.However, the overall quantity of evidence was limited.Based on the findings from this review, future research should address the following: ( 1 This review highlighted a lack of consistency in how polypharmacy is defined for individuals with SCI/D, with inconsistencies identified in both the number and type of medications included.Not unique to SCI/D research, there is a lack of a consistent definition for polypharmacy in the general literature as demonstrated by the following three definitions: (1) being on 10 or more drug classes [18]; (2) the use of five or more medications including prescription medications, natural health products and over-thecounter medications [20]; and (3) the use of nine or more prescription medications [37].Similarly, a 2017 systematic review of the definitions of polypharmacy among any population found substantial variability, with differences in the number of medications, the duration of medication therapy, the healthcare setting and the appropriateness of medications [10].In the general literature, variability in definitions for polypharmacy is not uncommon because different definitions often stem from the availability of data (e.g., administrative data often lacks over-the-counter claims or natural health products).However, inconsistent definitions are problematic when studying populations with small incidence such as SCI/D, with an estimated annual global incidence between 40 and 80 cases per million people [43].Different definitions of polypharmacy may lead to differences in the reported rates of polypharmacy among persons with SCI/D and associated risk factors/outcomes.This places greater importance on the collaboration of researchers and using common definitions when conducting research among persons with SCI/D, when possible.Based on the results of this scoping review, there is a need to develop a standardized definition for polypharmacy within SCI/D research that includes the types and number of medications.
Another significant gap this review highlighted was information on the appropriateness of the multiple medications or polypharmacy.Appropriate polypharmacy can be defined as, 'prescribing for an individual for complex conditions or for multiple conditions in circumstances where medicines use has been optimised and where the medicines are prescribed according to best evidence' (p.ix) [44].When polypharmacy is appropriate, it can improve quality of life and extend life expectancy [44].Conversely, inappropriate or problematic polypharmacy can be defined as, "the prescribing of multiple medications inappropriately, or where the intended benefit of the medication is not realized" (p.ix) [44].Components of problematic polypharmacy can include, but are not limited to: hazardous drug interactions, unacceptable medication burden on the patient and medications prescribed to treat side effects of other medications where alternatives are available [44].Similar to other populations with multimorbidity, there is a paucity of research on appropriate and inappropriate polypharmacy among persons with SCI/D.Further, there is a significant lack of evidence-based guidelines for patients with multimorbidity and polypharmacy, as the majority of evidence is with single-disease models [45].Based on these findings, it is important for future research to explore the appropriateness of polypharmacy among persons with SCI/D.
Another gap identified in this scoping review was research examining factors associated with polypharmacy and clinical outcomes of polypharmacy.The majority of studies that identified factors related to polypharmacy had large sample sizes but either employed a retrospective or cross-sectional research design.Based on these findings, future studies should use prospective, longitudinal study designs to identify if, and how, medication use and polypharmacy is affected by different factors such as level of injury, severity of injury, sex, age, type of injury, duration of injury, etc.It is important to know what factors are related to inappropriate polypharmacy in order to identify those at risk and explore potential alternatives.Despite the limited research, three factors associated with polypharmacy were identified in more than one study and included: older age [18,19,30], higher level of injury [19,30] and greater severity of injury [19,29].Older age is likely a factor related to polypharmacy because of the higher prevalence of chronic conditions and multimorbidity among older adults [46][47][48][49].Persons with non-traumatic spinal cord dysfunction are typically older than those with traumatic injuries [50][51][52][53], which results in a greater risk for having multimorbidity.The high prevalence of chronic conditions and multimorbidity often leads to an increase in medication use to manage these conditions [44].Similarly, a higher level of injury and greater severity of injury are likely to have higher rates of medication use and polypharmacy.Those who have higher and more severe injuries will likely have greater number of secondary complications that require medications (e.g., urinary tract infections, osteoporosis, diabetes, etc.) [54].
With respect to clinical outcomes, only half of the included articles reported outcomes of polypharmacy; all of which had a negative impact on health outcomes (e.g., drug-related problems, antibiotic-associated diarrhea, constipation, mortality and increased hospital admissions due to pressure ulcers).Similar negative outcomes have been reported in studies with older adults, as well as a variety of other outcomes including: increased healthcare costs [55,56], adverse drug events [57], medication nonadherence [14], decreased functional status [58,59], cognitive impairment [60] and falls [61,62].Despite these negative outcomes, the use of multiple medications and polypharmacy can also be beneficial among elderly patients in decreasing symptoms, reducing disease complications, slowing disease progression and improving quality of life [63,64].Further, polypharmacy has been linked to both positive and negative health outcomes in older adults, with the majority of recent research focusing on negative health outcomes [65][66][67].However, there is currently a lack of research among persons with SCI/D pertaining to the potential harms and benefits of polypharmacy highlighting the need for future research to prospectively study both potential clinical outcomes of polypharmacy.
A major gap identified in this review was the lack of patient, caregiver and clinician perspectives on the topic of polypharmacy.Only one study examined beliefs about medications, with a paucity of qualitative research exploring the knowledge, attitudes or experiences of persons with SCI/D.Attitudes, perceptions and experiences of medication use and polypharmacy have been explored in other populations such as older adults [68][69][70][71].Some common perceptions and experiences of the participants in these studies include the following: a willingness to stop one of more of their medications [68,69], concerns with side effects, long-term use and potential interactions [68][69][70][71], and a dependency or trust in their healthcare professional [70,71].Understanding an individual's knowledge, attitudes and beliefs toward medications is critical, as these are important factors that should play a role in what treatments the individual is offered [72].In order to optimize patient outcomes, it is important for healthcare providers to have more explicit conversations with patients and their caregivers about medications and medication management.Likewise, persons with SCI/D are encouraged to engage their healthcare providers in these conversations.Based on the findings of this scoping review, no qualitative studies have been conducted, to our knowledge, that explore the experiences of persons with SCI/D relating to the use of medications, thus reinforcing the need for future research to qualitatively explore the attitudes, beliefs and experiences with medication use and polypharmacy in this population.
Furthermore, there was a lack of consistency in reporting the sociodemographic and clinical characteristics of the study population, with the majority of studies not reporting the participant's ethnicity, income, level of education, household composition or employment status.Additionally, half of the studies (n ¼ 9) failed to report the type of injury as traumatic or non-traumatic.Among persons with SCI/D, managing secondary complications and comorbidities with prescription medications is a standard of care [9].Differing injury etiologies may lead to differences in medication use, as persons with non-traumatic injuries are often older and experience a greater number of secondary complications and comorbidities [73].Further, knowing the participants' income level or socioeconomic status is important because of its relationship with medication adherence, as persons with low income have been reported to have high rates of cost-related nonadherence [74][75][76][77][78][79][80].In order to improve rehabilitation outcomes and meet the needs of the patient, healthcare providers should be aware of financial resources when prescribing.Based on the implications that sociodemographic and clinical characteristics may have on one's medication use, it is recommended that future research explores potential differences in medication use between persons with traumatic and non-traumatic injuries and reports population characteristics that may lead to differences in medication use.Exploring these differences is important for understanding what works, and for whom, in order to tailor medication regimens for optimal outcomes.

Limitations
The included articles were not assessed for quality; however, the objective of this scoping review was to map the scope of the literature on polypharmacy in persons with SCI/D.It is also possible that relevant articles were missed, as only articles published between January 1, 1990 and July 31, 2018 were included in this review.However, in order to minimize the potential for missing relevant articles, an exhaustive search for articles took place.In addition to developing the search strategy in consultation with a librarian, gray literature and the reference lists of included articles were also searched.As our search strategy was in English, it is possible that articles written in different languages were missed.Despite only searching in English, this review was not limited to articles published in English.

Conclusions
This scoping review mapped the range of available literature on polypharmacy among persons with SCI/D.This review revealed a number of gaps, which led to the following suggestions for future research: (1) establish a standardized definition for polypharmacy; (2) explore the appropriateness of polypharmacy among persons with SCI/D; (3) explore factors that are related to polypharmacy in persons with SCI/D with prospective study designs; (4) examine positive and negative clinical outcomes of polypharmacy; (5) explore knowledge, attitudes, beliefs and experiences of persons with SCI/D, caregivers and clinicians relating to polypharmacy; and (6) describe the sociodemographic and clinical characteristics of the population.The importance of developing a standardized definition for polypharmacy and reporting the sociodemographic and clinical characteristics of the study population was highlighted.Focusing research in the recommended areas will allow for improvements in understanding factors and clinical outcomes related to polypharmacy and patient perspectives of their experiences being on multiple medications.

Figure 1 .
Figure 1.PRISMA flow diagram of included articles.

Figure 2 .
Figure 2. Factors related to polypharmacy identified in five studies.
) establish a standardized definition for polypharmacy; (2) explore the appropriateness of polypharmacy among persons with SCI/D and compare the extent of polypharmacy to other adults without SCI/D; (3) explore factors that are related to polypharmacy in persons with SCI/D with prospective study designs; (4) examine positive and negative clinical outcomes of polypharmacy; (5) explore knowledge, attitudes, beliefs and experiences of persons with SCI/D, caregivers, and clinicians relating to polypharmacy; and (6) describe the sociodemographic and clinical characteristics of the study population.

Table 1 .
Characteristics of studies included in the scoping review (n

Table 2 .
Definitions, factors and outcomes of polypharmacy in studies included in the scoping review (n a On demand medications are medications that are used as needed.AIS: American Spinal Injury Association Impairment Scale; NR: not reported; SD: standard deviation.