Social disconnectedness and perceived social isolation in persons with spinal cord injury/dysfunction living in the community: A scoping review

Context: Persons with spinal cord injury/dysfunction (SCI/D) are particularly at risk for social disconnectedness and/or perceived social isolation, which are key components to overall well-being. However, there is limited evidence that aims to understand these phenomena in this population. Objective: To investigate what is known about social disconnectedness and perceived social isolation for adults with SCI/D living in the community. Methods: A scoping review was conducted. A computer assisted search of four online databases was completed on all articles published until May 18th, 2021. Gray literature and key agencies were also searched for relevant documents. Two key concepts (spinal cord injury/dysfunction and social disconnectedness/perceived social isolation) and associated terms were used. All articles were double screened for inclusion by two reviewers. Results: Following deduplication, title and abstract screening and full-text screening, 37 articles met the criteria for inclusion in this review. Most of the included articles were published in North America, used quantitative methods and a cross-sectional design. Articles identified several factors that influenced social disconnectedness and perceived social isolation including participant characteristics, social support, psychological well-being, participation in activities, the built environment and physical health. Conclusion: The current literature suggests that issues with social disconnectedness and perceived social isolation exist for persons living with SCI/D in the community. More work is needed to better understand the relationship between these constructs. A more nuanced understanding can inform targeted interventions to help mitigate the impact of these phenomena on the SCI/D population.


Introduction
Social isolation is a growing issue affecting the health and wellbeing in the general population, with the COVID-19 pandemic further exacerbating this issue significantly. [1][2][3][4] To better understand the effects of social isolation on health and well-being, recent work has delineated between objective markers of social isolation (social disconnectedness) and subjective perceptions of isolation ( perceived social isolation, PSI). 5 Social disconnectedness refers to the objective lack of contact with others, which includes having a small social network size that consists of a limited number and type of different 'actors' that may not be seen frequently. As well, it may include a lack of engagement in a variety of activities. 5 PSI is defined as the subjective interpretation of isolation and typically manifests as feelings of loneliness and not belonging. 5 The distinction between social disconnectedness and PSI have been made because there is evidence that they do not always correlate strongly with one another. 5,6 As well, each have been shown to uniquely contribute to different health conditions. For instance, social disconnectedness may increase the risk for heightened inflammatory response, 7 and poor mental health 3 while PSI is associated with an increased risk of cardiovascular disease, 8,9 sleep issues/disorders (10), infections (11), a heightened inflammatory response to stress (12), and depression (13). However, both have been shown to contribute to premature mortality. [14][15][16] People with spinal cord injury/dysfunction (SCI/D) may be particularly vulnerable to being socially disconnected and experiencing PSI due to their mobility limitations and health issues (i.e. secondary health conditions). With regard to social disconnectedness, there are studies showcasing that this population has small social networks. 17,18 For instance, Guilcher and colleagues found in a qualitative study that people with SCI/D (n = 14 people with SCI/D; n = 14 care providers) had small social networks but these networks provided strong levels of social support. 17 Similarly, Guilcher et al. 18 quantitatively mapped the social networks of people with SCI/D in another study (N = 170) that confirmed these earlier qualitative findings. In addition to these small social networks, which puts people with SCI/D at risk for being socially disconnected, there is evidence showing this population experiences high levels of PSI (25-55%). [18][19][20][21][22] For instance, Robinson-Whelan et al. (21) found that 40% of people with SCI (N = 175) reported high levels of PSI via the UCLA Loneliness Scale.
To date, much of the work in this field has focused on the barriers that prevent persons with SCI/D from participating in their community and have not necessarily measured social disconnectedness and/or PSI per se. Furthermore, it is not well understood how being socially disconnected or experiencing PSI influences important outcomes post-SCI/D, such as their ability to participate in the community, their mental wellbeing or on their quality of life (QoL). For instance, a recent qualitative study of 30 persons with SCI/D highlighted that even though participants were engaging in social activities following their injury, they still experienced PSI in that they no longer felt connected to those around them. 19 Similarly, a study examining both social disconnectedness and PSI on QoL found that different aspects of social networks and PSI influenced physical and mental health-related QoL and life satisfaction while others did not. 20 There are also questions on the relationship between social disconnectedness and PSI in SCI/D, with some studies finding a strong association between these two constructs, 23 and others finding weak ones. 20 Overall, there are many questions on how to conceptualize and measure social disconnectedness and PSI post-SCI/D, as well as how they influence physical, mental and social health outcomes for this population.
To develop a better understanding of how these constructs are described and studied in the SCI/D literature, the current scoping review was undertaken to answer the following question: what is known about social disconnectedness and PSI for adults with SCI/D living in the community? The sub-questions of interest were: (1) What outcome measures have been used to quantify social disconnectedness and PSI?; (2) What are the characteristics of persons with SCI/D experiencing versus not experiencing social disconnectedness?; (3) What are the characteristics of persons with SCI/D experiencing or not experiencing PSI?; and (4) What are the factors that contribute to or mitigate social disconnectedness or PSI? Hence, this scoping review aimed to identify what gaps exist in the SCI/D literature with respect to social disconnectedness and PSI, and what is required to help mitigate the impacts of these phenomena on overall health and wellbeing for this population.

Methods & materials Protocol and registration
A scoping review was undertaken to examine social disconnectedness and PSI in the SCI/D literature because this type of review is well-suited for examining the extent, range and nature of research activity conducted on a certain topic and ideal for identifying gaps in the existing literature. 24 The scoping review framework originally introduced by Arksey and O'Malley 24 and further developed by Levac et al. 25 was used. In this framework, six stages are outlined, which include: (1) Identifying the research question; (2) Identifying relevant studies; (3) Study selection; (4) Charting the data; (5) Collating, summarizing, and reporting the results; and (6) Consultation. 25 For the purposes of this scoping review, the optional sixth stage was not conducted. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines were followed for this review 26 and the protocol was registered with OSF Registries (osf.io/rx3nb).

Eligibility criteria
Articles were included in this review if they were quantitative, qualitative, or mixed methods studies, as well as knowledge synthesis or protocol papers. Articles must have included adults (18 years or older) with an adultonset traumatic or non-traumatic SCI/D living either at home or in a congregate setting in the community. Additionally, included studies must have explicitly investigated or explored social disconnectedness or PSI as outcomes or phenomena of interest (see Table  1 for operational definitions). Given the variability in terminology used in the literature to describe social disconnectedness and PSI (e.g. social isolation, loneliness), we linked the construct to these main headings while reporting the original term used by the study authors in brackets.
The exclusion criteria were: (1) studies using a quantitative case study design; (2) paediatric onset-SCI/D since it is likely that experiences between this population and those who acquire their SCI/D later in life will differ 27 ; (3) studies reporting on people living in institutional settings (e.g. hospital, long term care, complex care settings); (4) conference publication (e.g. abstracts, posters); and (5) studies that did not explicitly investigate or explore social disconnectedness or PSI. For example, if an article was describing a related concept such as social participation, it needed to be in relation to social disconnectedness (i.e. social network characteristics) and/or PSI (i.e. measure of loneliness).

Information sources
A computer-assisted literature search was conducted on articles published up until May 18th, 2021. Four online databases were used: Medline (OVID), EMBASE (OVID), CINAHL (EBSCO) and PsycINFO (OVID). Gray literature was also retrieved using the Canadian Agency for Drugs and Technologies in Health (CADTH) checklist 28

and the University of Toronto
Libraries Gray Literature Documentation template (adapted from Godin et al.) 29 Gray literature was included in order to avoid publication bias and ensure thoroughness of the review. 30

Search
The search strategy used for this review was developed in consultation with an expert librarian at the University of Toronto. Two key concepts (spinal cord injury and social disconnectedness/PSI) and associated key words were used (see Appendix A for Medline search strategy). The search strategy was adapted for the remaining databases.

Selection of sources of evidence
Using the search strategy, a list of articles was obtained from various sources. Deduplication of records was conducted in EndNote using the process described by Bramer et al., 31 which helped to streamline the article review process. The resulting list of articles was uploaded to Covidence, which is an online software used for article screening. The reviewers (SRC and LC) met to discuss the inclusion criteria followed by the completion of a pilot-test on 10 randomly selected articles. The initial pilot-test did not yield sufficient interrater agreement (>75%) 32 so a second pilot-test was conducted once the disagreements were resolved, and the inclusion criteria was refined. This second pilot-test resulted in an interrater agreement of 100%, which allowed the title and abstract screening to continue. Two trained reviewers (SRC and LC) double screened all articles for inclusion using the previously described criteria.
Following title and abstract screening, 10 randomly selected full text articles were selected for a pilot-test. The pilot-test resulted in an interrater agreement of 80%, which was sufficient to proceed with this phase of the screening. Data extraction was conducted following the completion of the full-text screening. The same two trained reviewers (SRC and LC) double screened all articles for inclusion using the inclusion criteria noted above. Table 1 Operational definitions of social disconnectedness & perceived social isolation. 5 These definitions were selected based on previously published literature in SCI/D. [18][19][20] Key Term Definition

Social Disconnectedness
The objective lack of contact with others. 5 Components can include size of social networks, range of social networks (i.e. type of 'actors' within the network), frequency of contact with members of social networks. 5

Perceived Social Isolation
The subjective feelings of loneliness and not belonging. This can include a shortfall in one's social resources such as companionship and support. 5 Data items and charting process Key data (i.e. study information, participant characteristics, outcome measures used to collect social disconnectedness and PSI, and results related to these constructs) from the articles that met the inclusion criteria were extracted using Microsoft Excel. Data extraction was conducted by two members of the research team (SRC and LC) following a pilot of two articles. Each member extracted one article, which was then checked by the other member to ensure consistency and accuracy of the extracted data. Once this pilot test was completed, the remaining articles were divided between the two members for independent extraction. A quality assessment of the articles was not conducted for the purposes of this review, and it is not a requirement. 33

Synthesis of results
Results specific to social disconnectedness or PSI were compared across the relevant articles. Descriptive numerical analysis and thematic analysis were used to analyze the extracted information. The descriptive numerical analysis included counting and summarizing the number of articles that met a certain criterion (e.g. how many studies were conducted in a particular country). The findings aligned with our guiding research question and sub-questions, with the subheadings that are included under the 'Factors Influencing Social Disconnectedness and PSI' section being derived by grouping together similar topics across the included studies. Once articles reporting on similar outcomes were collated, each group was given a category by the authors and results were summarized and compared. The extracted data was also used to identify gaps in the literature and develop suggestions for future research.

Results
The resulting search yielded 3,053 records (see Fig. 1 for PRISMA diagram). Following deduplication, 1,993 articles were screened during the title and abstract review. Once the initial screening stage was completed, 83 full texts were identified for review. However, five full texts could not be retrieved, therefore 78 articles were assessed for eligibility. Thirty-eight full text articles were found to have met the inclusion criteria; however, one article was a systematic review and was not included in the extraction. Instead, the references used in the literature review were hand searched to identify any relevant articles that were not already included in this review.
Rates of PSI were described by nine studies, [18][19][20][21][22]34,35,47,62 with rates ranging from 0% to approximately 55% when measured at one time in relatively small sample sizes (n = 8 to n = 175). [18][19][20][21][22]34,35 In one study examining PSI, as conceptualized as loneliness over a 12-month period, it was reported that 55% (n = 123) of participants were experiencing PSI over that time. 62 In addition, one article noted that none of the eight participants (all men) who participated in their study reported feelings of PSI. 35 Social disconnectednessoutcome measures For social disconnectedness, both qualitative and quantitative data collection methods were used, with quantitative methods being used more frequently (n = 6). 18,20,23,51,54,63 These included: social network variables (e.g. size, composition and frequency of contact) (n = 3) 18,20,63 ; the Tetrafigap survey (n = 1) 51 ; the Life Satisfaction Questionnaire -Revised (which also had items relevant to PSI and included in the section below); the Duke Social Support Index (n = 1) 54 ; and Craig Handicap Assessment and Reporting Technique (n = 1) 23 (see Table 3 for specific outcome measure/ instrument information).

Sociodemographic characteristics
Several sociodemographic characteristics were reported to be associated with feelings of PSI in persons with SCI/D (e.g. sex, gender, level of education, race, age, income, living situation and geographical location). Mixed findings were suggested for sex and gender, depending on whether the study used quantitative or qualitative measures. For example, quantitative measures (e.g. UCLA Loneliness Scale, Patient Reported Outcome Measure Information System and the Life Satisfaction Questionnaire -Revised) reported no differences in PSI (measured as loneliness) by sex 18 or gender. 23 However, Cimino et al. 19 reported that women often described PSI and seemed to be more dependent on social connections to mitigate social isolation compared to men.
With respect to education level, three articles reported mixed findings. Two studies suggested a higher level of education was associated with less feelings of PSI, 21,52 while the third did not report a significant association between education and PSI. 18 Race was another sociodemographic variable that appeared to influence PSI, although reports were conflicting. One article by Newman et al. 23 suggested that White and Black participants with SCI were more likely to report lower levels of PSI compared to other races (other races not specified by authors), while Robinson-Whelen et al. 21 reported that those who identified as Hispanic had lower levels of PSI compared to those who did not.
Qualitative and quantitative evidence differed when reporting the impact of age on feelings of PSI. For example, two qualitative articles reported that feelings of PSI increased as participants aged and were able to do less (e.g. in their home, recreationally). 56,57 In a quantitative article, the authors reported no impact of age at injury on PSI. 21 Conflicting evidence was also reported when examining the association of income and PSI. One article suggested an association between higher income and lower PSI, 21 while another did not find PSI to differ significantly by income. 18 Relationship status was reported to influence levels of Table 3 Explanations of outcome measures/instruments used by included articles.

Outcome Measure Purpose of Measure
Tetrafigap Survey 51 Self-administered questionnaire that includes a question asking individuals about their social network as defined by the number of contacts per week. Life Satisfaction Questionnaire -Revised 65,66 Relevant items measure social contacts and participation in social activities. These include marital status, number of individuals in the household, and frequency of "people coming to see you" Duke Social Support Index 67 A brief measure of social support that includes items on social network size, social interaction, social satisfaction, and instrumental support. Craig Handicap Assessment and Reporting Technique 68 A quantitative measure that is used to measure societal participation following SCI. It also evaluates dimensions of social participation. UCLA Loneliness Scale 69 Survey used to assess loneliness which includes items about lack of companionship, feeling left out and feeling isolated. Life Situation Questionnaire 70,71 A multidimensional measure that is used to measure life satisfaction and life problems. The life problem scales include items on social isolation. Massachusetts Survey of Secondary Conditions 72 Instrument used to understand information about health status, secondary conditions, and a range of health-related quality of life measure. It includes scales that assess social participation and social isolation. Social Health Items from the Patient Reported Outcome Measure Information and Life Satisfaction Questionnaire (PROMIS) -Revised 73 Selected short-form scales from the Social Relationships subdomain used to evaluate perceptions of social isolation. Items from the following item banks were used: (a) social isolation, (b) companionship, (c) emotional support, and (d) instrumental support. Social Isolation Index 74 A three-item index used to measure social isolation that assess contact with friends and relatives. PSI, with those who had a partner indicating less PSI. 18,21,38,64 Similarly, those who were living alone were reported to have more problems with PSI than those living with others. 18,64 Three studies looked a geographical location (e.g. urban or rural) and PSI. 18,19,48 Two qualitative articles suggested that those living in rural communities 19 or away from family and friends had increased perceptions of PSI. 48 However, quantitative evidence suggests that there was no association between PSI and geographical location. 18 Sociodemographic variables were also reported to influence social disconnectedness in two studies. 18,23 For example, chronological age, marital status (i.e. single), and living situation (i.e. living alone or with a smaller number of other people) were reported to experience social disconnectedness. 23 However, level of education, geographical location, and income level were posited to not be influenced by social network size. 18 Injury characteristics Five studies reported time post SCI/D influenced PSI. 19,21,23,52,61 These studies posited that the longer a person was post-injury, the fewer problems they faced with social isolation. 19,21,23,52,61 Conversely, rates of PSI were posited by Li et al. (52) to decrease at a greater rate for those who were at a shorter time post SCI/D compared to those who had their injury for a longer period. Two articles reported that level of injury was not associated with PSI. 18,21 However, one article suggested that those with greater mobility impairments had greater levels of PSI. 21 No articles investigated the relationship between etiology of SCI/D and PSI.
With respect to social disconnectedness only two studies compared injury related characteristics. 18,23 First, Newman et al. 23 suggested that those with higher level injuries showed increased levels of social disconnectedness. This contrasts with Guilcher et al. 18 who reported that level of injury did not influence the size of participants' social networks. Finally, in terms of time post injury, Newman et al. 23 reported that those who had a longer time post injury had less problems with social disconnectedness.

Social support
Social support was the most common construct that was studied in relation to PSI in people with SCI/D (n = 14). 19,23,36,37,40,41,43,48,[49][50][51][52][53][54][55][56][57][58]53,61,64 With respect to PSI and frequency of visits, Tzonichaki et al. 64 claimed that the number of visits a person with SCI/D received was associated with decreased feelings of PSI. Peer support also seemed to be important to persons with SCI/D as noted by articles included in this review. 19,40,48,50,61 The process of replacing lost social connections was reported to be associated with feeling less PSI. 19 Replacing lost connections often included connecting with more people living with SCI/D (19). New social connections with peers with SCI/D were reported to help improve feelings of PSI. 19,40,48,61 Participants in these studies expressed the value of being able to relate to others who understood the everyday challenges of living with an SCI. 23,40 Positive role models were also valued by persons with SCI/D in another study, as these individuals were able to help facilitate acceptance and motivation to help strengthen self-esteem and improve problems with PSI. 50 Furthermore, peer support was important for participants in Dickson et al.'s study, 39 where a loss of sense of camaraderie was reported by participants following discharge from in-patient rehabilitation back home which resulted in profound feelings of loneliness. Two additional studies suggested that PSI was exacerbated by interactions with able-bodied individuals as participants felt their able-bodied peers did not understand what they were going through following their injury. 19,43 Negative attention from persons in society also resulted in feelings of PSI (50).
Participants in five studies spoke about the influence of family support and PSI. 19,36,43,50,53 Family breakdowns and family behavior in public was reported to contribute to increased feelings of social isolation. 36,53 For example, participants in the article by Boracco et al. 36 spoke about not wanting to go into the public, thus increasing feelings of PSI, because their family members were vocal about issues with accessibility. However, in this same intervention study, participants noted that community-based therapeutic recreation program specialists helped to mitigate these family conflicts, resulting in an increased ability to establish social connections, thus reducing feelings of PSI. 36,53 In two additional studies, participants felt as though they were a burden to their family due to their injury, which resulted in exacerbated PSI. 19,43 Conversely, Lovfenmark et al. 50 claimed that when participants felt supported by their family, they had less problems with PSI. In addition to peer and family support, interactions with health care professionals also helped to alleviate feelings of social isolation. 19,53,58 PSI and social support also appeared to have a reciprocal relationship as reported by Bohanske et al. 37 In their study, the authors posited that as PSI increased, participants exhibited decreasing interest in being around other people (e.g. trying to include others in their activities, trying to belong and to be with people, trying to be personal and intimate with others, and trying to express friendly and affectionate feelings towards others). 37 Cimino et al. Social disconnectedness and perceived social isolation Associations between social disconnectedness and social support were reported by two studies. 23,41 An increase in social disconnectedness was claimed to be linked to a lack of social support and a gradual fading away of friends. 41 Furthermore, PSI was posited to be influenced by the amount and frequency of social contacts. 23 Psychological well-being Psychological well-being in persons with SCI/D was reported by several articles to influence feelings of PSI. [19][20][21][22]46,50,57,[62][63][64] For example, an increase in PSI was reported to be associated with a decrease in life satisfaction, [20][21][22]62,64 an increase in depression, 21,46 and problems with mental health. 20,63 Furthermore, higher self-esteem was reported by Tzonichaki et al. 64 to result in lower levels of PSI as measured by a loneliness scale.
With respect to social disconnectedness, there were three studies that reported on psychological wellbeing 18,20,63 Two studies examined network intimacy. 18,20 Network intimacy is a component of social disconnectedness and is a marker of the strength or quality of a relationship and how connected someone feels to a particular actor in their social network. 75 These two articles claimed that greater network intimacy was associated with greater life satisfaction (20) and lower feelings of PSI (i.e. loneliness). 18 Within a person with SCI/D's social network, a higher frequency of contact was potentially linked to feelings of PSI 18 but was not reported to be associated with improvements in vitality or mental health. 63 Finally, social network size was also not associated with mental health or vitality 63 as well as PSI. 18 Participation in activities Five studies reported that PSI was associated with participation in various activities. 19,22,50,56,58 Being able to return to previous activities (e.g. work or recreational activities), 19 being able to increase leisure time physical activity, 22 and being able to do an occupation in a new way all contributed improving PSI. 22,56 However, in a study by Pentland et al. 56 participants spoke about how having to do less overtime or changing how they participated led to increased feelings of PSI. In another qualitative study by Cimino et al., 19 participants highlighted how self-advocating allowed them to increase their social connections and social participation, which alleviated PSI.
Social disconnectedness and its relationship with participation was described by three studies. 23,42,55 Participating in work positively influenced social connectedness by providing a sense of community outside of the workplace and providing opportunities for more social connection. 42 Furthermore, the quality of the relationships in the workplace was posited to be influenced by persons with SCI/D's desire to work. 55 Finally, in a study by Newman et al., 23 the authors reported that being less likely to go out was associated with social disconnectedness.

Built environment
The built environment was reported to influence PSI for people with SCI/D. Physical barriers that impacted participation in meaningful activities either in the community or at home were reported to be associated with increased PSI in five studies. 19,35,50,59,64 Only one study examined the connection between social disconnectedness and accessibility. 41 This study reported that issues with accessibility increased problems with social disconnectedness in people with SCI/D. 41 Physical health Only four of the included studies sought to understand the association between indicators of physical health and PSI. 35,44,49,60 Components of physical health that were suggested to be associated with loneliness included loss of sexual function, 60 uncontrolled bladder disorders, 60 symptoms of non-traumatic SCI, 44,49 and reduction in function. 35 There were only two studies that examined physical health and social disconnectedness. 20,51 One study by Le Fort et al. (51) demonstrated that a smaller social network was associated with presence of pressure ulcers, while Hitzig et al. 20 suggested that a higher level of network intimacy was linked to better physical health.
Other factors Several other factors described in four studies were suggested to be associated with feelings of PSI. 19,35,45,50 Lack of financial security was posited by two articles to increase feelings of PSI. 35,50 With respect to technology and PSI, two studies suggested that it was ineffective at improving feelings of social isolation. 19,45 Limited access to health resources was claimed to be associated with increased feelings of PSI in a qualitative study by Cimino et al. 19 In contrast to PSI, participants in Houlihan et al.'s 45 intervention study spoke about how email made it easier to stay in touch with friends and family which increased social connectedness. Sexuality and PSI were reported by only one study. 57 In their all-women qualitative study, Pentland et al. 57 reported that women experienced more PSI when they felt as though their sexuality and intimate relationships were being ignored.

Discussion
The main goal of this scoping review was to understand what literature is currently available on social disconnectedness and PSI in persons with SCI/D living in the community. From our review, 29 articles were identified, with the majority focusing on PSI and only ten explicitly studying social disconnectedness. Furthermore, only a few (n = 10) used theories or operational definitions to conceptualize social disconnectedness or PSI. Of the five articles that used theories to conceptualize PSI, none used the same theory. With respect to social disconnectedness, studies did not measure all the concepts that are associated with this construct (e.g. only measured network size and not network intimacy or frequency of contact). As a result, the findings reported in this review appear to be inconsistent, which is partly exacerbated by the use of measures that may not fully align with each articles' conceptual underpinnings. Ultimately, these inconsistencies make it difficult to interpret and compare findings across studies and may be responsible for inconsistencies reported in the literature (e.g. Hitzig et al. 20 vs Newman et al.). 23 For example, Hitzig et al. 20 found low correlations between PSI and social disconnectedness while Newman et al. 23 found high correlations. The conflicts may be due to the differences in the subjective measures used for these two studies (e.g. UCLA-3 vs PROMIS Social Health Item banks). While both measures are used to understand PSI, they target different areas (see Table 3), which highlights the inherit difficulties with measuring PSI as a concept. Few studies used PSI specific measures (n = 9) but instead opted for more generic measures that included a single question on PSI or related domains (e.g. life satisfaction or situation questionnaire). Overall, there is limited availability of validated quantitative measures of PSI and social disconnectedness for persons with SCI/D. While the UCLA-3 is a common measure that is used and validated in this population, it is a generic measure and may not be able to capture the nuances of living with SCI/D. A single measure that is specific to persons with SCI/D is needed that can be used across studies to generate a more thorough and more comparable understanding of social disconnectedness and PSI among persons with SCI/D. To provide clarity for future work, we propose that the approach taken by Cornwell and Waite 5 (see Table 1) be used. This approach considers both the subjective and objective components of social isolation. It is important to measure both components given that persons with SCI/D are clearly at risk for issues with PSI, while less is known about social disconnectedness according to this review. Due to discrepancies in terminology in the available literature, articles may have collected aspects of this construct including number and types of contacts but did not directly call it social disconnectedness (e.g. marital status, family membership or informal caregiving). When discussing PSI, recent studies suggest that 25-55% of persons with SCI/D experience issues with PSI. [18][19][20][21][22]34 Rates of PSI in this group are similar to those found in other groups of persons with physical disability such as lower limb loss (32% reported issues with PSI) 76 or stroke (44% reported issues with PSI). 77 By examining social isolation in this manner, we will be better positioned to use measures that align with these constructs which will lead to a better understanding of subjective and objective social isolation.
Given the findings of this review, more research is necessary to better understand how social disconnectedness influences health and wellbeing following SCI/ D. There is some evidence presented by articles in this review that a connection exists between social disconnectedness and health (e.g. mental, physical). Connections between social disconnectedness and health have been identified in other populations including older adults in the general population. [78][79][80] For example, recent literature suggests that those with lower levels of social disconnectedness had more problems with their mental (e.g. depression and anxiety), 78,79 and physical health (e.g. physical functioning and pulmonary functioning). 79,80 However, more work is needed to establish the extent to which social disconnectedness and mental health and physical health are related in persons with SCI/D. Future research should aim to use validated measures as described above to understand not only the influence of social disconnectedness on aspects of health, but also whether correlations exist between social disconnectedness and PSI. Further research is needed to explore the important context of a person's social disconnectedness and PSI as this information may help clinicians identify persons at risk and management options.
Another gap in the literature identified by this review involves the lack of reporting on etiology of SCI/D and how it may influence social disconnectedness and PSI. It is unclear from the currently available literature whether differences in etiology do not exist, were not reported, or were simply not studied. There is some evidence to suggest that etiology of SCI/D influence factors associated with social disconnectedness and PSI such as physical health and mental health. For example, those with traumatic injuries tend to have more severe neurological impairments 81 and more problems with independent ambulation 82 than those with non-traumatic injuries. Results from the current review suggest that having functional limitations was posited to be associated with greater levels of PSI 21 and problems with social disconnectedness including frequency of contact and participation in social activities. 23 When looking at mental health, a recent study suggested that those with a non-traumatic etiology had a higher prevalence of depression than the traumatic group. 83 The authors of this study suggest that because non-traumatic participants were significantly older, there may be a link to the comorbid medical conditions and thus an increase in risk of depression. 83 Furthermore, the nature of non-traumatic SCI where the prognosis is less certain due to the degenerative nature of these condition may result in a longer adjustment period than those with traumatic etiology who often get a diagnosis soon after their initial injury. 83 This in conjunction with the fact that depression has been associated with PSI in this review 21,69 may put those with non-traumatic SCI/D at an increased risk for experiencing PSI. While direct information regarding the influence of etiology is lacking in the literature, it is important to get a more nuanced understanding of risk factors for problems with social disconnectedness and loneliness to develop specific interventions and address etiology related concerns.
Given the current state of the COVID-19 pandemic, it is interesting to note that no studies that were included at the time of this review (May 2021 to February 2022) mentioned social disconnectedness and PSI in relation to SCI/D persons and the pandemic. Eleven studies had publication dates from the beginning of the pandemic to the search period. Given the impact the pandemic has had on mental health in the general population, 84 it is reasonable to expect that those effects were felt to an equal or greater extent by the SCI/D population. Articles included in this review identified that social disconnectedness and PSI are experienced by this population, therefore it is conceivable that COVID-19 may further exacerbate these phenomena given community lockdowns and physical distancing protocols. On the other hand, there may be a protective effect of quality of connections discussed by several of the included articles. 18,19 For example, Guilcher et al. suggested that although persons with SCI/D have smaller social networks, the quality of those connections influence PSI. 18 It is possible that these high-quality connections could have been maintained during the pandemic and had a protective effect. Further studies are needed understand the impact of the pandemic on the SCI/D population to understand how social disconnectedness and PSI may have been experienced and what the implications are for support programming during and beyond the pandemic.

Limitations
Given the nature of scoping reviews, human error may have prevented all relevant literature from being included in this review (e.g. exclusion of a relevant article). To minimize the influence of human error, a rigorous selection process (including double screening throughout the whole screening process, identifying relevant gray literature to avoid publication bias as well as hand reference searches) was followed to ensure that unintended omissions were minimal. Although a quality assessment is not required when conducting a scoping review, 33 our findings did identify methodological implications in the included studies that may have increased bias. For example, some of the findings highlighted in this review may have come from articles that have methodological issues. Future work should assess the quality of articles in order to ensure that stronger conclusions can be drawn from the available literature on social disconnectedness and PSI. Limitations also exist due to the nature of the inconsistent use of terminology used by the articles in this review. These inconsistencies create problems when attempting to compare and contrast the findings and caution should be taken when drawing conclusions based on the literature. In addition to the inconsistent use of terminology, the way "quality of contact" has been described in this review may lead to some confusion. Quality of connection can be easily construed as a subjective measure, while it has been used as an objective measure of social disconnectedness here. However, this categorization is aligned with other previously published research that includes quality of social network and the connections within the network as part of the objective understanding of social isolation. 17,18,20

Conclusion
The currently available literature suggests that issues with social disconnectedness and PSI exist for persons living with SCI/D in the community. However, more work is needed to better understand the relationship between these constructs and related domains (e.g. physical health, mental health, etc.) as well as whether these two phenomena are related in persons with SCI/ D. By developing a better understanding of these constructs, clinicians and researchers will be better able to develop key interventions to help mitigate the impact of these on the SCI/D population. In terms of social disconnectedness, the currently available literature is limited, and it appears that there is inconsistency in the way this construct is used. It is important to understand this construct since preliminary evidence suggests that persons with SCI/D may face challenges with this objective measure. Finally, etiology of SCI/D and the COVID-19 pandemic and their influence on social disconnectedness and PSI to ensure that adequate care is provided to those who need it.