A cross-sectional comparison of patient characteristics across healthcare settings using the minimal dataset for adults with acquired brain injury (MDS-ABI)

ABSTRACT Acquired Brain Injury (ABI) can have lifelong consequences and accordingly, persons with ABI often receive various types of healthcare. Facilities have their own preferences towards measurement instruments used to evaluate patients, impeding data comparison across healthcare settings. In this cross-sectional study, we used the previously developed minimal dataset for persons with ABI (MDS-ABI) to uniformly document and compare characteristics and outcomes of ABI patients in three healthcare settings: (1) residential care (n = 21), (2) non-residential care (n = 80), and (3) no ABI-related care (n = 22). Overall, patients of residential care settings had the lowest functional outcome compared with patients in the remaining groups. Nonetheless, all groups showed substantial disabilities within numerous life domains, indicating that the consequences of ABI are widespread among patients within and outside of healthcare facilities. These results demonstrate the need for a broad measurement of the potential consequences of ABI. The MDS-ABI covers twelve life domains most frequently affected by ABI and therefore helps to better recognize the consequences of ABI. In research contexts, implementation of the MDS-ABI allows for direct comparison of research findings. Future directions should be aimed at further implementation of the MDS-ABI to guide clinical decision-making and assist in identifying treatment goals specific to each healthcare setting.


Introduction
Acquired brain injury (ABI) is an umbrella term encompassing all forms of brain injury acquired after birth, potentially leading to deficits in various life domains.As medical techniques are advancing, mortality rates of ABI are declining (Gerber et al., 2013;Kunst et al., 2011).Subsequently, there are many persons living with the consequences of ABI, for which they often receive some type of health care.Since ABI can lead to chronic disabilities in many different areas of functioning and is therefore considered a lifelong condition (Corrigan & Hammond, 2013), care for this particular group is provided within a variety of settings.These settings include hospitals for the treatment of medical consequences, rehabilitation facilities for regaining function and community reintegration, home care or community support for daily problems arising from the consequences of ABI, and residential care for those who are no longer able to live independently.Many of the healthcare facilities that deliver these types of health care have their own preferences towards measurement instruments for assessment, monitoring and outcomeimpeding the comparison of data across healthcare sectors.
As brain injury patients often move between healthcare settings, standardization of instrument use could be used to promote data sharing between healthcare institutions.By sharing data, static information (such as date and cause of injury) does not need to be collected with each clinical transition, which often is experienced negatively by patients.Moreover, standardization can improve clinicians' understanding of available data during transfer when they are familiar with the instruments in use at the previous institution.In case of expected changes over time, for instance, as a result of recovery, degeneration, or when evaluating treatments, clinicians can use the same instruments to take an additional measurement to evaluate change and adjust their treatments accordingly.Further, research investigating the consequences of ABI could be progressed through systematic data collection.This would not only enable uniform data collection but would also facilitate findable, accessible, interoperable and reusable (FAIR) data (Wilkinson et al., 2016).
One approach to obtaining FAIR data is the development of minimal datasets.Minimal datasets, or core outcome sets, are defined as sets of standardized measurement instruments used to collect the minimum amount of data that is crucial for obtaining a global image of a patient.Ideally, sets should be used across all healthcare sectors and disciplines and in every stage of the health condition thus allowing for cross-sectional comparison.Researchers have previously developed tools for streamlining data collection within specific patient groups other than ABI (Haywood et al., 2018;Köke et al., 2017;Lutomski et al., 2013).Likewise, some initiatives have been undertaken for streamlining data collection within the field of ABI.Some of the first initiatives aimed at standardizing data collection in the field of ABI are the International Classification of Functioning, Disability and Health (ICF) Core Sets for stroke (Geyh et al., 2004) and TBI (Laxe et al., 2013).ICF Core Sets prescribe a number of ICF domains that at a minimum should be assessed after a health condition.However, the operationalization of these domains is not further specified and currently, no ICF Core Set for ABI in general exists.Further, there is little overlap in the domains that should be assessed after stroke and TBI (van Heugten, 2017), despite research suggesting that these injuries are associated with similar outcomes (Verberne et al., 2021).Therefore, previously we developed a minimal dataset for people with ABI (MDS-ABI) (Domensino, van Haastregt, et al., 2020;Domensino, Winkens, et al., 2020), which can be integrated into research projects as well as clinical practice.
The MDS-ABI was developed through multiple steps.In the first preparational step, we established a set of criteria for potential measurement instruments (e.g., freely available and applicable to all persons with ABI), and preselected measurement instruments complying to these criteria.Subsequently, a Delphi study was conducted to obtain consensus on the domains and instruments to be included in the MDS-ABI (Domensino, Winkens, et al., 2020).Over three rounds, an expert panel on developing/evaluating measurement instruments used within ABI care, identified a set of twelve essential outcome domains (demographics, injury characteristics, comorbidity, cognitive functioning, emotional functioning, energy, mobility, self-care, communication, participation, social support and quality of life), together covering all levels of the ICF model (World Health Organization, 2001).Further, experts expressed which of the preselected measurement instruments they found suitable for measuring the domains, or proposed different instruments.Eventually, they reached consensus on which instrumens should be included, leading to a draft version of the MDS-ABI.As a third step, the draft version was tested for usability in healthcare practice in a feasibility study within twelve healthcare organizations (Domensino, van Haastregt, et al., 2020).This study led to recommendations for further development of the MDS-ABI, based on which small adjustments were made to improve the lay-out and readability of the MDS-ABI.Participating clinicians also remarked that self-report outcome ratings of persons with ABI can be less reliable in the case of low levels of cognitive functioning or impaired self-awareness, and impossible in the case of (partial) loss of consciousness.Therefore, as a final step, a proxy-reported version of the self-reported questionnaires of the MDS-ABI was developed for patients who were incapable of accurately completing these questionnaires.Further, instruments for measuring caregiver burden and sense of competence were added to provide context to the overall status of a person with ABI.The aforementioned procedure led to the final version which is ready for use in studies investigating the outcomes after ABI.
Previous initiatives studied the characteristics and outcomes of persons with ABI within one specific health care setting, for example, disability care (Domensino et al., 2021;Regionaal Beraad Gehandicaptenzorg, 2006) and mental health care (van Heugten et al., 2012).However, no comparison has been made of persons with ABI between various types of healthcare settings.Information resulting from such a comparison could provide insights into the overall characteristics of people with ABI and the differences between patient groups.This in turn may help inform clinical decision-making and assist treatment design.Moreover, this information can identify potential treatment goals that are specific to each healthcare setting that persons with ABI receive care from.Therefore, we employed the MDS-ABI in a cross-sectional study among people with ABI within a variety of healthcare settings, and among persons with ABI living at home without receiving any treatment related to their brain injury.With this inventory, we aim to describe the characteristics of people with ABI and explore the differences between patients in various (healthcare) settings.Ultimately, with this study, we aim to demonstrate the potential of the MDS-ABI for clinical outcome measurement across these settings.

Design and participants
Participants were sampled within two types of healthcare settings as well as a community setting.The first type concerned institutions that provided ambulatory, inpatient, or outpatient non-residential care: hospitals, rehabilitation centres, and mental healthcare institutions.As these non-residential care settings all provide a similar type of care primarily aimed at the treatment of the consequences of ABI, participants originating from these settings were grouped as "non-residential care".The second type of healthcare settings concerned care institutions aiming to provide care and support in a living facility, which were therefore grouped as "residential care".Additionally, we included ABI patients who lived at home and received no care related to their ABI.This sampling strategy resulted in three care groups: patients living at home independently without receiving care related to their ABI ("no care"), patients receiving outpatient or temporary inpatient ABI care ("non-residential care"), or residing in a healthcare facility as a result of their injury ("residential care").
Persons of at least 18 years old who had sustained any type of medically confirmed acquired brain injury (stroke, traumatic brain injury, brain infections, brain tumours, hypoxia or intoxications) were eligible for participation.Furthermore, persons with ABI needed to be competent to judge their will to participate and to provide informed consent.There were no restrictions with regard to injury duration.Informal caregivers needed to have a personal relationship with a person with ABI, for whom they provide unpaid care (Wilkinson et al., 2016) and needed to be at least 18 years old.They could, for instance, be a relative, partner, friend, or neighbour of the persons with ABI included in the current study.There were no requirements for the quantity of provided care.
We employed a convenience sampling strategy which is described in detail under "procedure".Due to the explorative nature of our study, no sample size calculation was performed.This study was approved by the Medical Ethical Committee of Maastricht University (reference number METC 2018-0506).All participants provided written informed consent.

Included measurements
The final MDS-ABI, which is used in the present study, was divided into four parts.Part A is completed by the clinician or researcher/research assistant and consists of medical (record) information and a cognitive screening test.Part B is completed by the person with ABI and consists of demographic and patient-reported questionnaires.Part C contains proxy-reported questions about the person with ABI and is completed by a friend, family member, or healthcare professional who is able to make an estimation of the functioning, such as a nurse in the case of residential care.Part D is completed by the primary informal caregiver of the person with ABI and contains self-reported measures on the characteristics and status of the caregiver.As this part addresses informal caregiver burden, it is never completed by a healthcare professional.Therefore, part C and D can in some cases be completed by different informants.Table 1 displays all domains and measurement instruments that were included in the MDS-ABI, as well as a short description of each outcome.

Procedure
Data for the current study were collected in two separate procedures.The first part of the dataset concerned a secondary analysis of the data originating from a previous study into the feasibility of the MDS-ABI, in which only Part A and B of the MDS-ABI were included.In this study, thirteen healthcare professionals administered the MDS-ABI to 3-5 of their patients in order to judge its clinical usability, which resulted in MDS-ABI data for fifty persons with ABI.The full procedure of this study is described elsewhere (Domensino, van Haastregt, et al., 2020).The second part of the dataset was specifically collected for the current study.To this end, we contacted clinicians who work with persons with ABI in either (1) healthcare settings aimed at non-residential care or (2) residential care, using a convenience sampling method.Participating clinicians were asked to administer the MDS-ABI to 5-10 of their patients.To maximize sample representativeness, clinicians were instructed to select a representative sample of their patient population (e.g., in terms of injury type, injury severity, and demographic characteristics).Barthel Index (Mahoney & Barthel, 1965) Global assessment of mobility and selfcare.Total scores range from 0 to 20.
Higher scores indicate higher levels of functional independence.

Activities & participation
Part B: Patient-reported Demographic characteristics Age Gender Education Living situation

Personal factors
Emotional functioning HADS (Zigmond & Snaith, 1983) Screening questionnaire for anxiety and depression.The questionnaire is divided into two 7-item subscales for anxiety and depression.Higher scores represent higher levels of emotional complaints.Subscale scores of ≥ 8 indicate a probable anxiety disorder or clinical depression.

Body structures & functions
Energy and fatigue FSS (Krupp, 1989) Screening questionnaire consisting of 9 items on a 7-point Likert scale.
Higher scores indicate elevated levels of fatigue.A mean score of ≥ 4 indicates a substantial impact of fatigue on daily life.

Body structures & functions
Societal participation USER-P (Post et al., 2010) Questionnaire for societal participation consists of three subscales (frequency-, restriction-and satisfaction of participation) with a total length of 31 items.All subscale total scores range from 0 to 100; higher scores indicate better outcomes (higher frequency, less experienced restrictions and higher levels of satisfaction).

Activities & participation (Continued)
Additionally, advertisements on social media pages for persons with ABI were used to reach persons with ABI who live at home without receiving any care related to their ABI.After having expressed their interest in the study by email or telephone, these persons were contacted by the research team for an appointment to administer the MDS-ABI at their homes or at the research institute.
All parts of the MDS-ABI were administered once for every participant.Part A was completed by the clinician or researcher/research assistants based on patient file information or discussing the subject with the participant.Part B was completed by the participants independently or with help from a caregiver or clinician.Part C was completed by someone who was close to the participant and able to make an estimation of his/her functioning, such as a family member or nurse (in case of institutional care).Part D was completed by the primary informal caregiver of the participant.Part C and D were only administered when appropriate proxy-reporters and/or informal caregivers were present and willing to participate.

Analyses
Data that were collected with Part C (proxy-reported module) and Part D (caregiver module) were only available for a subset of the participants (n = 50).Questions on comorbid health problems and communication difficulties were revised based on results from the feasibility study.Therefore, data on these questions were dichotomized and then merged with the data from the subset of participants from the current study for statistical purposes.Descriptive statistics (means, medians, and frequencies) were used for all patient characteristics and outcome measures that form part of the MDS-ABI.Type of care (no care, non-residential care, or residential care) was used as the dependent (grouping) variable.To examine differences between participants over care settings, one-way ANOVA or nonparametric Kruskal-Wallis tests and χ 2 -tests or Fisher's exact tests (FET) were used when appropriate, depending on whether assumptions were met.Games-Howell post-hoc tests or nonparametric Dunn-Bonferroni pairwise comparisons were used to evaluate combinations of between-group differences.All analyses were performed using IBM SPSS version 24.0 (Corporations, 2016).Alpha was set at 0.05.

Sample characteristics
In total, data were available for 123 persons with ABI (n = 50 feasibility study participants and n = 73 additional inclusions), and 50 of their caregivers.Twentytwo (17.9%) participants lived at home independently without receiving care related to their ABI ("no care").80 (65%) participants received outpatient or temporary inpatient ABI care ("non-residential care").Twenty-one (17.1%) participants resided in a healthcare facility as a result of their injury ("residential care").Demographic and injury-related characteristics are presented in Table 2.The mean age of the sample was 54 years old (SD = 15.4,range 19-98).A slight majority of the participants were male (52.5%).Most participants lived together with others (26.4% with a partner, 23.1% with their partner and children).The most frequently reported cause of ABI was an ischaemic stroke (45.1%).Participants had incurred their injury on average 5.9 years before inclusion (range 0-35.9).20.2% of the participants had more than one  5).

Demographic and injury-related characteristics
Participants in the no-care group were more often female (63.3%), whereas most participants in a residential care setting were male (84.2%), p = .005.Compared with the no-care group and non-residential care group, participants in residential settings more frequently lived alone (52.6%, as compared to 31.8% and 37.5%, respectively).With regard to the type of brain injury, stroke was more prevalent in the non-residential care group (58.2%) compared with the no-care group (13.6%) or residential care group (28.6%).Patients in a residential setting had a fewer time between inclusion and acquiring their brain injury (M = 5.3, SD = 6.2, p < .001)compared with patients without care (M = 8.2, SD = 4.2, p < .001).Comorbid conditions were less common in the no-care group (36.4%) than in the non-residential care (81%) and residential care (78.9%) groups, p < .001.There were no significant group differences for participant age and the presence of previous brain injury.

Care characteristics
Compared with the non-residential and residential care groups, participants in the no-care group less frequently had had a hospital admission (68.2%, compared to 93.6% and 90.5%, respectively).Patients in a residential setting had significantly higher hospital admission times (M = 64.7,SD = 56.2) compared with   patients without care (M = 20.6,SD = 19.9)p = .021,and patients who received non-residential care (M = 19.1,SD = 26.6),p < .001.Participants in a residential setting were least likely to have been discharged home after hospital admission (5.6%) compared with the no-care group (68.2%) and the non-residential care group (48.6%).

Functional outcomes
Cognitive impairment was highest in the residential care group (81%), compared with the non-residential care group (49.4%) and the no-care group (22.7%).Patients in the no-care group had significantly higher functional independence levels (M = 19.8,SD = 0.5) compared with patients within residential care settings (M = 16.0,SD = 5.2), p = .002.Prevalence of communication problems was highest (95.5%) in the no-care group compared with the non-residential care group (41.0%) and the residential care group (47.6%).Patients within a residential care setting engaged significantly less frequently in societal participation (M = 22.2, SD = 11.3)than patients in the no-care group (M = 33.5,SD = 10.4), p = .009.There were no group differences with respect to (proxyreported) emotional functioning, fatigue, participation restrictions and satisfaction, and social support.

Caregiver characteristics and outcomes
Caregivers of patients in residential facilities were significantly older (M = 61.8,SD = 15.1)than caregivers of no care participants (M = 42.3,SD = 16.8),p = .014.Sample size was insufficient for groupwise comparisons on living situation and relationship to the person with ABI.No significant differences on group level existed for caregiver gender, caregiver sense of competence or caregiver burden.

Discussion
This study explored the usability of the MDS-ABI as a clinical outcome measure by comparing characteristics of persons with ABI across different healthcare settings.These settings included patients living at home independently without receiving care related to their ABI ("no care"), patients receiving outpatient or temporary inpatient ABI care ("non-residential care"), and patients residing in a healthcare facility as a result of their injury ("residential care").Demographic and injury-related characteristics, as well as functional outcomes within a variety of life domains, were collected using the MDS-ABI.Analyses showed that stroke was the most common cause of injury among participants.This especially was the case for the "non-residential care" group, which could be a reflection of well-organized care for persons with stroke.Stroke care in the Netherlands is largely integrated (Vat et al., 2016), meaning that all providers work together during the acute phase, over the course of rehabilitation, and in the chronic phase after the injury to provide coherent and patient centred care (i.e., stroke service).As a result, strokes are better recognized, acute medical care is well organized and patients with stroke receive standard follow-up care, which is not always the case for patients with other types of ABI such as TBI (Ribbers, 2007).There were high levels of comorbid health conditions, especially among patients who received temporary or residential care.This finding is in line with previous research (Jackson et al., 2020) and may underlie lower levels of ADL independence within these groups.Patients who lived at home without receiving care related to their brain injury had suffered their brain injury longest ago, which may be an indication of adaptation to the injury.This does not necessarily imply that patients experience no more limitations in their everyday lives, but that they have found a way to live with these consequences (Stiekema et al., 2020).
Despite the limited sample size of this study, we found a lower level of overall functioning among patients in residential care settings, which corresponds to the findings of previous research on the living situation of persons with ABI (Chen et al., 2014).Additionally, communication problems were most prevalent among the no-care group.As communication was measured using a selfreported screening question, this finding does not indicate a higher level of impairment within this group, but rather displays an increased level of experienced communication restrictions.This could be explained by the fact that community-dwelling persons with ABI might depend most upon their communicational abilities.As this subgroup also shows the highest level of societal participation, they might be confronted with their inabilities more often than participants in other groups.Further, no differences between patient groups regarding emotional functioning, fatigue, participation restrictions and satisfaction, or social support were found.Although no previous comparison on these outcomes over healthcare settings has been made, high levels of fatigue, depression and limitations to societal participation have been found in studies investigating patient characteristics within one setting (Juengst et al., 2013;Ouellet & Morin, 2006;van Markus-Doornbosch et al., 2020;White et al., 2012).Moreover, fatigue, emotional functioning and societal participation may be interrelated (Chen & Marsh, 2018;Juengst et al., 2013;Ouellet & Morin, 2006;van Markus-Doornbosch et al., 2020) and mediated by social support (Johnson & Ditchman, 2020;Leach et al., 1994).The fact that these outcomes are relevant to patients in all healthcare settings underlines the importance of measuring functional outcomes on several domains for all patients with brain injury.
Next to patient characteristics, the MDS-ABI also includes information on the characteristics of Informal caregivers of persons with ABI.We found that caregivers are most frequently cohabiting spouses.Overall, they feel competent to provide informal care to their partner or relative, however, some of the caregivers experience a disproportionate burden.Interestingly, this proportion did not differ over healthcare groups, meaning that healthcare burden might still be present even when daily care is outsourced to healthcare professionals.
With this inventory, we demonstrated the potential of the MDS-ABI for clinical as well as research purposes in the field of ABI.By covering all chapters of the ICF model, the MDS-ABI provides a comprehensive health summary that can be used to compare patient status over healthcare settings and disciplines or monitor changes in healthcare status as patients move through the healthcare chain.The fact that this study showed considerable disability among patients in all care groups is an important argument to keep following up on their wellbeing using a standardized set of measurement instruments such as the MDS-ABI.In the context of scientific research, the MDS-ABI can be implemented in research projects to provide a broad overview of (confounding) outcomes and allow for a direct comparison to previous research findings.Data from this study can be considered a foundation of an international database that can be used to answer secondary research questions.
This study has several limitations.Firstly, this study was underpowered to support the number of analyses that were performed.Although multiple testing corrections were applied to Tukey's post-hoc tests, conclusions should be interpreted with caution.As we needed to use nonparametric tests for a proportion of the analyses, further large-scale research is needed to support these findings.Due to the convenience sampling strategy of this study, the "non-residential care" group was disproportionately large and consisted of three subgroups (patients who were temporarily admitted to a hospital or rehabilitation centre, outpatients of hospitals or rehabilitation centres and patients who received community-based care), which are likely to show between-group differences.Similarly, it cannot be assumed that our subsamples are representative of the population they are supposed to resemble and a selection bias may have been introduced by the clinicians who selected participating patients.However, we tried to control for this source of potential bias by instructing caregivers to select a representative sample of their patient group, and by sampling patients from various healthcare institutions.With this study, we aimed to provide a first description of patient characteristics (demographic, injury-related, and functional outcome), investigate differences between these healthcare groups, and to demonstrate the application of the MDS-ABI across healthcare settings.In the near future, the MDS-ABI could be employed in large-scaled studies to study between-group differences with more certainty and assess which factors determine the best-suited type of care for each patient.
Another factor that might have influenced the data that were used in the current study is the fact that a proportion of the data was collected during governmental restrictions as a result of the COVID-19 pandemic.This could have influenced the estimates of participation as measured with the USER-P, and may explain why participants expressed low levels of satisfaction with their participation in society as compared with the validation sample of the USER-P (M = 60.1 vs. 69.2,respectively).
In general, the use of minimal datasets has the potential to progress research by facilitating data sharing through findable, accessible, interoperable and reusable data (Wilkinson et al., 2016).However, in order for MDS-ABI data to be truly FAIR, future developments should be aimed at providing an international data platform.This online platform can be used as a public data repository that can be supplied with anonymized research data from any research groups that implemented the MDS-ABI.Similarly, the data platform needs to be findable for other researchers who wish to use data within the database.In order for the data to be accessible, but simultaneously protect participant privacy, a commission responsible for reviewing data applications needs to be installed.Interoperability of data can be aided by universal and international codebooks.Finally, the reusability of MDS-ABI data can be further progressed by registering metadata that is associated with the study in which the MDS-ABI was implemented, such as setting, design and administration method.
This study serves as a basis for large-scale future research aimed at investigating whether information related to demographic characteristics, care that was provided after acquiring the injury, functional outcome, and caregiver characteristics are determinants of care indication after ABI.Guidelines for treatment after ABI recommend to assess common impairments on all levels of the ICF model before determining whether patients can be discharged home or, for instance, to a rehabilitation centre.The MDS-ABI can be used to measure these impairments over a multitude of factors and in different stages after the injury.Eventually, this comprehensive measurement approach can lead to timely recognition of potential treatment goals and improve healthcare for persons with ABI.Future research should therefore also be aimed at investigating the factors within the MDS-ABI that are predictive of care needs throughout the full care trajectory (from acute hospital care to long-term support) using prospective longitudinal designs.In order to improve uptake, future directions should be aimed at the implementation of the MDS-ABI in ABI research by funding bodies, as is currently the case for a national research programme into the long-term consequences of ABI.To expand the MDS-ABI database and allow for large-scale studies, researchers who use the MDS-ABI are therefore encouraged to share their data through personal correspondence.

Conclusion
The MDS-ABI is a tool for obtaining a broad but concise image on the health status of ABI patients regardless of the setting in which they reside or receive care.This study explored the characteristics of persons with ABI in a variety of healthcare settings, thereby demonstrating the potential of comparing patient groups and combining data following the FAIR data principles.Results showed that patients within residential settings experienced the highest level of functional disability after ABI.Patients of non-residential and residential healthcare institutions dealt with comorbid health conditions more frequently than patients who lived at home without any healthcare related to the injury.Despite these between-group differences, there was considerable disability over the different outcome domains within each care group.These data can be used to understand the consequences of ABI and aid in further shaping healthcare aimed at improving the outcomes of this lifelong health condition.Future directions need to be aimed at developing a data sharing platform for the MDS-ABI in order to establish a large-scale dataset that can be used to answer future research questions.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Table 1 .
Domains and measurement instruments in the MDS-ABI.
Note: ICF, International Classification of Functioning, Disability and Health; MoCA, Montreal Cognitive Assessment; HADS, Hospital Anxiety and Depression Scale; FSS, Fatigue Severity Scale; USER-P; Utrecht Scale for Evaluation of Rehabilitation -Participation; NPI-Q, Neuropsychiatric Inventory -Questionnaire; CSI, Caregiver Strain Index.

Table 2 .
Demographic and injury characteristics of total sample and per care group.
a F-test.b X 2 .c Kruskal-Wallis test.d Fisher's Exact Test.

Table 3 .
Care characteristics of total sample and per care group.
c Kruskal-Wallis test.d Fisher's Exact test.NEUROPSYCHOLOGICAL REHABILITATION

Table 4 .
Functional outcomes of total sample and per care group.

Table 5 .
Characteristics and outcomes of informal caregivers (n = 50).Caregiver Strain Index.Bold values denote statistical significance at the p < 0.05 level.
a F-test.d Fisher's Exact Test.