|Home | About | Journals | Submit | Contact Us | Français|
More than 5.3 Million Americans are living with long-term disability following TBI and approximately 40% of TBI survivors report at least one unmet need at 1 year post-injury. The totality of the problem of TBI may therefore put increased responsibilities on the significant other and family. The purpose of this work was to conduct an integrative review of the literature to identify available instruments that might be useful to researchers and clinicians interested in the effects of TBI on family functioning. A review of the literature was undertaken using CINAHL Plus, Family Systems Abstracts, and Pubmed from 1998–2008. 35 articles were identified the initial search and 8 were excluded leaving 27 articles for full review and analysis. Conceptual and methodological issues identified across the studies resulted in an inability to recommend any of the instruments used in the present studies for use without further study. These issues identified included: a lack of conceptual framework for construct validity, variability in injury characteristics, issues with sampling methodology, a lack of longitudinal designs, comparison group issues and an inability to compare instruments across studies.
More than 1.4 Million people sustain a traumatic brain injury (TBI) in the US each year and more than 5.3 Million Americans are living with long-term disability following TBI (Injury Fact Book, 2002). Traumatic brain injury may result in persistent cognitive and communication problems that vary depending on the location and severity of the injury. Symptoms commonly include: difficulty concentrating, impaired judgment, impaired memory, problems with decision making and problem solving, word-finding difficulties and inappropriate social behavior. Approximately 40% of TBI survivors report at least one unmet service need at 1 year post injury (Heinemann, Sokol, Garvin, & Bode, 2002). In order to attempt to meet unmet needs following injury in TBI patients, increased responsibilities may be placed on the significant other/family than prior to injury.
Head trauma can affect anyone at any age but older adults are particularly vulnerable to injuries from falls and traffic accidents. In persons 65 years of age and older, traumatic brain injury (TBI) is responsible for over 80,000 emergency department visits each year and adults age 75 years or older have the highest rates of TBI-related hospitalization, thus they are a large and growing population within the TBI community (Langlois, Rutland-Brown, & Thomas, 2004; Department of Health and Human Services., 2000).
When one partner in an intimate relationship experiences a sudden injury resulting in temporary or permanent disability, such as a TBI, the relationship is stressed by the events. In some cases; however, these stressors strain the couple’s relationship to the breaking point, and may result in higher rates of separation and divorce among TBI survivors (Wood & Yurdakul, 1997). Research to date has primarily focused on younger and middle aged adults and little is known about older adults following TBI. A single qualitative study was identified from the literature about the experience of using the multiple case study approach Layman and colleagues (2005) focused on the experience of older partners of persons with TBI. The authors identified Relatedness and Relationship Persistence as the two primary themes of interest and noted that there was an inability of couples to discern normal aging changes versus TBI-related changes (Layman, Dijkers, & Ashman, 2005).
Thus, given the paucity of available research for caregivers/family member functioning of older adult TBI survivors, it is clearly an area warranting further attention. Well-validated instruments of caregiver role functioning are available from the geriatric literatures and include the Kingston Caregiver Stress Scale (Hopkins, Killik, & Day, 2007) the Modified Caregiver Strain Index (Thornton & Travis, 2003) the Zarit Caregiver Burden Scale (Zarit, Reever, & Bach-Peterson, 1980). However, these were originally developed for use in caregivers of persons with dementia, not TBI. As to date there is only a single article available specific to older adults TBI survivors and family functioning, a review of the literature was consequently undertaken to identify available instruments used to assess caregiver/family member functioning of TBI survivors of all ages.
Is there an optimal measure of family functioning identified from the TBI literature for use in future studies of TBI survivor family member/caregivers? Is this identified measure appropriate for use with older adult TBI survivor family member/caregivers?
To answer the research questions of interest, a review of the literature was undertaken using the following databases: CINAHL Plus, Family Systems Abstracts, and Pubmed. The databases were searched from 1998–2008 using the MeSH terms: Brain Injuries; caregivers; questionnaires and the following restrictions: English; Research Articles; Full Text Available.
From the initial search, 35 full-text articles were identified. Following a review of the article abstracts, 8 were excluded from the initial search because they were a) qualitative interviews and/or b) studies that used caregiver factors to predict of TBI survivor outcome and this was not the focus of research question. 27 articles remained for full review and analysis (See Table 1).
The unit of analysis in the majority of studies (n=21) was the individual caregiver or family member. Very few studies of caregivers have focused on a particular type of individual eg. spouse in these studies. Thus there is a high degree of variability in the type of family members represented within studies. Some studies have focused on the dyad (n=4) of the TBI survivor and the caregiver/family member (Carnevale, Anselmi, Busichio, & Millis, 2002; Ergh, Rapport, Coleman, & Hanks, 2002; Ponsford, Olver, Ponsford, & Nelms, 2003; Wells, Dywan, & Dumas, 2005); but only a single group (Gan, Campbell, Gemeinhardt, & McFadden, 2006; Gan & Schuller, 2002) has focused on the family system, which is a framework that centers on the family as a whole and the interactions within the family, rather than an individual member. Using a theoretical approach that views the family holistically, is logical as the sequelae of TBI can be far reaching beyond any individual (Gan & Schuller, 2002).
From this review, there were conceptual and methodological issues identified across the studies. This resulted in the inability to recommend any of the instruments used in the present studies for use without further study. These identified issues included: a lack of conceptual framework for construct validity, variability in injury characteristics, issues with sampling methodology, a lack of longitudinal designs, comparison group issues and an inability to compare instruments across studies.
There was a lack of an explicit conceptual framework present in the majority of studies (Table 1 and Table 2). This absence of a clear framework was then manifested in a wide range of concepts of family functioning presented by various authors (See Table 2) across studies as outcomes of interest. The concepts presented in the various papers could be classified as having positive, negative, or neutral connotations for functioning. This lack of conceptual clarity likely influenced other issues such as design, comparison and interpretation (Hutchison, 1999; O'Reilly, 1988).
Many issues frequently encountered in the cross- sectional studies centered around the sampling methods (e.g. multiple injury severities, the wide range of time since injury in many cross sectional studies, and inclusion of child/parent/siblings in single studies). Injury severity may play a large role in the type of sequelae that result following TBI and thus influence the caregiver/family’s function over time. The studies reviewed in this paper often included wide ranges in brain injury severity indicators such as post-traumatic amnesia and length of unconsciousness (e.g. Wells et al., 2005) which may have influenced these but were not used as covariates in the analyses. Time since injury also varied widely in the majority of the cross-sectional designs; in some cases from a few months to up to 30 or 40 years post-injury (Katz, Kravetz, & Grynbaum, 2005); Wells et al., 2005). While this may have allowed for larger sample sizes, it is not representative of a population, thus the ability to draw any real inference or to identify an effective intervention is significantly reduced. In only a few studies was the cross-sectional study designed to assure sampling a temporally similar group of caregivers (Kolakowsky-Hayner, Miner, & Kreutzer, 2001; Marsh, Kersel, Havill, & Sleigh, 1998a, 1998b; McPherson, Pentland, & McNaughton, 2000). Lastly, for several of the studies (e.g. (Marwit & Kaye, 2006; McPherson, Pentland, & McNaughton, 2000), the inclusion criteria were family members or caregivers of persons with acquired brain injury which is not exclusive to TBI, but also includes chronic or pathological injury such as stroke. The use of these broad inclusion criteria was stated by the authors to be deliberate to increase sample size (Murray, Maslany, & Jeffery, 2006). For most analyses; however, this actually represents multiple perspectives of multiple types of recovery experiences, rather than a single uniform analysis.
Often the samples selected were not an optimal match for the research question of interest; frequently the samples were convenience samples from secondary sources (See Table 1-Study Design; Sample) or from patients presenting in clinic. While the stated purpose of many studies was to elucidate family members’ or caregiver’s experiences, they did not approach this from multiple persons within a single family, so the study could only really attempt to describe the experience of a single family member’s experience. Also, in using a convenience sample often the defining characteristics were ill-defined, e.g. in many samples “Frequent close contact” was required, but this was not explicitly defined and could vary widely based on the subjects’ interpretation; this differed from other studies in which the subjects were required to reside in the same household to define “family”. This could have greatly influenced the types of responses.
In many of the studies, despite the use of a cross-sectional design and the use of convenience sampling, the numbers obtained were relatively small. The sample sizes varied from 28 to 249. In the larger sample sizes, these were generally mixed populations of various types of family members, with wide age ranges, various types of injury (mild and severe in same group), so again the issue of multiple perspectives of multiple types of recovery experiences pooled together is raised as study limitations.
The majority of studies (n=23) available were cross-sectional designs. A single study (Marsh, Kersel, Havill, & Sleigh, 2002) has examined the individual experience of individual primary caregivers of adults discharged from rehabilitation services participating in the TBI model systems program. The authors reported that there was an adaptation of the caregiver in the period from 6 months to one year post-injury. They also found that social isolation and behavioral problems of the TBI patient were predictive of burden. Limitations in the cross-sectional designs again include the times chosen to report outcomes of family members/caregivers varied widely based again primarily on convenience and included unusual timepoints e.g. 40 years post-injury (See Table 1).
In several studies the comparison group selected included rehabilitation professionals (Man, 2002) or professional caregivers (Godfrey et al., 2003). The selection or inclusion of these groups is not particularly informative in describing caregiver functioning as defined by the investigators. Frequently, the comparison group selected was convenient, but not concordant with the research question of interest. An additional issue was that researchers made temporal assertions based on comparisons of differing times since injury of various families/caregivers in cross-sectional studies. These are not valid comparisons to make.
Many researchers sought to develop or validate their own family/caregiver functioning instrument for use in TBI and specifically developed the instrument for the study reported. Seven of the 27 studies reviewed were testing new instruments, often with insufficient data provided on reliability and validity. Across the 27 studies more than 50 different instruments (See Table 1-Instruments) were used to measure family/caregiver functioning and only rarely (e.g. Beck Depression Scale, Family Needs Questionnaire, Caregiver Appraisal Scale) was an instrument used in more than one study, thus there is an inability to compare instruments across studies of family functioning in TBI.
In future cross-sectional studies, it will be particularly useful to evaluate a particular family member’s/caregiver’s perspective of the recovery experience of the same level of TBI severity at a similar time-point post-injury. In designing or reporting future studies, when data is obtained from intake assessments, more information would be particularly useful in interpreting findings in relationship to generalizability to those persons who do not seek treatment for TBI or qualify for rehabilitation services. This work would be better done with a clear time of assessment post-injury defined (e.g. 2 years post injury when most plasticity and recovery has occurred or within 1 year of injury when the adjustment and service use is really the greatest). In particular, a prospective longitudinal study that enroll families near the time of injury and includes an assessment of pre-injury functioning as a baseline measure would be especially useful.
This review is limited in that only published articles available on-line in English were reviewed from particular databases and gray literature (e.g. dissertations) were not included, thus some bias may have been introduced. Attempts to reduce bias in this review were maintained via the use of clear questions to guide the literature review, a threshold for inclusion of studies and systematic methods evaluate the research literature.
The lack of conceptual clarity within the field of family functioning in TBI has resulted in a lack of consistent use of terminology and has led to the use of more than 50 instruments across various studies and the continual development of additional, yet poorly justified tools. There is a clear need to conduct an evolutionary concept analysis of family functioning in TBI and to gain a lucid, comprehensible definition of the idea prior to continuing additional work in this area. Based upon this review, the author was unable to answer the second research question since there was insufficient data to be able to recommend any TBI-specific instrument for use with caregivers/family members, let alone recommend its’ with use of older populations. The use of well-validated family function and caregiving instruments from other fields, such as geriatrics, is recommended in the interim for ongoing and planned family TBI research with older adults. These studies should focus on longitudinal analyses of a family functioning within the context of a focused TBI population (e.g. mild brain injury). Additionally, these studies should account for differences in family development in their sampling structures (e.g. children of parents with TBI should be examined separately from spousal issues) until there is a clear understanding of these issues. Once a better understanding of family functioning within the family system has been obtained, comparison across these populations can occur.
The author thanks Dr. Karen Schepp for thoughtful discussions regarding this work. This work was funded, in part, by the Building Geriatric Academic Nursing Capacity Program, John A. Hartford Foundation and the NIH Roadmap for Medical Research, KL2RR025015-01.