Abstract
The experience of clinically significant anxiety and anxiety disorders represent significant and often debilitating problems for many residents in long-term care (LTC) settings. However, anxiety problems often go undetected and untreated in this growing population. The purposes of this paper are to examine the prevalence and impact of anxiety problems among residents in LTC facilities, describe the efficacy of the current instruments that are used to detect anxiety in these settings, and provide clinical guidance for the thorough assessment of anxiety. Regarding measurement tools, the GAI, GAI-SF, GAS-LTC, and the BADS are the only self-report measures designed for older adults that have been successfully validated for use with older adults living in LTC settings. Clinicians should focus on ways to educate LTC directors and staff to emphasize the importance of screening and assessing for anxiety, using validated measures, to improve the assessment and treatment of anxiety in their residents.
Anxiety disorders are within the top three most common mental disorders among older adults (Bryant et al., 2008; Edelstein et al., 2011; McCombe et al., 2018). Although many older adults are able to live independently well into late adulthood, physical and cognitive deficits may make this impossible for some notably frail and ill older adults, leading to long-term care (LTC) institutional placement. Anxiety disorders and subsyndromal anxiety symptoms represent significant challenges among many older LTC residents, yet anxiety symptoms and disorders often go undetected and untreated in this growing population (Creighton et al., 2018). One leading reason for this lack of detection and lack of subsequent intervention is that there are insufficient screening and assessment instruments for anxiety specifically designed and validated for older adults in diverse LTC settings (Gerolimatos & Edelstein, 2012). Nursing homes and assisted living communities often house residents with significant cognitive, social, and physical impairments which may impact their scores on traditional anxiety questionnaires, especially those that were not developed with older respondents in mind. As such, the use of properly validated measures for detecting and measuring anxiety in LTC settings is vital. The purposes of this paper are to provide evidence-based clinical guidance for professionals working within LTC settings, by furnishing information related to the prevalence and presentation of anxiety disorders for older adults residing in LTC, and to describe the efficacy of current instruments that are used to detect anxiety in these settings. This is not a comprehensive review of the literature on the topic of anxiety in LTC settings. Rather, our aim is to condense, summarize, and highlight key information from the literature that is most relevant for clinicians working in these settings. We conclude by offering additional practical considerations and suggestions for detecting anxiety in LTC settings. The guidance in this article is derived from our combined professional clinical experiences and from empirical evidence gathered by the authors.
Anxiety in Long-Term Care Residents
Older adults may choose to leave independent living situations for many reasons, but usually make the decision (or are forced to relocate) as sensory, cognitive, or physical challenges begin to accumulate and significantly interfere with their ability to independently complete the essential activities of daily living (ADLs; Segal et al., 2018). In cases such as these, many older adults subsequently enter supportive LTC living situations. LTC includes an array of supportive settings, including assisted living, skilled nursing homes, home health care, and adult day programs, which all differ in the degree of assistance provided to residents (Segal et al., 2018) and in the setting where care is delivered. In some cases, older adults may dwell within continuing care retirement communities (CCRCs), which are facilities with two or more types of living arrangements (e.g., independent apartment living, assisted living, skilled nursing home). These joint facilities allow older adults to age in-place and transition to a higher level of care (i.e., assisted living to skilled nursing care) when physical or cognitive burdens become too high. Most commonly, moves into LTC settings are due to cognitive impairment—even if a person’s physical health is in reasonably good condition—as cognitive impairment undermines the ability to complete basic ADLs (Segal et al., 2018). The U.S. Department of Health and Human Services estimates that 70% of people who live past age 65 will need long-term services and supports, and that 48% will spend some time in formal LTC (Johnson, 2019). The sheer number of older people who will utilize LTC services is expected to increase substantially over the next few decades due to the projected increase in number of people aged 65 and older in the United States (Vespa et al., 2020).
A common mental health problem among residents in LTC settings is anxiety. In fact, in their comprehensive review of anxiety prevalence rates in older adulthood, Creighton et al. (2016) estimate that 3%–20% of nursing home residents experience at least subsyndromal anxiety, with many having symptoms severe enough to warrant a formal anxiety disorder diagnosis. This is comparable to point-prevalence rates of anxiety disorders in community-dwelling older adults, which range from 1.4% to 17% (Creighton et al., 2016). Moreover, it is estimated that about 24% of community-dwelling older adults experience subsyndromal anxiety that is still meaningful and impacts quality of life, but is below a diagnostic threshold (Bryant et al., 2008; Edelstein et al., 2011). These estimates are somewhat variable, as methodology and sample sizes differ significantly across studies.
Older adults in LTC are at an elevated risk for the development of anxiety problems due to factors unique to these settings and individuals living in these settings, including higher levels of loneliness, increased frailty, and increased dependence on others for basic needs (Smalbrugge et al., 2005). Despite prevalence data suggesting that anxiety disorders are quite common for nursing home residents, formal diagnoses are rarely made in these settings (Creighton et al., 2018; Drageset et al., 2013). Reasons for this appear to be systemic, as many nursing homes operate on a medical model and prioritize medical care, often focusing less on quality of life (Segal et al., 2018). Moreover, staff training in mental health and psychological services provided to residents in nursing homes are generally considered to be inadequate (Grabowski et al., 2010; Molinari et al., 2008). Nursing homes also tend be understaffed and underfunded, especially those that service lower-income individuals, with some suggesting that Medicaid does not cover the full cost of care (American Health Care Association, 2018; Shipman & Hooten, 2007). This makes it difficult for nursing home staff to have the time or support to effectively screen for and assess anxiety symptoms among residents.
With a sample of 1.3 million nursing home residents, Jester et al. (2018) found that the prevalence of anxiety disorders upon admission to nursing homes is lower with increasing age (e.g., 18–49: 18%, 50–64: 18%, 65–84: 15%, 85+: 13%). It is possible that older cohorts may have more medical or physical conditions (Jaul & Barron, 2017), which may cause nursing home staff to focus less on psychological risk factors, given the emphasis on the medical model within nursing homes. Additionally, there is evidence to suggest that anxiety disorders lessen in intensity with increased age (Nilsson et al., 2019), and therefore the oldest-old adults may be more likely to have subsyndromal anxiety symptoms. In some LTC settings, it is possible that mental disorders such as depression and anxiety in residents are viewed as less important to assess for and treat than medical disorders or diseases. However, anxiety symptoms are often quite debilitating for older adults in LTC and may co-occur with depression, which we discuss next.
Comorbid Anxiety and Depression
The interplay between anxiety and depression in older age is complex, but it is clear that they often co-occur at both subthreshold and diagnostic levels. Braam et al. (2014) suggest that the etiology of comorbid anxiety and depression in older age may be explained by: (a) true dual-diagnoses of depression and a specific anxiety disorder (e.g., panic disorder, specific phobia), (b) the presence of somatic symptoms that contribute to the development of both anxiety and depression, (c) increased anxiety symptomatology during a depressive state due to feelings of loss of control, or (d) a combination of the previously mentioned. This heuristic may describe why community-dwelling samples with depressive disorder diagnoses are found to have high rates of anxiety symptomatology as well.
In a population study examining the prevalence of comorbid anxiety and depression in community-dwelling older adults, Kvaal et al. (2008) found that 2% of older adults had comorbid anxiety and depressive disorders, and it was more common to have subthreshold symptoms of both anxiety and depression (14%). Anxiety symptomatology appears to be more common in older adults with a depressive disorder (87%) compared to those with subthreshold depression (67%; Braam et al., 2014). Additionally, comorbid anxiety and depression in older adults differs across anxiety disorder diagnoses. For example, Canuto et al. (2018) found that of seven anxiety disorders assessed (agoraphobia, generalized anxiety disorder [GAD], obsessive-compulsive disorder, panic disorder, posttraumatic stress disorder, social phobia, specific phobia), only panic disorder and specific phobia were related to major depression in community-dwelling older adults. Comorbid anxiety and depression may also be compounded by cognitive decline. Bierman et al. (2007) found an increased prevalence of comorbid anxiety and depressive symptomatology in those with moderate (9.5%) or poor (8.3%) cognitive functioning in comparison to those with good cognitive functioning (4.3%) when measured by the Mini Mental State Exam (MMSE). These findings showcase the unique presentation of anxiety and depression in older age when examining community-dwelling populations.
In nursing home settings, comorbid anxiety and depression may be misreported in medical records. Drageset et al. (2013) estimated the true prevalence of comorbid anxiety and depression to be twice as great (10%) as what was reported in medical records (5%) in a sample of 227 residents. Goyal et al. (2018) reported a prevalence rate of comorbid anxiety and depression at 17.6% at baseline and at 19.5% 12 months later in nursing home residents. Moreover, the remission rate for nursing home residents with both anxiety and depression was low at 16.7% in comparison to residents with only an anxiety disorder (57.6%) or only a depressive disorder (55.6%), suggesting long-term persistence of the psychiatric comorbidity. In a practical sense regarding assessment, due to this high comorbidity we suggest that LTC residents who are depressed should be assessed for anxiety, and conversely, that LTC residents who are anxious should be assessed for depression.
Subsyndromal Anxiety
A recent study was the first in many years to systematically examine the prevalence of anxiety disorders and comorbid depression in nursing home settings. Creighton et al. (2018) examined a sample of 180 nursing home residents and found prevalence rates to be at 19.4% for diagnosable anxiety disorders and 11.7% for subthreshold anxiety disorders. The most common threshold anxiety disorder among this population was GAD, with 11.1% of participants meeting criteria and 3.3% having subthreshold GAD symptoms. The prevalence of participants with any threshold anxiety disorder was within the upper estimates reported by previous studies (Smalbrugge et al., 2005). Despite finding a significant number of nursing home residents to be living with either threshold or subthreshold anxiety symptoms, less than half of these participants had any kind of anxiety symptom indicated in their medical records (Creighton et al., 2018). Furthermore, only 8.6% of those with threshold anxiety symptoms were receiving any kind of psychological treatment, such as psychotherapy, for their anxiety. No participants with subthreshold anxiety symptoms were receiving psychological treatment (Creighton et al., 2018). This highlights the significant under-detection and lack of treatment for nursing home residents who are experiencing clinically relevant anxiety.
A possible reason for more people meeting the threshold for anxiety disorders rather than subthreshold anxiety in the above study may be the use of the newer Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) criteria, which removed the GAD requirement for the individual to consider the anxiety to be excessive or disproportionate (American Psychiatric Association, 2013; Creighton et al., 2018). Nursing home residents often have less insight due to cognitive impairment, and therefore may not recognize the significant impact of their anxiety on their quality of life (Creighton et al., 2018). Thus, removal of this criterion may have allowed more residents to meet full criteria for diagnosis. Moreover, nursing home settings may also exacerbate existing anxiety, so that residents who may have been at a subthreshold level in the community may experience an increase in symptom burden in the nursing home, allowing them to meet the threshold for an anxiety disorder. In light of the burgeoning literature on comorbid anxiety and depression in older age, the Advisory Committee to the Lifespan Disorders Work Group suggested revisions to the DSM-5. The committee recommended development of additional age-sensitive guidance on the assessment of comorbid anxiety and depression, the assessment of anxiety in light of comorbid (nonpsychiatric) medical conditions, the assessment of anxiety in the context of Alzheimer’s disease and related dementias, and thoughtful development of age-specific instruments to reliably and validly measure anxiety (Mohlman et al., 2012).
Health Effects of Anxiety and Depression in LTC
The evidence that anxiety is prevalent within LTC settings is compelling, and is made even more so when we consider how anxiety negatively impacts the health and well-being of older adults. Several studies have demonstrated associations between anxiety and several medical comorbidities for older adults, including gastrointestinal problems (El-Gabalawy et al., 2011), diabetes (Smith et al., 2013), musculoskeletal pain (de Koning et al., 2018), respiratory problems (Panagioti et al., 2014; Vögele & von Leupoldt, 2008), and cardiovascular disease (Celano et al., 2016; Smoller et al., 2007; Tully et al., 2008). These disorders have overlapping symptoms with anxiety, including restlessness, poor sleep, fatigue, shortness of breath, increased heart rate, upset stomach, and muscle pain and tightness. This symptom overlap creates difficulty with diagnosis and determining whether the symptom is due to anxiety or another medical condition. Anxiety has also been identified as a predictor for increased risk of death following heart surgery (Tully et al., 2008), and panic attacks specifically are associated with greater cardiovascular comorbidities and greater likelihood of mortality related to cardiovascular conditions (Smoller et al., 2007).
Studies also suggest that anxiety in older adults is a risk factor for the development of cognitive decline and dementia (Gallacher et al., 2009; Palmer et al., 2007). Several medications which are often given to older adults to manage other comorbid health conditions may also contribute to the development or maintenance of anxiety symptoms and disorders, including certain antidepressants, antispasmodics, and muscle relaxants, among others. These medical conditions, medications, and other considerations which may contribute to anxiety symptoms in older adults are listed in Table 1. Indeed, understanding the role of medications and medical conditions is a critical aspect of differential diagnosis. While these results have largely been found in community-dwelling older adults, residents in LTC likely experience similar cognitive and health outcomes related to anxiety.
Medications and Medical Conditions That Might Cause or Exacerbate Anxiety.
*A differential diagnosis of all related psychiatric (e.g., depression, psychotic disorders, personality disorders) and neurological disorders (e.g., Alzheimer’s disease and related dementias, Parkinson’s disease, sensory loss) should be investigated.
Within LTC, the health effects of anxiety disorders and symptoms have been studied less frequently than the health effects of depression. Results from two studies indicate that anxiety symptoms can be especially debilitating for older adults in LTC given their increased risk of physical and cognitive decline, which may impair their ability to cope with and manage anxiety symptoms (Drageset et al., 2013; Smalbrugge et al., 2006). Residents of LTC with anxiety symptoms have significantly impaired quality of life and present a much more substantial caregiving burden than their nonanxious counterparts (Drageset et al., 2013). Older adults in LTC with anxiety are also at a higher risk for hospitalization and dementia (Drageset et al., 2013), which significantly increases health care costs for the residents and the LTC facility. Anxiety also puts nursing home residents at greater risk of death (Drageset et al., 2013).
Smalbrugge et al. (2006) examined the impact of depression and anxiety on well-being and the use of health care services in 14 nursing homes in the Netherlands. Findings suggest that residents with depression and/or anxiety (this variable included those with comorbid anxiety and depression, pure depression, and pure anxiety) rated their well-being to be 25% lower compared to residents without depression or anxiety, as measured by the Philadelphia Geriatric Center Morale Scale. Additionally, residents with depression and/or anxiety required less assistance with ADLs but had higher use of medical specialist consultation and received more medications (Smalbrugge et al., 2006). Although studies on the impact of anxiety among LTC residents are relatively scarce, the findings of these studies demonstrate that anxiety negatively impacts health outcomes, increases risk of death, and increases health care costs. These findings highlight the urgent need for improved assessment of anxiety in these settings, so that individuals who are experiencing anxiety in LTC can be more effectively identified and treated.
Assessment of Anxiety in Older Adults
With the increasing numbers of older adults in nursing homes and other LTC settings, high prevalence rates of anxiety symptoms, and the significant negative impact of anxiety on well-being and quality of life comes the crucial need for better assessment and treatment of anxiety in this population. Psychological assessment for mental disorders among older adults is particularly challenging, as symptoms for many mental disorders often present differently among older people than the young adults for whom assessment measures and diagnostic criteria are largely based (Edelstein et al., 2008; Mohlman et al., 2012). Measure sensitivity is compromised when used on a population for which the measures were not designed for nor validated with, decreasing the ability of these measures to accurately detect anxiety. Ideally, researchers and clinicians would have measures specifically designed and validated for the assessment of anxiety in LTC residents. Currently, there are several options for assessing anxiety in LTC that were designed for use with older adults, as well as a few designed specifically for residents of hospitals and LTC settings. The most common self-report measures for assessing anxiety in LTC are discussed next, as well as the efficacy of their use in LTC settings.
Geriatric Anxiety Inventory (GAI)
The Geriatric Anxiety Inventory (GAI) is a self-report measure consisting of 20 items, inquiring about various anxiety symptoms over the past week with a dichotomous “agree/disagree” format. The GAI was developed by Pachana et al. (2007) to help address the lack of anxiety assessment measures designed for older adults. Items for the GAI were developed from common items on existing anxiety scales and were designed to capture common themes across anxiety measures. Initial validation of this measure was conducted with two samples of healthy community-dwelling older adults (n = 452). The GAI showed excellent internal consistency, with a Cronbach’s α of .91, and was significantly correlated with other common measures of anxiety, including the Beck Anxiety Inventory (BAI), the Penn State Worry Questionnaire (PSWQ), the Geriatric Anxiety and Depression Scale (GADS), the State Trait Anxiety Inventory (STAI), and the Positive and Negative Affect Schedule. In the original validation study, Pachana et al. (2007) identified an optimal cut-score of 10/11 for identifying a potential anxiety disorder. Since its initial validation, the GAI has performed well and has been validated for use in various older adult populations (Gould et al., 2014; Johnco et al., 2015). Additionally, a 10-item version of the GAI, called the GAI short form (GAI-SF; Byrne & Pachana, 2011), was created for a more brief anxiety assessment. The GAI-SF performs similarly to the GAI among community-dwelling older adult samples (Byrne & Pachana, 2011).
Validation in LTC
To date, there have been three published studies examining the use of the GAI in LTC settings. Boddice et al. (2008) examined the psychometric properties of the GAI among a small sample of older adults residing in skilled nursing care. Participants completed the modified MMSE, the GAI, and the anxiety and depression components of the Composite International Diagnostic Interview (CIDI) as a diagnostic screen for anxiety disorders and depression. From this sample, all participants who were found to meet criteria for an anxiety disorder also scored above the cut-score on the GAI, and there was a significant positive association between the GAI and the CIDI.
Gerolimatos et al. (2013) evaluated the use of the GAI and the GAI-SF among 75 residents (age range: 52–94 years) of a state-operated nursing home. This study involved the review of residents’ medical records for completed GAI measures, as the nursing home had been using this measure for anxiety screening since 2008. If GAI measures were found, the residents then completed additional measures such as the Geriatric Depression Scale (GDS), the Adult Functional Adaptive Behavior Scale (AFABS), and the Executive Interview (EXIT). Diagnoses of anxiety disorders were made based on GAI scores and a clinical interview. Using the cutoff score for the GAI suggested by Byrne and Pachana (2011), 36 of 75 residents (48%) were determined to have clinically significant anxiety symptoms. Internal consistency for GAI scores was excellent in this sample, with a Cronbach’s α of .92. The GAI-SF had a slightly lower Cronbach’s α of .73, but still within acceptable range. Researchers assessed convergent validity with Pearson correlations between the GAI and the GAI-SF and found a significant positive correlation between the two measures, r = .89, which is to be expected since they share similar items. Divergent validity was supported by weak correlations between the GAI and GAI-SF with the AFABS (GAI: r = −.29; GAI-SF: r = −.24) and EXIT (GAI: r = .21; GAI-SF: r = .19). Gerolimatos et al. (2013) identified an optimal cut-score of 9 for the GAI when used within LTC settings.
Another study compared psychometric properties of the GAI, the Hospital Anxiety and Depression Scale (HADS), and the Rating Anxiety in Dementia scale (RAID) among 120 LTC residents (Creighton et al., 2019). The GAI demonstrated excellent internal consistent reliability (α = .95) and was positively associated with the RAID and the HADS, indicating good convergent validity. This study also examined how cognition impacts the psychometric properties of these scales by splitting participants into two groups of either no cognitive impairment or mild cognitive impairment based on scores on the MMSE. Researchers did not find any significant differences in the psychometric properties of the GAI between participants with normal cognition and those with mild cognitive impairment. From this study, Creighton et al. (2019) recommended an optimal cut-score of 9, consistent with Gerolimatos et al. (2013), as indicative of a likely anxiety disorder for residents of LTC. The sensitivity of this cut-score for the GAI in this study was 90%, with a specificity of 86.3%.
Critique
Whereas these studies provide strong evidence for the clinical utility of the GAI and GAI-SF in nursing homes, some study limitations need to be addressed. First, sample sizes in Gerolimatos et al. (2013) and Boddice et al. (2008) studies were fairly small, which limits the generalizability of the findings. Additionally, Boddice et al. (2008) did not include measures of divergent validity, nor was test–retest reliability examined. While Boddice et al. (2008) did include a diagnostic measure for anxiety disorders, they did not statistically examine or report the sensitivity and specificity of the GAI for diagnosing anxiety disorders in this sample. In the Gerolimatos et al. (2013) study, the sample was not randomly selected, but rather it only included patients in the nursing home who had previously been administered the GAI. This may have skewed the sample to more likely include residents with anxiety disorders or significant anxiety symptoms. Another issue is that the convergent validity was evaluated with two versions of the GAI, the GAI and the GAI-SF. While the convergent validity was excellent, this is to be expected as the items that make up the GAI-SF were derived from and are included on the full GAI. Evidence of convergent validity for this population may be more compelling if it had been evaluated by another common measure of anxiety, rather than the short form of the same measure.
Creighton et al. (2019) appeared to address some of the limitations, by including a larger sample size (n = 180) and including several measures of convergent validity. This study also included an objective measure of cognition, which showed that the GAI performs well among those with cognitive impairment. However, this study has some limitations as well, such as the lack of inclusion of a depression measure, which is important to include when examining anxiety symptoms given the strong symptom overlap of these constructs. Additionally, this study excluded participants with severe cognitive impairment that may have impacted their ability to answer the self-report questionnaire. Therefore, results cannot be generalized to more cognitively impaired individuals beyond mild cognitive impairment. Finally, the GAI tends to have a heavy content focus on symptoms of worry, which could be problematic in LTC because older adults may be more likely to endorse somatic anxiety symptoms rather than worry symptoms (Gerolimatos & Edelstein, 2012).
Discussion
Despite these limitations, the GAI appears to be a reliable and valid instrument for the assessment of anxiety in LTC. This measure consists of 20 items that tap in various anxiety symptoms, specifically worry symptoms. The response format is a simple, agree/disagree format, which may be preferable in LTC settings where higher rates of cognitive impairment may cause difficulty answering more complex answer formats, such as Likert-type scales. The GAI has strong psychometric properties across many older adult samples and may be preferred over measures with less empirical support. The GAI also has support for use with residents with at least mild cognitive impairment, which is advantageous when working with older adults in LTC.
Geriatric Anxiety Scale (GAS)
The Geriatric Anxiety Scale (GAS) is another commonly used assessment measure for anxiety in older adults. This measure was designed to assess anxiety severity specifically among older adults (Segal et al., 2010). There are three versions of this measure, the standard GAS (with 30 items), a shortened version with 10 items, called the GAS-10 (Mueller et al., 2015), and a 10-item version created for use in LTC, called the GAS-LTC (Pifer & Segal, 2020). Both the GAS and the GAS-10 have strong evidence of reliability and validity for use with diverse samples of community-dwelling, medically ill, and treatment-seeking older adults (Mueller et al., 2015; Segal & Mueller, 2019; Segal et al., 2010). The GAS-10 was created using item response theory (IRT) to aid in ease of use, as short forms are often preferred in busy clinical settings and they reduce the burden of administration and scoring time, as well as pose less burden on respondents (Mueller et al., 2015). Both the GAS (α = .93) and the GAS-10 (α = .89) had good to excellent internal consistency of scale scores within a sample of community-dwelling older adults, and they were highly positively correlated (r = .96, p < .001), as expected due to item overlap.
The GAS-LTC was recently created to help address the gap in assessment tools created and validated for older adults in LTC residents (Pifer & Segal, 2020). It was developed using the 10 items on the GAS-10, although several items were slightly modified to accommodate the lower level of cognitive functioning characteristic of many nursing home residents. The response format was also changed from an ordinal Likert-type scale to a dichotomous Yes/No response. Items tap into cognitive, affective, and somatic symptoms of anxiety, and higher scores indicate higher levels of anxiety.
Validation in LTC
The GAS-LTC is a recent measure; therefore, there is only one published study examining psychometric properties of this measure in LTC (Pifer & Segal, 2020). Initial examination of the psychometric properties for the GAS-LTC was examined with 66 LTC residents (M age = 84.4 years, range = 59 to 100 years; 74.2% women). Results showed good internal consistency (α = .80), good convergent validity with another measure of anxiety, the GAI (r = .70, p < .01) and with a measure of depression, the GDS-15 (r = .67, p < .01; Pifer & Segal, 2020). Due to the high levels of cognitive impairment in some LTC residents, the impact of greater subjective memory impairment on the reliability of the GAS-LTC scores was examined. Notably, the reliability coefficients did not significantly vary between groups with high (α = .75) and low (α = .77) self-reported memory impairment (Pifer & Segal, 2020). The GAS and the GAS-10 have not yet been specifically examined or validated for use in LTC settings.
Critique
Although the initial study of the GAS-LTC (Pifer & Segal, 2020) demonstrated promising initial psychometric support for the use of the measure in LTC settings, there are several limitations to address. First, this study did not include an objective measure of cognitive ability such as the required Brief Interview for Mental Status (BIMS) in the Minimum Data Set 3.0 for all certified nursing homes (Saliba et al., 2012). As such, it is unknown how increased levels of objectively measured cognitive impairment might impact the psychometric properties of this measure. This represents an area of further research of the GAS-LTC. Additionally, this study did not include formal anxiety disorder diagnoses of residents, so documentation about the sensitivity or specificity of the measure in its ability to detect anxiety disorders was not provided. Test–retest reliability was also not examined in this study; therefore, conclusions cannot be made about the temporal stability of the GAS-LTC. Finally, this study did not include subjective or objective health measures. Given that there are several items on the GAS-LTC inquiring about somatic symptoms of anxiety, researchers were unable to parse out if these symptoms may be related to medical conditions instead of anxiety. This is important, as older adults often have multiple chronic health conditions (as noted earlier), which may impact their responses on the GAS-LTC and other measures.
Discussion
Overall, preliminary results indicate that the GAS-LTC is a promising assessment measure of anxiety for older adults in LTC settings, and further evaluation of this measure appears warranted. The GAS measures are potentially advantageous compared to other measures of anxiety because they broadly assess several different domains of anxiety symptoms, including cognitive, affective, and somatic symptoms of anxiety (Segal et al., 2010). Like the GAI, the GAS-LTC has a simple yes/no response format that is suitable for older adults residing in LTC. At this time, it is recommended that clinicians use the GAS-LTC when assessing anxiety in LTC, rather than the GAS or the GAS-10, as they have not yet been validated in this setting.
Brief Anxiety and Depression Scale (BADS)
The Brief Anxiety and Depression Scale (BADS) is an eight-item self-report screening tool designed to detect anxiety and depression in older adult LTC residents. It was specifically created to identify residents who could benefit from a more comprehensive evaluation of mood (Mansbach et al., 2015). The response format is a graded three-point Likert-type scale, with response options “no,” “somewhat,” or “yes.” The BADS was designed to be quick (less than 3 min), easy to administer, and can be administered by both clinicians and paraprofessionals. This scale was developed using diagnostic criteria for major depressive disorder (MDD) and GAD, and also includes somatic and behavioral symptoms. The BADS has two distinct depression and anxiety factors, developed using a principal component analysis (PCA) analysis. This measure is not diagnostic, but rather was designed to alert clinicians that a further, more extensive mood assessment is warranted.
Validation in LTC
Initial validation of the BADS was conducted with 277 older adult LTC residents (Mansbach et al., 2015), with 179 residing in a skilled nursing facility, and 98 residing in assisted living communities. Reliability was found to be acceptable, with a Cronbach’s α of .75 for an overall factor (Mansbach et al., 2015). Convergent validity was examined with other measures of anxiety and depression. As expected, the BADS anxiety factor was strongly and positively correlated to another measure of anxiety, the GAD-7 (r = 0.76, p < .001). Discriminant validity was demonstrated with a near zero correlation with a measure of cognitive ability, as measured by the Brief Cognitive Assessment-Tool (r = .07, p < .001). The anxiety factor had a cutoff score of 4, with scores below 4 indicating no presence of GAD, with sensitivity of 73% and specificity of 81% for a score above 4 identifying probable GAD (Mansbach et al., 2015). To date, this is the only study examining the psychometric properties of the BADS in LTC settings or otherwise.
Critique
The first limitation to Mansbach et al. (2015) is that the researchers did not examine the Cronbach’s α for each factor, despite only using the factor scales for analyses of convergent and divergent validity. This is potentially problematic, as the subscales were used individually for analyses; however, the authors do not provide support for their individual reliability. The researchers also included a measure of objective cognition, but they did not analyze how cognitive ability might have impacted reliability and validity of the measure. Due to the high level of cognitive impairment in LTC settings, it is important to examine how cognition might impact a resident’s ability to complete the measure, and if the psychometric properties of this measure suffer when residents have higher levels of cognitive impairment. The BADS also has a three-point Likert-type scale response format, which may be more difficult for some older adults with cognitive impairment to complete. Finally, the BADS did not examine the test–retest reliability of this measure. Therefore one cannot draw conclusions about the temporal stability of this measure.
Discussion
The BADS is a fairly new screening tool for anxiety and depression in older adults and it has several positive features, including the less than 3-min administration, great sensitivity and specificity for detection of anxiety and depressive disorders, and solid evidence for convergent and divergent validity. The measure includes items for both depression and anxiety, which makes sense given the strong overlap between these two constructs. This measure may be especially helpful to clinicians who are concerned about both anxiety and depression in a resident. However, perhaps for patients who have already completed depression measures or if depression is not a concern, it may be more useful for clinicians to use an assessment measure that examines anxiety on its own, such as the GAI or the GAS-LTC. This would reduce redundant information and ensure that the assessment captures primarily information about the resident’s anxiety. At this time, there has been little focus on continued validation of this measure since its initial validation in 2015. This appears to be a promising measure of anxiety and depression screening in LTC; however, additional studies are needed to support and extend the findings of this initial study.
Hospital Anxiety and Depression Scale (HADS)
The Hospital Anxiety and Depression Scale (HADS) is a 14-item self-report measure that includes seven items each for the anxiety subscale (HADS-A) and for the depression subscale (HADS-D). Items are rated on an ordinal Likert-type scale of 0–3 with higher scores indicating greater anxiety or depression. The score ranges on the HADS subscales are: 0–7, normal; 8–10, mild disorder; 11–14, moderate disorder; and 15–21 severe disorder. The HADS has well-established psychometric properties (Herrmann, 1997), although a systematic review of this measure revealed its underlying structure to be somewhat inconsistent across samples (Norton et al., 2013). Despite the HADS not being developed specifically for older adults, Creighton et al. (2019) identified the HADS as one of the second most commonly used anxiety assessments in the literature for older adult LTC residents. The HADS has demonstrated good psychometric properties in community-dwelling older adults and hospitalized older adults (Dozeman et al., 2012; Drageset et al., 2013). An optimal cut-score of 8 has been identified as indicative of a possible anxiety disorder on the HADS-A (Dozeman et al., 2012).
Validation in LTC
Haugan and Drageset (2014) examined the psychometric properties of the HADS in nursing home settings using a confirmatory factor analysis. This study examined two samples, one with significant dementia symptoms and one without such symptoms. Results showed that the HADS-A had poor Cronbach’s α values in both samples (those with significant dementia, α = .40; those without significant dementia, α = .56), indicating poor internal consistency. Therefore, the reliability of this measure for nursing home residents is questionable. Construct validity was also found to be problematic in the nursing home population, with poor model fit in the group without dementia, and poor reliability in both samples. Additionally, convergent validity was not established with another established anxiety measure. Rather, it was supported by good correlations with other constructs that could be related to anxiety, such as quality of life, sense of coherence, and social support. In other studies, the HADS-A showed adequate to excellent internal consistency (Drageset et al., 2013; Haugan, 2015).
Creighton et al. (2019) examined the psychometric properties of the HADS-A within a sample of 180 older adults LTC residents, the HADS-A demonstrated good internal consistent reliability (α = .80) and was positively associated with the RAID and the GAI, indicating good convergent validity. The psychometric properties of the HADS-A were not diminished when used with residents who had mild cognitive impairment compared to normal cognition on the MMSE. From this study, Creighton et al. (2019) recommended an optimal cut-score of 6, which resulted in a sensitivity of 90% and specificity of 80%. This cut-score is different than the recommended cut-score of 8 from previous studies (Dozeman et al., 2012), as this cut-score produced less desirable sensitivity and specificity indices.
Critique
Despite having positive psychometric support in other hospital and rehabilitation settings, it appears that the HADS may not perform satisfactorily within LTC settings. The HADS-A demonstrated low reliability in one sample (Haugan & Drageset, 2014), and the convergent and divergent validity of this measure within LTC have only been examined in one published study (Creighton et al., 2019). In a study comparing the HADS-A with other measures of anxiety in LTC settings, the HADS-A performed less well than the GAI on measures of internal consistency. The HADS also includes items that may not be relevant to older adults, as it was designed with a young and middle-aged adult sample (Zigmond & Snaith, 1983). Like the BADS, the HADS response format is also a three-point Likert-type scale, which may be confusing for older adults with cognitive impairment. Also similar to the BADS, this measure assesses both anxiety and depression, which may be preferred if depression is a concern, but may be less appealing if anxiety is the sole concern for that resident.
Discussion
The HADS is a well-known measure that is often used to assess anxiety and depression in hospital and rehabilitation settings. Results from these studies highlight the need to evaluate psychometric properties of well-validated measures when using them in unique settings for which they were not originally designed. As such, the HADS needs further validation and examination to gain a better picture of its performance in LTC settings, as results across studies have been inconsistent. Overall, it is recommended that clinicians opt for anxiety assessments that have more support for their use among older adults in LTC, such as the GAI, the GAS-LTC, and the BADS over the HADS.
Clinical Issues for the Assessment of Anxiety in Nursing Home Residents
Regarding the assessment of anxiety in new nursing home residents, there are several pragmatic issues to consider, and these are discussed next:
A comprehensive biopsychosocial evaluation should be conducted to assure the proper diagnosis, rule-outs, determinants, and lifespan course of anxiety. This should include an extensive psychiatric interview and history, particularly regarding whether the person has experienced anxiety symptoms prior to the current episode. In general, without treatment, anxiety is a chronic condition. Most people who are anxious in old age were anxious when they were younger (Lenze & Wetherell, 2011). Is this a first episode of anxiety or were there prior ones? If there were prior ones, what interventions or supports were effective in reducing anxiety?
The evaluation should specifically determine the circumstances that precipitated the person’s admission, and whether the admission was particularly stressful. Was this a voluntary or involuntary admission? Did the family pressure the new resident to be admitted? Does the resident agree with the need for nursing home care?
Did an interpersonal loss such as the death of a caregiving spouse precipitate admission? Is the resident grief-stricken? In this scenario, clinicians should carefully assess for Persistent Complex Bereavement Disorder (Complicated Grief)—a DSM-5 diagnosis in which extreme feelings of sadness, yearning, and preoccupation with the deceased exceed 12 months (American Psychiatric Association, 2013). Additional consideration should be given to grief-stricken residents who may be experiencing traumatic bereavement due to the nature of the death (e.g., suicide, homicide, mutilating injury).
Do family members and friends visit on a regular basis? Are some residents feeling anxious because they are lonely? Has the person ever had close relationships with family or friends? Does the resident feel abandoned by those who were relied on for direction in their life? Does the resident feel frightened because of a perceived lack of social support due to lifelong dependent personality traits?
Has the new resident received a recent physical examination? Is the resident health literate and knowledgeable about his/her medical status and treatments? Is the resident in pain? What are the resident’s comorbid medical conditions that could cause anxiety-like symptoms (e.g., chronic obstructive pulmonary disease)? Likewise, what medications are the resident taking? Has the resident’s medication regimen changed? Are there inappropriate medication combinations that should be reviewed? As noted earlier, many medications and medical conditions are known to cause or exacerbate anxiety symptoms. As such, a thorough assessment in this area is critically important (Table 1).
What is the resident’s cognitive status? Did a recent medical event (e.g., stroke) impair the ability to cope with such a major life transition as nursing home placement? How has the resident dealt with prior life transitions? Has the person been able to resolve prior challenging life circumstances such as taking a new job, initiation of new relationships, or moving to a new environment? Is the ability to manage or cope with anxiety by lifestyle habits such as smoking, drinking alcohol, eating fatty foods, and/or exercising inhibited by the rules of the nursing home?
If not declared incompetent, is the resident aware that s/he can leave at any time and is not incarcerated or committed? Is the resident cognitively impaired and not fully oriented to the rules of the nursing home and the rights of residents?
What is the new resident’s personality structure? Is the person resilient or generally inflexible and limited (i.e., lacking adequate coping strategies to manage novel stressful situations)? Does the person value independence or is the resident comfortable with being dependent on staff? If dependent, is this a lifelong personality trait or temporary, precipitated by abrupt life changes? Is the person introverted and uncomfortable having a roommate, or extroverted and contented in a communal environment?
Is the new resident always very anxious or does the person become anxious only at certain times (e.g., shift change; time of day; when left alone; with certain staff members, etc.)? A thorough behavioral assessment of the anxiety symptoms can reveal patterns that trigger anxiety that may be incorporated into a treatment plan.
What kinds of trauma has the person experienced throughout life (if any)? What old traumas might be reactivated by the nursing home environment? For example, has the person been in combat, in a major car accident, or been sexually abused? See item 3 for a discussion of grief-related trauma.
What are the resident’s expectations of life in the nursing home? Is the nursing home merely a place to die? Is the resident afraid of being abused or taken advantage of? Is the nursing home administration and staff sensitive to the cultural aspects of care? For example, does an African American resident not feel comfortable with the ethnic mix of residents in the nursing home? Do residents on the LGBT spectrum become anxious because of perceived homophobic staff or resident attitudes?
There is a definite need for an anxiety screening test well-validated for use in LTC settings. Such an instrument might be included as part of an overall assessment of anxiety and mental health issues for LTC residents, and perhaps broaden the scope of the federally mandated Preadmission Screening and Resident Review (PASRR) to screen those with a serious mental illness. This evaluation should help assure that those residents with anxiety and/or comorbid depression will live in an environment that is least restrictive, and if admitted to an LTC setting that their mental health issues will be addressed in a comprehensive treatment plan encompassing both medical and mental health concerns.
Policy Implications
The accurate assessment of mental health disorders in LTC is of the utmost importance to improve older adults’ independence and psychosocial well-being. As such, we briefly describe some policy issues that are relevant. According to Title II of the Americans with Disabilities Act, adults with mental health disorders must be allowed to dwell in the least restrictive setting possible as determined by the U.S. Supreme Court Olmstead vs L.C. ruling of 1999. Under perfect circumstances, the PASRR Level 1 and Level 2 forms should prevent the admission to nursing homes for adults with mental health disorders in absence of a clinically-meaningful reason for placement (e.g., rehabilitation, severe functional limitations, cognitive impairment). However, research has suggested that PASRR compliance is inadequate (Linkins et al., 2006; O’Connor et al., 2011), and this may ultimately jeopardize the universal desire to age-in-place. For older adults with mental health disorders, community living with supports and services is certainly possible, with assisted living communities (e.g., “Limited Mental Health” licensure as found in Florida) as an appropriate placement if significant changes in health occur.
In order to comply with these nondiscriminatory policies, the Office for Civil Rights of the U.S. Department of Health and Human Services (2016) further developed Section Q of the Minimum Data Set 3.0 resident assessment questionnaire to specifically address active discharge planning and the integration of Local Contact Agencies. Currently, the Minimum Data Set 3.0 includes other validated surveys such as the Brief Interview for Mental Status for cognition (BIMS; Saliba et al., 2012) and the Patient Health Questionnaire-9 for depression (PHQ-9; Saliba et al., 2012), but no such measure for anxiety. To improve the detection of mental health disorders and deliver appropriate treatment in LTC settings, we suggest incorporating a measure of anxiety. With the new Behavioral Health deficiencies developed by Centers for Medicare & Medicaid Services (CMS) in November 2017 (§483.40 Behavioral Health Services; Centers for Medicare & Medicaid Services, 2017), nursing homes with residents who have untreated anxiety disorders or with insufficiently trained staff may be more likely to be cited and fined. We hope that greater stringency and vigilance will lead to improved health outcomes for residents with mental health disorders.
Conclusions and Recommendations
As we have underscored, anxiety is a significant and common problem among older adults in general, and is common among those living in LTC settings as well. Anxiety among nursing home residents is related to poorer quality of life, increased caregiving burden, increased risk of hospitalization, dementia, and death (Drageset et al., 2011, 2013). Although self-reported symptoms of anxiety are common among nursing home residents, actual diagnosis and treatment of anxiety problems are uncommon. This may be due to many factors, including lack of appropriate screening and assessment for anxiety, lack of specific measures designed for use in this population, and psychometric limitations of older measures that are used in LTC settings.
As highlighted, there are several good options for self-report screening and assessment measures in LTC, such as the GAI (and GAI-SF), GAS-LTC, and the BADS. These are the only self-report measures designed for older adults that have been successfully validated for use with older adults living in LTC settings, and study results provide support for the use of these measures in this population. Despite some obvious overlap among these measures, each measure has its own nuances, advantages, and limitations that researchers and clinicians should carefully consider when deciding on the best tool for their purposes. For example, the GAS-LTC measures the broadest array of anxiety symptoms, but formal cutoff scores are not yet established. The BADS screens for both depression and anxiety, so it may be preferable when both depression and anxiety are a concern. Finally, the GAI and the GAI-SF represent good choices to screen for anxiety, especially when the clinician or researcher wants a strong assessment of worry symptoms, with empirically supported cut-off scores.
As noted earlier, some commonly used measures of anxiety have either not been examined for the validity of their use in LTC or they produce questionable reliability and validity results when examined, such as the HADS. Therefore, it is recommended that researchers and clinicians working in LTC settings use measures such as the GAS-LTC, GAI, GAI-SF, or the BADS which all have at least preliminary evidence of strong psychometric properties for their use in LTC. An important future research direction is to directly compare these measures in a diverse sample of LTC residents to more directly gage relative strengths and weaknesses, including measurement of the impact of cognition on the psychometric properties of the measures. Given that both pharmacotherapy and cognitive behavioral therapy (CBT) are moderately effective in the treatment of late-life anxiety in community-dwelling older adults (Miloyan et al., 2014), it behooves clinicians and researchers to determine how best to modify intervention protocols to address the unique needs of LTC residents with anxiety. Finally, consideration for residents’ medical comorbidities and medication regimen must be given. Clinicians should focus on ways to educate LTC directors and staff to emphasize the importance of screening for anxiety to improve the assessment and treatment of anxiety symptoms and disorders in their residents. The time is ripe to better serve older residents with anxiety symptoms and disorders in diverse LTC settings.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Biographies