Factors associated with frequent use of emergency-department services in a geriatric population: a systematic review

Population aging significantly impacts the use of healthcare services, notably emergency departments (EDs). A certain proportion of seniors are considered frequent ED users [1], which are a minority of patients using a major proportion of ED services over a given period [2]. Frequent users aged 65 years or older represent as little as 6% of all ED patients but can account for up to 28% of its accesses [3, 4]. Frequent ED use increases the risk of adverse effects such as hospitalizations, functional decline, complications related to treatment and procedures, and suboptimal follow-up [5, 6]. Older adults visiting EDs are more prone to experience misalignment between their medical needs and their use of health care. Unmet medical needs can lead older adults to use these services instead of more adequate ones [7].

Currently, there is a lack of consensus regarding the definition of frequent ED users. Definitions generally range from patients with two to 12 and more ED visits per year [3, 8]. However, the most common definition for ED frequent use is four or more visits within a year period [2, 9]. Many reasons have been raised to explain seniors’ frequent ED use, including avoidable visits for nonurgent problems [10, 11]. Nevertheless, seniors present more vulnerability factors, as well as more chronic conditions and complex medical needs [5]. The ED is at the core of healthcare for acutely ill seniors, who tend to present a higher priority score than younger adults [12].

Frequent geriatric ED users are far from being a homogeneous population. Identifying their numerous individual characteristics could improve our understanding of how their medical and social needs, as well as their healthcare use, might be best managed and assisted. Even if many studies, including scoping and systematic reviews, have examined frequent ED use among the adult population, few seem to focus on older adults [1, 13]. McCusker et al. (2003) aimed at identifying determinants of ED use among older people. Their main results report that need factors figure among the main determinants of older adults’ ED use. These factors include perceived health status, specific diagnoses (heart disease, diabetes, psychiatric disorders), and composite measures of comorbidity [14]. No recent review has specifically examined frequent ED use among older adults. This systematic review aimed at identifying variables associated with frequent ED use by older adults.

A systematic review was conducted in accordance with the Joanna Briggs Institute (JBI) guidelines [15]. Reporting was made in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Checklist, presented in Additional file 1 [16].

This systematic review identified factors most often associated with frequent ED use by older adults: a high number of previous hospital and ED admissions, living in a rural area adjacent to an urban center, low income (Medicaid beneficiary), high number of prescribed drugs, and history of heart disease. Variables associated with a lack of social support yielded mitigated results, while others—such as comorbidity, dementia, and primary-care-related variables—returned few results or were not accounted for.

First, the results emphasize the impact of past ED visits and hospitalizations. Directly related to seniors’ health, they were among the most cited risk factors [14, 28]. As an example, Andersen’s Behavioral Model of Health Services Use is commonly used to account for the use of healthcare services among the older population. It indicates that need factors—including past use of healthcare services—are the most important variables influencing healthcare use [14, 29]. Notions related to the length of stay in hospital EDs, while known as having deleterious effects, were not, however, considered in the included studies [30]. As an example, older adults with frailty or severe health conditions are more prone to longer hospital stays. This puts them at greater risk of functional decline, associated with ED returns and rehospitalizations [31].

Our results also indicate residence location as being a demographic variable associated with frequent ED use. Older adults living in rural areas tend to report poorer health status and more healthcare issues than their peers in urban areas [32]. Indeed, people from adjacent rural areas tend to have more ED visits primarily because of the limited availability of care services. In addition, since service availability is generally limited, older adults in remote rural areas use fewer ED services. The availability of healthcare services also varies according to the type of rural area [32].

Being a Medicaid beneficiary came out as an important variable. Nevertheless, it received scant attention in studies looking at ED use by older adults. Income information is not usually provided in databases, while Medicaid insurance status—used as a proxy for socioeconomic status [14, 33]—can only be considered in studies conducted in the United States. Otherwise, income can be considered an indicator of unmet health needs; some authors indicate that these needs should be considered instead. Indeed, seniors with lower incomes or on Medicaid insurance might be more likely to report such unmet needs and to turn to healthcare services such as EDs in fulfilling them [34].

In addition, the risk of using several drugs simultaneously is increased as a result of multiple chronic conditions, which frequently occurs in older adults [35]. Polypharmacy is a proxy for comorbidity severity and can therefore be considered a variable associated with frequent ED use by older adults [36]. Indeed, multimorbidity is associated with medication intake, and the number of medications tends to be proportional to health status [37]. Conversely, having comorbidities and being a major user of health services exposes patients to polypharmacy, among other things [36]. Only two studies adjusted for case mix [23, 24], and none considered a composite measure of comorbidity. Controlling for these would yield a more accurate representation of the population of frequent geriatric ED users. Moreover, morbidity clustering is considered a better predictor of healthcare-services use than a specific disease [38]. In our results, heart disease was the most important specific disease associated with frequent use. Indeed, older adults with such conditions are considered at greater risk of health deterioration. They are then more likely to use EDs and become frequent users of these services [28].

For its part, social support yielded mitigated results. A systematic review by Valtorta et al. (2018) pointed out that there is no clear evidence of a relationship between social support and ED use by older adults. This can be explained by the multidimensional aspects of this concept and the many variables used to measure it (e.g., homelessness, dwelling type, social network, or perceived social network). Social support is related to social functioning, a broader concept covering, among other forms of social participation, social networks, social resources, and social relationships [39]. Pinsonnault et al. (2009) stated that assessing social functioning might be useful in documenting the needs for interventions in older adults [40]. In fact, subsequent studies should then attempt to consider all facets of social functioning to better represent its effects [39].

Additionally, some other important variables received only little or no consideration. First, none of the studies included seniors’ specific living environments. Because of their lower health status, older adults living in private residential-care facilities are more predisposed to ED visits and hospitalization, compared to those living at home or in an independent-living facility [41]. Moreover, variables related to community health services, such as home healthcare, were not considered. If home healthcare can contribute to improving older adults’ health management, this clientele is considered at higher risk of ED visits, related to their higher multimorbidity level [42, 43].

Little consideration was given to variables related to mental health, which were reported as one of the main reasons older adults visit EDs [44]. The mental-disorder category includes various conditions, and some, particularly dementia, would benefit from being explored independently. According to an integrative review by Hunt (2018), individuals with dementia had consistently higher rates of ED visits [45]. Dementia induces vulnerability and implies greater comorbidity, hospitalization episodes, and mortality rates among older adults [35]. Mental-health considerations would therefore be no less important than physical disease in analyzing frequent ED use by older adults.

Variables related to primary-care settings variables were also understudied. Adequate primary-care follow-up can help prevent complications, as well as a certain proportion of ED visits by older adults [46]. As reported in our results, having a primary-care provider plays a role in managing seniors’ health issues and positively affects their ED use [47]. Beyond that, however, care must be accessible and ongoing. For instance, in the United States, states with high rates of ED visits tend to report gaps in access to primary care [48]. Patients with limited continuity of care tend to report more visits to these services, as well as an increased risk of ED visits and hospitalizations [49]. So, unmet health needs may persist even with a higher number of primary-care visits. A study by Horney et al. (2012) points out that older adults who were frequent ED users had more prior visits to primary-care physicians [50]. The number of visits might then be more important than the fact of having a primary-care provider. Therefore, measuring continuity and access to primary care should be considered.

This study has clinical and policy implications. Indeed, pooling the most recent information–even if scarce–on this population is necessary to monitor trends in frequent ED use and to help implement accurate interventions. In fact, case management strategies, additional support systems, and better access to care would have the benefit of improving the health of frequent geriatric users of ED and may contribute to lower their use of these services. As suggested by Pines et al. (2012), creating an effective categorization of frequent users could represent a promising avenue, making it easier to compare and group studies. Categorization would also bring relevant information to existing risk-assessment methods, direct the development of new ones, and raise situations where better care coordination is needed.

This systematic review identified the main factors associated with frequent ED use by older adults: a high number of past hospital and ED admissions, living in a rural area adjacent to an urban center, low income, high number of prescribed drugs, and a history of heart disease. Further studies should bridge the gap in understanding and give a more global portrait by adding important personal variables such as dementia, organizational variables such as use of community and primary care, and contextual variables such as social and economic frailty.