Case management for frequent users of the emergency department: study protocol of a randomised controlled trial

Individuals attending emergency departments (ED) on a regular basis account for a disproportionally high number of all ED visits. LaCalle and Rabin [1] in their systematic review found that patients visiting an ED four or more times per year accounted for 4.5%–8% of all ED patients and 21%–28% of all ED visits. Emergency department frequent users (ED-FUs) attend the emergency department on multiple occasions; however, definitions and threshold numbers of visits vary across studies. According to Locker [2], the definition of five attendances or more per year corresponds to a non-random event and should be used to allow better comparisons between studies. ED-FUs present a higher rate of morbidity and mortality than less frequent ED users [3‐7], are more at risk of drug and alcohol abuse [5, 7‐9], often present mental health issues [3, 5, 6, 10], are more likely to visit for complications and exacerbations of chronic conditions [10, 11], and are often homeless, uninsured, and from low socio-economic levels [3, 12‐14]. The majority of them believe that their complaints require immediate attention [1], and thus they constitute a significant burden on hospitals due to multiple visits and the number of problems they bring to the ED.

ED-FUs contribute significantly to ED overcrowding and extended waiting times, often due to inappropriate visits to the unit [15]. Overcrowding is detrimental to the quality of care in EDs. However, the severity of the reason for consultation at the ED is often controversial [1]. Indeed, several studies show that ED-FUs have non-emergency conditions [10, 16‐18] and could receive better care in settings other than an ED [19, 20], which is not designed to provide continuous care to patients with non-emergency, chronic conditions. In addition, the numerous issues that ED-FUs have are not easily addressed by simply providing care alone. Appropriate and consistent medical and social services are needed for such vulnerable populations.

In response to these concerns, several institutions worldwide (e.g. in the United States, Canada, Sweden, the United Kingdom, the Netherlands, Spain, and Australia) [9, 12, 21‐31] have introduced specific interventions for ED-FUs aimed at reducing the number of their visits, treating their medical co-morbidities, and/or addressing their social needs. Interventions vary, according to a recent systematic review of the literature by our research team that identified different types of interventions aimed at improving the management of adult ED-FUs [32] and at assessing interventions’ effectiveness. Most of the studies describe interventions referring to and/or inspired by case management (CM) [9, 12, 25, 29‐31, 33].

One of the most common interventions consists of CM multidisciplinary teams composed of nurses, psychologists, and possibly physicians [27, 34‐39]; this approach can help address complex situations and scenarios. Team members from different professional backgrounds, such as psychiatrists and health educators might complement the team, depending on the specific CM project. Coordination and organizational care tasks are often allocated to a case manager [37] who guides patients through the care process and provides social support. Care is generally considered as a continuous integration of medical and social dimensions. It is commonly patient-centered and holistic in nature, and takes patient empowerment [27, 35, 36] into account. Moreover, the locus of intervention is not limited to the hospital, and often extends into the community.

CM is a highly flexible and dynamic process and mainly depends on patient needs; the order of individual steps is often not constricted. In fact, its dynamic condition emphasizes that sometimes several steps take place simultaneously, or that the case manager has to return to a previous step. Based on the literature, this can be summarized in five steps [27, 38‐43]: identification, assessment/reassessment, planning, implementation, and evaluation/monitoring. The Behavioural Model for Vulnerable Populations [44] provides a theoretical framework for understanding how CM might improve the care of vulnerable patients; this theoretical framework suggests that the use of health services is a function of:

CM guarantees that issues in each of these domains are addressed.

Interventions aimed at improving ED-FU management have had positive outcomes: some of the interventions evaluated have been effective in reducing emergency department use [9, 12, 21, 24, 26, 29, 31]. Cost-reduction analyses are also promising: Wassmer anticipated reductions in cost even when partially based on modeling estimates [31]; two other studies showed the effects of clinical case management on hospital services and its cost effectiveness [12, 29]. Some interventions have had positive effects on social outcomes [12], such as a significant reduction in homelessness [25, 29]. A positive effect on social outcomes is essential, as the link between social problems and health has been demonstrated by many authors [45]. Finally, clinical outcomes were assessed in three studies [12, 25, 29]; one of them demonstrating a positive effect in reducing alcohol and drug use [12].

In the literature, interventions aimed at improving the management of ED-FUs have demonstrated several positive outcomes, but there are still some knowledge gaps:

– There is only one randomised controlled trial (RCT) showing a significant reduction in ED use by FUs compared to patients receiving standard care [29].

– The threshold for number of visits varies across the three existing RCT [22, 29, 30]; only one is based on the definition of five or more attendances per year, corresponding to more than known random events [29].

– Cost reductions were demonstrated in three studies [12, 29, 31], but only one is an RCT [29], and the other two did not contain a control group.

– Patient baseline characteristics and health-care specificities shown in 11 studies included in a systematic review by Althaus and al. [32] were only relevant within the country in which each study was conducted (the US, Sweden, Canada, Australia, and the UK).

Because of the existence of the knowledge gaps mentioned above in a topic that is of the utmost importance for patients, clinicians, and policymakers, with this trial we would like to demonstrate that by establishing locally a model of care for these patients, we can decrease the use of the health-care system, improve these patients’ quality of life, and reduce costs consequent on frequent use.

This study is coordinated with recent local research projects dedicated to assessing profiles and improving healthcare for ED-FUs, who are considered to be a highly vulnerable subgroup and a proxy for vulnerable populations in general.

At the Lausanne University Hospital ED, in 2008–2009, ED-FUs accounted for 4.4% of ED patients and 12.1% (n = 5,813) of all ED visits (n = 48,117) [46]. A retrospective chart review case–control study, conducted in this hospital between April 2008 and March 2009 by Bieler et al. [46], demonstrated that social (i.e. homelessness, institutionalization, unemployment, or dependence on government welfare) and specific medical vulnerability factors (i.e. ED primary diagnosis of substance abuse and the use of five or more clinical departments in the 12 previous months) increased the risk of ED use among 719 patients. A combination of social and medical factors was markedly associated with frequent ED use, as FUs were 10 times more likely to have three of them (of a total of eight factors; 95% CI = 5.1 to 19.6). This result is confirmed by Althaus et al. [56] in a retrospective chart review on hyperfrequent users (12 attendances or more during a year): they were 13 times more likely than non-FUs (65.5 vs 5.0%) to present three or more of the risk factors of vulnerability that Bieler et al. referred to [46] and 2.2 times more likely than FUs (62.5 vs 28.4%). Finally, unpublished, local, prospective, cross-sectional data (Bodenmann P. et al., in progress) obtained between November 2009 and June 2010 has demonstrated differences between 226 FUs and 173 infrequent users. FUs were more often younger with a mean age of 51 vs 56 in infrequent users, and the former had experienced five to 18 admissions in the previous 12 months. They cumulated vulnerabilities in terms of somatic problems, mental diseases, risk behavioral indicators, and unfavorable social determinants of health.

Taking care of a growing number of vulnerable patients requires specific interventions. A systematic review of the effectiveness of interventions targeting ED-FUs concluded that such interventions may reduce ED use and that CM, the most frequently described intervention, seemed to improve social and clinical outcomes and reduce ED costs in different studies [32]. Three studies [12, 29, 31], from which one RCT [29], concluded that CM could contribute to the reduction of ED use and of consequent costs, while two of these studies [12, 31] found additionaly that CM could also lead to positive social outcomes. However, patterns of care that have succeeded elsewhere have to be tested in local or national settings before being introduced into a new context of care among local patients. A mixed methodology using quantitative and medico-economic analysis is needed.

Responding to the knowledge gaps in the literature [57, 58] and following our local studies through different observational designs, our hypothesis is that CM leads to reduced ED use by ED-FUs through a better orientation in the health-care system, improves their quality of life, and is more cost-effective than is standard emergency care alone provided by nurses and physicians serving ED-FUs. Positive findings would constitute a strong incentive to replicate these studies on a larger scale, in a multicenter study with more extensive follow-up procedures. Positive findings would also suggest that specific populations need specific care, and would have major implications for healthcare quality and costs. Finally, the total number of ED visits in Switzerland is around 1.3 million per year [59] and has been steadily growing. If our intervention results in a reduction in the number of ED visits, the impact at the national level could be significant.