A qualitative study on conveyance decision-making during emergency call outs to people with dementia: the HOMEWARD project

It is estimated that around 850,000 people in England are living with dementia [1]. Dementia is a progressive and irreversible condition resulting in a decline of cognitive, functional, behavioural and psychological abilities, and tends to be associated with a loss of independent living and social interaction [2]. An increasing number of emergency medical services (EMS) calls are to older adults, and research indicates that 14% of older adults who make an EMS call have recognised cognitive impairment consistent with dementia; the true number is possibly higher due to under-diagnosis [3].

Dementia can be a significant barrier to clinical assessment in the emergency care setting; confusion resulting from dementia may contribute to inaccuracies in the medical or medication history, and may limit an individual’s ability to comprehend questions or follow instructions [4‐6]. It is often difficult to establish if confusion is due to pre-existing cognitive impairment, delirium associated with the event leading to the call, or a combination of the two [7, 8]. Nonetheless, paramedics are required to decide whether the patient can be safely treated and managed at home or in the community, or if conveyance to hospital is the most appropriate course of action. This decision may be further complicated if family members or carers express a preference for where the patient should be treated; emergency care is sometimes accessed by informal carers due to ‘desperation’ for health support [9] and unmet needs [10].

Unnecessary hospital conveyance may result in adverse outcomes for people living with dementia. Evidence suggests that people with a diagnosis of dementia are more likely to be admitted to hospital [11], and once admitted, they have poorer outcomes including: longer length of stay; higher rates of readmission; higher chance of discharge to a care home; higher mortality [12‐14]. However, EMS attendances to people with dementia are commonly due to a fall [15] and there is also evidence to suggest that not conveying older adults who have fallen can lead to a high rate of subsequent emergency healthcare contacts and an increased risk of death and hospitalisation [16]. Consequently, making appropriate conveyance decisions will improve patient outcomes.

The rising demands on prehospital and emergency care are well documented, and paramedics are increasingly required to make decisions as to whether to convey a patient to hospital or manage them solely at the scene [17, 18]. Research on conveyance decisions has highlighted the complexity of decision-making for paramedics [19], and suggests that a skilled workforce is key to managing diverse patient needs and reducing unnecessary conveyance to the emergency department [20]. However, to our knowledge no research has specifically examined the process of decision-making by paramedics in relation to people living with dementia.

This qualitative study was designed to investigate the factors that influence the decision-making process of paramedics during calls to older people with dementia, with a view to providing paramedics with adequate support to enhance care in this patient group.

Sixteen paramedics were recruited to the study, and their characteristics are shown in Table 1. The researcher observed 42 shifts over a 6 month period and attended 154 incidents. Sixty-eight of the 154 patients (41%) were aged 65 years or over. A diagnosis of dementia was established for eleven patients (13%) who were initially recruited to the observational phase of the study. Two were later excluded as further information became available during the incident that indicated they might not meet eligibility criteria. Analysis was therefore conducted on nine case studies. There were a number of additional cases where the paramedic and the researcher agreed that a patient was likely to have a cognitive impairment consistent with dementia but were unable to establish a diagnosis in accordance with the eligibility criteria. These patients were not recruited to the study.

This paper focuses on findings from the documentary data, observations and paramedic interviews. Four main themes emerged from the data concerning the way that paramedics make decisions during emergency calls to people with dementia: 1) Clinical condition; 2) Cognitive capacity; 3) Patient circumstances; 4) Professional influences. Each of these themes is discussed below, drawing on field notes recorded during nine emergency calls and data from interviews with nine paramedics, each of whom attended one of these events. Five patients were conveyed to hospital, two were treated at home and two did not require treatment. In one case, the paramedic provided advice and reassurance, and in the other a referral for social support was made. In the presentation of findings below, each call has been assigned a unique number (Table 2). Paramedics who participated in interviews are denoted by a research code, e.g. P1 or P2. Table 2 shows details of the ambulance calls.

The findings indicate that the clinical condition of patients with a diagnosis of dementia is the primary reason for conveying to an acute care setting following an EMS call. The conveyance decision was also influenced by the patient’s cognitive capacity, their personal and social circumstances and other professional influences, such as the availability of information from a GP or the opinion of a colleague. This, to some extent, reflects findings from other research concerning a more general patient population. A recent review by Ebben et al. [24] found that factors influencing a non-conveyance decision are related to the professional’s competence and experience, the patient’s health status and best interest, the healthcare system and the availability of decision support.

However, the focus of this study was decision-making for EMS calls to older people with dementia. The findings indicated that where the decision was not clear-cut, a diagnosis of dementia was likely to discourage conveyance. Paramedics expressed concern about removing people unnecessarily from their own familiar surroundings, aware that this can have a detrimental effect on people living with dementia. This may create professional challenges for the paramedics making these decisions as they need to weigh up the risks to the patient associated with conveyance to hospital with the risks associated with leaving the patient at home or in the community.

Despite the apparent preference for non-conveyance, five of the nine patients in this sample were conveyed to hospital. None of these patients had a condition that required immediate treatment, and it could therefore be reasoned that difficulties in the assessment of patients resulted in the decision to convey. One paramedic discussed difficulties relating to the assessment of people with dementia, particularly when communication is an issue and it is not easy to establish the presence of injury and the level of pain that the patient is experiencing. This is an important issue as commonly used pain assessment instruments rely on self-report and the communicative capacity of the patient [25, 26]. For patients with cognitive difficulties, pain can be assessed using observational techniques [27, 28], but these observational approaches require repeat assessments over time [29] and may not be well suited to the prehospital environment. Accepted methods for assessing pain in cognitively impaired adults in prehospital care have significant limitations [30], and there is a need for further work in this area.

Research has indicated that the decision to not convey a patient is a complex one, and is often negotiated between EMS staff, the patient and the patient’s family [19, 24]. Paramedics in this study were careful to familiarise themselves with relevant information concerning the patient’s medical history and level/type of support in order to inform their decision-making. Being a participant in a study examining decision-making is likely to have made this activity more pronounced, for example describing activity that is usually unspoken. Professional factors taken into account included views from colleagues and other practitioners, as well as their own professional obligations as registered healthcare practitioners. Additional information and expertise can strengthen the rationale for a decision. Previous research has found that there can be a mismatch between policy and practice in relation to non-conveyance decisions [19]. Indeed, the findings from this study indicate that the relationship between guidelines and the paramedics’ opinions is not always harmonious, and there is a fear of disciplinary action and litigation involved in making conveyance decisions. Decision-making is an iterative process involving weighing up the risks and benefits for patients, family and the healthcare system. These factors may favour a decision to convey to hospital, particularly when there is limited access to information or alternative services, even when this is not in the patient’s best interests.

There are a number of limitations to this study which affect the transferability of the findings. Data was collected by one researcher in one division of a single ambulance service. The availability of alternatives to hospital varies widely according to location, and a larger study is needed to determine the extent to which additional services affect decision-making. In addition, the sample size for the cases was small, and it is possible that data saturation was not reached [31]. The intention was to recruit 20 patients for case studies but despite adopting a number of strategies, such as targeting particular times of day and paramedics most likely to attend calls to older people, it was not possible to recruit the target number of patients within the allocated time and budget. The paramedics and the researcher were sometimes unable to be certain whether or not a patient had a diagnosis of dementia. When there was any uncertainty, the patient was excluded and this further reduced the sample size. There is also a risk of observation bias in this study; the presence of the researcher may have influenced the paramedic decision-making and the self-selecting participant sample are not necessarily representative of the wider paramedic profession. Finally, the methodology of observing shifts in the ambulance service was resource intensive, and alternative methods should be considered for future research. However, the use of both real-time observation and subsequent interview enhanced the quality of the study in methodological terms, allowing for a comprehensive exploration of the paramedics’ decision-making processes.

The study findings give rise to number of implications for research and future policy. The factors influencing conveyance decision-making for people with dementia are similar to those that guide decisions for other patient groups, and older adults in particular. However, the preference for avoiding unnecessary conveyance in people with dementia, combined with difficulties in obtaining an accurate history and assessing the condition of the patient, indicate that decision-making in this particular patient group is influenced by multiple factors and can be especially challenging. Paramedics rely heavily on information that may or may not be available or accessed on scene, such as informants, carer records and input from GPs. Further research is needed to find reliable ways of assessing patients, such as an evidence-based decision tool, and service changes are required to support access to information that can assist decision-making in people with dementia.