Article Text

Download PDFPDF

Home emergency response team for the seriously ill palliative care patient: feasibility and effectiveness
  1. Nicolas Deniau,
  2. Taraneh Shojaei,
  3. Alexandre Georges,
  4. Jean Danis,
  5. Georges Czapiuk,
  6. Stephane Mercier,
  7. Claudine Maari,
  8. Sylvain Pourchet,
  9. Elisabeth Balladur and
  10. Clement Leclaire
  1. Paris Public Hospital at Home (HAD AP-HP), Greater Paris University Hospitals, Assistance Publique—Hopitaux de Paris, Paris, Île-de-France, France
  1. Correspondence to Dr Clement Leclaire, Paris Public Hospital at Home (HAD AP-HP), Greater Paris University Hospitals, Assistance Publique - Hopitaux de Paris, Paris 75005, Île-de-France, France; clement.leclaire{at}


Objectives To characterise trajectories associated with a new team organisation combining critical care and palliative care approaches at home.

Methods We describe the pattern of an emergency response team 24/7 directed to patients with advanced illness presenting a distressing symptom at home, who wanted to stay at home and for whom hospitalisation was considered inappropriate by a shared medical decision-making process in an emergency situation. To assess preliminary impact of this Programme, we conducted a descriptive study on all consecutive patients receiving this intervention during the first year (between 6 September 2021 and 5 September 2022).

Results Among the 352 patients included, main advanced illnesses were cancer (41%), dementia (28%) or chronic organ failure (10%). They were critically ill with acute failures: respiratory (52%), neurological (48%) or circulatory (20%). Main distressing symptoms were breathlessness (43%) and pain (17%). Median response time from call to home-visit (IQR) was 140 (90–265) min. Median length of follow-up (IQR) was 4 (2–7) days. Main outcomes were death at home (72%), improvement (19%) or hospitalisation (9%) including three visits to emergency department (1%).

Conclusions Our study supports that shared decision-making process and urgent care at home are feasible and might prevent undesired hospitalisations.

  • Clinical decisions
  • End of life care
  • Home Care
  • Nursing Home care
  • Symptoms and symptom management
  • Transitional care

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Dying at home is an important concern of patients with advanced illness and may be hampered by sudden distressing symptoms, but the impact of a dedicated programme to increase goal-concordant care at home in emergency remains unclear.


  • During the first year of operation, among the 352 patients who received the intervention, only three (1%) were referred to the emergency department and 252 (72%) died at home. This preliminary study supports the view that a programme to promote reactive shared decision-making process and rapid care delivery at home might reduce undesired hospitalisations, increase deaths at home and be cost-effective.


  • A double observation was put forward here: first, besides anticipation, the rapid response capability could be a crucial component of an integrated palliative care approach. Second: combining curative and palliative treatments could help to overcome apprehension about loss of chance and support patients’ preferences and then improve goal-concordant care achievement.


The place where people die is influenced by allocated healthcare resources and policies.1 2 A majority of people prefer to receive care and die at home, and most people who die at home or in nursing home need palliative care3. Less than one-third of deaths occur at home4 5 because patients with an advanced illness presenting distressing symptoms have no realistic solution apart from being referred to the emergency department (ED).6 7 While ED is a common entry point into the healthcare system, it is accepted as a bad environment for ending life.7 8 Although data suggest that integrated palliative care at home can improve care, decrease costs and even extend life, it remains difficult to implement strategies for critical palliative situations.5 After a first experience during the COVID-19 pandemic,9 recognising the need of a better strategy for these patients, Assistance Publique–Hôpitaux de Paris (AP-HP) hypothesised that a dedicated organisation combining critical care and palliative care approaches might reduce undesired hospitalisations and increase deaths at home. In September 2021, AP-HP developed the PALLIDOM-Programme to deliver curative and palliative treatments at home within a few hours following the acute symptom. This Programme focused on seriously ill patients who wanted to stay at home and for whom hospitalisation was considered inappropriate by a shared medical decision-making process in an emergency situation. The home-based palliative care pathway in France involves multiple stakeholders aiming to proactively address urgent crisis situations. Nevertheless, not all palliative emergencies can be preempted. Currently, they are managed by the primary care pathway during sporadic interventions. Referrals to the PALLIDOM-Programme serve to provide support during the initial phase and, more importantly, to ensure follow-up over a period of several days.


To assess the feasibility of this new organisation and to describe the functioning of this 24/7 emergency response team, we conducted a descriptive study on every consecutive patient who received this intervention during the first year of operation (between 6 September 2021 and 5 September 2022). Our model covered a dense urban area (Paris, France) for a population-catchment of 3.5 million inhabitants and was based on three steps: triage, care delivery and follow-up. For the emergency shared decision-making process, a 24/7 hotline was held by a dedicated medical staff composed of physicians with dual experience in critical care and palliative care. They answered to all referrals for care from a medical doctor at the bedside or with a good knowledge of the patient presenting an acute deterioration. If appropriate, the second step was a home visit (by a team composed of a nurse and a physician) to evaluate, deliver care and support communication between patients and their families. Acute interventions allowed for the administration of curative medications via intravenous or subcutaneous routes (antibiotics or diuretics…) as well as the prescription of symptom-focused medications (opioids, anxiolytics and sedative-hypnotics). Oxygen was promptly delivered by a healthcare provider if deemed necessary. The point-of-care analysis was conducted by the referring primary care physician. Patients were followed-up by visits and telemedicine until improvement enabling discharge from the Programme, or a deterioration ending in death or hospitalisation. Our study aimed to determine the feasibility targets of the Programme, which include the correlation between supply and demand, financial viability and operational adaptability. For this purpose, the feasibility outcomes were, respectively, the uptake, cost comparison with that of a palliative care unit stay and diversity of the referrals to avoid the funnelling effect and ensure equity in access to healthcare. An independent institutional review board approved our study.


Shared decision-making process

During the 1-year period, the team received 501 call requests, referrals for care in emergency situations related to unpleasant symptoms among patients with advanced illness. Decision-making was based on medical deliberation about the appropriateness of a transfer to ED according to chronic frailty, acute severity and patients’ wishes. Patients who have received an intervention from the Programme at home are referred to as ‘admitted patients’. This term does not refer to traditional admission to a healthcare facility but to the framework of hospital-at-home. The final cohort included 352 patients (70% of referrals). This admission rate did not change over time although the number of referrals for care increased from 1.0 to 1.7 per day (three first months vs three last). Neither sociodemographic, clinical nor frailty characteristics had significant influence on this admission rate. Main causes of non-admission were the absence of need for technical care (consulting activity and advice for prescription), decision of hospitalisation or death before team arrival (respectively, 32%, 20% and 19%). The only factor significantly associated with admission was the documentation of former goals-of-care conversations (advance directives or advance care planning, p=0.02) (table 1).

Table 1

Patients characteristics, shared decision-making process and care delivery (N=352)

Characteristics of admitted patients

Among the 352 admitted patients, the median (IQR) age was 87.5 (76–94) years, 129 (37%) lived in nursing home and 223 (63%) lived at home. They were highly frail, and the main advanced illnesses were cancer (41%), dementia (28%) or terminal stages of organ system failures (10%). They were critically ill with acute life-threatening events: respiratory failure (52%), neurologic impairment (48%) or circulatory failure (20%). Main distressing symptoms were breathlessness (43%), pain (17%) and coma (15%). It was noted that 18 (5%) patients expressed a wish to die or to hasten death with two patients requesting euthanasia. Fourteen (4%) patients died during the admission visit. The mean cost per admitted-patient was 2665€.


Median response time from call to home visit (IQR) was 140 (90–265) min. Median length of follow-up (IQR) was 4 (2–7) days. Mean number (±SD) of team home visits during follow-up was 3.63 (±1.54) (13% during the night). The team provided both curative and palliative treatments. Continuous injectable infusion or patient-controlled analgesia were used in 51% of cases. Midazolam was used in 46% of cases, most of the time for anxiolysis, transitional or procedural sedation. It was used in 5% of cases for deep and continuous sedation for refractory symptoms. Main outcomes were death at home (72%), improvement (19%) or hospitalisation (9%), most often in a palliative care unit and including three visits to ED (1%).


With only 1% of visits to the ED and 72% of deaths at home, we can suppose that substantial undesired hospitalisations have been avoided. A double hypothesis has been put forward for interpreting these results: First, besides anticipation, the rapid response capability could be a crucial component of an integrated palliative care approach. Second, the delivery of care combining curative and palliative treatments could overcome apprehension about loss of chance and improve patient and caregiver acceptance of staying at home. Given the variety of clinical scenarios, we have chosen not to rely on an algorithm for the decision-making process but instead engage in a deliberation between the two physicians, centring on three key factors: chronic frailty (an incurable underlying illness), acute severity (organ failures) and the patient’s expressed preferences, either directly or through a substitute decision-maker or family member.

Many healthcare systems are oriented towards providing life-sustaining treatment as default,10 and even though most patients wish to avoid invasive procedures, many die after aggressive end-of-life care for lack of alternatives.4 5 10 Interestingly, our Programme is positioned as this alternative and has probably played a role in reducing hospitalisations at opposite ends of the hospital-care spectrum: critical care units and palliative care units.11

During follow-up, the incidence of ED visits and hospitalisations was very low. This Programme confirms that presence (24/7 assessment), competence (effective symptom management) and security (reproducible pattern) are identified as the palliative care core mechanism enabling patients to stay at home,12 supporting the recommendation that all critically ill patients, by definition, have a serious illness and, thus, have palliative care needs.11 Less aggressive care can decrease patient, family and professionals’ distress; and more generally decrease mistrust in the healthcare system.10

The funding for this Programme was obtained through the current financing model for hospital-at-home in France. Additional funding was mobilised by APHP to ensure the launch, given the unusual level of medicalisation in the current hospital-at-home model. The mean cost per admitted patient was 2665€. Assuming that these patients were admitted to a palliative care unit (mean cost for 3 days 6328 euros), the estimated savings could be of more than 50%, and over 1.2 million to healthcare insurance, but additional cost-effectiveness analysis is needed.5 13

Our study has some limitations: the absence of data about non-admitted patients’ trajectory, and about to which service patients would have been referred without the Programme. Moreover, these retrospective results contain subjective variables and offer a team-centred point-of-view. Our findings help to explore new strategies to engage society around a hospital-at-home solution for home-based end-of-life care,5 14 but need further research.


A programme to promote patients’ preferences and goal-concordant care at home is feasible in emergency. During the first year of operation, among the 352 patients who received the team visit, only 3 (1%) were referred to the ED and 252 (72%) died at home. This support the view that a dedicated programme for patients with advanced illness might improve goal-concordant care achievement and be cost-effective, but additional analyses are needed.

Ethics statements

Patient consent for publication



  • Contributors According to CRediT taxonomy: ND: conceptualisation, methodology, literature search, data analysis, writing. TS: conceptualisation, methodology, data collection, data analysis. AG: conceptualisation, methodology. JD: resources, visualisation, validation. GC: resources, visualisation, validation. SM: resources, visualisation, validation. CM: resources, visualisation, validation. SP: resources, visualisation, validation. EB: conceptualisation, project administration, supervision. CL: conceptualisation, literature search, data analysis, writing, supervision.

  • Funding This study was funded by Hôpitaux de Paris (AP-HP) (75005).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.