Hospital at Home in 10 years—Here, borrow my binoculars

Abstract

Hospital in the Home (HAH) refers to the delivery of acute hospital care to patients at home, or in their usual place of care, including nursing homes. HAH is an acute hospital substitution service—it delivers medical, nursing and allied health staff, skills, interventions and treatments, technologies and supervision usually found in hospitals to sick patients at home. The mission of HAH is to improve the lives of sick people who need hospitals. Patients consistently express a preference for the HAH model of care. HAH has developed an evidence base for safety and effectiveness over the last 30 years, but its role has become prominent since the COVID‐19 pandemic. In most health systems where HAH is successful, it is organised and funded as an acute hospital inpatient service. Progress continues to broaden the clinical applications of HAH and narrow the operational definition and processes. It is likely that in 10 years, HAH will form an essential part of the scope of an acute hospital, taking its place alongside the Emergency Department and Intensive Care.

Introduction

In 1950, an intuitive, natural experiment saw brilliant Scandinavian clinicians introduce a known intervention and apply it to hospital care in a way that had not been seen before [1, 2]. The introduction of Intensive Care Unit (ICU) changed hospitals. From humble, even completely impractical origins, a unique hospital service platform was built upon which further technologies and applications were developed, and these, in turn, also allowed for advances in multiple traditional specialties, both in surgery and medicine. However, it took decades for that platform to be adopted and disseminated around the world and for the basic unifying definition, structure and medical responsibility of the ICU to resemble what we now come to expect. It takes far longer to adopt a new hospital structure‐based innovation than any individual disease‐based treatment intervention delivered within those walls—whether a drug, vaccine or an implant. This is because hospitals deliver many pretested interventions, but the hospital itself struggles to test, innovate and change what it does. The work of Atul Gawande and others offers many examples of the difficulty healthcare institutions have in adopting proven innovative solutions [3, 4].

As a result, making predictions about the progress of any truly new idea that requires significant, perhaps even fundamental, hospital reorganisation in order to deliver it is a thankless task. Perhaps a contributing factor to that glacial pace of change is the tendency for involved clinicians at the early stage to shirk away from the task, at least until the scale of evidence is overwhelming. But ironically, without pursuing a positive vision of the future by planting seeds of support, the opportunities to gather even the buds of that evidence will never come.

Like ICU, Hospital at Home (HAH) has essentially undertaken a similar natural experiment—it has taken several basic hospital interventions and transplanted them outside the hospital, in ways that are initially humble, and probably very impractical, for the strong benefit of some sick people who need hospitals. Just as it is unimaginable today to have an acute hospital without an ICU, HAH will become an integral part of the organisation and service delivery profile of any acute hospital. In a decade, all hospitals will include HAH as an understood part of their basic scope of services.

Names are important. Hospital at Home (or Hospital in the Home, or Home Hospital) is already three decades old in some parts of the world. Although many alternatives have been posed and some used, there are complications in changing, or adopting multiple, names [5]. For the purpose of this paper, the acronym HAH will be used. It may be that HAH is not the most exciting choice for a name, and as search terms, ‘home’ and ‘hospital’ are too broad if used separately. However, changing the name threatens the loss of crucial experience. Curiously, and amusingly, one effect of the need to describe the control arm for research on HAH has been the emergence of a new name for its older alternative—BAM (for Bricks and Mortar), to identify non‐HAH hospital care [6].

Definition of Hospital at Home

The definition of HAH has been established in principle: acute hospital‐level care delivered to patients at home who would otherwise need to be in hospital [7]. This involves the use of hospital medical, nursing, pharmacy and other allied health staff, their skills and interventions delivered to patients who would otherwise need to be in hospital to receive them. The principle of substitution of hospital care can and should be extended to the classification of HAH patients as inpatients in every way. This is a hospital program, not a community health program. For 30 years in Australia, the funding of HAH comes through acute hospitals in the identical way the hospital receives activity funding for hospital services [8]. This principle has been adopted for HAH in Singapore, Canada and the United States [9]. At the time of writing, the United States has just extended the inpatient funding for HAH for 5 years, ensuring the ongoing investment in HAH adoption, growth and research [10]. In the other major established HAH health system, Spain is also organised as an inpatient service. In countries including the United Kingdom, Israel and Taiwan, there are variations in organisation, but HAH has commenced the journey [11, 12, 13]. Although there are many other important points to acknowledge, this founding principle remains important. Without this structure, hospitals would have no reason or interest to engage with it. The interventions delivered by HAH are firmly and exclusively hospital based. Access to necessary staff skills, ancillary services such as pathology and radiology, 24‐h access to care, and the ability to immediately access BAM care for patients who require escalation back to BAM are all hospital‐centred interventions and cannot be delivered by community health. If HAH exists in 10 years, it will be because it has retained its role within the acute hospital inpatient service. A drift to a community health service sphere of practice and funding will threaten it. Of course, every country has differing thresholds for inpatient care, and this will necessarily be reflected in their HAH patient population. Although the principles of the service are clear, the model by which the critical inputs are arranged and delivered still vary widely across hospitals, cities and countries. This is an ongoing problem. An outcome that may be demonstrated to be able to be delivered safely and effectively by a HAH unit in one place may not translate to another where the medical support, 24‐h on‐call, or other features are not included. As discussed later, this affects research and the interpretation and adoption of findings. One of the factors that will slow the advance of HAH is the failure to appreciate the current experience in the establishment and management of HAH and to resist the pressure to bend to local opinions or pressures. The role of national societies for HAH will assist, even to the extent of advising on optimum models or even accrediting such models. In 10 years, there should be broad similarity between the structure and organisation of HAH units. They should contain their own internal medical management so that patients transferred into HAH are managed by HAH physicians or perhaps HAH ‘hospitalists’ [14]. These physicians will assess and know the patients and their context, the treatment required, the nursing staff, the unit capability, the research evidence and are devoted to using all means to keep the patient safe, well, and… at home. Units managed by non‐HAH physicians will have an extremely low threshold to return patients to their familiar BAM ward, and the high rates of unplanned escalations will threaten the acceptability of the service among patients and hospitals alike.

HAH is not OPAT (Outpatient Antimicrobial Therapy) [15, 16]. They are not interchangeable terms. At best, OPAT was an attempt to acknowledge the desire for patients to not be confined to hospital for some acute care. However, OPAT stuck to the past rigid interpretations of acute hospital medicine, which previously defined anyone physically outside the hospital to be outpatients and to not be entitled to the additional support and supervision that comes with inpatient care. It pushes onus, risk and cost onto the patient and sometimes extends that risk to self‐administration of intravenous therapy. Some patients will accept those burdens in order to leave the hospital if they have no other choice, but that does not mean it is right to offer it. It also seeks to maintain the hospitals’ archaic self‐identity and not to challenge the way it works. HAH has rendered OPAT redundant.

Patient issues in HAH

The core mission of HAH is to improve the lives of sick people who need hospitals. Patient demand for HAH will grow over the next decade. A universal theme of HAH until now has been the willingness of patients with no prior knowledge of HAH to accept it when it is offered. HAH exists because it is a service that some acute patients prefer [17, 18, 19, 20, 21, 22], which can be challenging for those working in traditional hospitals. There are many reasons for this willingness to accept HAH. More patients are being kept alive through previously lethal conditions or combinations of conditions, but the treatments that achieve those results necessitate frequent or prolonged hospital stays with inputs from multiple hospital teams. Those admissions come with fear, discomfort and isolation. At some point, the trade‐off becomes so burdensome as to be questionable. The confidence offered by the ongoing link between the HAH and the same hospital, which provides its usual acute medical care, must not be underestimated. Patients themselves can also see that some of the interventions delivered during admission can be relatively routine once diagnoses are made and conditions settle, and this adds to frustration. Nowadays, even the most privileged hospitals can no longer match the comfort and familiarity of their patients’ homes. The reported high levels of patient satisfaction for HAH relating to increased personalised attention may seem counterintuitive given that ostensibly there is less overt physical supervision. This only highlights the brevity and distracted, protocol‐driven nature of modern hospital care. When an HAH clinician enters a patient's home, they have their undivided attention. Carers will always be more cautious and concerned than patients, but this often diminishes when the care is actually delivered as promised. It is imperative that the burden on carers should be limited and extend to no more than if they were helping the patient with a simple case of influenza. The avoidance of feared complications related to hospitalisation also forms part of the acceptance. Patients are aware of the risk of hospital‐acquired infections, and family members know that deconditioning and delirium can also occur. The technological capacity of homes and its operability by those who live in them is increasing, albeit variably. Any increase in technology is complicated by the cognitive capacity (or perhaps willingness) to use it. Some of the comfort of homes could also be due to the ability to escape from the enforced technological intrusions associated with BAM hospital care, so any enforcement of more remote monitoring in HAH may find those interventions rendered ineffective by the unwillingness of patients to adopt them because they do not fit with the notion of ‘home’, which is why they have chosen to be in HAH.

Clinical applications of HAH

HAH has made great advances into certain hospital areas and treatments. It is able to treat all acute, complex, multidrug‐resistant, hospital or community‐acquired infectious diseases using intravenous antibiotics, antifungals and antivirals [23, 24, 25, 26, 27, 28]. It is capable of managing cardiac failure through the use of intravenous diuretics and inotropes [29, 30, 31]. It can administer chemotherapies and treat the infectious and non‐infectious complications of cancer or cancer therapies: intravenous antimicrobials, blood and blood products, albumin, immunoglobulin, right up to involvement in stem cell transplantation and CART‐T therapy [32, 33]. Force of circumstance saw it rapidly develop its application in the management of COVID [34, 35, 36]. Respiratory conditions, such as bronchiectasis, acute COPD and flares of cystic fibrosis, are all capable of being managed in HAH [37, 38]. Conditions that require short term intravenous fluids, electrolytes or parenteral nutrition (PN)—such as dengue, acute gastroenteritis or post‐surgical or cancer‐related conditions—can be managed in HAH. Surgical infections, whether complication of surgery or conditions such as acute diverticulitis or cholecystitis, are capable of admission in HAH [39, 40, 41, 42, 43]. Surgical interventions are being performed on older patients with more comorbidities, and more complex surgeries continue to be performed, including transplantation, cancer treatment and advanced plastic surgery techniques. All of these require skilled hospital nursing and medical teams following the surgery to manage not just the complex drains and wounds but also often the medical management of the destabilised, elderly patient [44, 45]. HAH is being applied to both children and very elderly patients. All the principles and applications of HAH also apply to children, and the advantages flow to the entire family [46, 47]. The treatment of patients in residential aged care facilities leads to improvements for patients, families and facilities [48, 49, 50, 51, 52]. The principle is that the HAH cares for all the acute treatment, with none of it abrogated to the family or the facility staff, whereas the family or facility staff continue their deeply personal, normal daily attendant care. The management of some obstetric care has shifted in HAH, including the treatment of hyperemesis of pregnancy. Despite the introduction of NOACs shifting most inpatient care of acute thromboembolism to the outpatient sector, HAH still retains the ability to treat deep vein thrombosis and subsegmental pulmonary emboli. The extension of HAH principles in acute mental health admissions for certain patients is gaining traction [53]. The extent to which HAH may address some of the acute hospital needs of patients in rural centres is a challenging question, but its exploration is ongoing [54]. What should be evident from this list is that HAH has an extremely wide application across every hospital specialist discipline, but the suitable patient groups must be carefully selected. In the past, the only question a clinician (either in the emergency department or elsewhere) had to ask themselves was: Does the patient require admission to hospital? The same question in the future has two parts: Does the patient need admission to hospital?; and, if yes, can that admission be in HAH?

It is expected that HAH clinicians have confidence in the application of HAH in many acute conditions, but it is gratifying to note that the role of HAH is starting to be recognised by non‐HAH clinical specialities—with green shoots of consensus appearing regarding its role in some areas, including infectious diseases and heart failure [55, 56]. They are perhaps timid: heavily qualified and of the lowest priority, far short of what those of us involved in HAH know can be achieved, but it is a start. In 10 years, HAH will be included in many more hospital treatment protocols with more confidence and with HAH involvement appearing higher up the cascade of options.

Research in HAH

Research poses unique difficulties for HAH. HAH is a complex and expensive intervention. Despite efforts to achieve greater consistency, it is still heterogeneous and interpreted and applied differently between hospitals, states, provinces and countries. The intervention is not just the condition treated at home or the medications administered but also the model of care used for delivery, supervision and responsiveness to events. Research activity has been taken up by the leading individual clinicians within HAHs around the world who have undertaken extra work to conceive, prepare, conduct and report the research. They have done that and continue to do that, but the scale of such research generally remains local. The role of larger scale descriptive, quasi‐experimental or natural experiment designs has dominated HAH research, and despite academic concerns, they have demonstrated with durability and repeatability that HAH is safe, effective and acceptable [57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69]. These efforts have formed the beachhead that has allowed further interest, refinement of questions, and development of HAH and, in the context of the difficulties, have been extraordinarily successful. RCTs are difficult to undertake, except possibly at the initiation of a service, because hospitals, having funded and staffed a complex and expensive intervention such as HAH, want as many patients as possible treated in it as quickly as possible. It is difficult to convince anyone to deliberately halve throughput to undertake an RCT, and in any case, the timing of preparing such research with initiation of a new unit is precarious. This also leads to RCTs being generally small, if they are done. Moreover, as indicated, if such research tends to be conducted by new units, staff and leaders are less experienced, processes for assessment, delivery and supervision are less refined, and the understanding of the task is less formed. So the outcomes may not reflect those of a high functioning, experienced, cohesive unit that has been operating for many years. Most HAH research is also single‐site, with very few, if any, multi‐site studies. Worse still, most research papers do not include a comprehensive description of the HAH model used to deliver care, so the intervention can be a black box to the reader or to anyone seeking to replicate the intervention. Systematic reviews or meta‐analyses have reflected the generally positive findings of published research but are also limited [68, 69, 70]. The development of HAH‐specific outcome indicators is progressing, and now, there are outcomes that allow comparison between units. Moreover, problematic is the lack of dedicated researchers and research centres for HAH—but early adopters in the United States, such as Ariadne Labs, will see other centres follow. The challenge to demonstrate successful and measurable hospital substitution, as opposed to just HAH activity, remains a grail quest. HAH research must battle to be published. There was strong novelty interest in the HAH concept by journals, especially in its early efforts, so much research has, in fact, been published, but as the amount of research increases and becomes more specific [71], it is unlikely that general or subspecialist journals will continue to be interested. A new online journal dedicated to HAH has commenced publication this year: the Journal of Advanced Home Medicine (JAHM). It is a first for this area. There are national and international HAH scientific conferences, which are critical for sharing good research and experience, discussing issues of definition and design and energising clinicians working in a new field, who can experience isolation.

The often unstated barrier to the wider adoption of HAH within hospitals is the fear among hospital clinicians that HAH care is inferior, or riskier, than BAM care. HAH models should accept that they hold all‐hours responsibility for HAH patients and have clear responses for deterioration [72]. There have been strides made in advancing the definition of what constitutes safe outcomes and high‐quality HAH practice. Where the hospital outcomes for the treatment of specific conditions and specific treatments exist, they can be applied in HAH, because in most cases the treatments will be identical or similar. Where HAH applies innovative and interesting modifications to those specific treatment protocols, the usual outcomes can still be measured to see if they are appropriate. There are many emerging examples in HAH, especially in the field of infectious disease—for example, the use of twice‐daily cefazolin for the treatment of cellulitis or the use of cold packs to modify temperature to improve the stability of carbapenems. These changes can be tested using normal patient outcomes for the conditions in which they are applied. There are also generic measures of adverse hospital outcomes that can be directly applied to HAH care if HAH care remains hospital‐based: falls, drug errors, unexpected mortality and post‐discharge readmissions. Although adjustments for the likely less severe illness in HAH patients can be made, the comparisons are worthwhile. However, there is also a challenge in developing, applying and agreeing on quality measures that are specific to HAH care [73]. That challenge is being met, with exciting advances in defining and measuring outcomes—in particular, the application of unexpected escalations back to BAM from HAH. This indicator is only measurable and makes sense if the patient is still considered an inpatient and the HAH episode is unexpectedly ended (or interrupted) by a transfer back to BAM. Thus, these are not readmissions (because patients should not be ‘discharged’ to HAH) but unplanned interruptions to the HAH episode. The causes and outcomes of these escalations are specific to HAH and provide rich data and opportunity for improvement in assessment, care provision, communication and responsiveness to deterioration [74]. It will—finally—provide a basis for comparing (or collaborations between) HAH models, providers or countries. It is wholly appropriate that HAH care is judged as, and against, inpatient care, because its core mission is to substitute for inpatient care. To do that, HAH patients need to remain inpatients and remain within hospital databases. Published criticism of HAH is rare, so it is difficult to directly address concerns. Sachin Jain wrote of ‘toxic positivity’ towards HAH, demonstrating that when enthusiasm becomes an unchallenged mantra, it can alienate those who hold concerns [75]. One of Jain's criticisms is that HAH is a highly complex intervention—but no one within HAH thinks otherwise.

Targets for HAH activity

Scalability is a religious concept for many health executives and planners. HAH will not solve all hospital access problems. At present, up to about 10% of acute multi‐day inpatient hospital bed days have been reported through HAH [76]. Others estimate that up to 50% of hospital medical admissions might be eligible for HAH [77]. If anything between those two figures represents genuine BAM substitution, it is a significant achievement, without any expectations of further growth. Further, HAH adds tensile strength to hospitals—it has proven itself as one of very few viable options open to health systems in situations of extreme hospital demand, where rapid increases in hospital services are required. BAM hospitals are scalable, but only in the long term. This was clearly demonstrated during the COVID pandemic. During COVID, some Spanish hospitals saw rates of HAH approaching 50% of all inpatient activity [78]. Not only were COVID admissions managed in HAH, but non‐COVID episodes also contributed to dramatic scaling of services. Thus, HAH offers the first genuinely responsive alternative to BAM hospitals in the event of a pandemic, seasonal demand or disaster, whether to directly manage those involved or as a way to transfer care out of BAM. Current activity levels have occurred largely in the absence of remote patient monitoring (RPM). In addition, there are some demographic factors that mitigate HAH—particularly the increasing number of elderly people living alone. Living alone, with or without supports, might be possible for many elderly patients when they are well, but that may become difficult when they develop pneumonia or urinary sepsis. It may be that rapid and unlimited access to personal care at home may be necessary to extend HAH for that group. The setting of artificial activity targets for HAH by organisations needs to be carefully considered balancing resource allocation and executive commitment against the realities of cultural adoption of HAH in each institution.

The technologies that can be applied in HAH will expand over the coming decade. That is easy to predict in principle. It is harder to predict where those technologies will come from. It is likely that some of the interventions that can be applied in BAM hospitals will be quickly adopted in HAH. Some will be designed specifically for HAH. Some biomedical advances that improve survival will further strain hospital capacity, so hopefully the design of those advances will have HAH application built into their design from development. These technologies will extend the gains already made in better infusion devices, better intravenous access devices, drugs that require less frequent dosing, better point‐of‐care pathology and imaging [79, 80], and the potential for surgical techniques that might allow earlier transfer to HAH. The advance of HAH, however, will require the ongoing long‐term, assured stability of funding through acute hospital reimbursement, because the technology investment is based on levels of reimbursement related to hospital‐level care. It is easy to imagine that companies preparing new hospital technologies will now consider how those interventions can be delivered through HAH rather than traditional BAM wards. However, pari passu there will also be advances which will detract from hospital and, thus, HAH growth. For example, HAH had rapidly developed a very significant role in the hospital treatment of acute venous thromboembolism with the introduction of low molecular weight heparin in the 1990s, which supplanted continuous intravenous heparin infusions. This continued to grow until the development of oral NOACs, whereupon treatment of those conditions shifted to outpatient care, and now they have little profile within hospitals and HAH.

Economic considerations

If research into the clinical areas of HAH is complicated by the variety of models of HAH, then those variables are magnified in any economic analysis [81, 82, 83, 84, 85]. The control arm for HAH in an economic analysis of HAH is rightly the BAM ward bed. There are obvious inbuilt inefficiencies in HAH care if one considers the travelling and one‐to‐one care provided by usually higher qualified staff and the complex logistics enterprise required to get those staff, equipment, drugs, medical records and pathology to patients across a city, as well as the preparations for contingencies, RPM, urgent visits and a multitude of other tasks. Thus, the potential savings in recurrent direct costs, which decrease the cost to the provider of heating, cooling, lights, food, rent and so forth, may not amount to huge savings because in a hospital many of these costs are marginal. HAH savings have been reported across a wide range, but with varying HAH models. There are other assumptions about costs to HAH patients and carers that are less obvious—for example, there is a cost to families and carers for having their family member in a BAM hospital, and these are often not accurately quantified: travelling to and from the hospital, time away from other duties, car parking and so forth. Moreover, many of the infrastructure costs (heating, cooling, water, food, etc.) attributed to the family as indirect HAH costs would still be incurred even if the patient were in the hospital. But these academic issues are of less significance if we consider the economic elephant in the room. Any expansion of the mega‐hospital comes with the need for capital. So, although the extent of recurrent cost savings from HAH may be debated, the savings in the capital cost of building new BAM beds are obvious, enormous and growing. This is almost never considered in formal economic analyses, despite the fact that in most Western public hospital systems, BAM hospitals are operating at (or beyond) capacity and the system requires more acute beds. The cost of a single new bed in Australia is in the order of AUD$3–$4 million, and this generally excludes the land acquisition cost. It also takes up to 3–4 years to deliver a project of any scale, and that soaks executive attention and time. In addition, despite the growth in BAM beds, they are still inflexible in times of high demand. Not enough attention has been paid to the capital savings, and indeed the capital requirements, of a genuine substitution bed in HAH. This author estimates that a HAH bed (in a unit that is already established) requires approximately 5% of the capital of a BAM bed because an HAH bed still requires infrastructure and is still reliant on some services provided by the BAM hospital, and these should also be costed. When planners set targets for HAH beds, they often presume that the capital cost of HAH beds is zero, and that is incorrect—the budget for reaching any target requires investment and should reflect that cost. But to begrudge an alternative that costs 95% less to establish is genuine false economy.

Remote patient monitoring

The application of RPM has caused excitement, and opportunities for its application in HAH continue to be explored [85, 86]. However, legacy HAH services have demonstrated that positive outcomes can be achieved with basic technologies—more crucial is the adoption of hospital‐like monitoring that underpins any inpatient service: inclusion of patients into a clinical unit; rounding on patients at least daily; monitoring pathology and radiology; case conferences; and providing in‐house after‐hours care. Many HAH services have done that with only daily vital sign monitoring and delivered great care. Intermittent ad hoc vital sign monitoring in response to patient‐observed concerns offers a useful adjunct or replacement of an unscheduled staff visit, particularly after hours. RPM might offer referrers some additional security. Some have linked the development of HAH to the growth in RPM and telehealth (especially for medical review) [87, 88]. But this is not an inevitable association. Remote monitoring and telehealth will add greatly to the HAH toolbox, but it is a tool and not a platform in itself. Thus, like future drug developments, pump designs, point‐of‐care imaging and pathology, or venous access device improvements, remote monitoring will add to the ability of HAH to better care for some patients, some of the time. But the advance of HAH is not contingent on RPM or telehealth. There are several reasons to proceed cautiously along the path to RPM. There is very little (so far) published independent research. First, it must be admitted that HAH‐eligible patients in BAM wards are not monitored, and almost none receive medical care through telehealth in BAM. Moreover, there are very few, if any, objective measures of the clinical impact of RPM or telehealth. For the most part, RPM systems were built on ambitious business models that foresaw the continuous monitoring of vast swathes of elderly people in their homes with armies of remote staff, or preset alarms, pinging for attention. These systems have little application in HAH. Continuous monitoring is prone to interruptions and loss of contact and multiple false positive alarms. RPM is very expensive. Patients and carers prefer face‐to‐face visits [89]. So do HAH clinicians. Telehealth slows down decision‐making and staff visits. Complete removal of medical staff from undertaking home visits removes them from an understanding of the context of individual patients but also of general HAH care, limiting their ability to modify or extend care and limiting their ability to co‐operate and communicate with HAH nurses who are physically in the home [90, 91]. It is very likely that the future will see that HAH directly instructs developers of RPM to produce systems aligned to HAH needs that are more relevant than offerings designed for other purposes.

A pathway to the future

Patient demand, a definition, a name, remuneration pathway, research evidence and even cool technology do not guarantee a future. HAH needs clinicians to commit to it, as practitioners and as referrers. Every new or emerging medical specialty is at a gross disadvantage to the existing medical culture. The existing medical world maintains a clear taxonomy—a classification of subspecies that is accepted. Those species fill niches, and they draw talented people into those species as clinicians, researchers and organisers. Becoming a new species is extraordinarily difficult. Territories are marked and defended, and because medical training is long and difficult, there are incredible forces of inertia applied to ensure that structures that influence career decisions made a decade earlier are still in place when training is complete. So, the creation of a new medical subspecies initially requires existing clinicians to jump from their established genus, and that is not an easy task. In HAH, this has already started: Clinicians have come from a variety of backgrounds—including general physicians, geriatricians, emergency physicians, hospitalists, infectious disease specialists and family physicians. However, most have only had HAH as a small part of their roles, whereas the future will require a commitment to fully engage with HAH due to the demand and the skills required. Not only are adequate numbers required, but those clinicians must be of reputation, status and commitment so that they will contribute to the establishment of a new clinical area. The number of national representative clinical societies for HAH is growing: In addition to societies long established in Spain, Australia and France, there are now societies in the United Kingdom, United States, Taiwan, Portugal, Israel, Canada, South America and the Nordic countries. Postgraduate courses in HAH have started, developing and teaching a curriculum borne of clinical experience as well as academic evidence. These are all predictable and necessary but also very difficult steps on the winding path to clinical recognition, respect and hopefully a viable future for a new way of delivering medical care. It is the only way to attract talented young graduates for whom this represents a career within medicine or nursing.

We are in the midst of the creation of a new (in the modern sense) form of hospital care. It is exciting. In 10 years, every hospital will have a dedicated unit delivering HAH care. Those models will look broadly similar, with unit medical leaders and medical staff and senior nursing staff. These units may be accredited by a national HAH society. Patients will still be referred from the emergency department, the wards, and sometimes directly from the community. Patients will know about HAH and ask to be sent there. It will be staffed by specially trained HAH physicians and HAH nurses and offer high‐quality care at home to patients who would be in hospital, for a wide variety of acute hospital conditions. The evidence for the safe and effective treatment in HAH for conditions will be published by researchers and clinicians in peer‐reviewed journals. The HAH will operate 24 h a day and be able to respond to concerns. Technologies will be developed by companies interested in extending HAH care to many speciality applications. There will be a college (or speciality board) for HAH physicians. Universities will have conjoint professorial positions with hospitals for HAH research and evaluation. There will be at least one international journal dedicated to HAH care. There will be a specific, stable remuneration for HAH care. In 10 years, HAH will be accepted as an integral part of the culture of hospitals, part of their long history and mission to improve the lives of sick people.

Conflict of interest statement

Paid consultancies last 12 months: Kenes International, Mobile Radiology Australia, West Australian Health Department. Honorary: Journal of Advanced Home Medicine.