The notion that many people who were previously cared for in hospital could be cared for just as well (or even better) at home is now widely accepted. Indeed, it is a central plank of UK government policy. I want to gently question this idea – it is in danger of becoming something of a trope.
Before laying out my argument, I would like to recount an event – the death of Emperor Hirohito in 1989. After a long and eventful life, the man who signalled the end of World War Two was on multiple organ system support. He was as intensively supported as anyone in the world – perhaps more so if you read the account. Yet he was at home – there in the Imperial Palace. Likewise, King George VI had his pneumonectomy in Buckingham Palace. The point I am making is that anyone can be treated at home up to any intensity; whatever level of intensity required, it can be provided at home. However, it is not. This is for one reason and one reason only – cost.
As a result, we have hospitals that concentrate resources – plant and people that are in close proximity can treat people at much lower cost than when those facilities are scattered over a wide geographical area.
So, the issue of who should be treated where, is an economic one. It is preferable to treat people at home but only up to a point. It might be preferable to pay more to treat more people at home. But again, only up to a point.
In that case, where is the threshold?
The threshold falls where the marginal gain exceeds the value of the opportunity forgone by making that choice.
But how much more expensive is it to treat at home a patient who, in the counter-factual situation, would be tucked up in a hospital bed (where they would be intensively cared for, but nevertheless be at an increased risk of infection, confusion, alienation, dementia and so on)? This expense turns on two things – the relative cost difference and the relative health outcome difference.
Here we need to think of two very different scenarios – hospital at home after early discharge, and hospital at home for people at the point where they would otherwise be referred.
In the second scenario we must expect two things. First, the relative difference in cost between community and hospital is likely to be (much) higher than in the first scenario (of earlier discharge). Why? Because the intensity of care in much higher when a person is first admitted. This is where hospital specialists make a careful assessment – often many different specialists are involved. This is also when most tests are done and, importantly, where information is transmitted between different care givers. It is also the phase during which observation is most intense.
All of these activities can be delivered in the home care counter-factual. However, replicating all the above activities over a geographical area ratchets up the costs. Travel times are also much greater. Surges and troughs in demand mean either that some people will be neglected or that there will be unused capacity. All the above factors could be quantified by means of time-and-motion methods, but such studies would be very expensive if done properly. And they would tell only half the story. This is because hospital avoidance schemes cause supply-induced demand. In other words, a proportion of people cared for at home would not have been sent to hospital in the counterfactual scenario.[1]
Post-discharge hospital at home is an altogether different kettle of fish. Here the patient has already had the bulk of tests, diagnoses have been made and they have passed the critical phase where rapid deterioration is a real risk. Such patients can be safely monitored and cared for remotely.
What does the evidence say? To summarise a large literature in one phrase – hospital avoidance schemes are more expensive but safe and preferred by patients.[2]
My advice to policy makers is to concentrate on early discharge schemes and use hospitals to provide care for the acutely sick. My research suggests that it is hard to scale-up hospital avoidance schemes.[3] Hospital avoidance has been all the rage for ten years, yet our hospitals are busier than ever with very long waits. It will cost more to reduce pressure by substituting hospital provision with a system to safely avoid hospital admission in the first place.
So I recommend supply-side investments to increase front door capacity in hospitals and facilitate discharge pathways, as they are potentially cost-releasing. Fortunately, ARC WM researchers Sam Watson and Paul Ellis, supported by Alice Turner, are investigating a Birmingham-based initiative to increase hospital at home provision. This is a superb study because it is built around a service, not researcher-based intervention. This makes it much more sustainable and even scalable than an intervention dreamed up by researchers and then implemented with agreement of service managers who do not really feel they own the thing. In other words, opportunistic research promoted by ARC WM.[4] This large study funded by NIHR HS&DR is unusual because it is a cluster RCT and will harvest all its quantitative data from routine sources.
* Some people make a strict distinction between virtual ward and hospital-at-home, which, to me, seems unnecessary.
— Richard Lilford, NIHR ARC WM Director
References:
- Taylor SP, Golding L. Economic Considerations for Hospital at Home Programs: Beyond the Pandemic. J Gen Intern Med. 2021; 36(12): 3861-4.
- Knight T & Lasserson D. Hospital at home for acute medical illness: The 21st century acute medical unit for a changing population. J Intern Med. 2022; 291(4): 438-57.
- Apenteng P, Harris C, Bird P, et al. Interventions for assessment and medical care without hospital transfer for older people living with frailty: findings from a formative evaluation. [Preprint]. 2025.
- Lilford RJ. Reflections of an ARC Director 4: ARCs and Their Role in Service Evaluation. NIHR ARC WM News Blog. 2022; 4(5): 1-5.