Intended for healthcare professionals

Opinion Acute Perspective

David Oliver: With virtual wards, NHS England is overpromising once again

BMJ 2023; 383 doi: https://doi.org/10.1136/bmj.p2787 (Published 29 November 2023) Cite this as: BMJ 2023;383:p2787
  1. David Oliver, consultant in geriatrics and acute general medicine
  1. Berkshire
  1. davidoliver372{at}googlemail.com
    Follow David on Twitter @mancunianmedic

The term “virtual wards” gained fresh momentum in the NHS during the first waves of covid-19. Before the pandemic it had referred to community based, multidisciplinary teams case managing patients at high risk of acute hospital admission to prevent or respond to crises and keep them at home.12

It was then repurposed in 2020-21 to describe telephone and digital monitoring of symptoms and physiological oxygen observations for people mostly with respiratory covid-19, to help support them at home with a direct link to a secondary care clinical team.3 Some individual trusts456 that reported local success with this model published their data or extended it to other respiratory infections or exacerbations and were then singled out as exemplars by NHS England (NHSE), which in turn picked up on the idea as a way of reducing demand and activity in acute hospital sites by providing a safe, cost effective alternative.7

Having initially promised 7000 new virtual ward places in 2021, NHSE then set a target of 10 000 places in its 2023 Delivery Plan for Recovering Urgent and Emergency Care Services8 and provided an additional £450m over two years for local expansions. It now tells us that this target has been achieved and that more than 100 000 patients were treated in such wards in the year to March 2023, with politicians also claiming credit.9

NHSE also claims that most of these virtual wards are effectively using the “hospital at home” model—a different, long established, and better researched model10 than the one more narrowly focused on respiratory illness seen during the pandemic response, thereby muddying the terminology further. It also claims that the hospital at home model is helping to reduce admissions. I find some of the developments and conflations problematic for several reasons.

First, in public communications NHSE is now routinely using the terms “virtual ward” (in the covid era sense) and “hospital at home” almost interchangeably, even though they’ve never been synonymous. The 10 000 beds are inevitably a patchwork of different models depending on local practice, with the label applied indiscriminately to claim expansion of places and the additional central funding. NHSE lumps together examples based on single (often respiratory) conditions and more generic ones for admission avoidance or early discharge in frail older people, as well as clinical teams of very different sizes, skill sets, and medical oversight. Yet NHS Providers claimed in a recent article that patients on virtual wards received “consultant-led care, from multidisciplinary teams, overseen both locally and remotely.”11 Can we know that?

Feeding the numbers

Second, with a finite number of doctors, nurses, and allied health professionals, there’s a risk of robbing Peter to pay Paul: the staff in virtual wards will have to be taken from other parts of the system. The shift of care to community settings may place additional strain on existing primary and community care staff who are not part of the “virtual ward.” And that £450m could have been spent on existing community services to help patients flow out of acute hospitals. We now face seeing patients with ongoing peri-acute problems being supported to leave or stay out of hospital, while others with no acute medical needs are stranded there for want of underfunded health and care services, feeding the virtual ward numbers.

Third, much of the existing research evidence is patchy or inconclusive. Local service descriptions and good news stories are always interesting—but without rigorous, independent statistical and methodological evaluation and data they provide no basis for central agencies to set big targets and roll out the model nationally. A recent rapid evidence synthesis in Age and Ageing found low to moderate evidence that clinical outcomes including mortality were probably equivalent or better with the “hospital at home” model.12 Subsequent residential care admissions were “probably reduced.” Evidence of cost effectiveness showed “methodological issues,” meaning that the results were uncertain. Evidence on carers’ experiences was lacking.

A 2021 systematic review13 of the hospital at home model—both for admission avoidance and early supported discharge, which in turn included several earlier systematic reviews—found that “for suitable patients, [hospital at home] generally results in similar or improved clinical outcomes compared with inpatient treatment, and warrants greater attention in health systems facing capacity constraints and rising cost.” It also concluded that evidence was poor for carer experiences, cost effectiveness, and clinical complications, while many of the studies were small or methodologically problematic.

None of this means that virtual wards or hospital at home models are a bad thing. For the right selected patients they can provide a more acceptable alternative to hospital admission, often in line with patient preference and without the attendant risk and complications of admission to overcrowded hospitals. Two recent British Geriatrics Society position papers on care outside hospital and rapid community responses to older people have set this out clearly.1415 But none of it is a good enough basis for setting ambitious targets or assuming that such approaches are a silver bullet.

Failing to learn

Fourth, we’re failing to learn from history. For two decades we’ve heard the Department of Health and Social Care, NHSE, and government ministers announce the rollout of shiny new service models, based on limited peer reviewed evidence. On retrospective evaluation they found that the models had never delivered the promised scale of reductions in hospital attendance, costs, admissions, or bed days at a population level, no matter how good the benefits for selected patients.1617 Yet here we are again.

Finally, it’s a big assumption that all or most patients who are selected to be enrolled in a virtual ward or a hospital at home scheme would otherwise have been admitted. Perhaps in many cases they would still have been sent home, but now they have additional reassurance and safety netting to reduce the risk or anxiety of being at home.

By all means, provide more out-of-hospital care because patients like and value it—and because, for selected groups, the outcomes can be at least as good as admission. But don’t assume that it will solve the lack of capacity in acute hospitals, while ignoring the opportunity costs and taking decision making away from local health systems by setting their priorities for them.

Footnotes

References