Never in the field of healthcare was so much (recurrently) consumed by so few

First off, apologies for the dreadful bastardisation of Churchill. And while the current crises facing healthcare systems are incomparable to the existential threat of WWII, we are nonetheless barrelling, at an accelerating rate, towards disaster in the provision of healthcare in many countries. A new vernacular has emerged which symbolises the magnitude of the problem, where the terms ‘ambulance ramping’ and ‘corridor care’ are now commonplace. I’d recommend reading the Darzi report into the NHS for a reality check. Every single aspect of the system is performing somewhere on the spectrum between abysmal and dangerous and that’s before we get to the lack of integration. Avoidable morbidity and mortality on an industrial scale.

Incomprehensible amounts of money are invested in R&D to develop new tests and treatments for every imaginable ill. While this has extended lifespan and helped to treat and cure disease that previously killed us, it has (literally) come at a price that society can no longer afford. The irony of the monumental advancements in scientific discovery is that there are many more things that can be done to us at an ever increasing age. In most industries, technological advancement tends to reduce the cost of production. In healthcare, it just increases the volume of consumption. Anyone betting that AI, digital, personalised medicine, and all the other focus areas of investment will bend the proverbial cost curve are delusional (I’m sure there’s a digital start-up, powered by AI, with a sure-fire cure for delusional disorder).

Throwing more money at healthcare hasn’t addressed demand. In most major OECD countries, healthcare now encompasses more than 10% of GDP and growing. Even the woes of the NHS can’t be attributed to funding, which has increased in real terms, including on a per capita basis, despite the wilfully uninformed commentary routinely vomited up by idealogues:

Yet despite this additional funding we have woefully insufficient capacity to provide healthcare to the population and the problem is accelerating. In some countries, such as the UK, public finances are now so dire that the bailiffs (or bond markets) are knocking on the door. No one seems to have any real answers, least of all government, aside from pillaging the developing world of all it clinical resources to plug the holes in the dyke.

The complexity of healthcare systems structure, funding, and regulation is a major barrier. Expecting positive systemic change is, however, a futile exercise. Thus, the question to be asked, is what can practically be done now to make a material difference at scale? A good starting point is always to follow the money, though if we take the UK as an example, there isn’t even reliable data on expenditure categorisation. In a paper in May 2024, the Nuffield Trust wrote that:

“The Department of Health and Social Care (DHSC) released a breakdown of spending from NHS England’s commissioning budget, drawn from commissioner in-year performance reporting. While this purported to show how around 95% of the NHS budget for England had been spent in recent years, closer inspection begged more questions than answers. In particular, the 47% cash increase in funding reported in the figures for community health care between 2019/20 and 2022/23 did not marry up with experiences of people running and using such services, leading some to suspect a classification error.”

In any business or organisation, if you don’t have granular, real-time insight into your cost-base then you’re likely doomed. Not to worry, however, healthcare is only around 20% of UK government expenditure annually. I’m sure they have a really firm grip on the other 80%.

What we do know is that an increasing proportion of healthcare budgets are spent on hospitals, and not just because of Covid. As Nuffield’s own analysis demonstrated, more than 50% of the NHS opex budget is spent on acute hospital care, and that’s before adding-in opex on psych hospitals and hospital capex:

The same pattern is true in Australia and many other countries. Health systems are becoming more hospital centric despite rhetoric and aspiration to the contrary.

If we are to make meaningful impact at scale on healthcare, we therefore need to start with the drivers of hospital utilisation. There is a myriad of perspectives on this question, including factors such as aging and lifestyle, with many pointing to the relative underfunding of primary and community care as the main culprit. While there’s validity to this argument, it is worth first considering who is in hospital. This may seem like a straightforward question, but curiously it is one that’s rarely asked. Healthcare systems are (usually) good at measuring activity such as number and types of procedures and expenditure by modality. Few understand the patient population beyond basic demographics and diagnoses. And if this is not well understood, then the capacity to reduce hospital admissions is fundamentally compromised.

Using the NHS as an example, there is data showing 17.6 million admissions to hospital in 2023/24, though try finding any data on unique patient admissions. Was it one person admitted 17.6 million times (unlikely), 17.6 million people admitted once (unlikely), or somewhere in between? This is a critical question as it is much more manageable to impact a small number of intensive users than it is to impact an entire population that are infrequently and unpredictably admitted to hospital.

When looking at this in Australia a decade ago, we found that c. 3% of a large population (many millions of people) attended hospital at least twice per year and recurrently accounted for around one third of hospital admissions, bed days, and costs (this excludes chemotherapy, dialysis, endoscopy, maternity patients). The term ‘recurrently’ is key here. These are people not generally in the last 12-24 months of life but who frequently attended hospital over many years. They were overwhelmingly multimorbid and, interestingly, over a third were under the age of 60.

These people are also heavy users of outpatient services as well as imaging, pathology, pharma, devices, and all manner of other healthcare costs. Dimensionally, if 3% of the population recurrently consume a third of total healthcare services and costs in the UK that would mean c. £47,500 per annum per person expenditure on healthcare for these people.

The evidence on multimorbidity and consumption of healthcare in the UK supports the Australian analysis:

• According to Cassell et al (The epidemiology of multimorbidity in primary care) studied a random sample of 403,985 adult patients and found that 27.2% of the patients involved in the study had multimorbidity and these patients accounted for 52.9% of GP consultations, 78.7% of prescriptions, and 56.1% of hospital admissions.

• Stokes et al (Multimorbidity combinations, costs of hospital care and potentially preventable emergency admissions in England: A cohort study) studied data from 8,440,133 unique patients with an elective or emergency inpatient admission in 2017/18 and found that patients with no chronic conditions (42.4% of the sample) contributed to 23.3% of the total secondary care costs, patients with 1 chronic condition (25.8% of the sample) to 21.4% of the total costs, patients with multimorbidities (31.8% of the sample) to 55.3% of the total costs.

• Soley-Bori et al (Impact of multimorbidity on healthcare costs and utilisation) conducted a systematic review of the UK literature and found that patients with four or more chronic conditions have almost 15 times the odds of experiencing an unplanned potentially preventable hospitalisation. The combination of mental and physical conditions particularly increases the probability of unplanned hospital care to between 58% and 100%. In primary care, having multimorbidity more than doubles its expected use (Odds Ratio = 2.56) compared with having 0–1 morbidities.

• Soley-Bori et al (Disease patterns in high-cost individuals with multimorbidity) study of 386,238 GP patients in London found that 101,498 (26%) had multimorbidity. The high-cost group (n = 20,304) incurred 53% of total costs in general practice and had 6833 unique disease combinations.

While the consumption of so much by so few is hardly a revelation to people working in healthcare, it does beg the question as to why there’s been relatively little success in preventing hospitalisation among this cohort. Virtual hospital and homecare models have had an impact, though these are usually in the context of substituting for a physical admission or else post discharge follow-up to minimise the risk of readmission. The crucial element as is well known is primary care, yet the current strategy of funding general practice to manage multimorbid and complex patients has generally failed. While funding, resources, and time pressure are clearly issues, a critical problem is that the operating model for traditional general practice is ill-suited to managing this patient population. And nowhere has any GP service demonstrated capability to successfully operate a traditional high volume, low-intensity, reactive service concurrently with a low volume, high-intensity, proactive service at scale.

Indeed, one of the main reasons that there has been such a proliferation of dedicated primary care services in the US for elderly, chronic, and complex patients is the recognition that the model required to support these patients is so fundamentally different to traditional primary care that bespoke facilities, staff, technology, and delivery channels are essential to success. These dedicated models have proven successful at impacting utilisation and outcomes, generally reducing hospitalisation and ED attendances by around 50%. So much so, that the operators of these services work on risk-based funding models.

While countries such as the UK and Australia don’t have the funding flexibility afforded by Medicare Advantage in the US, it is my view that the establishment of dedicated primary care services for multimorbid and complex patients in these markets is long overdue, given the crises in ED and hospital services. This would entail a highly data driven, predictive, proactive, and preventative model and would, in each geography, require a SWAT team comprised of:

• Dedicated clinics (5-10 FTE GPs) with small patient panels (500 per FTE GP vs 1,500 in traditional general practice), meaning 2,500 – 5,000 patients per practice

• Chronic care nurses, care coordinators, social workers, allied health, and pharmacists to focus on care planning, coordination, social determinants, and polypharmacy

• Psychiatrists (given the high prevalence of physical and psychiatric comorbidity) and other sessional specialists

• Clinical homecare services to assess and regularly visit patients in the home

• Remote patient monitoring

• Patient transport as required to bring people into clinic

From a tech perspective, there are excellent systems around today for multidisciplinary and multichannel care management, integrated with a modern cloud PMS and all the digital patient and clinician facing tools required for engagement, efficiency, and efficacy. Critically, such systems would enable a data-led model using established and proven algorithms to predict deterioration and hospitalisation, to identify and close gaps in care, to monitor patients remotely, to deliver a plethora of digital and telehealth services, to enhance service delivery in real-time, and to demonstrate superior outcomes versus matched cohorts.

While I fully realise that such a model would be fraught with politics and complicated in the UK by the GP Contract (and in Australia by the Medicare fee for service funding model), this ultimately would be positive for GPs. Not only would it free-up significant capacity by moving a small number (c. 3%) of very high utilising patients to a dedicated service, it would also enable those with an interest in this area to spend part/full-time working in this environment. Moreover, if we could reduce hospitalisation and ED presentation rates by half in this cohort, that would eliminate c. 16% of current hospital admissions. Great for the system and even greater for the patient that avoid hospital and those on waiting lists. The prize is significant.

If this sounds of interest, please feel free to reach out.