General practice is rightly venerated as the bedrock of the NHS, providing care from cradle to grave, offering succour to the needy and comfort to the terminally ill, as it has done for the past 67 years.

But is this model sustainable in the modern world? I do not believe it is. An ageing population, the increasing complexity of patients’ conditions and the expectation of a growing range of services are putting intolerable pressures on general practice.

GP consultations have risen by more than 70% in England since the mid-1990s, and are projected to rise still further. Yet attendances at A&E have also soared, provoking criticism about GPs’ failure to adapt to changing demands. Meanwhile the proportion of doctors going into general practice in England has declined from 44% in 2003 to 33% in 2013.

We know that different people want different things from their GP yet we persist with a one-size-fits-all model. People in work want to pop into the surgery to pick up a prescription quickly and conveniently. Older people want continuity and a focus on their social needs. Those with long-term conditions want coordinated, efficiently planned care and support to look after themselves.

In 2013, I chaired the London Health Commission, established by Boris Johnson to examine the health of the capital, and a part of our task was to examine primary care. We concluded that to meet these contrasting demands, GP care should be more personal, planned around groups of patients with similar needs rather than around professionals with similar skills.

Elderly people with multiple chronic conditions have the greatest needs and are the biggest users of the NHS. We suggested their health could be improved by joint teams working across specialities and organisational boundaries which would provide care holistically, considering all aspects of their situation.

As Labour minister of health in 2008, I recommended bringing GPs together with hospital specialists and social care in polyclinics in my NHS Next Stage Review, as part of a wider drive to improve quality. But the idea foundered under opposition from the BMA and others.

Now attempts to introduce integrated care are being made in the so-called vanguard GP practices, launched by NHS England last year. Six of the 29 vanguards are focusing on care homes, linking with social services, housing and local authorities to offer better, joined up care.

In Gateshead, where GPs have been working for two years in care homes, hospital admissions are down by 9%. Now they have been designated as a vanguard project, they expect to go further in keeping people well where they live.

But to see a more fully realised example of this approach, you need to cross the Atlantic. It seems unlikely that the US would have anything to teach the NHS about primary care. But a family-owned company in Miami, Florida, has pioneered an approach that is attracting attention from across the US and further afield.

It operates 37 neighbourhood medical centres across six US states providing care exclusively to moderate- to low-income elderly people with multiple chronic conditions – the most expensive patients in the world’s most expensive healthcare system.

The strategy is this: provide the patients with intensive primary care and in that way keep them healthy and out of (expensive) hospital. Patients are seen once a month on average, normally by the same doctor, and have longer appointments than in comparable practices. There is a pharmacy on site so medicines are dispensed on the spot and patients can be shown how to take them and encouraged to discuss side effects. Medication adherence – a constant challenge in the elderly – is high. Vans are available to take patients to and from the clinics to overcome problems of access.

By focusing exclusively on the care of the elderly – the average patient is aged 72 – ChenMed has introduced a form of specialist general practice. The results are striking: patients spend 38% less time in hospital yet have higher ratings for care quality, health outcomes and patient satisfaction than comparable centres. Spending is 15% lower – some of the saving from reduced hospital stays goes on increased investment in primary care.

For the patients the care is free – funded by Medicare, the public health programme for the elderly. The budget is higher than in the UK, reflecting the greater expense of the US health system. Under the Medicare Advantage scheme, US providers such as ChenMed are allocated a national average of $9,600 (£6,700) per year per patient – compared with the estimated £4,000 annual cost for elderly patients in the UK with at least one long-term condition.

The key to the success of the scheme is the low doctor-patient ratio – one to 400 patients compared with one to 1,500 in the UK – which allows ChenMed doctors time to build a personal relationship with their patients. That presents a workforce challenge for the UK – it means training more GPs. But, in addition to the savings on hospital costs, there may be additional benefits. GPs caring exclusively for a healthier, younger population may cope with 2,000 or 3,000 patients or more.

Providing pro-active rather than re-active care improves health, avoids crises and saves costs. The challenge will be to persuade GPs wedded to the cradle-to-grave model that specialisation and joint working is the way forward. We know all parts of the NHS must modernise and the bedrock cannot be exempt.