What is the biggest black box in the NHS? By which I mean, where is the worst ratio of cash to clarity about what taxpayers are getting for their money?

Not hospital services, with their waiting times standards, external inspections, patient choice and tariff payments. Arguably not GP services, even with their alleged hidden local cartels and lowball pay-for-performance contracts. Nor mental health services, despite their geographical block contracts - still intact 17 years into the NHS's purchaser-provider split.

No, the winners of this competition are the community health services: the nursing services, community hospitals, the staff of primary care trusts' provider arms. This is not to deny that they are critical for a well functioning NHS. It's just to affirm that with a price tag of at least£10bn a year - up to£20bn depending on your definition - they cost more than all GP services combined, but are subject to almost no meaningful measurement or commissioning. This means community health services are one of the most opaque and loosely accountable parts of the NHS.

How then to shine some light into the black box? For payment by results technocrats, the answer is said to lie in developing measurement systems for community tariffs - by 2012, perhaps. This is more or less what this same cadre of informaticists were saying when I started work for the NHS 20 years ago.

At that time, a woman called Edith Korner had been appointed by the government to devise new activity measures for community health services. Then - as now - the concern was that typically only half of community health professionals' time was being spent with patients and it wasn't necessarily the right patients who were being targeted for the right level of care.

Two decades later and little has changed. So forgive me for being a tad sceptical about leaving it to the information techies. Perhaps this is why others advocate the more radical alternative of extending individual budgets from social care into NHS care. That way, they argue, users/patients will be able to get a personal and accurate grip on what's being bought in the community health services, even if the NHS commissioner cannot.

Hence recent reports that the government will soon announce individual budgets in the form of vouchers for patients with long-term conditions such as multiple sclerosis or motor neurone disease. These would mean "patients can shop around for care, arrange visits when they want, swap one type of treatment for another or buy their services from the voluntary and private sector", according to an interview with health secretary Alan Johnson. As a Sunday Telegraph editorial put it: "At last, a radical policy has emerged from the post-Blair Labour government. If the implementation matches Mr Johnson's rhetoric, it will mark a revolution in the NHS."

But will it - or more precisely, can it? First, the easy answer. For long-term conditions with predictable costs, with healthcare needs that blur into social care (eg chiropody, physiotherapy or the management of incontinence), where the patient is able to monitor the value of what is being provided and where the supply side is readily contestable, then not only can individual budgets or vouchers work, but they could in many cases be converted into social services-style direct cash payments for the patient to deploy.

Now for the more complex reality. Take the case of a patient with diabetes. Attempting to carve out a budget - let alone a direct payment - specifically for diabetes runs up against all kinds of conceptual and practical problems. Does the budget cover all conceivable complications of diabetes? For example, diabetes is implicated in much peripheral vascular disease, but not all peripheral vascular disease is caused by diabetes. Or to take another example, people with kidney disease may often have diabetes but not all diabetes causes kidney disease. So there is the difficulty of circumscribing budgetary scope, attributing causation and dealing with the effects of co-morbidity.

There is, however, a more elegant solution that avoids the pitfalls of trying to carve out disease-specific health budgets. And that is to allow carved-out individual budgets for the whole of a patient's healthcare needs for one year.

The advantage of this approach is that it is not necessary to try to prospectively partition the likely components of a person's healthcare consumption between disease states. So while having one of a pre-specified list of long-term conditions may qualify you for this new type of commissioning arrangement, your annual person-specific risk-adjusted budget would be set holistically to cover all aspects of your healthcare.

Who would hold this budget? Patients probably need a commissioner to aggregate risk on their behalf. But that does not mean the local PCT has to do this. Instead, eligible patients could choose a third-party commissioner to manage their individual commissioning budget, based partly on who would do a better job of unleashing value from the billions being spent on community healthcare.

In time you would expect to see specialised health and social care commissioning organisations develop. They would receive capitated payments, incentivising them to support people with long term conditions, for example, to avoid hospital or nursing home admissions.

Fortunately there are several international models for what this looks like. Adapting them for the NHS could mean stimulating new models of care co-ordination and delivery, particularly for those patients who drive most cost, genuine patient choice of commissioner and needed change in community health services. All those in favour?

What do you think of Simon's proposal? Email your comments to hsjfeedback@emap.com or post them online by clicking the "post your response" button below.