Managing demand in the downturn – thinking the unthinkable, held last Thursday, was the latest in the NHS Confederation’s policy salon series looked at what demand management may be possible in recessionary times. The previous session was on Payment By Results in the economic downturn.

The starting point for this session was the Social Market Foundation’s recent From Feast To Famine report, which we discussed here.

The report’s lead author suggested that the three categories of potential action to address the NHS funding shortfall are:
1. raise taxes (believed to be unfeasible)
2. become more efficient – spend better, making across-the-board efficiency savings, something at which the NHS has not previously been good
3. manage demand – should be on menu of options

Can demand be reduced?
It was proposed that healthcare is not immune to the economic rule that demand and use respond to price. The use of healthcare does not always result in increased or maintained health (given the harm issues thought to result from 1 in 10 healthcare procedures). The RAND experiment was cited as showing that higher co-payment reduces utilisation, possibly without affecting health.

This led to two further questions:
what are the most effective demand management tools?
is charging the answer? (the decision is in an NHS context a political move, not just policy lever; and the report’s proposals was to use charging for GP appointments mainly as a ‘nudge’ (from behavioural economics), not a revenue source)

The SMF report’s ideas were charaterised as deliberately challenging in order to stimulate debate and expand the limits of what is deemed possible. The suggestions is that the system would remain totally free for those on lowest incomes.

Those above the lowest income level would pay an initial charge of £20 for each GP visits (capped at £100 a year), wth no charge for follow-up visits related to a diagnosed comdition.

The report proposed that the NHS should be fair, but that doesn’t mean free – the ‘free’ idea prevents honest discussion about limits of healthcare. This makes NHS GP access a good that is no longer priced at zero – sending potential users a signal that appointments are a valuable resource, to be used carefully.

Healthcare is, the lead author added, a supplier-driven good, admitting that most spending happens after GPs refer patients into the system. Nevertheless, he suggested that the level of (over-)reaction to the proposal suggests that something is wrong in public debate on healthy policy.

Looking at evidence of user charges as blunt instruments
The second speaker, from LSE Health, explored what evidence-based information was available in three areas:
1. looking at efficiency arguments for user charges
2. considering more sophisticated approaches to user charges to manage demand
3. reviewing the SMF proposal.

In the efficiency argument, it is evident that for some healthcare of low value, the cost of production outweighs any benefit (and some can be harmful as well as ineffective, such as antibiotic prescribing for viruses). Conversely, some is high-value and its use outweighs costs. The aim should be to stop use of low-value services and ensure access to high-value.

Economic theory confirms that user changes mean people paying have to think twice. This is based on a number of assumptions: that people have good information on the product to be consumed and can use that information; and that patients' decision-making is rational.

In healthcare, the speaker added that these do not always hold and may not for some groups. Research evidence on user charges finds that they encourage people to forego use of low-value care - but also high-value care. User charges are a very blunt instrument, which actually have potential to introduce inefficiencies into use of services.

Another potential introduction of inefficiency with user charges arises through the ‘waterbed effect’, where pushing demand down in one part of a constrained system causes it to pop up somewhere else. Thus more people may go to A&E as a result of user charges for GP access, leading to no net saving and possibly higher costs overall.

Value-based cost-sharing (VBCS) offers a more sophisticated approach to managing demand. It can be targeted to low-value care and abolished for high-value. The LSE Health speaker’s view was that VBUC offers less risk to efficiency and equity. Three examples how it has been used are:
1. in the US and France, VBCS is used to direct away from branded to generic drugs
2. in France and Germany, Best Value User Charges are used to get people to register with GPs and get GP referrals to specialists (the tradition in both countries is one of direct access to specialists). E10 in Germany, in Fr 50% of cost (30% if registered)
3. in the US and Germany, VBCS encourages to enrol in disease management programmes

The speaker noted that VBCS have to be very carefully designed to avoid worsening matters, and certainly create higher admin and transaction costs in the short term. They should be used if a system must introduce charges, but after having exhausted alternative approaches.

Is there a role for VBCS in the NHS? It already has effective value-based supply-side controls – generic substitution by PCTs and pharmacists, and the tradition of GP registration. The SMF report proposes putting the incentive not on the commissioner or provider but on the patient.

The second speaker concluded that VBCS does not seem very viable or feasible in the NHS, particularly since GP appointments are high-value healthcare in temrs of being demonstrably efficient, associated with better outcomes and ppotentially preventative of wasteful spending.

Charging at point of use has no effect on the price of services or the intensity of their use once a patient is in the system. Equity issues were also raised: introducing formal charges might undermine the middle-class taxpayers’ sense of inclusion and commitment, especially if insurers develop policies to cover all or most of such charges and private GPs draw people and funding away from the NHS (the ‘bourgeois bale-out’ thesis).

The second speaker concluded that there is no evidence user charges enhance efficiency, and are more likely to prove inefficient as a mostly indiscriminate blunt tool. The NHS’s challenge in primary care is one of managing supply, not demand.

German lessons - sprachen sie co-payment?
The third speaker, from an independent consultancy, reviewed the impact of introducing GP charges in Germany. Though the co-paty there was effective in reducing demand, it had consequences elsewhere in the system.

Research data on the German experiment found that 27% of those surveyed reported having delayed a GP visit; 18% avoided one; 25% made an additional visit (in the German approach, one payment makes GP visits free for a quarter of the year); and 30% reported no change to their behaviour. 20% of over-60s reported having avoided or delayed a visit. The poorest avoided visits most, 2/3 of poorest reporting having done so.

This is clearly perverse in terms of healthcare need. People who perceived a GP visit as hard to fund (whatever their actual income) also don’t go.

Assessing foregone care, studies in Germany and Greece found that those in the in lowest income group are twice as likely to not go to see a GP.

Charging affects behaviour, clearly – is it the right affect on behaviour and what the are long-term cost consequences of the poorest people with long-term conditions not seeing GP while the affluent ‘worried well’ are going?

The third speaker proposed alternative approaches to introducing GP charging, pointing out that recent events have proven that when the NHS focuses on stuff, it can actually do it. The Better Care, Better Value generic prescribing move gave GPs data on their statin prescribing (they’d had it forever, but done little with it), but in the following two years, with dashboards and “a bit of PCTs shouting at them”, generic use rose from 60% to 75%. Pfizer were reported to have calculated that this cost them £0.25 billion. The key to success was to be transparent with data: within three months, every PCT increased generic statin prescribing.

The opportunity of financial crisis creates common acknowledgment of the problem. Aligning management behaviour and information about outcomes can affect NHS behaviour fast.

Another approach could be to intrduce co-payments for people with long-term conditions who don’t participate in disease management programmes (which the NHS does not do systematically, being less proactive than Australia, Canada and America - insurance-based systems tend to do disease management “not because it’s adorable, but because it lowers the cost of healthcare”.

In Medicare, a benchmark control group of patients with heart failure were compared against a group given a disease management programme and nurse coaching. Over two years, the intervention group halved acute admissions, and there was a 15% drop in their treatment costs.

General discussion
It was clear that the NHS supply-side, and the primary-secondary interface, are where big savings lie. One participant observed that in NHS policy discussion, user charging falls into the same camp as acute reconfiguration – a diversion from real issues.

Another suggested that for much of the last 10 years, “I think we have greatly reduced the standing, responsibility and capability of the GP service. The key question now is how to free up GP time and resources in a cost-constrained service to deliver disease management programmes? By a shift resources out of the acute sector, or by revising the quality and outcomes framework (QOF) to focus more on case management? The time of clinicians in primary care is limited, and introducing case management is a huge new job – but it’s got to be done”.

Noting David Colin-Thome’s recent comments on PBC ("the corpse is not for resuscitation!”), one observation was that the “corpse is still getting the pound notes”.

A primary care specialist noted that attention needs to focus on GP referral, citing their experience of a practice’s weekly review by two external clinicians empowered to prevent referrals they deem unnecessary. Their intervention causes irritation, but is thought more effective than internal reviews (which can often tend to the diplomatic).

In terms of the potentially complex and opportunity-costly mechanism of user charging, the possible effects of user payments for ‘health-positive’ behaviours were raised – without any clear consensus.

In the same ilk, a suggestion was raised that there could be a charge on doctors who are outlier non-prescribers of generic substitute drugs – without any clear consensus.

The challenge of introducing payment for inappropriate attendances, whether in primary care or A&E, sits on the subjective assessment of what is appropriate. In passing, it was observed that the strongest correlation for growth in A&E attendance was the introduction of the 4-hour waiting time target (the new GP contract, devolving out-of-hours, had much less clear impact).

Quip of the day was from a chief executive, who observed that general public fondness for the NHS is on a level with that for the late Diana, Princess of Wales – on an emotional basis rather than a rational one.

Ambitious demand management plans under way in London aim to divert resources from high-cost acute care into lower-cost primary care: “this can save proper money and take out the hospital cacti”. The tactic is one of managing demand by managing supply. This requires clinical commissioning, and GPs are the gatekeepers.

London (which spends £16 billion of the NHS’s £109 billion budget a year) is focusing on areas where there has been success in developing out-of-hours care – asthma and respiratory disease in children, where community programmes have lowered admissions and mortality. Paediatrics in the capital generally has been managing with fewer beds and admissions: all these projects are based on close partnerships between consultants and GPs and joint programmes. They also requite the right alignment of institutional arrangement and incentives.

One challenge is how to integrate the introduction of disease management programmes better into primary care and general practice, otherwise they will become “a deflection system that doesn't deflect and a double-cost – like NHS Walk-In Centres”.

A management consultant suggested that a number of health economies which in good faith planned to do these things – develop community services, and reduce acute provision – ended up runing the numbers and found that it would result in the loss of consultant-level jobs in the acute sector and the creation of a few more community nurse jobs. The prospect of medical unemployment of up to 10% of some acute trusts (particularly of senior medics) led to a “you can’t do this” impasse.

Some clinical commissioners have capped activity and are paying more, with HRG4 inflation.

It was suggested that the payment system needs to be more stable; increased sophistication tends to increase price. The NHS also needs to develop a balanced systems to get away from a win-lose mentality.

One participant’s challenge was, “we’re basically saying there’s nothing terribly wrong with how the NHS works, except that it costs too much and is terribly inefficient. GP gatekeeping drags people off the street and pushes them though the system. Ours are now the most expensive doctors in the world per hour of delivery related to clinical training”

Putting in more absolute and proscriptive referral rules was discussed, but any such rules woud have to be “too complicated. We know almost everything we need to know – there’s no magic bullet; everything works a bit; and we have to keep going with a bit of everything”.

It was suggested that it would be nice to hear about areas of community management progress other than on COPD (the most common example – although even this is by no means universally maximised).

One challenge was thought to be the relatively few patients eligible for CDM programmes per practice – one SHA chief executive was reported to have ruefully observed that they had enough GPs in their area to “man-mark” the chronically ill patients. However, the dissipated numbers of patients risks seeing the establishment of an inefficient ‘cottage industry”.