21 JULY 2007, Page 16

MATTHEW PARRIS

Road congestion and casualty waiting times are explained by my Rut Theory of Queueing Afriend twisted his knee badly playing football last week. In considerable pain next morning and able to bend the knee only with difficulty he contemplated going to an Accident and Emergency unit at a London hospital. The alternative was to assume his injury was what he took it to be — a twisted knee, no more — and that there was no point in queuing for many hours only to be told to bandage it up, take a painkiller and anti-inflammatory tablets, borrow a pair of crutches and try to rest the knee as much as possible. Such things he could organise without specialist advice.

But he opted to go to A& E after work. For two reasons. First, what if it were more serious than it looked? Second, maybe the queue would only be for an hour or two at most, and he would be able to join a group of us for dinner later that evening, as planned.

In the event the queue was for four and a half hours, the advice was as outlined above, and he missed dinner.

That mirrors what happened to me under a Tory government when I suspected I'd cracked a couple of ribs (and indeed had); wondered whether to bother with A & E as there's nothing you can do with cracked ribs but wait for them to heal; and decided on balance to go to St Thomas's Hospital because (a) what if a cracked rib were to threaten a lung? And (b) Maybe I could probably be out of there in less than three hours and return to work after lunch? The wait was about four and a half hours, I missed a whole day's work, and there was in fact no need to have gone at all.

The throughput of A& E units over the last decade is said to have increased substantially but, though the government does claim that waiting times have gone down, this is not the experience of most I speak to. Most seem stuck in a range between about three and five hours, though of course there will be surges at times like Saturday nights.

By logic which I hope will soon become apparent, I take you next from hospital wards to London's roads. The average speed of vehicular traffic in inner London has remained more or less within a band between eight and 12 miles per hour for about a century. During that time, however, the volume of traffic has increased enormously and horses and carriages have disappeared.

I link roads to A & E wards because they are both examples of a mismatch between demand and supply which is remedied by the interposition of a queue. I know perfectly well that this field has been studied by experts for years, that there will have been scores of learned dissertations on the subject, and that someone will now write to me sarcastically congratulating me on reinventing the wheel. But academic wisdom too often gathers dust while the rest of us blunder around, and perhaps it is worthwhile for ignorant columnists like me to go clodhopping into subjects we only partially understand, in an attempt to begin conversations between experts and the interested public.

So here goes. I wonder whether, in the topography of the interface between supply and demand, there are hidden depressions — ruts, we might call them — in which the balance found between supply and demand tends to get stuck. We might call it the Rut Theory of Queuing. Take London traffic congestion: essentially a queue for limited road space. Average speeds are stuck in a rut, oscillating around the ten-miles-per-hour mark. Doubtless this is because whenever speeds fall much below 8 mph a number of people who can walk (at up to 5 mph) start opting to walk rather than ride. So traffic runs more freely. But as average speeds increase cars and taxis re-enter the system and clog it up again.

The policy implication of this is that if you want to increase average speeds, there isn't much point increasing the capacity of London's roads; you must find ways of removing some of the traffic — which is what the Congestion Charge aims to do.

The application of my Rut Theory to A & E may be less obvious, but relates, likewise, to human psychology and choice. The first, cheapest and quickest judge of whether an injury needs expert attention is the injured individual. He will often be in two minds. The longer he expects to wait, the less inclined he will be to join the queue. That does not in itself produce any `rut' but only means that freeing up A & E wards is an uphill battle because as queues grow shorter, would-be customers multiply. Health managers know this.

But I wonder whether, as (while bringing waiting times down) we dip below four hours, we pass a psychologically important point after which there will be a surge of patient demand. As hospitals reduced waiting times from (say) ten hours to four they would pick up new customers only gradually, most (unless truly frightened by their injuries) being almost as unwilling to wait five hours as seven, or ten.

But somewhere short of four hours we reach an anticipated wait which it feels possible to build into one's day without abandoning everything and clearing the diary. One might call in to the hospital very early on the way to work, take an extended lunch break, or leave work a bit early and still hope to be home for a late supper. One might take a taxi to the hospital after a pub fight and still expect to sleep in one's own bed that night.

My guess is that the NHS has found that rut for the A & E queue somewhere short of four hours, and if they dip below it, will find an accelerating demand for treatment, as hundreds of thousands of black eyes, wrenched knees, burnt arms or cut faces seem worth expert inspection, 'just to be on the safe side'.

This, after all, is one of the few hospital queues to which the general public can admit itself without a doctor's or consultant's referral. Such referrals act as a filter for unwarranted requests for treatment: there will not be many joining the queue for cancer or heart surgery just because waiting times are dropping. Here there is no rut, and the NHS can take it all the way to zero. But the concept of `drop-in' treatment at A & E is a potent recruiting-sergeant for new customers.

The policy implications are the same as for congested roads: where there is a hidden potential surge in demand waiting round the corner, there is no point in increasing supply. Unless we are prepared to levy a charge, the queue is a necessary constraint on demand. That's not as comfortable an argument to make in the provision of health care it is in the provision of roads, but it remains true.