Underplanning is in the health ministry’s DNA

The saga of insufficient hospital beds continues. Our public hospitals are so full, they have to hire extra space from other institutions. The alternative would be to turn away emergency patients, again. Here’s a headline from the Straits Times, 30 August 2011:

The problem of insufficient public hospital capacity is a long-standing one, dating back several years. Even when the new Khoo Teck Puat Hopsital in Yishun opened (March 2010), and the Ministry of Health public relations went into dizzy-spin mode saying that the new facility should banish the problem into the dustbins of history, I was sceptical.

Sure enough, the new hospital was filled up before one could blink (see my July 2010 article Bed crunch continues even as new hospital opens) and by earlier this year, more crisis stories reappeared (see my June 2011 article Hospital bed supply trailing far behind increase in elderly numbers).

The latest news story reports that the situation is expected to be so dire,

The crunch has led Health Minister Gan Kim Yong to float the idea of bringing forward the opening of Sengkang Hospital, scheduled for 2020.

— Straits Times, 30 August 2011, Public hospitals ‘borrowing’ ward space, by Salma Khalik

This new hospital is in addition to the Ng Teng Fong Hopsital now under construction in Jurong East and expected to be ready by 2015. It had earlier been touted that these two planned hospitals should more than cope with the expected increase in need, claims repeated as recently as earlier this year. The latest news that they can’t come on stream soon enough is ample proof that the ministry’s planning process is flawed.

In Hospital bed supply trailing far behind increase in elderly numbers I had already pointed out that something about the planning projections makes no sense. The planners used the assumption that the demand for hospital beds increases about 1 percent per annum. I took a quick look at demographic data from the 2010 census and showed that the population of elderly is growing by 4 percent per year. I am sure the civil servants at the Health Ministry can see the same numbers but somewhere along the line, the implications stemming from the data might have been rejected out of hand, perhaps because they conflicted with wishful thinking. You get a glimpse of that in the same Straits Times story:

But [Health Minister Gan Kim Yong] told The Straits Times yesterday that simply adding more beds to the system alone will not be a viable solution in the long run.

It does not take into account how patients prefer to be cared for in the community instead of in hospitals, Singapore’s limited land space, and how with an ageing population, beds will simply not be enough no matter how many hospitals are built.

— ibid.

You see it? A dogmatic resistance to recognise the implications of data.

To be fair, the minister is exploring other ways:

More innovative ways to manage the health-care system must be found, he said, adding that he was reviewing the health-care masterplan.

One solution is to use general practitioners and polyclinics more intensively to keep people healthy – and out of hospitals. Another is to develop home care services, he said.

— ibid.

However, ideal though it may be, this is an untested model for healthcare delivery. Even if it eventually works, it’s going to take years to create the system and debug it. But most crucially, the concept seems to confuse keeping people healthy with coping with the ill. Keeping people healthy may prolong lives and improve the quality of lives. But ultimately, we will all get seriously ill. If we remain healthy, we postpone this from age 75 to age 85, but in the end we will still be hospitalised.

Alternative strategies may dampen demand somewhat but the rapid growth in the numbers of elderly will still mean that hospital capacity has to increase substantially and rapidly. There is no wishing this away.

* * * * *

How is it that barely a year after inaugurating a brand new hospital (Khoo Teck Puat) the public hospital system is facing a bed shortage overall? According to the long-term development plan, another hospital should not be needed until about 2015 when the Ng Teng Fong is ready. So why is there a bed crunch now?

It’s hard to know what went wrong with the planning system, but I can take a stab at it. Perhaps this was what happened:

I’ve noticed in previous news stories that the ministry refers to historical hospital admissions data when justifying its planning numbers. This suggests to me that civil servants treat historical data with respect to bed-days as indicative of demand, projecting (the blue line in the rough illustration above) into the future. It is likely that on this basis, they drew up plans for new hospitals.

In the illustration below, total planned bed capacity is the thin black line that steps up with the addition of a hypothetical New Hospital P in Year 8 and New Hospital Q around Year 13. The plan is approved, and everybody is congratulated for assuring the country of sufficient bed capacity to cope with growing demand in the years ahead.

Three years later, in Year 10, after New Hospital P has come into operation, the public is screaming again. Hospitals are operating at close to 100 percent of capacity, and patients are being turned away.

Why did the turquoise bars shoot up in Year 8 and after? — You might think it’s an important question, but I suspect it is not one that triggers much soul-searching in government. This is because there is an ideological tendency to fall back on the dictum, borne out of the government’s anathema to socialism, that when it comes to public goods, demand will always increase to fill supply. (You want evidence? See the sentence from the Straits Times report cited above and attributed to Gan Kim Yong: “beds will simply not be enough no matter how many hospitals are built”.) So the official attitude is: Nothing you can do about the constant cry for more beds. If you succumb, you’re only feeding the beast of social expectations.

In other words, the first instinct of the government is to dismiss the phenomenon of capacity crunch as no more than a symptom of the moral failure of citizens in their greed for public resources. This dismissal continues until the political price becomes unbearable, and then something is done.

While that explanation — that when it comes to public goods, demand will always increase to fill supply — may have a ring of truth about it, it is actually flawed. The dictum may be true when the public good is free or cheap, e.g. use of roads. But hospitalisation is a costly thing. Nor is it something people can choose to consume by themselves. They first need to fall ill (as if people like to be ill?) and a doctor needs to consider the illness severe enough to warrant warding.

So why did the bars shoot up?

The error came about when planners used historical bed-use data as an indicator of demand. This is only meaningful if there is surplus capacity in the system. In years when there is no surplus capacity, the bed-use data represents the supply limit, not demand.

In fact, in those years, hospitals would have been discharging patients early to make way for new incoming patients. New emergency cases would have been waiting in corridors for lack of beds, and elective surgery would  have been postponed. In other words, true demand was higher than supply.

In the graph above, unfilled demand is represented by the faint-coloured bars. As soon as a new hospital opens up, that unfilled demand has somewhere to go, and that’s how a new hospital fills up so quickly.

Thus, the blue line that had been used for planning new hospitals was inaccurate as a projection of demand growth. Instead it should have been the pink line:

Undeniably, it is very hard to “count” demand, unlike hospital admissions and bed stays. By definition they are patients turned away, or hospital stays cut short. How many extra bed-days would have been involved if they had not been turned away or discharged early? It is hard to know, and most certainly there’d be no records. Estimating demand is not a science but an art.

So, even if a civil servant had the foresight to know that the blue line projection is probably misleading, it would be extremely difficult for him to make a case for an accelerated building program based on the pink line projection. He would be asked: Where’s the hard data to support the pink line? Especially if he is faced with a political bias against unnecessary spending, and a cultural reluctance to point out to ministers that they are over-optimistically reliant on untested ideas of community-based healthcare, the civil servant’s chances of success are poor. Resistant to contrarian ideas, the ministry is institutionally programmed to lurch from crisis to crisis.

* * * * *

I am not at all underestimating how difficult it is to get planning right. Ask any corporation about to launch a new product, and they’d tell you estimating demand is often guesswork at best. Even for established products, estimating next year’s sales is hard enough.

Moreover, building forward capacity for the healthcare system is not just a matter of construction. There’s an even bigger problem of where one would find the needed doctors, nurses and other personnel. Then there is the question of costs and subsidies for patients. . . .

But the starting point must be to get the projections right. In this respect — to have a brand new hospital full within 12 months, and the whole system groaning under the weight of unforeseen demand with no new hospital for 3 – 4 more years to come — the record is rather less than sterling.