Integrating the fractured and fractious components of health and social care systems seems to be everybody’s current favoured “solution” for healthcare problems, but it’s hard to make happen. We now have evidence that it may literally take an earthquake or some other natural disaster to make it happen.

Conceptually it’s easy to see why integration makes sense. Increasingly most of healthcare is about people with many problems not one, and they are shuttled among providers who communicate poorly and have different cultures and incentives. Errors, costs, and frustration rise exponentially.

But trying to put hospitals, primary care, community services, mental health services, social care, and ideally housing, education, and benefits together into one seamless system is akin to simultaneously solving several Rubik cubes on a unicycle while reciting Paradise Lost.

It doesn’t happen.

But remarkably in Canterbury, New Zealand it has. Nick Mays and Judith Smith have tried in a BMJ editorial to unravel why. One factor is politicians managing to resist the almost irresistible urge to restructure the system. Another is that GPs have organised themselves into a network called Pegasus. I’ve met those people, and they are impressive. Removal of competition among providers seems to have been important, and crucially there have been new funds to invest in community services. “Alliance contracting” with sharing of financial risk and reward has also been helpful.

But would all these be enough to bring about true integration? I’m sceptical, and then I read: “A final factor impossible to quantify is the effect of the 2011 Christchurch earthquake.” It got rid of 100 hospital beds, brought people together, and quickened plans that “had been long in discussion.”

Being hit by an earthquake is to be attacked by common enemy. Suddenly everybody is on the same side. The squabbles, deceptions, and evasions that usually stop integration let the side down. They become unacceptable, unpatriotic.

After the second world war, the destroyed economies of Germany and Japan roared ahead and soon overtook the sclerotic victors like Britain and France. I was told when a student at the Stanford Business School that this was because of the total destruction of “distributive coalitions”— monopolies, cartels, political groupings, professional organisations, trade unions, and the many other organisations that make up an overdeveloped society.

The distributive coalitions of Canterbury would not have been destroyed by the earthquake, but they would have been less willing to defend their vested interests—and the destruction of 100 hospital beds, presumably a significant proportion in a thinly populated country, would have provided impetus for integration because it depends on shifting resources from hospitals to the community.

As I contemplate the Byzantine complexity of the English health and social care system—with bodies like clinical senates and health and wellbeing boards, which hardly know what they are supposed to do—I fear that we need an earthquake, preferably one that selectively destroys underperforming hospitals, but doesn’t kill anybody.

Competing interest: RS is an unpaid member of the Whole Person Care Commission; chair (with equity) of Patients Know Best, which promotes integration of health and social care systems through provision of patient held records; and an employee (with shares and stock options) of UnitedHealth, which is active in the English NHS, although he has no responsibility for what happens in Britain.

Richard Smith was the editor of the BMJ until 2004 and is director of the United Health Group’s chronic disease initiative.