IT is too early to be absolutely clear about the cause the Berkshire train crash on Saturday.

Our sympathies and prayers go to the families of those killed in the crash, and the dozens of passengers who were injured. Our praise and gratitude go to the emergency services for their tireless and skilled work in the rescue operation.

Whatever the outcome of the investigations, it is vital that lessons are learned from the tragedy.

We must never be complacent about safety. While level crossings in Britain have a comparatively good record compared to the rest of the world, we must never overlook opportunities to improve them.

This incident could have been avoided, by replacing all level crossings with bridges or tunnels.

To do so, however, would be a huge, lengthy and expensive engineering task. And, whether we like it or not, cost effectiveness is a necessary consideration.

A factor in this particular incident appears to have been the speed of the train.

The very least we can now expect is a detailed evaluation of level crossings on high-speed stretches of the rail network.

Automatic half barrier crossings do appear to offer a rather flimsy protection and security.

Urgent consideration must be given, whatever the cost, to either replacing these barriers on lines where trains travel at high speeds, or else introducing a better advance warning system to alert train drivers.

There is no need to wait for the outcome of investigations before setting about such an important task.