The clinical data revolution came closer this week with the unveiling of the approach for improving quality and a survey on what to include in quality accounts.

The intention is that a selection of local, regional and national quality measures will drive quality, safety and choice, with the first accounts being published in April 2010.

The data tsunami about to crash upon the shore of the NHS has the potential to do more for service improvement than any restructuring or political fad. But to make the quality accounts work, the government has to avoid four pitfalls.

First, the indicators chosen need to be logical, coherent and manageable. The batch of 400 up for consultation is a random assortment of all the information currently picked up by NHS IT systems. Lord Darzi himself stresses they are only part of the picture and that others, such as the royal colleges' national databases, are "just as exciting". Order and simplicity will aid their effectiveness as a management tool and, crucially, make them more intelligible to patients.

Second, ministers must resist the temptation to allow these indicators to mutate into targets. The government professes a desire to keep nationally determined indicators to a minimum, with the main work being focused regionally and locally.

Pious words

But to make these pious words a reality ministers must resist ratcheting up the number of central indicators and using them as a stick to beat the NHS with every time a particular part of the health service hits a media storm. If the government's soundbite first, engage brain later response to the tragedy of Baby P is anything to go by, the omens are not good.

If de facto targets emerge, this sort of mission creep would stoke clinical resistance. Quality accounts won't work unless the overwhelming majority of clinicians are on board.

Third, commissioners need to play a leading part in developing the local and regional indicators. The consultation document contains the usual platitudes about commissioners - "a central role in driving and achieving the vision" no less. But this is followed up with precious little detail, and the overall impression is that commissioners will be shouting from the sidelines while providers run the show.

Finally, it is more important to get quality accounts right than get them quickly. If staff perceive the final plan to be rushed and poorly thought through, hard-won ground of managerial and clinical buy-in will be quickly lost. Getting all this ready by 1 April would be helpful, but not at the expense of - quality, of course.