Provisional hospital data used in a criticised Home Office press release has been revised.

The adjustment to the data, formally released by the NHS on Friday, was predicted by health officials in December as they tried to prevent use of the statistics in a release about the impact of the government's ten Tackling Knives Action Plan (TKAP) areas.

Emails published earlier this month showed discussions between Downing Street and the Department of Health about whether or not to use unverified admissions numbers showing a 27 per cent drop in the number of knife crime victims treated in the 10 TKAPs.

One unnamed NHS Information Centre (IC) statistician explained that "these are provisional data for 2007/08 and 2008/09 and therefore they are potentially inaccurate and may possibly give the wrong impression". A senior health statistician warned that it would look as though the department had "cherrypicked the good news", but Downing Street was reported to be "adamant about the need to publish this statistic", another email said.

The verified data has now been released and shows that the actual decrease in admissions was smaller than originally thought: 18 per cent between July and September, and eight per cent annually.

Liberal Democrat home affairs spokesman Chris Huhne said: "This should be a warning to ministers to wait until figures are properly cooked, and not attempt to cook them up first."

Home secretary Jacqui Smith has apologised for the early release of the hospital admissions figures against the wishes of the NHS IC, and the incident prompted a cross-Whitehall review of how data is handled.