A blunder has seen the number of workplace deaths over-estimated by up to 30 per cent.
Statistics New Zealand put a freeze on its release of work-related injury indicators, which tot up fatal and non-fatal injuries in the workplace, because of "quality concerns".
The number of reported deaths from 2000 till 2011 was too high because in cases of multiple injuries the fatality was counted more than once.
The error was discovered before a scheduled release in November, and Statistics NZ said its previous figures should not be used until they were reviewed and revised.
"We're completing our analysis and will be releasing the updated information on May 14," a spokesman said.
"Until we complete that analysis, we cannot give exact figures."
The Government's statisticians have also reviewed the definition of fatal and serious non-fatal injuries.
It will now exclude occupational disease or illness, bystanders, unpaid workers and volunteers, workers commuting to and from work, suicide or self-harm, and those working for a New Zealand company overseas.
It will also no longer include injuries with a claim made to ACC on a farm when the claim cannot be identified as occurring at work.
The indicators used to be collated by the University of Otago on behalf of ACC, but Statistics NZ took over the work in 2011.
Earlier this week, a report criticised New Zealand's workplace safety culture, which it said was "not fit for purpose".
The Independent Taskforce on Workplace Health and Safety found that about one in 10 workers were harmed at work in some way every year, with about 26,000 injuries resulting in compensation claims.
The Engineering, Printing and Manufacturing Union's assistant national secretary, Ged O'Connell, said workplace accident rates in New Zealand were 20 per cent to 25 per cent higher than in Australia and Britain.
"The current health and safety system is quite dysfunctional," he said.
- © Fairfax NZ News