Figures that suggested hundreds of babies being prescribed antidepressants were a coding error, says drug agency Pharmac.
It has re-analysed the figures it gave to the United Future Party, and found only 91 prescriptions for preschool children.
Assuming one prescription is written each month, the agency said this
translated to fewer than 10 children under six receiving antidepressants in the 2006/07 financial year.
None of the children was under a year old.
The re-analysis also found that approximately 14,000 prescriptions were written for children between the ages of 6 and 17 - a number consistent with the original figures.
"Re-examination of the data shows a picture more in line with what we would expect in clinical practice," said Pharmac's medical director Dr Peter Moodie.
"A very, very small number of pre-school children do appear to be prescribed antidepressants, and this is a far more acceptable pattern."
Pharmac's figures had earlier shown 1229 antidepressant prescriptions for children aged five and under - a high figure which left Pharmac and other medical authorities mystified and concerned.
"It's now clear that we made an error in interpreting the figures for preschoolers and are sorry for that," said Dr Moodie.
"We now realise the impact a small coding error, which could have been made at a number of levels of data entry, can have on accuracy when small numbers of prescriptions are involved."
In its re-analysis, the agency cross-referenced data using patients' National Health Index numbers, on which it had earlier based its figures, with prescriptions from pharmacists which were coded by patient age bands.
The results showed that the data they had earlier based their information on had over-reported the number of children.
Dr Moodie said that overall, the two sets of data for age matched one another 99.6 per cent of the time.
But with a total of 1.125 million prescriptions for antidepressants in the last financial year, a 0.4 per cent discrepancy in the age of patients - while small in percentage terms - equated to around 4000, resulting in a large impact on the small subset of prescriptions for young children.
Dr Moodie maintained that the overall data was accurate.
