Ministers have pledged that no one will wait more than 18 weeks from GP referral by the end of 2008. Data released by the DoH last week show that 48 per cent of patients are already being treated within this timescale.
But over a third of referral-to-treatment times have been excluded from the data.
The DoH estimates that about 330,000 patients per month are admitted to hospital - the point at which the referral-to-treatment clock stops.
This is known as a 'completed pathway'.
However, the NHS reported only 282,000 completed pathways for March 2007, the month measured. Of these, 74,000 were excluded because it had 'not been possible to link back to the appropriate start of the pathway', leaving data on just 208,000, or 63 per cent, of referrals.
GPC negotiator Dr Richard Vautrey said: 'We can't rely on this figure.
It is like putting your finger in the air and making your best guess.
'We have known for ages the complete inadequacy of the way this data is recorded. Practice-based commissioners know data from hospitals on waits and other things are very unreliable.'
GPC negotiator Dr Peter Holden said the data contained 'too many variables'.
'I'm not happy with what I see - I think there is rather more distance to go than is being portrayed,' he said.
Asked whether PCTs had deliberately excluded cases that missed the 18-week target, NHS Alliance GMS contract lead Dr David Jenner said: 'It is possible that some of this is conspiracy, but it is most likely a cock-up.'
Nick Chapman, DoH national director of the 18-week programme, said: 'I think the data is a reasonable representation of the proportion of patients treated in 18 weeks. I have no evidence to suggest the figures are biased.'
The PCT with the highest proportion of patients treated in 18 weeks was Leicester City, with 98 per cent. However, the trust was unable to pinpoint when referral occurred for 47 per cent of patients. Swindon PCT had the lowest proportion of patients treated in 18 weeks, at 22 per cent.
Heywood, Middleton and Rochdale PCT had the least complete data - it could only specify referral-to-treatment time for 12 per cent of procedures.
Referral-to-treatment times excluded from the data: 37%