A system designed to highlight dangerous side-effects of medicines is not working, claim researchers.
"Yellow Cards" on Seroxat were studied
Doctors, and other health professionals are supposed to notify watchdogs if a patient reports an "adverse effect".
However, a researcher from the University of Oxford says that this information is wasted because they are not analysed properly.
He said emails to the BBC's "Panorama" after a programme about one drug contained more useful information.
The "Yellow Card" reporting scheme has been described as the "cornerstone" of the government's efforts to spot early warning signs that a drug might have previously-unknown hazardous side-effects.
Any health professional can send a report, and even coroners and drug companies themselves are expected to pass on details if a problem is drawn to their attention.
Dr Andrew Herxheimer wanted to test whether, in the case of Seroxat, a widely-used antidepressant, the scheme was set up to respond to reports of side-effects.
There have been widespread concerns that if a patient stops taking the drug suddenly - or changes dose - they are vulnerable to suicidal feelings.
Campaigners have called for warnings to be added to patient leaflets, although GSK, which makes the drug, insists that it is safe.
The Medicines and Healthcare Products Regulatory Agency currently advises that there is insufficient evidence of a causal link between the drug and suicidal behaviour, although it now says it should not be given to children.
Dr Herxheimer, aided by Charles Medawar, from Social Audit Ltd, which has campaigned on antidepressant safety, examined more than 1,000 yellow card reports about Seroxat.
Trained staff are supposed to interpret these short messages and sort each type of side-effect into coded groups - so that the scale of the problem, if any, can be assessed.
He found that while many of the forms contained details which might raise the suspicion that the drug was involved in suicidal thoughts or behaviour - but that they were often classified into different headings, lessening their impact.
While some were classed as "suicide" or "parasuicide" (attempted suicide), others were labelled as "non-accidental overdose" or something else entirely.
Some of the Seroxat side-effects reported were completely wrongly labelled - for example, a sensation akin to an electric shock in the head, described by a number of the drug's users, was often labelled just so, suggesting that the patient had actually been exposed to an electric current.
Dr Herxheimer said that the findings threw into question the value of the present yellow card scheme, and questioned whether the data produced by the scheme could effectively highlight problems in a timely way.
He described it as "chaotic and misconceived".
"Its value is limited by the emphasis on numbers not words; the focus on rare and 'interesting' adverse drug reactions, rather than the generality of drug-induced problems.
"There is secrecy that obstructs wider access to anonymised data, and lack of input from users of medicines themselves."
In contrast, he said that similar analysis of more than 1,000 emails sent to the BBC's Panorama programme in a wake of a report on Seroxat suggested that far more useful information could be derived if, as in other countries, patients were encouraged to send details of serious side-effects rather than their doctor or pharmacists.
The scheme is currently under review by the Department of Health, and a report on its workings is due to be published before the end of the year.
A spokesman for the department said that there were other ways in which problems with drugs could come to light.
She added: "Patients' views and experiences can make an important contribution.
"Since April, a pilot scheme enabling patients to report reactions through NHS Direct has been increasing patient involvement, and is currently being evaluated.
"The safety of the Selective Serotonin Reuptake Inhibitors (SSRIs) such as Seroxat has been closely monitored by the MHRA and the Committee on Safety of Medicines
(CSM) since first authorisation.
"An Expert Working Group on their safety has been set up, and all available data including the views of key stakeholders such as Mr Medawar and Dr Herxheimer are being considered."
The study was published in the International Journal of Risk and Safety in Medicine.