Tuesday, April 27, 1999 Published at 01:28 GMT 02:28 UK
Quality measures urged in screening programme
Abnormal cells can be difficult to detect
Improved quality procedures could reduce the number people falsely given the all-clear following a smear test, according to a report.
False negative results occur when a screener working in a laboratory misses pre-cancerous cells in a smear test.
They cause untold stress as women who believe that they are healthy can suddenly find themselves diagnosed with advanced cancer.
Current estimates put the rate of false negatives at between 7% and 15% of all smears.
The NHS Cervical Screening Programme aims to catch signs of cancer as early as possible, as the earlier a cancer is treated, the better a patient's chance of long-term survival.
But a number of high profile cases where many women have received false negative results has led to doubts being raised about accuracy within the programme.
Professor Dulcie Colman, of the department of histopathology and cytology at Hammersmith Hospital in London, led the study.
The authors looked at 150 abnormal smears - including 50 that had been passed as false negatives - and re-assessed them.
They used a special microscope that generates a computer image of the sample, taking into account the number of cells and their size.
Their results showed that the cancerous cells on the false negative samples were much harder to detect than those on the samples correctly identified as abnormal - the true positive samples.
False negative smears contained about 10 times fewer abnormal cells than the true positive smears, and the cells tended to be more unevenly distributed across the slide.
However, the authors - while pointing out that such slides are difficult to diagnose correctly - suggest that quality control measures could improve rate of identification.
At the moment, the researchers said, screeners have to examine too many slides in too short a space of time to ensure accuracy.
A heavy workload was one of the elements implicated in the Kent and Canterbury screening scandal.
In that instance, eight women died, 30 needed hysterectomies and more than 90,000 women had their tests re-examined after mistakes at the Kent and Canterbury Hospital between 1989 and 1995.
The authors also criticise the checking procedures for smears. It is called rapid review, and as such involves a rapid scan of the slide rather than a detailed examination.
They said: "On the rare occasion when the image is one of a small, severely dyskaryotic [cancerous] cell, the previous high level of exposure means that the cells are ignored, with disastrous consequences for the woman from whom the smear was obtained."
They conclude that the processes involved in the screening of cervical smears need to be re-evaluated.
Last week a report in The Lancet medical journal suggested that a semi-automated computer system developed to analyse cervical smears could decrease the number of screening errors and improve productivity.
They tested an automated system, called Papnet, which uses artificial intelligence to identify abnormal smears and check negative slides.