Dr. Angela Raffle and colleagues analysed the screening records of 350,000 women over 20 years and modelled cases of cervical cancer and deaths with and without screening.
Their results suggest that 8 out of every 10 women with "high grade" cell change will not go on to develop cancer, but all need to be treated.
For each death prevented at least 150 women have abnormal test results, over 80 women are referred to specialist clinics for investigation, and over 50 have treatment. At least one woman will die of cervical cancer despite participating in screening.
These findings have important implications, say the authors. For instance, even women who receive results of borderline changes can become anxious about their risk of cancer, so it is essential to change people's perception of the meaning of an abnormal screening result.
Furthermore, because of the resources involved and the potential to do harm, it is in the public's interest to control the introduction of cancer screening programmes that have not been adequately evaluated or quality assured, they conclude.