Mammography screening has little or no effect on breast cancer deaths, Swedish data indicateBMJ 2012; 345 doi: https://doi.org/10.1136/bmj.e4847 (Published 17 July 2012) Cite this as: BMJ 2012;345:e4847
Mammography screening has little or no effect on mortality from breast cancer in women aged 40-69, shows an analysis of Swedish figures published this week. But commentators argue that trends in breast cancer deaths cannot be used to evaluate screening in isolation from developments in diagnostic methods and treatment.
Regular mammography screening for women aged 40-69 years was introduced gradually in Sweden from 1974, achieving nationwide coverage across all 21 counties in 1997. Attendance for mammography screening is now among the highest recorded in any country, with 75% to 85% of eligible women attending screening regularly.
Researchers led by Philippe Autier, director of the International Prevention Research Institute in Lyon, France, hypothesised that the gradual introduction of screening into different Swedish counties would be reflected in county specific mortality over the past 20 years. They analysed data from the Swedish Board of Health and Welfare from 1960 to 2009 to assess trends in breast cancer mortality in women aged 40 and older by the county where they lived.1
The group expected to see a gradual reduction in breast cancer mortality after regular screening was introduced. But instead the results showed that breast cancer mortality had already started to fall in 1972, before mammography screening was introduced. Deaths then continued to fall at a similar rate to that in the pre-screening period.
Breast cancer death rates in Swedish women aged 40 years and over fell by 0.98% each year between 1997 and 2009. The rate fell continuously in 14 of the 21 Swedish counties. It declined sharply in three counties during or soon after the implementation of screening, fell sharply in two counties at least five years after screening began, and increased in two counties after screening began.
Overall, the decline in breast cancer mortality in counties that implemented screening in 1974-8 was similar in the 18 years after the programmes were introduced to the downward trend before they started.
“It seems paradoxical that the downward trends in breast cancer mortality in Sweden have evolved practically as if screening had never existed,” said Autier. “Swedish breast cancer mortality statistics are consistent with studies [in Norway and Denmark] that show limited or no impact of screening on mortality from breast cancer.”2 3 4
In an accompanying editorial Michael Vannier, from the University of Chicago Medical Center, argued that the limitations of treatment have a greater bearing on mortality from breast cancer than the shortcomings of screening.5 Using mortality to evaluate the effect of screening is simplistic, when diagnosis and treatment also have major effects on cancer outcomes, he wrote. “Even if screening were 100% successful, if diagnosis and treatment were delayed or ineffective, the number of deaths might not be affected.” He recommended targeting screening at subgroups according to risk and suggested that better diagnostic tools would help to improve the cost-benefit balance of breast cancer screening.
Nereo Segnan and colleagues from the cancer epidemiology unit at the ASO San Giovanni Battista University Hospital in Turin, Italy, said in a second editorial, “It’s time to move beyond the apparently never ending debate about the extent to which screening for breast cancer in the 1970s to 1990s has reduced mortality from breast cancer—as if it was isolated from the rest of health care.”6 Current priorities should be to increase the accuracy of screening and the appropriateness and availability of treatment, they concluded.
Cite this as: BMJ 2012;345:e4847