Although improvements in survival rates are widely used to promote increases in cancer screening, they are a poor indicator of whether screening tests actually save lives.
But the reason survival improvements are observed is often the result of lead-time bias — 1 of the major biases that pervade cancer screening statistics, along with overdiagnosis bias and length-time bias.
Lead-time bias arises when increased screening uptake increases the measured survival time of a patient without changing the course of the disease or prolonging time until death. This bias may come into play when a cancer diagnosis occurs prior to when the patient experiences any symptoms related to the malignancy.
Recent studies suggest that some misunderstanding around lead-time bias continues to persist in the literature.
Lead-time bias is often overlooked, said My Catarina von Euler-Chelpin, PhD, an assistant professor at the University of Copenhagen’s Institute of Public Health, Denmark. “I think that there are lots of mistakes being committed … it can really skew the results.”
To Dr Euler-Chelpin, one example was provided by a recent study that compared differences in mortality and survival between men and women diagnosed with breast cancer that were recorded in the US National Cancer Database between 2004 and 2014.1 The study found that 3-year survival, 5-year survival, and overall survival was significantly higher for women compared with men.
In a commentary on the paper, Dr Euler-Chelpin noted that the study did not consider lead-time bias, which is a particular issue when comparing 2 cohorts with different levels of screening.2 During the study period, more than 60% of women older than 40 years had undergone screening for breast cancer, indicating that a large proportion of diagnoses in women would have been made at screening.
As a result of the screening, some of the women may have received a cancer diagnosis several years earlier than they normally would have, according to some models of breast tumor growth.3 “And then when compared to men, their survival will look longer, even if they don’t necessarily survive better from the cancer,” Dr Euler-Chelpin said. “Part of the prolonged survival will be a screening effect (ie, lead-time bias).”
It’s well established that breast cancer screening through mammography significantly decreases mortality rates by enabling earlier treatment. However, it is difficult to disentangle how much of this survival enhancement revealed in screening studies is due to early treatment of the disease once it is detected, and how much of this improved survival is really just a result of lead-time bias. “Early detection lowers the mortality rate, so … it probably increases survival, but as we can’t measure the lead time, we don’t know how much is explained by lead time [bias],” Dr Euler-Chelpin explained in an email.