2023-10-03 20:04:01
A machine learning model equipped solely with data on people’s age, duration of smoking, and number of cigarettes smoked per day can predict lung cancer risk and identify who needs lung cancer screening, according to a new study published in the journal ‘PLOS Medicine‘.
Lung cancer is the most common cause of cancer death worldwide and if not detected early, survival is low.
It is estimated that screening people at highest risk for lung cancer could reduce lung cancer deaths by almost a quarter, but the ideal way to determine the high-risk population is unclear.
The current standard-of-care model of lung cancer risk requires 17 variables, few of which are routinely available in electronic medical records.
In the new study, the researchers used data from 216,714 smokers in the study cohort. UK Biobank and 26,616 smokers who participated in the US national lung screening trial to develop new models of lung cancer risk.
A machine learning model used three predictors (age, smoking duration, and pack-years) to estimate the odds of people developing lung cancer and dying from lung cancer in the next five years.
The researchers coordinated by Thomas Callender, of the University College London(UK), tested the new model on a third data set, from the US Prostate, Lung, Colorectal and Ovarian Screening Trial.
The model predicted lung cancer incidence with a sensitivity of 83.9% and lung cancer deaths with a sensitivity of 85.5%. All versions of the model had higher sensitivity than currently used risk prediction formulas with equivalent specificity.
«We know that screening those who have a high chance of developing lung cancer can save lives. “With machine learning, we have been able to substantially simplify the way we determine who is at high risk, presenting an approach that could be an exciting step in the direction of widespread implementation of personalized screening to detect many diseases early,” adds Callender. .
However, in statements to Science Media CentreIsabel Portillo, specialist in Preventive Medicine and Public Health and coordinator of the early cancer diagnosis working group at the Spanish Society of Epidemiologypoints out that the study does not add value to what has already been published on known risk factors.
Although the authors present it as a prediction tool, they add, “its applicability at a practical level (health professionals) is debatable in the case of indicating screening to a person (CT-low dose). It should be adjusted to be a tool to help both professionals and patients (Framingham Cardiovascular Risk Study). “Biomarkers, environmental and work factors must be considered.”
#predicts #screened #lung #cancer