MARK ALBEEK: Development of Risk Prediction Tools for Early Lung Cancer Detection

Project description

Background

Lung cancer remains the leading cause of cancer-related death worldwide, including in Sweden, primarily due to its often late-stage diagnosis. A growing proportion of these cases are being identified in never-smokers, particularly women. Early detection is crucial for improving survival rates, making primary health care essential in recognizing initial symptoms. However, the early symptoms of lung cancer are nonspecific and common among patients without the disease, complicating the ability of general practitioners to identify individuals with an elevated risk of having lung cancer. A deeper understanding of the patterns and interactions among various pre-diagnostic symptoms is therefore necessary.

Objective

The purpose of this project is to enhance early detection strategies for lung cancer in primary health care settings by developing and evaluating risk prediction tools. This will be achieved through a multi-phase approach involving an interactive electronic questionnaire. Additionally, we aim to investigate the questionnaires’ ability to predict the occurrence of lung cancer in people with different smoking status, including both patients with suspected lung cancer and population controls. The overall aim is to create a basis for the development of evidence-based risk prediction tools, that can be used for decision support in general practitioners work with early detection of lung cancer.

Method

The project is structured in three phases.

• Phase I: Qualitative study using think-aloud methodology to evaluate the clarity and usability of the questionnaire.
• Phase II: Feasibility study to assess the inclusion process and address practical challenges in conducting a multicenter case-control study.
• Phase III: Multicenter case-control study using the questionnaire, utilizing machine learning to analyze collected data and identify predictive symptom patterns, stratified by smoking status.

The questionnaire collects extensive data on symptoms and health changes. This tool is expected to capture differences in symptom presentation between individuals with elevated and non-elevated risk of having lung cancer.

The project will explore whether integrating comorbidity information (such as chronic diagnoses), visit frequency, and air pollution data from the residential area, can enhance the predictive capability of the tool. Additionally, home-based blood sampling using blood cards will be implemented to analyze biomarkers for lung cancer, examining whether these biomarkers can improve the model’s ability to identify individuals at elevated risk for the disease.