Lung cancer is one of the most commonly diagnosed cancer types. It does not usually cause noticeable symptoms until it has spread to other parts of the lungs or around the body. As a result, the chances of surviving lung cancer are low.
Having a way to indicate that a patient may be at a higher risk of lung cancer and thus being able to refer them for more testing would help improve the speed at which is a patient is diagnosed and therefore allow for earlier access to treatment.
This study will look at the recording of several symptoms/tests prior to a lung cancer diagnosis to asses their potential in indicating the onset of lung cancer. The study will achieve this by comparing two groups of patients up to a 2-year period before a lung cancer diagnosis. The first group of patients will be those who have received a lung cancer diagnosis and the second group will be patients with similar population characteristics but who have not had a lung cancer diagnosis.
Compare symptoms and blood test results captured by general practitioners in the two years prior to cancer diagnosis in individuals who developed lung cancer and those that did not.
This will be a nested case-control study using data collected within the United Kingdom's Clinical Practice Research Datalink (CPRD). Cases (lung cancer patients) will be matched to up 4 controls with no record of cancer by sex, age, general practitioner (GP) practice and year of registration. The two-year period prior to cancer diagnosis will be stratified into 2-month time windows, and suspected cancer symptoms as well as blood test (basophil, eosinophil, lymphocyte, monocyte, neutrophil platelet and C-reactive protein) results will be identified for cases and controls. Proportions of patients reporting symptoms and mean test results for all individuals having a test in the time window will be compared between cases and controls, and multivariable logistic regression will be used to assess the association between blood tests and symptoms and lung cancer diagnosis.
Appropriate descriptive statistics will be used to summarise relevant clinical characteristics (Section O). These will be categorised to reflect lung cancer status.
Univariable and multivariable logistic regression models will be used to estimate odds ratios comparing the odds of each event between cases and controls.
Data sets(s) to be used
- CPRD GOLD
Health Outcomes to be Measured:
• Smoking status
• Shortness of breath
• Chalson Comorbidity index
• Chest pain
• Weight loss
• Basophil count
• Eosinophil count
• Lymphocyte count
• Monocyte count
• Neutrophil count
• Platelet count
• C-reactive protein (CRP)