The primary analysis compared the rate of stage III/IV lung cancer within 2 years of randomisation between the intervention and control arms. The analyses followed the intention-to-treat principle. Cox proportional hazards models were used to estimate the hazard ratio. One participant who withdrew consent for use of their data was excluded from analysis. The models were adjusted for age, sex, smoking history, socioeconomic status and general practice.

Similar methodology was used to analyse the secondary outcomes of mortality rates. Further analysis compared the outcomes of those in the intervention arm with a positive test, those in the intervention arm with a negative test and those in the control arm. Comparisons of proportions were carried out using Fisher's exact test due to the small number of events. Poisson regression models (adjusting for follow-up time when necessary) were used to investigate other clinical outcomes. Specificity and sensitivity were calculated, but these are estimated values as the true figures are not estimable for early stage and late stage separately. This is because the test-positives received a more intensive intervention than the test-negatives and in a prospective study cancer status is unknown most of the time. The full statistical analysis plan can be found in the supplementary material.

A within-trial model-based cost-effectiveness analysis was conducted, estimating the cost per stage I/II lung cancer case detected comparing the intervention with the control arm. Diagnostic costs were included for all groups. A model-based approach was taken for two reasons: 1) prevalence of lung cancer during the trial was different between arms (our model assumed the same prevalence in both arms) and 2) data about resource use for detection were only available for test-positive participants (n=598) during the trial, therefore resource use was modelled. Full assumptions and parameters used in the model are presented in the supplementary material. Briefly, detection resources comprised the EarlyCDT-Lung test, monitoring tests and confirmatory diagnostic tests. The outcome was number of stage I/II lung cancers detected within the 2-year follow-up. Treatment costs are not included in this within-trial cost-effectiveness analysis.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.