Evaluating the follow-up chest x-ray to detect lung cancer in patients with pneumonia

Research question What is the prevalence of underlying lung cancers following pneumonia and how effective are follow-up chest x-rays at detecting them? Background In the UK, over 220,000 people are diagnosed with pneumonia every year.

Pneumonia survivors have long-term mortality rates three times higher than the general population, with cancer the cause of death in over a quarter.

The prevalence of an underlying lung cancer in pneumonia survivors has been reported as high as 9%, although estimates vary from 0.8-9%.

Current UK guidelines recommend a follow-up chest radiograph (CXR) be performed 6 weeks after recovery, to exclude an underlying lung cancer that may have been masked by the pneumonia. However, the evidence supporting this recommendation is weak and international guidelines are contradictory.

Aim To assess the role of CXR in detecting underlying lung cancer in people with pneumonia, to inform radiological follow-up of pneumonia survivors.

Objectives Determine the prevalence and stage of lung cancer diagnoses after pneumonia, using national primary and secondary care data. Determine the diagnostic accuracy of follow-up CXRs in detecting lung cancer post-pneumonia.

Methods Workstream 1: Using the Clinical Practice Research Datalink (CPRD) primary care database, we will identify patients diagnosed with pneumonia in primary or secondary care.

Through patient-level linkage to the National Cancer Registration and Analysis Service (NCRAS), we will identify patients who were subsequently diagnosed with lung cancer and gather information on the timing and stage of their cancer diagnosis.

Workstream 2: We will embed a prospective observational study within an existing randomised controlled trial of patients hospitalised with pneumonia.

The ASPECT trial is currently recruiting and aims to enrol over 20,000 people with pneumonia admitted to one of 60 hospitals across England over the next 2-years.

Using existing operational and ethical frameworks, we will review routinely performed follow-up CXRs to determine the diagnostic accuracy for detection of underlying lung cancer. Timelines for delivery The two workstreams will run simultaneously with a total duration of 36 months (m).

Workstream 1: Ethical approval and obtaining data linkage from CPRD (6m), data analysis and interpretation (12m), reporting and dissemination (6m) Workstream 2: Contracting with sites (6m), identification of recruitment and data collection of ASPECT participants (6m), data collection and analysis (18m), and reporting/dissemination (6m).

Anticipated impact This project will provide an accurate estimate of lung cancer prevalence in pneumonia survivors, and provide the largest prospective assessment of the diagnostic accuracy of CXR in detecting lung cancer after pneumonia.

These outputs will advance the evidence base regarding optimal follow up of patients with pneumonia and provide data to determine whether CXR is sufficient to exclude lung cancer.

Dissemination We will publish our results in high-impact, peer-reviewed journals and facilitate their incorporation into clinical guidelines, using established links with acute care, respiratory and primary care organisations.

With the support of our PPI group, we will prepare a plain English summary and develop accessible infographics of the findings, published on Trust, University and other websites with links disseminated by social media.