Use this URL to cite or link to this record in EThOS:
Title: Symptomatic diagnosis of lung cancer in a population referred to lung-shadow clinic with high rates of chronic respiratory diseases : a feasibility study
Author: Shim, Joanna
ISNI:       0000 0004 5370 1134
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
n the UK, 86% of lung cancer (LC) patients are diagnosed when curative treatment is not possible. Government guidelines recommend urgent chest X-ray referrals for patients presented with any 1 of 10 suggested LC symptoms. Little evidence currently supports these recommendations. Thus, the need for prospective studies to identify the predictive values of symptoms for LC diagnosis. This study aimed to investigate the feasibility of a prospective study to identify symptoms that predict LC in a secondary care population with high rates of chronic respiratory disease by investigating 1) the content validity and acceptability to this population of a patient-completed questionnaire, and 2) identifying patient-elicited symptoms that predict LC. The Identifying Symptoms that Predict Chest and Respiratory Disease (IPCARD) questionnaire was used to prospectively collect symptom, risk and comorbidity data in a chest clinic population investigated for LC (Patients aged ≥40). LC was identified six months following recruitment. Semi-structured and cognitive interviews explored the acceptability and content validity of the IPCARD questionnaire in this population. Multiple logistic regression was used to identify symptoms independently associated with LC in the clinic population, and in a COPD sub-group; at two levels of entry criteria (p<0.05 and p<0.15). 359 patients (70% of those eligible) completed the IPCARD questionnaire, and 77 (21.4%) were diagnosed with LC. Qualitative research indicated that participants found the IPCARD questionnaire acceptable, and its content validity was established in this secondary care population. Cough and breathing changes first indicated in the last three months, predicted LC in this referred population (p<0.05). In the COPD sub-group, the symptom descriptor, unable to get enough air predicted LC (p<0.05). At the relaxed criteria (p<0.15), five symptoms predicted LC in the full clinic population; a hard/harsh cough without phlegm, increasing chest infections, and weight gain (last 12 months) were added to the previous model. The COPD sub-group at p<0.15, breathing changes experienced (last three months), breathing changes first indicated within the last three months, unable to get enough air, and wheezing sensation, were predictors. Based on the more rigorous entry criteria (p<0.05), the diagnostic criteria for the COPD sub-group (positive likelihood ratio (+LR)=1.91; Area under curve (AUC)=0.739) performed slightly better than the criteria for the full population (+LR=1.49; AUC=0.729) (at optimal cut-off). The better performance of the COPD–specific model supports the need for an adequately powered study to investigate the predictive values of LC symptoms in homogeneous populations with specific respiratory diseases.
Supervisor: Brindle, Lucy ; George, Stephen Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: RC0254 Neoplasms. Tumors. Oncology (including Cancer)