Skip to main navigationSkip to main content
The University of Southampton
Health SciencesOur research

Research project: Using a participant-completed questionnaire to identify symptoms that predict lung cancer: A feasibility study (Current CI: Brindle)

Currently Active: 

A feasibility study is required to establish the content validity of a symptom, risk and co-morbidity questionnaire in a population of CXR attendees, to identify adjustments required to achieve acceptable response and item completion rates, to provide preliminary estimates of the discriminatory value of the questionnaire and to inform the design of a prospective study of symptoms in a CXR population.

Researchers/CLRN staff recruited consecutive eligible attendees (aged >=40, referred by GPs for CXR) at 3 Chest X-Ray centres (2 Birmingham, 1 Southampton) during 3 episodes of recruitment. Adjustments were made to the questionnaire before each iteration of recruitment if necessary to improve content validity, completion and response rates.

The questionnaire included some open ended questions during the feasibility phase to identify symptom experiences not accurately recorded by the fixed response items and to facilitate purposeful sampling for interview and changes to the questionnaire. Findings from qualitative analyses informed the conversion of open questions to fixed response items where appropriate.

Respondents were purposefully selected for interview on the basis of their questionnaire responses to include a range of symptom experiences and co-morbidities. In-depth and cognitive interviewing were used to explore symptom experiences and to investigate interpretation of the questionnaire items and reasons for missing data.

Participants' CXR results, referral requests, results of further investigations for those with an abnormal CXR, and cancer diagnoses were extracted from hospital records. LC diagnosis were identified 6 months after initial CXR. Data completion and response rates were calculated with 95% CIs.

Exploratory multivariable logistic regression analyses will be used to model the relationship between: (i) symptoms and lung cancer status, (ii) symptoms, epidemiological risk factors and lung cancer status, (iii) symptoms, epidemiological risk factors and abnormal CXR result. Positive and negative predictive values with their associated 95% confidence levels will be calculated for the models.

Descriptive statistics will be used to explore the distribution of participants and reasons for referral by age, sex and socio-economic status.

Conferences and events

James L, Clifford S, Parker L, Dowswell G,  Brindle L.
"Using a participant-completed questionnaire (IPCARD) to identify symptoms that predict lung cancer: A feasibility study" Poster at National School of Primary Care Showcase, Royal College of Physicians, London, September 2012

Project team

Lucy Brindle and Sue Wilson (leads)
Southampton researchers:
Liz James
Birmingham research team:
George Dowswell
Sue Clifford
Laura Parker

Project funder


Associated research themes

Questionnaire development and evaluation
Predictive values of symptoms for abnormal chest x-ray
Early lung cancer diagnosis
Symptom epidemiology

Related research groups

Complex Healthcare Processes
Share this research project Share this on Facebook Share this on Twitter Share this on Weibo
Privacy Settings