Variations in GPs’ Decisions to Investigate Suspected Lung Cancer: A Factorial Experiment Using Multimedia Vignettes

Jessica Sheringham1,Rachel Sequeira1,Jonathan Myles2,Willie Hamilton3,Joe McDonnell4,Judith Offman2,Stephen Duffy2,Rosalind Raine0

1UCL, London, UK,2QMUL, London, UK,3University of Exeter, Exeter, UK,4London Borough of Waltham Forest, London, UK

Presenting date: Tuesday 3 November
Presenting time: 16.15 - 16.30


Patients with symptoms suggestive of lung cancer commonly present to primary care but diagnostic delays are well documented (1-3) and it is unclear how general practitioners (GPs) distinguish which patients require further investigation. This factorial experiment examined how patients' clinical and socio-demographic characteristics influence GPs' decisions to initiate lung cancer investigations.


A multimedia interactive website simulated key features of GP consultations using actors (‘patients'). A national sample of GPs made management decisions online for six ‘patients' randomly selected from 36 vignettes, with clinical and socio-demographic characteristics systematically varying across three levels of cancer risk. In low-risk vignettes (positive predictive value [PPV] <1.2%) investigation by the GP (i.e. chest X-ray ordered or respiratory physician referral) was not indicated, in medium-risk (PPV=1.7-2.5%) investigation could be appropriate, in high-risk vignettes (PPV>3%) investigation was definitely indicated. Each ‘patient' had two lung cancer-related symptoms; one volunteered and the other elicited if GPs specifically asked. Variations in investigation likelihood were examined by 'patient' characteristics using multilevel logistic regression.


227 GPs completed 1356 vignettes of which 1348 were included in analysis. GPs investigated lung cancer in 74% (1000/1348) of cases. Investigation likelihood did not increase with cancer risk. Investigations were more likely when GPs sought information on symptoms that ‘patients' had but did not volunteer (adjusted odds ratio [AOR] 3.18 95% CI 2.27-4.70) but GPs omitted to seek this information in 42% (570/1348) of cases. GPs were less likely to investigate older than younger ‘patients' (AOR 0.52 95% CI 0.39-0.70) and Black ‘patients' compared with White (AOR 0.68 95% CI 0.48-0.95).



When GPs explicitly seek relevant clinical information, most act on it appropriately, but inequalities in cancer investigation by age and ethnicity remain.