Using socio-demographic and early clinical features in general practice to identify people with lung cancer earlier

Abstract

IntroductionIn the UK, most people with lung cancer are diagnosed at a late stage when curative treatment is not possible. To aid earlier detection, the sociodemographic and early clinical features predictive of lung cancer need to be identified. Methods We studied 12 074 cases of lung cancer and 120 731 controls in a large general practice database. Logistic regression analyses were used to identify the socio-demographic and clinical features associated with cancer up to 2 years before diagnosis. A risk prediction model was developed using variables that were independently associated with lung cancer up to 4 months before diagnosis. The model performance was assessed in an independent dataset of 1 826 293 patients from the same database. Discrimination was assessed by means of a receiver operating characteristic (ROC) curve. Results Clinical and socio-demographic features that were independently associated with lung cancer were patients' age, sex, socioeconomic status and smoking history. From 4 to 12 months before diagnosis, the frequency of consultations and symptom records of cough, haemoptysis, dyspnoea, weight loss, lower respiratory tract infections, non-specific chest infections, chest pain, hoarseness, upper respiratory tract infections and chronic obstructive pulmonary disease were also independently predictive of lung cancer. On validation, the model performed well with an area under the ROC curve of 0.88. Conclusions This new model performed substantially better than the current National Institute for Health and Clinical Excellence referral guidelines and all comparable models. It has the potential to predict lung cancer cases sufficiently early to make detection at a curable stage more likely by allowing general practitioners to better risk stratify their patients. A clinical trial is needed to quantify the absolute benefits to patients and the cost effectiveness of this model in practice.

Lung cancer screening: what is new since the NLST results?

“…A history or diagnosis of emphysema should be an interesting additional criterion for better defining at-risk populations [31][32][33]. Although many authors have reported their risk scores, most were validated retrospectively [34,35]. An exception is the UKLS trial's LLP score, which was tested in a prospective trial [13,36].…”

Prediction of risk of lung cancer in populations and in pulmonary nodules: Significant progress to drive changes in paradigms

“…These models show AUC values of 0.88-0.92 for predicting lung cancer [23,24]. One model was developed from data recorded between 4 and 12 months prior to the diagnosis of lung cancer, thus reflecting factors that might help to bring the diagnosis forward [23]. Employing this model would reduce the number of chest X-rays required to diagnose one lung cancer from 421 using conventional referral criteria to 119.…”

“…In the UK this has been possible by using large primary care datasets of routinely collected data. These models show AUC values of 0.88-0.92 for predicting lung cancer [23,24]. One model was developed from data recorded between 4 and 12 months prior to the diagnosis of lung cancer, thus reflecting factors that might help to bring the diagnosis forward [23].…”

Benefits and Risks of Lung Cancer Screening

“…Therefore, the use of an accurate model that incorporates additional risk factors to select persons for screening is likely to be cost-effective and will reduce harm to people with the least risk of lung cancer [33,51]. Individualized risk estimation has been developed in several models [52,53,54]. The PLCO cancer screening trial lung cancer risk model was developed from the largest dataset known to date.…”