Brief Report: Impact of PLCOm2012 Ethnicity Model Variations on Eligibility for the Alberta Lung Cancer Screening Program

Abstract

Introduction

Screening with low-dose computed tomography reduces lung cancer mortality among individuals at high risk of lung cancer. Three PLCOm2012 model variations for predicting lung cancer risk have been developed: the original model with six race or ethnicity categories (6Race), a model with three race or ethnicity categories (3Race), and a model excluding race and ethnicity (NoRace). We evaluated the impact of model selection on screening eligibility for the Alberta Lung Cancer Screening Program.

Methods

Retrospective analysis of 5961 individuals contacting the Alberta Lung Cancer Screening Program. For each participant, we calculated their 6-year risk of lung cancer using the three PLCOm2012 models and compared their eligibility using a 1.5% 6-year lung cancer risk threshold.

Results

Of the participants, 70% identified as White, 9% Asian, 4% Indigenous, 0.5% Black, 0.4% Hispanic, and 17% not reported. Furthermore, 67% had a risk more than or equal to 1.5% risk across all three models, whereas 30% had less than 1.5% risk across all models. Using the NoRace model, 31% of Indigenous and 25% of Black individuals were classified below the 1.5% threshold but would have been eligible using 3Race or 6Race. Using the 6Race model, 18% of Asian and 22% of Hispanic individuals were below 1.5% but were above 1.5% in either the NoRace or 3Race model.

Conclusions

Choice of PLCOm2012 model can have a considerable impact on screening eligibility, particularly for ethnic minorities. Using the 3Race model increases screening eligibility for Indigenous, Black, Hispanic, and Asian individuals. Canadian-based data validating and quantifying the impact of race and ethnicity on lung cancer risk are urgently needed.