OBJECTIVES : The number of incidentally detected, indeterminate, non-calcified lung nodules is increasing, adding to the abundance of extraneous invasive diagnostic procedures. The introduction of a biomarker as an adjunctive test for physician assessment of malignancy has the potential to reduce unnecessary invasive diagnostic workup. We undertook this study to quantify patient harms, cost-effectiveness and change in procedure-related management with the use of a biomarker in altering physician-assessed pre-test risk of malignancy in incidentally detected indeterminate lung nodules.

METHODS : We modified a previously published Markov model with Monte Carlo simulations for incidentally detected lung nodules and imputed a novel set of empirically-based parametric functions derived from a referral hospital setting to assign the initial procedure distribution (serial computed tomography (CT), biopsy, or surgery) based on a clinically assessed pre-test risk (clinical scenario) or biomarker post-test risk (biomarker scenario). We set the base cost of a biomarker at $750. One-way sensitivity analysis was used to evaluate the impact of test cost, biomarker test performance, compliance to the biomarker result and other key inputs. The primary outcome was impact on harms and the incremental cost-effectiveness ratio. Costs were estimated from a payer perspective.

RESULTS : The biomarker strategy resulted in a relative decrease of 31% and 36% for unnecessary biopsies and surgeries for benign nodules, respectively. The biomarker strategy dominated the clinical scenario, resulting in a per-person cost savings of $873 and QALY gain of 0.0984. One-way sensitivity analysis across a range of biomarker performance characteristics, physician adherence and other inputs did not alter these findings.

CONCLUSIONS : Based on our adaptation of a previously published Markov model and empirically-derived risk-based initial procedure assignment, a biomarker strategy may reduce unnecessary invasive procedures while providing cost-savings.