Predictors and disparities in surveillance of incidental lung nodules in a diverse urban population: a retrospective observational study

Lung cancer remains the leading cause of cancer-related mortality worldwide, with late-stage diagnosis contributing to poor prognosis. Incidental lung nodules (ILNs) are frequently detected during routine imaging, with approximately 1.6 million new cases identified annually in the United States. Although the majority of ILNs are benign, up to 6% may represent early-stage lung cancer. Despite established follow-up protocols, adherence remains suboptimal, with over 60% of ILNs not receiving appropriate surveillance. This study aimed to assess follow-up rates for ILNs at a large urban hospital system and to identify factors influencing adherence, including demographic, socioeconomic, and clinical variables. A retrospective cohort study was conducted at Mount Sinai Morningside and Mount Sinai West hospitals (New York, NY). Radiology reports from January 1, 2022, to June 30, 2022, were analyzed using Natural Language Processing (NLP) to identify patients with ILNs. Eligible patients underwent systematic chart review to extract demographic data, insurance status, smoking history, comorbidities, and follow-up details. The primary outcome was the rate of adherence to follow-up recommendations, with secondary analyses exploring factors affecting the follow-up rate. An integrated multivariable model combined demographic/clinical and imaging predictors to assess associations after cross-domain adjustment. Among 3,925 identified cases, 1,266 patients met inclusion criteria. The median patient age was 70 years (IQR: 54–86), with a slight female predominance (57.5%). Racial distribution included White (43.4%), Black (22.5%), Asian (5.4%), Other (26.1%) patients and not reported (2.6%). The overall follow-up adherence rate was 41.3%, with 5.7% of those completing follow-up diagnosed with lung cancer. In demographic and clinical models, Black and Other-race patients demonstrated lower odds of follow-up. Although these associations were attenuated in the fully adjusted, integrated multivariable model, the direction of effect persisted (Black: aOR 0.774, 95% CI 0.559–1.070; Other: aOR 0.779, 95% CI 0.574–1.056). Medicaid coverage and uninsured status were similarly associated with reduced follow-up rates. Imaging obtained in the emergency department (aOR 0.451, 95% CI 0.318–0.635) or during inpatient admission (aOR 0.531, 95% CI 0.355–0.783) was independently associated with lower odds of follow-up, whereas imaging performed during standard working hours and documented emphysema were associated with higher odds of follow-up. Follow-up adherence for ILNs remains suboptimal. Racial differences were attenuated after integrated adjustment, although lower follow-up among Black and Other patients persisted directionally. Imaging setting remained an important determinant of follow-up. Structured tracking, patient navigation, and standardized reporting may improve surveillance and reduce disparities in lung cancer care. Given the potential for early lung cancer detection, targeted interventions—such as structured tracking systems, patient navigation programs, and standardized reporting protocols—are needed to improve follow-up rates and mitigate healthcare disparities. Addressing these gaps is critical to reducing lung cancer-related morbidity and mortality.