Abstract
BACKGROUND: Reducing health disparities is crucial for improving health outcomes. In cancer care, disparities persist across regions, socioeconomic groups, and racial populations. To better understand health disparities in lung cancer, the leading cause of cancer mortality in the U.S., we developed a novel method to visualize healthcare disparities by analyzing the sequence of care received, referred to as care paths. This approach aims to identify how variations in care paths among different patient groups are linked to poorer outcomes. METHODS: Using Latent Class Growth Analysis on visit sequences of lung cancer patients, we grouped patients into three classes. Then, we employed hazard modeling to predict adverse outcome probabilities for each class. FINDINGS: We identified three classes within our lung cancer cohort (N = 729) using the heterogeneity in their healthcare utilization patterns during 2016 and 2017. The results indicate differences between racial and urban/rural distributions across the classes (p = 0·004 and <0·0005, respectively). Black patients consistently had higher Social Deprivation Index (SDI) scores compared to Whites within each class, with significantly greater SDI in Classes 2 and 3 (p < 0·05 for both comparisons). Rural patients had significantly higher SDI scores than urban patients (p < 0·05 for each class). The area under the risk trajectory curve indicated greater total longitudinal risk of adverse outcome was larger for Black and urban patients in each class than their White and rural counterparts. INTERPRETATION: Our research indicates that Black individuals experienced less favorable adverse outcome risk trajectories compared to White patients based on their care path sequences. Rural patients demonstrated better outcomes than urban patients despite exposure to more social deprivation.