Go to DBLP homepageA Topological Data Analysis of Un met Health Care Needs Among Injured Patients
Abstract: Despite universal health care in Canada, research has shown persistent disparities in access to care for various sociodemographic groups. Most research has focused on access to primary care for particular sociodemographic groups, with much less attention paid to the effect of injury on unmet healthcare needs. The current study aimed to address this knowledge gap by evaluating and comparing unmet healthcare needs between injured patients and non-injured individuals. The primary study objective was to segment the population and explore differences in unmet healthcare needs among patient clusters. U sing the Canadian Community Health Survey, a study cohort of injured and non-injured subjects was created, restricted to those who responded to questions regarding unmet healthcare needs. Determinants of access related to predisposing characteristics (age, sex, marital status, immigration status, race), enabling characteristics (income, education, having a healthcare provider), and need-based factors (mental health comorbidities) were included as features. Topological Data Analysis (TDA) was used to identify the complex interactions between features, with respect to the underlying shape of the dataset, to generate patient clusters with respect to unmet health care needs. Cluster boundaries were defined using Louvain community detection. Statistical differences between features were tested across clusters. The study included 2,777,268 injured patients and 12,031,031 non-injured patients. TDA discovered three distinct phenotypes within the injured patient population, each exhibiting unique patterns of unmet healthcare needs. These clusters showed differences across the maj ority of features, indicating that this segmentation method effectively distinguished between injured patients with varying levels of unmet healthcare needs. This study is the first national analysis of injured Canadians reporting unmet healthcare needs. Our results indicate substantial inequalities in healthcare services for injured patients.
Loading