Domestic violence is a significant public health issue. Routine health screening is a cost-effective strategy to reduce the harmful physical and mental consequences after the violence occurs. However, existing research finds consistently low rates of violence screening offered by healthcare providers, e.g., nurses, nurse practitioners, physicians. There is a critical need for research that helps understand how providers’ screening behavior is impacted by their intentions of screening, demographics, workload, and practice environment of the health center which they work in. However, much of the existing research in this field is empirical and qualitative. Two recent studies, Health Care Providers (HCP) study and American Association of Nurse Practitioners (AANP), involved quantitative data collected to measure providers’ screening behavior and intention, and multi-level impacting factors including providers’ demographics and work experience at individual level and characteristics of practice environment at organizational level. The data have many challenging properties such as multi-level and multi-source, which make existing statistical machine learning and data analytics models inadequate. The proposed research develops a novel system-level multi-center structural equation model (SEM) to fuse AANP and HCP data and integrated both datasets to examine how providers’ behavior are related to their intentions, demographics, workload, and practice environment. The innovative multi-center SEM enables joint estimation across multiple datasets and leverage information collected from multi-sources, which greatly improves power in statistical analysis. The knowledge and insights generated from our study could facilitate the design and optimization of health professional training, workload assignment, and practice environment, which leads to improved providers’ performance and quality of care in healthcare systems. 

Keywords

Healthcare performance improvement, quality of care, structural equation modeling, multi-center data analytics