Background: The Mission Act provides improved Veteran access to care both within the Veterans Administration (VA) and community systems. An underlying assumption is that faster care with more choices results in better care. However, care fragmentation is associated with increased length of stay, readmissions, and mortality. Postoperative complications and readmissions are higher in minority and low socioeconomic status (SES) patients. Low SES is also associated with frailty, one of the best predictors of 30-day postoperative complications and hospital readmissions. Despite having a profound influence on health outcomes, social risk factors are absent from risk adjustment for VA quality measures, further exacerbating disparities in minority and low SES populations. This strategy may further constrain resources to care for vulnerable populations, as many Veterans are economically disadvantaged and potentially adding avoidable costs to care delivery. Another major issue is care fragmentation. Nevertheless, the impact of non-VA care and care fragmentation is absent in performance metrics. Our goal is to identify social risk factors and levels of care fragmentation that affect surgical outcomes to inform VA quality metric policy and institutional resource allocation. We improve upon current practice by joining surgical outcomes data with 1) VA/Centers for Medicare & Medicaid Services (CMS) claims data, 2) VA fee-basis files to identify encounters outside of the VA health system and 3) using more granular proxy social risk factors and neighborhood disadvantage. Significance/Impact: Our significance is modeling surgical outcomes using social risk factors, rurality, living in a disadvantaged neighborhood and care fragmentation to identify factors contributing to health care disparities and to inform VA policy. The impact is to develop quality metrics using social risk factors and care fragmentation. HSR&D priority areas: Rural Health, Health Equity, Health Care Value and Health Care Informatics. Innovation: Joining diverse data sources to develop predictive models using both traditional parametric methods and exploratory machine learning techniques to provide clinicians and administrators with outcomes and economic analyses necessary to change institutional practices to benefit our most vulnerable Veterans. Specific Aims: Aim 1: Identify factors affecting surgical outcomes by assessing the contributions of ethnicity, race, SES, place of residence and care fragmentation to surgical complications, readmissions and mortality Hypothesis: Using ethnic/racial minority status, SES, place of residence and care fragmentation will identify important risk factors for postoperative complications, readmissions, and mortality Aim 2: Assess the impact of social risk factors and care fragmentation on hospital performance metrics for readmissions and mortality Hypothesis: Including social risk factors and care fragmentation in risk adjustment models significantly changes VA hospital performance rankings with respect to readmissions and mortality Aim 3: Determine the relationship of place of residence, care fragmentation, SES and minority status to acute and long-term VA surgical health care utilization to inform VA resource allocation Hypothesis: Low SES, rurality, care fragmentation and minority status are associated with higher VA resource utilization Methodology: Quantitative analyses using traditional parametric and exploratory machine learning techniques performed on diverse datasets to develop predictive models of surgical outcomes using care fragmentation, rurality and social risk factors risk adjusted for medical comorbidities and applied to VA quality metrics. Implementation/Next Steps: Deployment of quality metric models using social risk factors and care fragmentation within the VA system. Adjusting resource allocation to account for social risk factors.
External Links for this Project
Grant Number: I01HX003095-01A2
None at this time.
TRL - Applied/Translational
Outcomes - Patient, Socioeconomic Factors
None at this time.