Essays on Examining Health Workforce Policies Using an Equity Framework
Open AccessEssay 1: Examining changes in the nursing workforce in response to hospital financial resources before and during COVID-19Objective. The purpose of this study is to examine whether hospital financial resource levels, using current ratio and days cash on hand, were associated with nurse staffing levels prior to and during the COVID-19 pandemic, 2019-2021. Data Sources. For our analysis, we used data collected from a sample of Premier Inc. member hospitals across the country from January 2019 through December 2021. The data combined different databases that contained facility-level information, including nursing hours and other direct care providers worked hours, facility financials, and patient counts, as well as expected mortality among COVID and non-COVID patients. Study Design. We employed a multivariate random effects model of total nurse hours per patient on two financial measures, current ratio and days cash on hand, and included time interactions with the financial variables to account for any time-varying effects. Results. We find that in the pre-COVID period, hospital resources were not associated with nurse staffing levels, when using both current ratios and days cash on hand. During the early COVID periods, our results show that hospitals increased their total nurse staffing on average by 1.329 (p<0.01) hours per patient day when modelling the regression with current ratio. Interestingly, in the context of this overall staffing increase, we find that during the early period of COVID, for every increase in a hospital’s current ratio, there was an average decrease of 0.334 (p<0.01) total nurse hours per patient days. Conclusions. Our findings suggest that hospitals with greater financial resources do not necessarily invest more in nurse staffing. The null finding between financial resource and staffing in the pre-COVID period, and the surprising inverse relationship with current ratio and staffing in the early COVID periods illustrate that staffing decision models for hospitals remain a black box. Given that now more than $23 billion have been distributed through Phase 4 of the Provider Relief Fund and American Rescue Plan Rural payments, it is important for policymakers and administrators of these programs to explore ways to tie some portion of subsidies to staffing.Essay 2: Public health workforce and community health outcomesObjective. The need for a robust public health system has become an increasingly clear due to the COVID-19 pandemic. Public health provides essential services that protect communities and improve health. Research on the public health workforce has been limited in frequency and scope. A few studies have examined characteristics of the workforce and suggest top executive degree types (specifically a registered nurse degree), years of work experience, education levels, and types and relative staffing of public health workforce are associated with public health agency performance and health outcomes. Initial work at the state-level and in one state at county-level suggests public health workforce staffing is associated with community-level mortality outcomes. This study examines the change in local health department (LHD) staffing between 2013 and 2019 across US counties, as well as the association between public health workforce staffing and community-level health outcomes. Data Sources. For our analysis, we used cross-sectional surveys collected by the National Association of County and City Health Officials for the years 2013, 2016, and 2019. The survey data included staffing, financials, policy, and governance information on LHDs across the country. We combined this data with county health data from the Center for Disease Control (CDC) Wide-ranging Online Data for Epidemiologic Research (WONDER) as our measures of community health outcomes using age-adjusted mortality rates. Study Design. We employed a multivariate regression model for panel data to estimate the association between LHD staffing and community health outcomes at the county level, using public health governance as an instrumental variable. Principal Findings. Our study finds that on average, counties had 6.09 fewer LHD FTEs from 2013 (N=2,519) to 2019 (N=1,762). Across the study period, we find that mortality rates decreased between 0.523 and 3.19 percent for every 10 percent increase in a county’s LHD FTE. Conclusions. Our findings suggest that public health staffing are valuable investments for LHDs to improve community health outcomes. Especially in the context of the Center for Disease Control awarding $3.2 billion to state, local and territorial jurisdictions across the United States to build the public health workforce, our study provides evidence for decision-making in how to appropriately invest to better population health. Essay 3: Identifying the behavioral health workforce for Medicaid patients with schizophrenia using claims dataObjective. This study attempts to identity the current workforce trends and future workforce needs for patients with schizophrenia by first identifying patients with schizophrenia, then tracking their utilization patterns, and finally, estimating the workforce composition needed to deliver care to this population. Using Medicaid claims data from 2016 through 2019, I identify how patients transition between different treatment services, and use these service utilization patterns to map out the behavioral health workforce needs. Policy that aims to improve access to care for this patient population must consider how to build a workforce that can provide the service, and how to improve patient engagement with the services. Data Source. 2016 through 2019 Transformed Medicaid Statistical Information System (T-MSIS) inpatient, other services, and prescription claim files, as well as the T-MSIS beneficiary demographic and eligibility files. Study Design. There are three components to this study. The first is the identification of patients with schizophrenia using claims data. The case-finding algorithm will determine the incidence rate of schizophrenia within the Medicaid population in 2019, as well as those with any prior schizophrenia diagnosis in 2019, and any prior diagnosis from 2016 through 2019. The second is tracking utilization patterns of the identified population in order to find which modes of delivery are effective at increasing LAI use, and which modalities are being the most heavily utilized. These two utilization pattern findings will allow for better insight on the workforce composition required to service the current demand of services, and how the workforce pipeline could better shift to support the areas that are producing positive patient outcomes. The final part of the study will explore this by using the service utilization patterns to calculate the workforce configuration needed to service the population. Results. I find that in 2019, there were 93,435 Medicaid patients that were newly diagnosed with schizophrenia and 855,838 patients with any diagnosis in prior years (2016 through 2019). Among both the newly diagnosed and anytime diagnosed population, I find that most patients used emergency room services as their primary point of care. When examining the workforce composition of providers delivering services to Medicaid patients with schizophrenia in the Medicaid, most providers in the ER are not behavioral health specialists (such as psychiatrists). Patient utilization patterns also revealed that services related to coordinated specialty care (CSC) and assertive community treatment (ACT) associated with greater use of long-acting injectables, a proxy for quality of care for this patient population. Further, I find there is misalignment in the workforce participating in Medicaid and the workforce that is found in the Medicaid claims data interacting with patients with schizophrenia. Despite identifying 410,757 counselors and 280,975 social workers that had Medicaid participation, the providers associated with the billing of services for patients with schizophrenia were predominantly psychiatrists and emergency physicians. Conclusion. These findings provide important groundwork to better understand how to improve the provision of care to Medicaid patients with serious mental illness. Specifically, recognizing the burden on the emergency department where providers are not specialized in psychiatric care may lead to patients to enter a cycle of services that does not reflect the longitudinal care they require. Moreover, the finding that the workforce needed to provide services being utilized does not align with the Medicaid claims data points towards billing patterns that are not supporting what is expected or desired, that providers such as counselors and social workers are not engaging with Medicaid patients with schizophrenia. This is in the context of finding high presence of counselors and social workers with Medicaid participation in the data, but none that are engaging with patients with schizophrenia. Though there are data limitations to using claims files, this study importantly demonstrates that the workforce required to provide high quality of care to patients with schizophrenia must rely on greater engagement of counselors and social workers. To build a stronger workforce that can support the needs of patients with schizophrenia, policy should not only focus on encourage the use of effective treatment, but must also focus on how to get patients to the appropriate providers.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.