Emergency department visit frequency and health care costs following implementation of an integrated practice unit for frequent utilizers.

OBJECTIVES: The integrated practice unit (IPU) aims to improve care for patients with complex medical and social needs through care coordination, medication reconciliation, and connection to community resources. This study examined the effects of IPU enrollment on emergency department (ED) utilization and health care costs among frequent ED utilizers with complex needs. METHODS: We extracted electronic health records (EHR) data from patients in a large health care system who had at least four distinct ED visits within any 6-month period between March 1, 2018, and May 30, 2021. Interrupted time series (ITS) analyses were performed to evaluate the impact of IPU enrollment on monthly ED visits and health care costs. A control group was matched to IPU patients using a propensity score at a 3:1 ratio. RESULTS: We analyzed EHRs of 775 IPU patients with a control group of 2325 patients (mean [±SD] age 43.6 [±17]; 45.8% female; 50.9% White, 42.3% Black). In the single ITS analysis, IPU enrollment was associated with a decrease of 0.24 ED visits (p < 0.001) and a cost reduction of $466.37 (p = 0.040) in the first month, followed by decreases of 0.11 ED visits (p < 0.001) and $417.61 in costs (p < 0.001) each month over the subsequent year. Our main results showed that, compared to the matched control group, IPU patients experienced 0.20 more ED visits (p < 0.001) after their fourth ED visit within 6 months, offset by a reduction of 0.02 visits (p < 0.001) each month over the next year. No significant immediate or sustained increase in costs was observed for IPU-enrolled patients compared to the control group. CONCLUSIONS: This quasi-experimental study of frequent ED utilizers demonstrated an initial increase in ED visits following IPU enrollment, followed by a reduction in ED utilization over subsequent 12 months without increasing costs, supporting IPU's effectiveness in managing patients with complex needs and limited access to care.

Isolating the Drivers of Racial Inequities in Prostate Cancer Treatment.

BACKGROUND: Black individuals in the United States are less likely than White individuals to receive curative therapies despite a 2-fold higher risk of prostate cancer death. While research has described treatment inequities, few studies have investigated underlying causes. METHODS: We analyzed a cohort of 40,137 Medicare beneficiaries (66 and older) linked to the Surveillance Epidemiology and End Results (SEER) cancer registry who had clinically significant, non-metastatic (cT1-4N0M0, grade group 2-5) prostate cancer (diagnosed 2010-2015). Using the Kitagawa-Oaxaca-Blinder decomposition, we assessed the contributions of patient health and health care delivery on the racial difference in localized prostate cancer treatments (radical prostatectomy or radiation). Patient health consisted of comorbid diagnoses, tumor characteristics, SEER site, diagnosis year, and age. Health care delivery was captured as a prediction model with these health variables as predictors of treatment, reflecting current treatment patterns. RESULTS: A total of 72.1% and 78.6% of Black and White patients received definitive treatment, respectively, a difference of 6.5 percentage points. An estimated 15% [95% confidence interval (CI): 6-24] of this treatment difference was explained by measured differences in patient health, leaving the remaining estimated 85% (95% CI: 74-94) attributable to a potentially broad range of health care delivery factors. Limitations included insufficient data to explore how specific health care delivery factors, including structural racism and social determinants, impact differential treatment. CONCLUSIONS: Our results show the inadequacy of patient health differences as an explanation of the treatment inequity. IMPACT: Investing in studies and interventions that support equitable health care delivery for Black individuals with prostate cancer will contribute to improved outcomes.

Impact of the Affordable Care Act on participation in the Supplemental Nutrition Assistance Program among low-income older Medicare beneficiaries.

BACKGROUND: The Affordable Care Act (ACA) provisions, especially Medicaid expansion, are believed to have "spillover effects," such as boosting participation in the Supplemental Nutrition Assistance Program (SNAP) among eligible individuals in the United States (US). However, little empirical evidence exists about the impact of the ACA, with its focus on the dual eligible population, on SNAP participation. The current study investigates whether the ACA, under an explicit policy aim of enhancing the interface between Medicare and Medicaid, has improved participation in the SNAP among low-income older Medicare beneficiaries. METHODS: We extracted 2009 through 2018 data from the US Medical Expenditure Panel Survey (MEPS) for low-income (≤ %138 Federal Poverty Level [FPL]) older Medicare beneficiaries (n = 50,466; aged ≥ 65), and low-income (≤ %138 FPL) younger adults (aged 20 to < 65 years, n = 190,443). MEPS respondents of > %138 FPL incomes, younger Medicare and Medicaid beneficiaries, and older adults without Medicare were excluded from this study. Using a quasi-experimental comparative interrupted time-series design, we examined (1) whether ACA's support for the Medicare-Medicaid dual-eligible program, through facilitating the online Medicaid application process, was associated with an increase in SNAP uptake among low-income older Medicare beneficiaries, and (2) in the instance of an association, to assess the magnitude of SNAP uptake that can be explicitly attributed to the policy's implementation. The outcome, SNAP participation, was measured annually from 2009 through 2018. The year 2014 was set as the intervention point when the Medicare-Medicaid Coordination Office started facilitating Medicaid applications online for eligible Medicare beneficiaries. RESULTS: Overall, the change in the probability of SNAP enrollment from the pre- to post-intervention period was 17.4 percentage points higher among low-income older Medicare enrollees, compared to similarly low-income, SNAP-eligible, younger adults (ß = 0.174, P < .001). This boost in SNAP uptake was significant and more apparent among older White (ß = 0.137, P = .049), Asians (ß = 0.408, P = .047), and all non-Hispanic adults (ß = 0.030, P < .001). CONCLUSIONS: The ACA had a positive, measurable effect on SNAP participation among older Medicare beneficiaries. Policymakers should consider additional approaches that link enrollment to multiple programs to increase SNAP participation. Further, there may be a need for additional, targeted efforts to address structural barriers to uptake among African Americans and Hispanics.

An evaluation of copy and paste events in electronic notes of patients with hospital acquired conditions.

BACKGROUND: The increased use of the copy and paste function (CPF) in Electronic Health Records (EHRs) has raised concerns about possible clinician miscommunication and adverse patient outcomes. OBJECTIVE: This study investigated the prevalence and extent of CPF in the EHRs of patients diagnosed with Hospital-acquired Conditions (HACs). We also examined the association between the use of CPF and patient characteristics. MATERIALS AND METHODS: The prevalence and extent of CPF were investigated using electronic clinical notes of 50 patients hospitalized with HACs between 2017 and 2021 at a large academic medical center. Study patients were adults aged 21 and older with a length of stay greater than three days. ANOVA analysis was used to examine the differences in CPF use between patients with different characteristics. RESULTS: A total of 7,844 clinical notes across seven note types are compared in the study. The mean patient age was 63.7, with an average length of stay of 15.6 days. 54% of Discharge Summaries, 53% of Consults, and 47% of history and physical (H&P) notes had duplications with the same type of notes. In the Discharge Summary, ED notes, and Plan of Care, duplications accounted for 40% or higher of the full text. H&P and Consults, H&P and Discharge Summary, and Discharge Summary and Consults were more likely to have duplications than between other types of notes. Duplications accounted for 15.5% of the information provided in H&P and Consults. The prevalence of CPF was higher in the Discharge Summary of patients who were younger, female, and had longer hospital stays. CONCLUSION: Both prevalence and extent of duplication were high in the Discharge Summary, Consults, and H&P notes of patients with HACs. Future studies are needed to examine the intention and appropriateness of CPF use and its impact on patient outcomes.

A Scoping Review of the Use of Machine Learning in Health Economics and Outcomes Research: Part 1-Data From Wearable Devices.

OBJECTIVES: With the emerging use of machine learning (ML) techniques, there has been particular interest in using wearable data for health economics and outcomes research (HEOR). We aimed to understand the emerging patterns of how ML has been applied to wearable data in HEOR. METHODS: We identified studies published in PubMed between January 2016 and March 2021. Studies that included at least 1 HEOR-related Medical Subject Headings term, applied an ML, and used wearable data were eligible for inclusion. Two reviewers abstracted information including ML application types and data on which ML was applied and analyzed them using descriptive analyses. RESULTS: A total of 148 studies were identified from PubMed, among which 32 studies met the inclusion criteria. There has been an increase over time in the number of ML studies using wearable data. ML has been more frequently used for monitoring events in real time (78%) than to predict future events (22%). There has been a wide range of outcomes examined, ranging from general physical or mental health (24%) to more disease-specific outcomes (eg, disease incidence [19%] and progression [13%]) and treatment-related outcomes (eg, treatment adherence [9%] and outcomes [9%]). Data for ML models were more often derived from wearable devices with specific medical purposes (60%) than those without (40%). CONCLUSION: There has been a wide range of applications of ML to wearable data. Both medical and nonmedical wearable devices have been used as a data source, showing the potential for providing rich data for ML studies in HEOR.

Inclusion of nonrandomized studies of interventions in systematic reviews of interventions: updated guidance from the Agency for Health Care Research and Quality Effective Health Care program.

OBJECTIVES: We developed guidance to inform decisions regarding the inclusion of nonrandomized studies of interventions (NRSIs) in systematic reviews (SRs) of the effects of interventions. STUDY DESIGN AND SETTING: The guidance workgroup comprised SR experts and used an informal consensus generation method. RESULTS: Instead of recommending NRSI inclusion only if randomized controlled trials (RCTs) are insufficient to address the SR key question, different topics may require different decisions regarding NRSI inclusion. We identified important considerations to inform such decisions from topic refinement through protocol development. During topic scoping and refinement, considerations were related to the clinical decisional dilemma, adequacy of RCTs to address the key questions, risk of bias in NRSIs, and the extent to which NRSIs are likely to complement RCTs. When NRSIs are included, during SR team formation, familiarity with topic-specific data sources and advanced analytic methods for NRSIs should be considered. During protocol development, the decision regarding NRSI inclusion or exclusion should be justified, and potential implications explained. When NRSIs are included, the protocol should describe the processes for synthesizing evidence from RCTs and NRSIs and determining the overall strength of evidence. CONCLUSION: We identified specific considerations for decisions regarding NRSI inclusion in SRs and highlight the importance of flexibility and transparency.

Hospital value-based payment programs and disparity in the United States: A review of current evidence and future perspectives.

Beginning in the early 2010s, an array of Value-Based Purchasing (VBP) programs has been developed in the United States (U.S.) to contain costs and improve health care quality. Despite documented successes in these efforts in some instances, there have been growing concerns about the programs' unintended consequences for health care disparities due to their built-in biases against health care organizations that serve a disproportionate share of disadvantaged patient populations. We explore the effects of three Medicare hospital VBP programs on health and health care disparities in the U.S. by reviewing their designs, implementation history, and evidence on health care disparities. The available empirical evidence thus far suggests varied impacts of hospital VBP programs on health care disparities. Most of the reviewed studies in this paper demonstrate that hospital VBP programs have the tendency to exacerbate health care disparities, while a few others found evidence of little or no worsening impacts on disparities. We discuss several policy options and recommendations which include various reform approaches and specific programs ranging from those addressing upstream structural barriers to health care access, to health care delivery strategies that target service utilization and health outcomes of vulnerable populations under the VBP programs. Future studies are needed to produce more explicit, conclusive, and consistent evidence on the impacts of hospital VBP programs on disparities.

A Scoping Review of the Use of Machine Learning in Health Economics and Outcomes Research: Part 2-Data From Nonwearables.

OBJECTIVES: Despite the increasing interest in applying machine learning (ML) methods in health economics and outcomes research (HEOR), stakeholders face uncertainties in when and how ML can be used. We reviewed the recent applications of ML in HEOR. METHODS: We searched PubMed for studies published between January 2020 and March 2021 and randomly chose 20% of the identified studies for the sake of manageability. Studies that were in HEOR and applied an ML technique were included. Studies related to wearable devices were excluded. We abstracted information on the ML applications, data types, and ML methods and analyzed it using descriptive statistics. RESULTS: We retrieved 805 articles, of which 161 (20%) were randomly chosen. Ninety-two of the random sample met the eligibility criteria. We found that ML was primarily used for predicting future events (86%) rather than current events (14%). The most common response variables were clinical events or disease incidence (42%) and treatment outcomes (22%). ML was less used to predict economic outcomes such as health resource utilization (16%) or costs (3%). Although electronic medical records (35%) were frequently used for model development, claims data were used less frequently (9%). Tree-based methods (eg, random forests and boosting) were the most commonly used ML methods (31%). CONCLUSIONS: The use of ML techniques in HEOR is growing rapidly, but there remain opportunities to apply them to predict economic outcomes, especially using claims databases, which could inform the development of cost-effectiveness models.

SARS-CoV-2 infection and COVID-19 severity in individuals with prior seasonal coronavirus infection.

We show that individuals with documented history of seasonal coronavirus have a similar SARS-CoV-2 infection rate and COVID-19 severity as those with no prior history of seasonal coronavirus. Our findings suggest prior infection with seasonal coronavirus does not provide immunity to subsequent infection with SARS-CoV-2.

Piloting a model-to-data approach to enable predictive analytics in health care through patient mortality prediction.

OBJECTIVE: The development of predictive models for clinical application requires the availability of electronic health record (EHR) data, which is complicated by patient privacy concerns. We showcase the "Model to Data" (MTD) approach as a new mechanism to make private clinical data available for the development of predictive models. Under this framework, we eliminate researchers' direct interaction with patient data by delivering containerized models to the EHR data. MATERIALS AND METHODS: We operationalize the MTD framework using the Synapse collaboration platform and an on-premises secure computing environment at the University of Washington hosting EHR data. Containerized mortality prediction models developed by a model developer, were delivered to the University of Washington via Synapse, where the models were trained and evaluated. Model performance metrics were returned to the model developer. RESULTS: The model developer was able to develop 3 mortality prediction models under the MTD framework using simple demographic features (area under the receiver-operating characteristic curve [AUROC], 0.693), demographics and 5 common chronic diseases (AUROC, 0.861), and the 1000 most common features from the EHR's condition/procedure/drug domains (AUROC, 0.921). DISCUSSION: We demonstrate the feasibility of the MTD framework to facilitate the development of predictive models on private EHR data, enabled by common data models and containerization software. We identify challenges that both the model developer and the health system information technology group encountered and propose future efforts to improve implementation. CONCLUSIONS: The MTD framework lowers the barrier of access to EHR data and can accelerate the development and evaluation of clinical prediction models.

A predictive tool for identification of SARS-CoV-2 PCR-negative emergency department patients using routine test results.

BACKGROUND: Testing for COVID-19 remains limited in the United States and across the world. Poor allocation of limited testing resources leads to misutilization of health system resources, which complementary rapid testing tools could ameliorate. OBJECTIVE: To predict SARS-CoV-2 PCR positivity based on complete blood count components and patient sex. STUDY DESIGN: A retrospective case-control design for collection of data and a logistic regression prediction model was used. Participants were emergency department patients > 18 years old who had concurrent complete blood counts and SARS-CoV-2 PCR testing. 33 confirmed SARS-CoV-2 PCR positive and 357 negative patients at Stanford Health Care were used for model training. Validation cohorts consisted of emergency department patients > 18 years old who had concurrent complete blood counts and SARS-CoV-2 PCR testing in Northern California (41 PCR positive, 495 PCR negative), Seattle, Washington (40 PCR positive, 306 PCR negative), Chicago, Illinois (245 PCR positive, 1015 PCR negative), and South Korea (9 PCR positive, 236 PCR negative). RESULTS: A decision support tool that utilizes components of complete blood count and patient sex for prediction of SARS-CoV-2 PCR positivity demonstrated a C-statistic of 78 %, an optimized sensitivity of 93 %, and generalizability to other emergency department populations. By restricting PCR testing to predicted positive patients in a hypothetical scenario of 1000 patients requiring testing but testing resources limited to 60 % of patients, this tool would allow a 33 % increase in properly allocated resources. CONCLUSIONS: A prediction tool based on complete blood count results can better allocate SARS-CoV-2 testing and other health care resources such as personal protective equipment during a pandemic surge.

Evaluation of the secondary use of electronic health records to detect seasonal, holiday-related, and rare events related to traumatic injury and poisoning.

BACKGROUND: The increasing adoption of electronic health record (EHR) systems enables automated, large scale, and meaningful analysis of regional population health. We explored how EHR systems could inform surveillance of trauma-related emergency department visits arising from seasonal, holiday-related, and rare environmental events. METHODS: We analyzed temporal variation in diagnosis codes over 24 years of trauma visit data at the three hospitals in the University of Washington Medicine system in Seattle, Washington, USA. We identified seasons and days in which specific codes and categories of codes were statistically enriched, meaning that a significantly greater than average proportion of trauma visits included a given diagnosis code during that time period. RESULTS: We confirmed known seasonal patterns in emergency department visits for trauma. As expected, cold weather-related incidents (e.g. frostbite, snowboarding injury) were enriched in the winter, whereas fair weather-related incidents (e.g. bug bites, boating accidents, bicycle accidents) were enriched in the spring and summer. Our analysis of specific days of the year found that holidays were enriched for alcohol poisoning, assaults, and firework accidents. We also detected one time regional events such as the 2001 Nisqually earthquake and the 2006 Hanukkah Eve Windstorm. CONCLUSIONS: Though EHR systems were developed to prioritize operational rather than analytic priorities and have consequent limitations for surveillance, our EHR enrichment analysis nonetheless re-identified expected temporal population health patterns. EHRs are potentially a valuable source of information to inform public health policy, both in retrospective analysis and in a surveillance capacity.

Electronic medical records and medical procedure choice: Evidence from cesarean sections.

This paper examines how hospital adoption of electronic medical records (EMRs) impacts medical procedure choice in the context of cesarean section deliveries. It provides a unique contribution by tying the literature on EMR diffusion to the literature on the utilization of expensive medical technology and provider practice style. Exploiting within-hospital variation in three types of EMR adoption, we find that computerized physician order entry, an advanced EMR system that typically incorporates decision support, reduces C-section rates for low-risk mothers by 2.5%. Obstetric-specific EMR systems and physician documentation have no statistically significant effect on C-section rates. In addition, we find that the computerized practitioner order entry effect occurs predominantly in hospitals that were already performing fewer C-sections and does not change the behavior of already high-intensity providers.