Predicting pressure injury risk in hospitalised patients using machine learning with electronic health records: a US multilevel cohort study.

OBJECTIVE: To predict the risk of hospital-acquired pressure injury using machine learning compared with standard care. DESIGN: We obtained electronic health records (EHRs) to structure a multilevel cohort of hospitalised patients at risk for pressure injury and then calibrate a machine learning model to predict future pressure injury risk. Optimisation methods combined with multilevel logistic regression were used to develop a predictive algorithm of patient-specific shifts in risk over time. Machine learning methods were tested, including random forests, to identify predictive features for the algorithm. We reported the results of the regression approach as well as the area under the receiver operating characteristics (ROC) curve for predictive models. SETTING: Hospitalised inpatients. PARTICIPANTS: EHRs of 35 001 hospitalisations over 5 years across 2 academic hospitals. MAIN OUTCOME MEASURE: Longitudinal shifts in pressure injury risk. RESULTS: The predictive algorithm with features generated by machine learning achieved significantly improved prediction of pressure injury risk (p<0.001) with an area under the ROC curve of 0.72; whereas standard care only achieved an area under the ROC curve of 0.52. At a specificity of 0.50, the predictive algorithm achieved a sensitivity of 0.75. CONCLUSIONS: These data could help hospitals conserve resources within a critical period of patient vulnerability of hospital-acquired pressure injury which is not reimbursed by US Medicare; thus, conserving between 30 000 and 90 000 labour-hours per year in an average 500-bed hospital. Hospitals can use this predictive algorithm to initiate a quality improvement programme for pressure injury prevention and further customise the algorithm to patient-specific variation by facility.

Building for Value: A Foundational Structure to Support Population Health.

The journey to value relies heavily on a strong foundation in population health and on supporting systems of care. However, as the Centers for Medicare & Medicaid Services and commercial insurers rethink reimbursements to achieve cost savings, both patients and payments to health care organizations are at risk. The case for value-based care is ever stronger yet health systems will have to mature their culture, population health infrastructure, technologies and analytics capabilities, and leadership and management systems. In this article, the authors describe the functional organizational structure of the clinical transformation team responsible for population health in the University Hospitals Accountable Care Organizations (ACO). Based on their experiences building and evolving population health for the University Hospitals ACO, the authors layout the 3 pillars supporting their structure, including operations, clinical design, and data and analytics, and key areas of focus for each pillar.

A Cost-Utility Analysis of Remote Pulse-Oximetry Monitoring of Patients With COVID-19.

OBJECTIVES: Since 2020, COVID-19 has infected tens of millions and caused hundreds of thousands of fatalities in the United States. Infection waves lead to increased emergency department utilization and critical care admission for patients with respiratory distress. Although many individuals develop symptoms necessitating a ventilator, some patients with COVID-19 can remain at home to mitigate hospital overcrowding. Remote pulse-oximetry (pulse-ox) monitoring of moderately ill patients with COVID-19 can be used to monitor symptom escalation and trigger hospital visits, as needed. METHODS: We analyzed the cost-utility of remote pulse-ox monitoring using a Markov model with a 3-week time horizon and daily cycles from a US health sector perspective. Costs (US dollar 2020) and outcomes were derived from the University Hospitals' real-world evidence and published literature. Costs and quality-adjusted life-years (QALYs) were used to determine the incremental cost-effectiveness ratio at a cost-effectiveness threshold of $100 000 per QALY. We assessed model uncertainty using univariate and probabilistic sensitivity analyses. RESULTS: Model results demonstrated that remote monitoring dominates current standard care, by reducing costs ($11 472 saved) and improving outcomes (0.013 QALYs gained). There were 87% fewer hospitalizations and 77% fewer deaths among patients with access to remote pulse-ox monitoring. The incremental cost-effectiveness ratio was not sensitive to uncertainty ranges in the model. CONCLUSIONS: Patient with COVID-19 remote pulse-ox monitoring increases the specificity of those requiring follow-up care for escalating symptoms. We recommend remote monitoring adoption across health systems to economically manage COVID-19 volume surges, maintain patients' comfort, reduce community infection spread, and carefully monitor needs of multiple individuals from one location by trained experts.

The Unrecognized Impact of Anxiety in Complex and Costly Patients.

Strategies to reduce suffering and expense for complex and costly patients have met with limited success. This may be due to both the ongoing dependence on transactional relationships and the failure to recognize anxiety spectrum disorders as a primary driver of medical complexity. The authors describe an emerging current of thought regarding a universal approach to the conceptualization of anxiety disorders and extend it for application to medical complexity. Using 4 cases, they illustrate distinct anxiety-complexity patterns and describe how a relational intervention untangled and identified treatment targets within that process, with excellent results for patients, providers, and payors. They go on to propose future directions and implications of this intervention.

Achieving Large-Scale Quality Improvement in Primary Care Annual Wellness Visits and Hierarchical Condition Coding.

BACKGROUND: Completion of Medicare Annual Wellness Visits (AWV) and documentation of Hierarchical Condition Categories (HCC) are important metrics in accountable care organizations (ACO) with quality and financial implications. To improve performance in large healthcare organizations, quality improvement (QI) efforts need to be scaled up in a way that is feasible within available system-wide resources. OBJECTIVE: We describe a 3-year effort using a multifaceted QI framework called the fractal management system for AWV and HCC performance. DESIGN: Pre-post evaluation of a multi-level, health system-wide QI management system intervention between 2018 and 2020. The system provided project management, coaching, communications, feedback of performance, and health informatics. PARTICIPANTS: The intervention was delivered to all 97 primary care practices within an Ohio-based accountable care organization, comprising 72,603 attributed Medicare and Medicare Advantage patients as of 2018. Eighty-nine of these practices were included in the analysis. APPROACH: AWV completion was defined as percent of eligible patients with a documented AWV during the calendar year. HCC completion was defined as documented reassessment of all prior-year HCC conditions. KEY RESULTS: AWV completion at the practice level increased from 23.7% (SD .14) in 2018 to 34.9% (SD .18) in 2019, and 59.8% (SD .17) in 2020. This was a statistically significant effect of time on AWV completion rates overall (F[2, 87] = 164.43, p < .000). More than half (56.2%) of practices met or exceeded the 60% goal in 2020. Practice-level HCC completion tracking started in 2019 (M = 75.9%, SD 7.4%) and increased in 2020 (M = 79.7%, SD 7.1%); t(172) = 2.0, p < .001. CONCLUSIONS: AWV and HCC performance goals were met in 2020, despite service disruptions due to COVID-19. The QI approach we used is applicable to other problems and other large healthcare systems.

Digital Health: Unlocking Value in a Post-Pandemic World.

The COVID-19 pandemic has forever changed health care, spurring a revolution in digital health technologies. Across the world, hundreds of thousands of health care systems are considering a central question: how do we connect with our patients? Digital health has been used as a stopgap in many cases to continue the essential functions of health systems. As the post-pandemic world and our "new normal" come into focus, further needs will have to be met with a digital patient interaction, with an eye toward value transformation. One barrier to fully leveraging digital tools is the lack of a framework for classifying the type of digital health care. This can limit our ability to design, deploy, evaluate, and communicate through digital means. This article presents 3 categories of digital health and their relationships to value metrics: (1) telehealth or direct care delivery, (2) digital access tools, and (3) digital monitoring. An evidence-based discussion reveals past successes, current promises, and future challenges in reducing defects in value through digital care. In the coming years, value transformation will become more crucial to the success of health care systems. By using the taxonomy in this article, health systems can better implement digital tools with a value-driven purpose. Defining the role of digital health in the post-pandemic world is needed to assist health systems and practices to build a bridge to value-based care.

Stratifying for Value: An Updated Population Health Risk Stratification Approach.

Most risk stratification approaches attempt to predict clinical outcomes rather than value. For a provider organization or health system to have financial success in value-based contracting, future risk models must analyze costs as well as disease burden. The purpose of this study was to create a customized risk stratification algorithm that considered a patient's medical spend alongside disease burden while delivering a scoring system that improves the efficiency of a care coordination program. The authors focused on University Hospitals (UH) Health System's Accountable Care Organization population of 554,805 because this patient cohort is engaged with UH's primary care network and has the most robust data. The 5-category risk algorithm was found to be meaningful and impactful after integrating the foundation of the Minnesota Tiering system with an expanded comorbidity list and weighting the result by the previous 12 months of medical spend. This new technique can identify patients in need of intensive care coordination. The complex risk tier of the stratification system reduces the number of patients from 551,045 to 27,552, or 5% of the patient population, and accounts for 67.9% ($1,107,822,887) of total annual medical spend. Expanding care coordination efforts to patients in the top 2 tiers would account for 15% of the patients and 83.2% ($1,357,545,872) of annual medical spend. The novelty of the new approach allows clinical teams to focus intense resources on a smaller sample of the patient population and to identify chronic conditions contributing to costs, and feel confident that they have greater explanatory power regarding value.

Social Factors Predictive of Intensive Care Utilization in Technology-Dependent Children, a Retrospective Multicenter Cohort Study.

Objective: Technology-dependent children with medical complexity (CMC) are frequently admitted to the pediatric intensive care unit (PICU). The social risk factors for high PICU utilization in these children are not well described. The objective of this study was to describe the relationship between race, ethnicity, insurance status, estimated household income, and PICU admission following the placement of a tracheostomy and/or gastrostomy (GT) in CMC. Study Design: This was a retrospective multicenter study of children <19 years requiring tracheostomy and/or GT placement discharged from a hospital contributing to the Pediatric Health Information System (PHIS) database between January 2016 and March 2019. Primary predictors included estimated household income, insurance status, and race/ethnicity. Additional predictor variables collected included patient age, sex, number of chronic complex conditions (CCC), history of prematurity, and discharge disposition following index hospitalization. The primary outcome was need for PICU readmission within 30 days of hospital discharge. Secondary outcomes included repeated PICU admissions and total hospital costs within 1 year of tracheostomy and/or GT placement. Results: Patients requiring a PICU readmission within 30 days of index hospitalization for tracheostomy or GT placement accounted for 6% of the 20,085 included subjects. In multivariate analyses, public insurance [OR 1.28 (95% C.I. 1.12-1.47), p < 0.001] was associated with PICU readmission within 30 days of hospital discharge while living below the federal poverty threshold (FPT) was associated with a lower odds of 30-day PICU readmission [OR 0.7 (95% C.I. 0.51-0.95), p = 0.0267]. Over 20% (n = 4,197) of children required multiple (>1) PICU admissions within one year from index hospitalization. In multivariate analysis, Black children [OR 1.20 (95% C.I. 1.10-1.32), p < 0.001] and those with public insurance [OR 1.34 (95% C.I. 1.24-1.46), p < 0.001] had higher odds of multiple PICU admissions. Social risk factors were not associated with total hospital costs accrued within 1 year of tracheostomy and/or GT placement. Conclusions: In a multicenter cohort study, Black children and those with public insurance had higher PICU utilization following tracheostomy and/or GT placement. Future research should target improving healthcare outcomes in these high-risk populations.

Using Economic Evaluation to Illustrate Value of Care for Improving Patient Safety and Quality: Choosing the Right Method.

ABSTRACT: To scale and sustain successful quality improvement (QI) interventions, it is recommended for health system leaders to calculate the economic and financial sustainability of the intervention. Many methods of economic evaluation exist, and the type of method depends on the audience: providers, researchers, and hospital executives. This is a primer to introduce cost-effectiveness analysis, budget impact analysis, and return on investment calculation as 3 distinct methods for each stakeholder needing a measurement of the value of QI at the health system level. Using cases for the QI of hospital-acquired condition rates (e.g., pressure injuries), this primer proceeds stepwise through each method beginning from the same starting point of constructing a model so that the repetition of steps is minimized and thereby capturing the attention of all intended audiences.

Prioritizing Health Care Solutions for Pressure Ulcers Using the Quality Function Deployment Process.

Reducing the incidence and morbidity of pressure ulcers remains a leading national priority in patient safety. However, the optimal strategy for a hospital or health system to address this safety goal is not straightforward given the number and complexity of available solutions. Leveraging techniques from systems engineering, such as the quality function deployment process, may provide a transparent and objective way to address this challenge. A detailed and practical application of quality function deployment is presented that demonstrates the value of applying engineering practices for prioritizing solutions for pressures ulcers specifically and can easily be adapted to other conditions.

Adverse Effects of the Medicare PSI-90 Hospital Penalty System on Revenue-Neutral Hospital-Acquired Conditions.

OBJECTIVE: Health systems are grappling with improving the quality and safety of health care. By setting clear expectations, there is an opportunity to configure care models to decrease the risk of adverse events and promote the quality of care. The US Centers for Medicare and Medicaid Services have used Patient Safety Indicator 90 (PSI90), a composite rate of hospital-acquired conditions (HACs), to adjust payments and score hospitals on quality since 2015. However, PSI90 may be associated with adverse prioritization for preventing some conditions over others.Our objective was to evaluate the time-dependent rates of HACs between 2013 and 2016 to assess the association of funding models on adverse events, particularly pressure injury. METHODS: We analyzed a retrospective observational cohort of patients hospitalized in US Academic Medical Centers observed by the Vizient CDB/RM pre-post PSI90 implementation. Changes in HAC component rates of PSI90 between 2013 and 2016 were measured longitudinally using mixed-effects negative binomial regression modeling. RESULTS: Regardless of whether the composite measure of patient outcomes was PSI90 or all HACs, in general, there was significant decrease after PSI90 was implemented, reflecting an association between PSI90 and CMS reimbursement policy. However, pressure injury rates increased by 29.4% (SE = 0.08; P < 0.05) during this time frame, the only HAC observed to increase related to PSI90. CONCLUSIONS: Patient safety in hospitals will only thoroughly improve when hospitals are fully incentivized to practice prevention of all HACs rather than work around the harms that result from failed prevention efforts.

Value of hospital resources for effective pressure injury prevention: a cost-effectiveness analysis.

OBJECTIVE: Hospital-acquired pressure injuries are localised skin injuries that cause significant mortality and are costly. Nursing best practices prevent pressure injuries, including time-consuming, complex tasks that lack payment incentives. The Braden Scale is an evidence-based stratification tool nurses use daily to assess pressure-injury risk. Our objective was to analyse the cost-utility of performing repeated risk-assessment for pressure-injury prevention in all patients or high-risk groups. DESIGN: Cost-utility analysis using Markov modelling from US societal and healthcare sector perspectives within a 1-year time horizon. SETTING: Patient-level longitudinal data on 34 787 encounters from an academic hospital electronic health record (EHR) between 2011 and 2014, including daily Braden scores. Supervised machine learning simulated age-adjusted transition probabilities between risk levels and pressure injuries. PARTICIPANTS: Hospitalised adults with Braden scores classified into five risk levels: very high risk (6-9), high risk (10-11), moderate risk (12-14), at-risk (15-18), minimal risk (19-23). INTERVENTIONS: Standard care, repeated risk assessment in all risk levels or only repeated risk assessment in high-risk strata based on machine-learning simulations. MAIN OUTCOME MEASURES: Costs (2016 $US) of pressure-injury treatment and prevention, and quality-adjusted life years (QALYs) related to pressure injuries were weighted by transition probabilities to calculate the incremental cost-effectiveness ratio (ICER) at $100 000/QALY willingness-to-pay. Univariate and probabilistic sensitivity analyses tested model uncertainty. RESULTS: Simulating prevention for all patients yielded greater QALYs at higher cost from societal and healthcare sector perspectives, equating to ICERs of $2000/QALY and $2142/QALY, respectively. Risk-stratified follow-up in patients with Braden scores <15 dominated standard care. Prevention for all patients was cost-effective in >99% of probabilistic simulations. CONCLUSION: Our analysis using EHR data maintains that pressure-injury prevention for all inpatients is cost-effective. Hospitals should invest in nursing compliance with international prevention guidelines.