Diagnostic Safety and Quality Optimization in Sepsis

The Diagnostic Safety and Quality Optimization in Sepsis study is a multi-site study funded by the Agency for Health Care Research and Quality (AHRQ), which aims to improve the safety and quality of early sepsis diagnosis using principles and tools of safety and implementation sciences.

The study will conduct electronic surveys and in-depth interviews with organizational stakeholders from each hospital to collect barriers and facilitators of acute sepsis diagnosis. The retrospective sepsis data will be collected from electronic medical records of 20 hospitals located in different sites.

An organizational toolkit will be designed and disseminated to hospitals to guide implementation or maintenance of strategies aligned with organizational context to improve or sustain optimal sepsis diagnosis. The study will then evaluate the usability and utility of toolkit in sepsis diagnosis.

PI: Stephanie Taylor, MD, MS

Project Period: 5 years

Funded by: Agency for Health Research and Quality 

Sepsis Observational Data Project 

The sepsis observational data project aims to evaluate outcomes for patients and the delivery of care they receive across Michigan Medicine using retrospective health care data to answer questions about infectious diseases, sepsis and their treatment. 

PI: Stephanie Taylor, MD, MS

Sepsis Short Stay Unit

The study aims to investigate the prevalence, risk factors, practice variation, and outcomes for short stay admission among emergency department (ED) patients with sepsis. The study will use Hospital Medicine Safety Consortium data from 69 hospitals to identify patient and hospital factors associated with inpatient disposition and compare risk-adjusted in-hospital mortality and readmission after the initial short stay vs inpatient assignment from the ED. 

PI: Stephanie Taylor, MD, MS and David Paje, MD, MPH 

Funded by: Collaborative Quality Initiatives Operations

Effective Primary Care Practices that Enhance Recovery Trajectories after Pneumonia (EXPERT) 

The overarching goal of our work is to improve patient outcomes after pneumonia. The objective of this EXPERT study is to apply a novel strengths-based approach for transformational change to refine an intervention that helps primary care providers promote the recovery of pneumonia survivors.

The study will conduct quantitative analysis of electronic medical record data to identify and refine candidate components for a primary care intervention to improve post-pneumonia care and qualitative interviews with expert primary care clinicians and patients to identify enablers and barriers for the post-pneumonia intervention.  

PI: Stephanie Taylor, MD, MS

Funded by: NIH-National Heart, Lung, and Blood Institute 

Medicine Procedure Services Study

The Medicine Procedure Services (MPS) study is an observational cohort study which aims to understand the current clinical practice around these minor procedures (thoracentesis, paracentesis, and lumbar puncture) by describing patient and provider characteristics, and assessing relevant clinical and efficiency outcomes. Data are collected retrospectively through abstraction from patients’ electronic medical records.

The study will help appraise current practice and guide quality improvement efforts regarding the performance of minor medical procedures in the hospital setting.

PI: David Paje, MD, MPH

Hospitalist Detection of Adverse Drug Reactions Present on Admission to a Short Stay Medical

This project will seek to identify a baseline rate of detection of adverse drug reactions (ADR) occurring in adult patients admitted by Hospitalists to a Medical Short Stay Unit to improve the detection and management of ADRs. Data will be collected retrospectively through abstraction from patients’ electronic medical records.

PI: Jason Ham, MB, BCh, BAO

Geographic localization

Geographic localization (GL) of hospitalist teams to specific inpatient units has been associated with improvements in patient care. This study examines the mechanisms by which geographic localization of hospitalist teams to specific inpatient units affects patient outcomes, interdisciplinary collaboration, provider wellbeing, and operational efficiency within a large quaternary academic hospital, as well as the unintended consequences of localization. The study will use a mixed-method study design to compare the outcomes in a localized inpatient unit with those in a non-localized unit during the same timeframe.

PI: Irina Khurana, MD

Comparison Of Patient Level Outcomes of Those Patients Care for On Services Where Faculty Rotate For 1 Week at A Time vs 2 Weeks at A Time 

Traditionally at the University of Michigan, faculty rotate on our general medicine teaching services for two weeks at a time.  While faculty may have individual preferences about duration of time on service, little data exists around patient outcomes, educational outcomes, or burnout rates for faculty or learners on the team in order to identify the optimal duration for faculty to be assigned to a general medicine service. For the 2024-2025 academic year on our resident based general medicine services, two of the teams have faculty assigned in one-week intervals and two teams will have faculty assigned in two-week intervals. Therefore, the study aims to understand the impact of this pilot on patient outcomes, trainee evaluation completion rates by faculty, perceptions of faculty and trainees, and faculty well-being.   

PI: Kate Levy, MD

"HEAL: Hospitalists’ Equity and Advocacy Learning – Addressing Discrimination in Hospital Medicine"

Discrimination remains a significant issue for physicians, especially those in hospital medicine and internal medicine. While the negative impact of discrimination on physician well-being is well described, there is limited understanding of how to systematically prevent and mitigate discrimination effectively in Hospital Medicine. The study aims to conduct a survey with hospitalists to provide knowledge around the experience of discrimination in hospital medicine as well as best practices as it pertains to policies and resources that exist to mitigate/address discrimination within hospital medicine.

PI: Ruby Marr, MD

Enhancing Mortality Risk Documentation with MiCast 

Mortality risk assessment is crucial for predicting patient outcomes, improving healthcare quality, and ensuring patient safety. Improved documentation and coding can clarify increased patient complexity and length of stay, as well as boost hospital reimbursement. We initiated a hospital medicine (HM) division quality improvement (QI) project in July 2024 to promote and increase the use of MiCast for improved mortality risk documentation, and measure changes in observed to expected (O:E) mortality ratio after initiating the QI program. The study aims to understand the barriers and facilitators to using Mi-Cast to improve its uptake within the division. The data will be collected from hospitalist through electronic survey via REDCap. 

PI: Sri Ramani Peesapati, MD

Blood Culture Bottle Shortage Study

This study assesses the impact of a BPA on (1) the frequency of blood culture bottle ordering and (2) the unintended consequences of that BPA on length of stay, empiric antibiotic ordering and adherence to SEP-1 criteria. We will collect data from the electronic medical record pre- and post- intervention for comparison. The study will help appraise the balance of stewardship and clinical outcomes in the hospital setting. Data will be collected retrospectively through abstraction from patients’ electronic medical records.

PI: Robert Chang, MD

Code Status Preference and Quality of Goals of Care Documentation in Hospitalized Adults Prior To Decompensation and Transfer to the ICU 

Despite the known importance of patient’s goals of care, less is known about the quality of goals of care and resuscitative preference conversations prior to decompensation. The aim of this study is to examine the current quality of goals of care discussion documentation in adult patients admitted to the floor on general medicine services who decompensate and require transfer to the ICU. The study will systematically identify documented presence or absence of key information deemed important for goals of care conversations and evaluate patient level and system level factors, such as admitting team structure, associated with resuscitative preference and quality of discussions.  

PI: Emily Jacobson, MD

CLOSED STUDIES 

Impact of Using Bedside Mobility Assessment Tool on Clinical Outcomes in the Inpatient Setting  

The Bedside Mobility Assessment Tool (BMAT) was developed to help nurses improve patient mobility in an acute setting. It has been used to inform nursing plans that promote safe patient mobilization during hospitalization.

The purpose of this BMAT study is to evaluate if the use of BMAT on hospitalized older adult (≥65 years) patients is associated with improvements in clinical quality and efficiency outcomes, such as utilization of physical therapy consultation, length of stay, and the occurrence of falls, hospital-acquired pressure injury, and 15-day readmission. The study will collect pre- and post-intervention data retrospectively from patients’ electronic medical records.  

PI: Linda Wang, MD

A Pre-Post Study to Evaluate the Feasibility of Criteria-Based Discharge Communication (CBDC) Project Intervention

The purpose of the Criteria-Based Discharge Communication (CBDC) study is to evaluate the feasibility, acceptability, and usefulness of CBDC intervention. Physicians usually communicate with patients and families about discharge in a time-based fashion rather than in a criteria-based fashion. Preliminary and anecdotal data indicate that when physicians communicate with patients in a criteria-based fashion, shared expectations result in what must be and can reasonably be accomplished while inpatient and communication and patient satisfaction are improved.

The study will collect data from both patients and providers during pre- and post-implementation period. Additionally, the retrospective data will be collected from patients’ electronic medical charts.  

PI: Julieann Grant, MD

Priority Discharge Study

The Priority Discharge study aims to understand the key effectiveness/driving components of the priority discharge intervention.

The study uses a structured analysis of qualitative and quantitative data combined with a scoping review of priority discharge (PD) literature to identify core functions and forms of PD programs. The study conducts surveys with stakeholders involved in the PD program and performs scoping review of the PD program described in the literature to identify additional intervention data that may differ from our local context.

Data will be summarized to develop PD functions and form a matrix displaying a range of interventions that can support effective implementation of programs in diverse clinical settings. 

PI: Jigisha Patel, MD

Assessing Older Adult Patients' Perspectives on Mobility and Preparedness for Mobile App-Based Interventions (MAP) 

This study aims to identify patient perspectives including knowledge, attitudes, and behaviors/practice in hospital-associated deconditioning. The study aims to also assess digital literacy including comfort level using mobile devices in general and for health-related content. The study will conduct an electronic survey with hospitalized patients ≥ 65 years old. 

PI: David Paje, MD, MPH

Development and Feasibility of Mobile App (MoveRx) 

This study aims to develop, test, and implement a mobile device application (MoveRx) to prevent hospital-associated deconditioning. MoveRx will be designed to prompt and encourage patients to perform physical activities during hospitalization without creating a significant additional tasking burden on the health care team.

MoveRx will direct patients to perform individualized activities that are appropriate for their mobility level (assessed using AM-PAC 6 clicks and John Hopkins Highest Level of Mobility Scale). Patients will use their personal smartphone device with MoveRx installed. MoveRx notifications will remind the patient to complete three periods of ambulation activity per day, which provides a stimulus for behavioral regulation (choosing to be active) with a focus on maintaining strength of lower extremities. Patient ambulation goals will be based on Johns Hopkins Mobility Goal Calculator. This tool establishes an algorithm approach to setting daily mobility goals based on mobility limitation assessments. 

PI: David Paje, MD, MPH

Evaluating the Effects of Implementing AM-PAC “6-clicks” and JH-HLM on Hospital-Associated Deconditioning (MI-HAD) 

This is a retrospective observational study that aims to identify patterns and predictors of hospital-associated deconditioning. Hospitalization-associated disability manifests as the loss of ability to complete one of the basic ADLs needed to live independently without assistance: bathing, dressing, rising from bed or a chair, using the toilet, eating, or walking across a room.

The study will also assess the implementation of standardized nursing assessments of patient mobility (AM-PAC 6 clicks and JH-HLM) at Michigan Medicine University Hospital by measuring compliance rates at each nursing unit and compare high-performing units (units in the highest tertile of compliance) and low-performing units (units in the lowest tertile of compliance) by patient characteristics and unit structural and functional characteristics. 

PI: David Paje, MD, MPH

Tailoring a Physical Activity Behavior Nudge to Reduce Hospital-Acquired Deconditioning and Frailty Among Older Women 

The objective of this project is to refine our novel hospital mobilization behavior nudge, MoveRx by tailoring it to the specific needs and barriers of older women patients in order to reduce the development of Hospital Acquired Deconditioning and subsequent frailty. The study will use a participatory prototyping that introduces physical prototypes of interventions to participants to engage those participants in iterative improvements. The study hypothesize that a multilevel intervention optimized using this participatory strategy will mitigate patient and clinician barriers to achievement of adequate physical activity for women in the hospital.

PI: Stephanie Taylor, MD, MS

Funded by: Women’s Health Innovation

Non-Interventional, Observational Studies Using Retrospective Health Care Data to Evaluate Outcomes Related to Delivery of Hospital Care at Home (Care@Home LHS) 

This study aims to evaluate outcomes for patients and the delivery of care they receive across Michigan Medicine using retrospective health care data to answer questions about hospital care at home. Health care data will be accessed from databases which record health data on patients treated by U-M.

These data will contain protected health information (PHI) to allow for accurate data integration from different systems and will be de-identified to create the study database. All studies will be strictly retrospective and based on data collected by U-M for the purposes of clinical care. All publications will use de-identified and aggregate data to protect patient confidentiality. Survey or interview studies will follow best practices for human subject protection. 

PI: Stephanie Taylor, MD, MS and Anu Goyal, MBChB

Faculty Development Needs Assessment

This study's purpose is to understand the faculty development needs from the faculty's perspective to make changes to the quality of the local faculty development program and better shape the future of faculty development. This study will conduct surveys with hospital medicine faculties from different institutions and will analyze which development activities are preferred by the faculties with the goal of improving individual faculty's clinical quality, scholarly productivity, effectiveness as educators, wellness, and professional fulfillment 

PI: David Bozaan, MD

Utilizing a Logic Model to Enhance Best-Practices in Hospital Medicine Faculty Performance Evaluations 

This study aims to provide a standardized approach for using a logic model to guide best-practices in hospital medicine faculty performance evaluations. The study conducted an electronic survey with hospital medicine program leaders across the country. By leveraging the logic model, we intend to:

1. Identify the key activities of faculty evaluation.

2. Identify the mechanisms of change and intervention activities necessary to produce desired outcomes.

3. Identify any lessons learned from hospital medicine program leaders’ implementation of faculty evaluation processes.   

PI: Anu Goyal, MD

Implementation Research for Comprehensive Cardiometabolic Care in Guatemala

This research project focuses on implementing the WHO-recommended HEARTS Technical Package for Cardiovascular Disease Management in Guatemala. The aims are to identify barriers and facilitators to expanding an ongoing hypertension intervention to include diabetes and dyslipidemia, co-design an integrated treatment model using human-centered design, and evaluate its feasibility and acceptability through a mixed-methods cluster randomized pilot study.

The outcomes will inform a subsequent large-scale project to assess the effectiveness and implementation outcomes across Guatemala and other countries in Latin America.

PI: David Flood, MD

Brian Emmer Lab Research

The Brian Emmer Lab is a highly collaborative and multidisciplinary group of scientists who use modern genomic tools to functionally interrogate the molecular basis of health and disease. Areas of focus include atherosclerotic cardiovascular disease, lipoprotein metabolism, and acute lung injury and regeneration. The Emmer Lab uses a wide variety of experimental approaches with dedicated expertise in high-throughput CRISPR screening, massively parallel reporter assays, and deep mutational scanning. Learn about research in the Emmer Lab.

Utilizing Nucleic Acid Chemistry to Develop a Rapid and Universal Bacterial Identification System for Sepsis

The study team is developing a universal microbial detection system that can identify a bacterial pathogen within a few hours using PCR-based approaches. Without having prior knowledge of the infectious organism, the proposed system can determine both the identity and concentration of each pathogen in a clinical sample without the need for culturing.

Our goal is to design a system that can surpass sequencing, multiplex PCR, and MALDI-TOF in terms of speed, universality, and cost. Rapid identification of a bacterial species can dramatically reduce inappropriate antibiotic usage and help combat growing antimicrobial resistance as well as improve patient outcomes. 

PI: Vishwaratn Asthana, MD, PhD

Funded by: Frankel Innovation Initiative

Using Vectorcardiography to Predict Heart Failure in Patients Following ST Elevation Myocardial Infarction

Modeling outcomes, such as onset of heart failure (HF) or mortality, in patients following ST Elevation Myocardial Infarction (STEMI) is challenging, but clinically very useful. The acute insult following a myocardial infarction and chronic degeneration seen in HF involves a similar process where a loss of cardiomyocytes and abnormal remodeling lead to pump failure. This process may alter the strength and direction of the heart’s net depolarization signal. We hypothesized that changes over time in unique parameters extracted using vectorcardiography (VCG) have the potential to predict outcomes in patients post-STEMI and could eventually be used as a non-invasive and cost-effective surveillance tool for characterizing the severity and progression of HF to guide evidence-based therapies.

The study identified 162 patients discharged from Michigan Medicine between 2016 and 2021 with a diagnosis of acute STEMI. For each patient, a single 12-lead ECG >1 week pre-STEMI and >1 week post-STEMI were collected. A set of unique VCG parameters were derived by analyzing features of the QRS complex. We used LASSO regression analysis incorporating clinical variables and VCG parameters to create a predictive model for HF, mortality, or the composite at 90, 180, and 365 days post-STEMI.

The VCG model is most predictive for HF onset at 90 days with a robust AUC. Variables from the HF model mitigating or driving risk, at a p < 0.05, were primarily parameters that assess the area swept by the depolarization vector.

PI: Vishwaratn Asthana, MD, PhD

Funded by: Frankel Cardiovascular Center (FCVC) Just Do It Award