Nephrology, Hypertension & Renal Transplantation 2022

PRN Blood Pressure Medical Use in VA Hospitals

Faculty Mentor’s Name: Dr. Muna Canales
Email: munacanalas@medicineufl.edu
Phone Number: (352) 672-7433
Project Category: Case Review
International Component or Travel: No

Research Project Description:

Study overview: As many as 7 out of 10 Veterans have high blood pressure while in the hospital. Many doctors and nurses treat this high blood pressure with quick-acting blood pressure medications (PRN BP), even if the patient has no symptoms. Currently, there are no guidelines to say that this is the right thing to do. Also, recent studies show that treating high blood pressure in the hospital with quick-acting medicines may lead to very low blood pressure, kidney failure, stroke, falls and other harms, especially in older patients with other chronic illnesses like Veterans. With this study, we will find out how often doctors treat high blood pressure in VA hospitals with quick-acting medicines and talk to doctors and nurses about what makes them decide to treat. We will use this information to, in later projects, develop a tool to help doctors and nurses use these medications the right way. We hope that our research will, in the long term, improve the quality, safety and cost of hospital care for our Veterans.

Objectives:
Aim 1: To identify the prevalence, sociodemographic and clinical predictors and impact of PRN BP use among Veterans hospitalized in VA hospitals nationally from FY16-FY20.
Aim 2: To understand provider, nursing and system-level factors motivating PRN blood pressure use in VA hospitals through qualitative research techniques

Research Plan: We will identify the source analytic cohort for Aim 1 of this proposal using information in the Corporate Data Warehouse (CDW) based upon the following criteria: Inclusion criteria: 1) VHA enrollees age 18-89; 2) hospitalized for ≥ 3 days at least once in FY16-20. Rationale: ≥3 days of hospitalization allows time to ascertain at least one of our outcomes Exclusion criteria: 1) Initial admission to ICU; 2) admission diagnosis code for stroke, fall, myocardial infarction, aortic dissection or HTN emergency, or diagnosis code or laboratory data indicating AKI; 3) prior diagnosis code for end stage kidney disease. For outcomes analysis, we further require a diagnosis of hypertension by diagnosis code. Data sources include the Corporate Data Warehouse (CDW), Pharmacy Benefits management (PBM) system, the Managerial Cost Accounting (MCA) System National Data Extracts (NDE), and the VA Vital Status File (VSF).
For Aim 2, we will perform 1 on 1 interviews with 32 providers and nurses at the Malcom Randall VAMC. Inclusion criteria: 1) Any clinical provider (physician or mid-level or house-staff) or nurse providing active inpatient non-ICU care at Malcom Randall VA hospital for ≥ 6 months prior to contact. Exclusion criteria: 1) Providers or nurses who are unwilling to participate or who do not wish to be audio-recorded if participating in the interview component.

Role of the Medical Student: In order to establish our case definition of PRN BP use and validate our algorithm to identify PRN BP use in the national VA database, the medical student (under PI supervision)
will randomly select a subset of 100 North Florida/South Georgia Veterans identified at the national level as having received PRN BP medication and review the Computerized Patient Record System to verify all information provided by the national database regarding PRN BP. We will create a 2 X 2 table for each group comparing the national data vs chart review data in order to compute the sensitivity, specificity, PPV, NPV and accuracy of the national VA data source to identify PRN vs no PRN BP use. Our goal is to have 90% accuracy for identification of PRN BP use (vs no PRN BP) from the national databse. Based upon the results of this analysis, we will optimize our query as needed in consultation with national data managers. If time allows, the medical student is welcome to participate in other components of the research planned, as outlined above.

Funding: This is a VA funded 18 month project projected to begin in April 2022.

Relevant publications for student review:

https://www.ahajournals.org/doi/epub/10.1161/HYPERTENSIONAHA.121.17279

Rastogi R, Sheehan MM, Hu B, Shaker V, Kojima L, Rothberg MB. Treatment and Outcomes of
Inpatient Hypertension Among Adults With Noncardiac Admissions. JAMA Intern Med. Mar
2021;181(3):345-352.

Integration of IntraOperative data with Preoperative data for improved Postoperative Sepsis Risk prediction

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The project aims at developing prediction models to help study patient risk scores for postoperative sepsis complication given preoperative medical history and intraoperative EMR responses.

Clinical phenotypes for AKI using clustering of time series

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The purpose of this study is to use clustering to predict pathophysiologic signatures and clinical trajectory during the early stages of acute kidney injury. We hypothesized that clustering analysis of electronic health record data would reveal distinct, predictable patient phenotypes with unique pathophysiological signatures and clinical trajectories.

ALPHA-Artificial intelligence Augmented Acute Illness Phenotypes

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Patients with acute illness in intensive care unit (ICU) have sudden dysfunction of multiple organ systems presenting as a heterogenous syndrome that encompasses a multidimensional array of clinical, socioeconomic and biological features. Different combinations of these features may naturally cluster into previously undescribed subsets or “clinical phenotypes” that may have different clinical trajectories. The digital data captured during routine ICU care can be utilized for understanding and classifying acute illness using unsupervised clustering approaches but the complexity and granularity of ICU data poses a problem for conventional clustering. We propose to utilize deep learning networks to learn non-linear mappings and transform ICU data into clustering-friendly representation.

ALPHA-Artificial intelligence Augmented Acute Kidney Injury Subphenotypes

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Acute kidney injury (AKI) is one of the most common complications among hospitalized patients and is associated with morbidity, mortality and increased health cost. There is vast heterogeneity in cause, severity, trajectory and outcomes of AKI that are influenced by inherent disease mechanisms, the patient’s response and the care process. This heterogeneous syndrome encompasses a vast, multidimensional array of clinical, socioeconomic and biological features. Different combinations of these features may naturally cluster into previously undescribed subsets or “clinical phenotypes” that may have different clinical trajectories. These phenotypes must be identifiable soon after AKI onset to guide treatment, and efforts to determine such phenotypes have remained limited. Paradoxically the abundant real-time physiologic, laboratory, clinical notes, and other clinical data are available in the electronic health records (EHR) but their magnitude and complexity often overwhelms physicians’ ability to comprehend, retain, and organize the information in an optimal and timely way. The overall objective of this application is to identify AKI subphenotypes by developing and validating a generalizable, reproducible and fair algorithm using multimodal clinical data and novel deep learning (DL) technologies.

Learning optimal treatment strategies for hypotension in critical care patients with acute kidney injury using artificial intelligence

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Acute kidney injury (AKI) affects nearly one-quarter of hospitalized patients worldwide and up to 60% of patients in the intensive care unit (ICU). The delayed or incomplete recovery of renal function confers increased risk for chronic critical illness with poor long-term survival and quality of life. Prevention, early diagnosis, and appropriate treatment with euvolemia, avoidance of nephrotoxic substances, and relief of obstructive uropathy have variable efficacy in improving patient outcomes. To optimize these management strategies and their early delivery, it is necessary to understand mechanisms that yield different trajectories of AKI and recovery among hospitalized patients. Hypotension is a well-recognized modifiable risk factor for development of AKI and complications and in critical care settings is a life-threatening emergency that must be treated early.

The pathophysiology of AKI and trajectory is influenced by patients’ preexisting health, the management of events, and especially those events relating to hypotension. The electronic health records (EHR) contain a wealth of clinical data that could be used to identify all dimensions of an AKI episode as well as key determinants of undesired outcomes. The abundant real-time physiologic, laboratory, and other clinical data are available in the EHR but their magnitude and complexity often overwhelms physicians’ ability to comprehend, retain, and organize the information in an optimal and timely way. Patients and physicians make essential decisions on which diagnostic and therapeutic interventions should be performed or deferred under time constraints and uncertainty regarding patients’ diagnoses and predicted response to treatment which may lead to cognitive and judgment errors. Current practices in the administration of intravenous fluids and vasopressors to treat hypotension for patients with AKI are suboptimal. Reinforcement learning (RL) is a subfield of artificial intelligence (AI) that identifies a sequence of decisions to increase the likelihood of favorable outcomes and has the potential to play complementary roles in delivering high-value care through sound judgment and optimal decision-making.

Defining Postoperative Intensity of Care

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

After major inpatient surgery, patients’ risk for critical illness and death, summarized as patient acuity, should match triage destination (i.e. intensive care unit (ICU) vs. ward) and the frequency of vital sign monitoring and laboratory testing, summarized as intensity of care. There is no validated, unifying definition for postoperative intensity of care. Our objective is to develop and validate postoperative intensity of care definitions. This project will test the hypothesis that postoperative intensity of care can be autonomously adjudicated in EHRs as optimal (low-acuity and low-intensity or high-acuity and high-intensity) or suboptimal, which may be insufficient (high-acuity and low-intensity), or excessive (low-acuity and high-intensity), and that suboptimal intensity of care will be common and associated with increased in-hospital and 1-year mortality, morbidity, and costs. Our approach will be to use retrospective preoperative, intraoperative, and postoperative EHR data from surgical inpatients at two institutions, classify high- and low-acuity with a random forest classifier that predicts postoperative critical illness and death, classify high- and low-intensity of care according to ICU vs. ward location and frequency of vital sign monitoring and laboratory test orders, and compare in-hospital and 1-year mortality, morbidity, and costs among patients receiving optimal, insufficient, and excessive intensity of care.

The Artificial Intelligence Learns Optimal Treatment Strategies for Hypotension in Surgery

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Patients and physicians make essential decisions on which diagnostic and therapeutic interventions should be performed or deferred under time constraints and uncertainty regarding patients’ diagnoses and predicted response to treatment which may lead to cognitive and judgment errors. We aim to develop a deep reinforcement learning model that would provide individualized and clinically interpretable treatment decisions that could help balance blood pressure during surgery and decrease postoperative complications, and improve patient outcomes in patients undergoing surgery. Current practices in the administration of intravenous fluids and vasopressors during surgery as a treatment of hypotension are suboptimal. Reinforcement learning have the potential to play complementary roles in delivering high-value surgical care through sound judgment and optimal decision-making. We aim to develop a deep RL model that would provide individualized and clinically interpretable treatment decisions that could help balance blood pressure during surgery and decrease postoperative complications in patients undergoing surgery.

Addressing Gender Disparities in Leaderships in Nephrology

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

Longstanding inequities in our field unfairly hinder career advancement for nonmale-identifying individuals and for individuals from underrepresented racial, ethnic, and socioeconomic groups. The project aims to review evidence for current state of the women and underrepresented in medicine (URiM) groups in leadership in nephrology.

Sepsis Computable Phenotypes

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The project aims at developing computable phenotype algorithms to automate sepsis diagnosis.

Validation of Acute Kidney Injury (AKI) Computable Phenotypes

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The project aims at validating computable phenotype algorithms to automate AKI diagnosis in a multicenter cohort.

The role of the medical student will be to review electronic health records to adjudicate AKI phenotyping as well as help prepare results and do literature review and write introduction and discussion parts of the manuscript.

Validation of acuity and acute brain dysfunction computable phenotypes

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

The project aims at validating computable phenotype algorithms to automate patient acuity and acute brain dysfunction

The role of the medical student will be to review electronic health records to adjudicate patient acuity and acute brain dysfunction as well as help prepare results and do literature review and write introduction and discussion parts of the manuscript.

Other Mentor: Dr. Tezcan Ozrazgat Baslanti (tezcan.ozrazgat@medicine.ufl.edu)

Automated acute kidney injury assessment using computable phenotypes for health care providers

Faculty Mentor’s Name: Dr. Azra Bihorac
Email: munacanalas@medicineufl.edu
Phone Number: (352) 294-8580
Project Category: Clinical
International Component or Travel: No

Research Project Description:

We propose a multidisciplinary team of healthcare workers, information technology specialists, engineers and patients to develop an automated algorithm that uses EHR to identify patients with AKI and their AKI severity in real time using changes in serum creatinine and to communicate that information to healthcare workers, patients and patients’ families. This proposal would increase communication effectiveness and timeliness and decrease harm due to any delay in detecting AKI or its progression.

The role of the medical student will be to assist with data analysis, preparation of results, literature review, and writing parts of the manuscript.

Other Mentor: Dr. Tezcan Ozrazgat Baslanti (tezcan.ozrazgat@medicine.ufl.edu)