The purpose of this K23 proposal is to provide Vincent Liu, MD, MS with the protected time and resources to pursue the additional training needed to reach his long-term goal of improving hospital outcomes through the innovative use of complex clinical data. Dr. Liu is a board-certified intensivist and Research Scientist at the Kaiser Permanente Division of Research (KP-DOR). The proposal builds on his background in critical care, traditional biostatistics, and the applied use of highly-detailed electronic medical record data. It also leverages the rich training environment afforded by KP-DOR, advanced didactic training in casual inference and informatics at Stanford University, and a cohesive mentorship team to improve our understanding and treatment of sepsis. Sepsis, the inflammatory response to infection, affects millions of patients worldwide each year and is the most expensive cause of hospitalization in the United States. In severe cases, more than one- third of sepsis patients die in the hospital and a growing population of sepsis survivors face increased mortality and disability rates after discharge. Recent randomized controlled trials (RCTs) have failed to identify efficacious therapies for sepsis and, more than a decade after its original description, we still await three multi- center RCTs designed to definitively demonstrate the benefits of a promising therapy-early goal-directed therapy (EGDT). In light of the challenges of designing and conducting RCTs in sepsis, novel approaches that leverage highly granular observational data from real-world practice are urgently needed to: (1) provide novel insight into sepsis disease mechanisms;(2) link the pathophysiology of sepsis to new interventions;and (3) offer improved disease prognostication for clinicians and patients. This project outlines such an approach. Its training aims are designed to expand skills and knowledge by: (1) optimizing our ability to predict adverse outcomes in sepsis patients after initial triage and treatment;(2) assessing whether EGDT can benefit patients with less severe sepsis;and (3) quantifying the specific impacts of acute organ dysfunction on long-term outcomes among sepsis survivors. The proposed work has high potential to make a significant clinical impact, as these aims will not only advance our understanding of disease mechanisms in sepsis, but also provide preliminary data on putative risk factors to target in future interventional trials. This work will also converge in a R01 proposal aimed at designing a real-time sepsis risk prediction tool. Furthermore, this K23 develops specific examples in sepsis whose logical extension to other critical care syndromes indicates applicability beyond the scope of this proposal and, hence, high public health significance. Importantly, the proposed work is realistic and feasible within the award period and will allow Dr. Liu to continue to build research skills, generate preliminary data, create additional collaborative relationships, and compete for R01 funding. In summary, this K23 award will support and accelerate the career development activities of Dr. Liu and allow him to successfully launch into the next phase of his career as an independent investigator.

Public Health Relevance

Sepsis affects millions of patients worldwide and is the most expensive cause of hospitalization in the United States. Despite this, we still lack a comprehensive understanding of how sepsis affects hospital and long-term mortality. This project is designed to develop advanced methods for using detailed electronic medical record data to improve our understanding and care of sepsis patients.

National Institute of Health (NIH)
National Institute of General Medical Sciences (NIGMS)
Mentored Patient-Oriented Research Career Development Award (K23)
Project #
Application #
Study Section
Biomedical Computing and Health Informatics Study Section (BCHI)
Program Officer
Dunsmore, Sarah
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Kaiser Foundation Research Institute
United States
Zip Code
Linnen, Daniel T; Kipnis, Patricia; Rondinelli, June et al. (2018) Risk Adjustment for Hospital Characteristics Reduces Unexplained Hospital Variation in Pressure Injury Risk. Nurs Res 67:314-323
Rondinelli, June; Zuniga, Stephen; Kipnis, Patricia et al. (2018) Hospital-Acquired Pressure Injury: Risk-Adjusted Comparisons in an Integrated Healthcare Delivery System. Nurs Res 67:16-25
Patel, Pankaj B; Vinson, David R; Gardner, Marla N et al. (2018) Impact of emergency physician-provided patient education about alternative care venues. Am J Manag Care 24:225-231
Liu, Vincent X (2018) Toward the ""Plateau of Productivity"": Enhancing the Value of Machine Learning in Critical Care. Crit Care Med 46:1196-1197
Mayhew, Michael B; Petersen, Brenden K; Sales, Ana Paula et al. (2018) Flexible, cluster-based analysis of the electronic medical record of sepsis with composite mixture models. J Biomed Inform 78:33-42
Schuler, Alejandro; Wulf, David A; Lu, Yun et al. (2018) The Impact of Acute Organ Dysfunction on Long-Term Survival in Sepsis. Crit Care Med 46:843-849
Walkey, Allan J; Shieh, Meng-Shiou; Liu, Vincent X et al. (2018) Mortality Measures to Profile Hospital Performance for Patients With Septic Shock. Crit Care Med 46:1247-1254
Liu, Vincent X; Escobar, Gabriel J; Chaudhary, Rakesh et al. (2018) Healthcare Utilization and Infection in the Week Prior to Sepsis Hospitalization. Crit Care Med 46:513-516
Escobar, Gabriel J; Baker, Jennifer M; Turk, Benjamin J et al. (2017) Comparing Hospital Processes and Outcomes in California Medicare Beneficiaries: Simulation Prompts Reconsideration. Perm J 21:
Liu, Vincent X; Rosas, Efren; Hwang, Judith et al. (2017) Enhanced Recovery After Surgery Program Implementation in 2 Surgical Populations in an Integrated Health Care Delivery System. JAMA Surg 152:e171032

Showing the most recent 10 out of 41 publications