About

Meeta Prasad Kerlin, MD, MSCE, is an Associate Professor of Medicine in the Department of Medicine (Pulmonary, Allergy and Critical Care) at the Hospital of the University of Pennsylvania. She also serves as an Associate Scholar at the Center for Clinical Epidemiology and Biostatistics and is the Vice Chair for Faculty Development in the Department of Medicine at the Perelman School of Medicine at the University of Pennsylvania. Dr. Kerlin's clinical interests are primarily in critical care, including the management of respiratory failure and sepsis, and she serves as an attending physician in the Medical Intensive Care Unit and on the Critical Care Outreach and Procedure Service of the Hospital of the University of Pennsylvania. Her research focuses on the intersection between medical education, organizational structure, and patient outcomes in intensive care units, with active projects evaluating physician staffing models and the role of physician factors in patient outcomes among patients with acute respiratory failure.

Research topics

• Medicine
• Emergency medicine
• Intensive care medicine
• Nursing
• Medical emergency

Selected publications

• Associations with Asthma Trigger Avoidance Behaviors and Related Advice-Giving Among Adults with Asthma

  Journal of Allergy and Clinical Immunology · 2026-02-01

  article
  PublisherDOI

• Nudging implementation of low tidal volume ventilation: a stepped wedge, cluster randomized trial

  Implementation Science · 2026-05-07

  articleOpen access1st authorCorresponding

  BACKGROUND: "Nudges" embedded in the electronic health record (EHR) facilitate desired decisions while preserving autonomy and may provide a scalable strategy to overcome the common implementation barrier of lack of knowledge about a best practice. We sought to test whether EHR-based nudges targeting two intensive care unit (ICU) clinician groups would safely increase evidence-based use of low tidal volume ventilation. METHODS: We performed a stepped-wedge, cluster randomized, hybrid type 3 effectiveness-implementation trial in 12 ICUs from February 2021 to May 2023 to test three nudges targeting clinicians responsible for order entry and respiratory therapists responsible for operationalizing orders and documentation. A default ventilation order auto-populated a low tidal volume setting; an accountable justification order required a free-text justification to order high tidal volume; and an accountable justification flowsheet required a free-text justification to document delivery of high tidal volume. ICUs were randomly assigned to launch one of the two order nudges on a pre-specified date, followed by the flowsheet nudge six months thereafter. The primary outcome was fidelity to low tidal volume ventilation, defined as percentage of time during the first 72 h of ventilation with low tidal volumes. For additional contextual inquiry, we conducted qualitative interviews with ICU clinicians regarding their perspectives on low tidal volume ventilation and study nudges. RESULTS: The primary analysis included 4412 patients. Unadjusted median fidelity to low tidal volume ventilation was 45.7%. Using multivariable mixed effects regression, marginal estimates of fidelity to low tidal volume ventilation ranged from 47.1% to 57.8% across study groups, with no significant differences after Holm adjustment for multiple comparisons. ICUs experienced variable changes with nudges in fidelity to low tidal volume ventilation. Clinician interviews revealed potential explanations for this variability, including the possibility of differential effects by experience level of clinicians and culture of interprofessional collaboration, and influence of the COVID-19 pandemic on familiarity with and use of low tidal volume ventilation. CONCLUSIONS: EHR-based default and accountable justification nudges did not increase utilization of low tidal volume ventilation in a broad population of mechanically ventilated patients; however, nudge effectiveness varied by ICU. TRIAL REGISTRATION: Clinicaltrials.gov, NCT04663802. Registered 10 December 2020, https://clinicaltrials.gov/study/NCT04663802.

  PublisherDOI

• The authors reply:

  Critical Care Medicine · 2025-09-01

  article
  PublisherDOI

• Quantifying Practice Variability to Inform the Design of Implementation Programs in Critical Care and Assess Their Impact

  CHEST Journal · 2025-09-10

  reviewOpen accessSenior author
  PublisherOA PDFDOI

• Evaluating AI-based comprehensive clinical decision support for sepsis and ARDS: protocol for a Clinician Turing Test

  BMJ Open · 2025-12-01

  articleOpen access

  INTRODUCTION: Few artificial intelligence (AI) clinical decision support systems (CDSSs) are ever evaluated in practice. Although some signal of clinical effectiveness may be needed to justify AI deployment and testing, such data are typically unavailable in early-stage research. This conundrum is especially relevant in the intensive care unit (ICU), where conditions like sepsis and acute respiratory distress syndrome (ARDS) require high-stakes decisions. Our group developed the AI ventilator assistant (AVA), a novel AI CDSS for patients with sepsis ARDS receiving invasive mechanical ventilation. But the promising results of predictive performance estimates are not sufficient to assess AVA's clinical safety and appropriateness prior to future evaluation and deployment. Therefore, we propose a Clinician Turing Test as a novel validation approach to determine whether clinicians can distinguish AVA-generated treatment recommendations from those enacted by real human clinicians. If AVA's recommendations are consistently indistinguishable from those of real clinicians, thereby 'passing' this Turing test, this would provide a strong preclinical signal of safety and appropriateness. METHODS AND ANALYSIS: This multisite, randomised, electronic, vignette-based Phase 1b study will use a Clinician Turing Test design. We aim to recruit 350 critical care clinicians, including physicians and advanced practice providers from six US hospitals. Participants will review nine clinical vignettes of patients with sepsis and ARDS derived from the Molecular Epidemiology of Severe Sepsis in the ICU cohort and an associated profile of a suggested treatment plan. For each participant-vignette combination, the source of the treatment profile will be randomly assigned (AI-generated by AVA vs the actually enacted treatment from real human clinicians) in a 1:1 allocation. The primary endpoint is the participants' accuracy in identifying whether a treatment profile was AI-generated or human-generated, assessed using equivalence testing through a mixed-effects logistic regression model with random effects for participants and vignettes. Secondarily, a fitted binary classifier will assess discrimination ability using the C-statistic. Secondary endpoints include clinicians' perceptions of the safety and appropriateness of the treatment profiles, confidence in distinguishing AI-generated and human-generated recommendations, interest in AI CDSSs for sepsis and ventilator management and the time to complete the survey. This novel Phase 1b design provides preliminary but essential information about an AI CDSS's clinical appropriateness without the risk or cost of actual deployment, thereby informing decisions about future clinical implementation and evaluation in real clinical environments. ETHICS AND DISSEMINATION: This protocol was approved by the Institutional Review Board of the University of Pennsylvania (Protocol #858201). Results are expected in 2026 and will be submitted for publication in peer-reviewed journals and presented at scientific conferences. TRIAL REGISTRATION NUMBER: NCT07025096.

  PublisherOA PDFDOI

• The Legacy of COVID-19 on Critical Care—Is There an Upside?

  JAMA Network Open · 2025-08-25

  articleOpen access1st authorCorresponding

  The world changed for critical care practitioners in March 2020.For many, it began with a warning from colleagues in Italy.A social media message, which went viral, warned, "The current situation is difficult to imagineOur hospitals are overwhelmed by COVID-19," and it cautioned, "Don't make the mistake to think that what is happening is happening in a 3rd world country." 1 It concluded, "My friends call me in tears because they see people dying in front of them and can only offer some oxygen.PLEASE STOP, READ THIS AGAIN AND THINK."Similar stories of fear and uncertainty, of hopelessness and exhaustion, emerged from subsequent epicenters.At the same time, the situation brought forth a sense of urgency to act, prepare, and, once the cases arrived, do our best to care for an unprecedented deluge of patients experiencing, and suffering from, severe respiratory failure.So many (too many) patients died, especially in those initial waves.For our international critical care community who lived through this time, what is the legacy of the pandemic?The story, as is often the case, is an evolving one.The first wave enacted a significant toll on the mental well-being of critical care health care professionals.In a cross-sectional survey of European intensivists, 46%, 30%, and 51% of physicians experienced symptoms of anxiety, depression, and severe burnout, respectively. 2 The impact of the pandemic worsened with

  PublisherOA PDFDOI

• Authors’ response

  Annals of Allergy Asthma & Immunology · 2025-06-01

  letter
  PublisherDOI

• Comparing Outcomes Amongst Mechanically Ventilated Patients Who Are Transferred vs Retained in a Large Health System

  CHEST Critical Care · 2025-07-30 · 1 citations

  articleOpen access

  <h3>Background</h3> Mechanically ventilated (MV) patients undergoing a transfer to a higher-resourced hospital have variable outcomes. We aimed to evaluate patient outcomes by transfer status and timing. <h3>Research Question</h3> What are the characteristics and outcomes of mechanically ventilated patients who are transferred from 1 ICU to another? Does timing of transfer affect these outcomes? <h3>Study Design and Methods</h3> In a retrospective observational study, we identified patients admitted on a ventilator to a Midwestern health system (7 local hospitals, 1 tertiary hospital) from March 2018 to December 2021. Exposures were transfer status (being transferred to a tertiary hospital or retained at a local hospital) and transfer timing (early transfers if transferred ≤ 2 days after admission or late transfers). Propensity score weighting was used to balance patient characteristics (age, sex, race/ethnicity, insurance, comorbidity, and primary diagnosis) and baseline clinical factors (admission Sequential Organ Failure Assessment score, COVID-19 test, days on mechanical ventilation, major surgery, and ICU use of dialysis, tracheostomy, and vasopressors). Associations with exposures were estimated through multinomial logistic regression for discharge destination or negative binomial regression for total length of stay (LOS). <h3>Results</h3> Of a total of 5,883 adult MV patients, 5,719 were retained and 164 were transfers (80 early transfers and 84 later transfers). Transfers were associated with longer LOS (33.4 days vs 10.2 days; incident rate ratio [IRR] = 3.11; 95% CI, 2.53-3.82; <i>P</i> < .001) and had an increased likelihood of discharge to other facilities (rehabilitation and nursing facilities) (OR = 2.62; 95% CI, 1.51-4.56; <i>P</i> < .01) than retained patients at the local hospitals. Additionally, early transfer was found to be associated with a shorter LOS than late transfers (19.6 days vs 47.7 days; IRR=0.41; 95% CI, 0.33-0.51; <i>P</i> < .01). <h3>Interpretation</h3> Our results show that MV patient transfers within our health system have lower likelihood of discharge to home and longer LOS when compared with MV patients retained at local hospitals. However, the time to transfer may be an important contributor to better outcomes.

  PublisherOA PDFDOI

• Racial Differences in Shared Decision-Making About Critical Illness

  UNC Libraries · 2025-04-11

  articleOpen accessSenior author

  Importance: Shared decision-making is the preferred method for evaluating complex tradeoffs in the care of patients with critical illness. However, it remains unknown whether critical care clinicians engage diverse patients and caregivers equitably in shared decision-making. Objective: To compare critical care clinicians' approaches to shared decision-making in recorded conversations with Black and White caregivers of patients with critical illness. Design, Setting, and Participants: This thematic analysis consisted of unstructured clinician-caregiver meetings audio-recorded during a randomized clinical trial of a decision aid about prolonged mechanical ventilation at 13 intensive care units in the US. Participants in meetings included critical care clinicians and Black or White caregivers of patients who underwent mechanical ventilation. The codebook included components of shared decision-making and known mechanisms of racial disparities in clinical communication. Analysts were blinded to caregiver race during coding. Patterns within and across racial groups were evaluated to identify themes. Data analysis was conducted between August 2021 and April 2023. Main Outcomes and Measures: The main outcomes were themes describing clinician behaviors varying by self-reported race of the caregivers. Results: The overall sample comprised 20 Black and 19 White caregivers for a total of 39 audio-recorded meetings with clinicians. The duration of meetings was similar for both Black and White caregivers (mean [SD], 23.9 [13.7] minutes vs 22.1 [11.2] minutes, respectively). Both Black and White caregivers were generally middle-aged (mean [SD] age, 47.6 [9.9] years vs 51.9 [8.8] years, respectively), female (15 [75.0%] vs 14 [73.7%], respectively), and possessed a high level of self-assessed health literacy, which was scored from 3 to 15 with lower scores indicating increasing health literacy (mean [SD], 5.8 [2.3] vs 5.3 [2.0], respectively). Clinicians conducting meetings with Black and White caregivers were generally young (mean [SD] age, 38.8 [6.6] years vs 37.9 [8.2] years, respectively), male (13 [72.2%] vs 12 [70.6%], respectively), and White (14 [77.8%] vs 17 [100%], respectively). Four variations in clinicians' shared decision-making behaviors by caregiver race were identified: (1) providing limited emotional support for Black caregivers, (2) failing to acknowledge trust and gratitude expressed by Black caregivers, (3) sharing limited medical information with Black caregivers, and (4) challenging Black caregivers' preferences for restorative care. These themes encompass both relational and informational aspects of shared decision-making. Conclusions and Relevance: The results of this thematic analysis showed that critical care clinicians missed opportunities to acknowledge emotions and value the knowledge of Black caregivers compared with White caregivers. These findings may inform future clinician-level interventions aimed at promoting equitable shared decision-making.

  PublisherDOI

• Development and External Validation of a Detection Model to Retrospectively Identify Patients With Acute Respiratory Distress Syndrome

  Critical Care Medicine · 2025-04-08 · 4 citations

  articleOpen access

  OBJECTIVE: The aim of this study was to develop and externally validate a machine-learning model that retrospectively identifies patients with acute respiratory distress syndrome (acute respiratory distress syndrome [ARDS]) using electronic health record (EHR) data. DESIGN: In this retrospective cohort study, ARDS was identified via physician-adjudication in three cohorts of patients with hypoxemic respiratory failure (training, internal validation, and external validation). Machine-learning models were trained to classify ARDS using vital signs, respiratory support, laboratory data, medications, chest radiology reports, and clinical notes. The best-performing models were assessed and internally and externally validated using the area under receiver-operating curve (AUROC), area under precision-recall curve, integrated calibration index (ICI), sensitivity, specificity, positive predictive value (PPV), and ARDS timing. PATIENTS: Patients with hypoxemic respiratory failure undergoing mechanical ventilation within two distinct health systems. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: There were 1,845 patients in the training cohort, 556 in the internal validation cohort, and 199 in the external validation cohort. ARDS prevalence was 19%, 17%, and 31%, respectively. Regularized logistic regression models analyzing structured data (EHR model) and structured data and radiology reports (EHR-radiology model) had the best performance. During internal and external validation, the EHR-radiology model had AUROC of 0.91 (95% CI, 0.88-0.93) and 0.88 (95% CI, 0.87-0.93), respectively. Externally, the ICI was 0.13 (95% CI, 0.08-0.18). At a specified model threshold, sensitivity and specificity were 80% (95% CI, 75%-98%), PPV was 64% (95% CI, 58%-71%), and the model identified patients with a median of 2.2 hours (interquartile range 0.2-18.6) after meeting Berlin ARDS criteria. CONCLUSIONS: Machine-learning models analyzing EHR data can retrospectively identify patients with ARDS across different institutions.

  PublisherOA PDFDOI

Recent grants

• The role of physician experience in outcomes of patients with acute respiratory f

  NIH · $741k · 2013–2018

Frequent coauthors

• Scott D. Halpern

  University of Pennsylvania

  93 shared

• Rachel Kohn

  University of Pennsylvania

  36 shared

• Gary E. Weissman

  California University of Pennsylvania

  28 shared

• Kelly C. Vranas

  Oregon Health & Science University

  28 shared

• C. Jessica Dine

  University of Pennsylvania

  25 shared

• Mark E. Mikkelsen

  Weill Cornell Medicine

  24 shared

• Michael O. Harhay

  19 shared

• Stephanie Maillie

  Hospital of the University of Pennsylvania

  18 shared