Grace Lee, MD

BACKGROUND: Using the data collected through incident reporting systems is challenging, as it is a large volume of primarily qualitative information. Large language models (LLMs), such as ChatGPT, provide novel capabilities in text summarization and labeling that could support safety data trending and early identification of opportunities to prevent patient harm. This study assessed the capability of a proprietary LLM (GPT-3.5) to automatically label a cross-sectional sample of real-world obstetric incident reports.METHODS: A sample of 370 incident reports submitted to inpatient obstetric units between December 2022 and May 2023 was extracted. Human-annotated labels were assigned by a clinician reviewer and considered gold standard. The LLM was prompted to label incident reports relying solely on its pretrained knowledge and information included in the prompt. Primary outcomes assessed were sensitivity, specificity, positive predictive value, and negative predictive value. A secondary outcome assessed the human-perceived quality of the model's justification for the label(s) applied.RESULTS: The LLM demonstrated the ability to label incident reports with high sensitivity and specificity. The model applied a total of 79 labels compared to the reviewer's 49 labels. Overall sensitivity for the model was 85.7%, and specificity was 97.9%. Positive and negative predictive values were 53.2% and 99.6%, respectively. For 60.8% of labels, the reviewer approved of the model's justification for applying the label.CONCLUSION: The proprietary LLM demonstrated the ability to label obstetric incident reports with high sensitivity and specificity. LLMs offer the potential to enable more efficient use of data from incident reporting systems.

View details for DOI 10.1016/j.jcjq.2024.08.001

View details for PubMedID 39256071

OBJECTIVES: This study seeks to identify demographic and clinical factors prompting clinician prescribing of nirmatrelvir/ritonavir to pediatric patients for management of coronavirus disease 2019 (COVID-19) infection.METHODS: Patients aged 12 to 17 years with a COVID-19 infection and nirmatrelvir/ritonavir prescription during an outpatient clinical encounter within a PEDSnet-affiliated institution between January 2022 and August 2023 were identified using electronic health record data. A multivariate logistic regression analysis was used to estimate odds of nirmatrelvir/ritonavir prescription after adjusting for various factors.RESULTS: A total of 20959 patients aged 12 to 17 years were diagnosed with a COVID-19 infection on the basis of an electronic health record-documented positive polymerase chain reaction or antigen test or diagnosis during an outpatient clinical visit. Of these patients, 408 received a nirmatrelvir/ritonavir prescription within 5 days of diagnosis. Higher odds of nirmatrelvir/ritonavir treatment were associated with having chronic or complex chronic disease (chronic: odds ratio [OR] 2.50 [95% confidence interval (CI) 1.83-3.38]; complex chronic: OR 2.21 [95% CI 1.58-3.08]). Among patients with chronic disease, each additional body system conferred 1.18 times higher odds of treatment (95% CI 1.10-1.26). Compared with non-Hispanic white patients, Hispanic patients (OR 0.61 [95% CI 0.44-0.83]) had lower odds of treatment.CONCLUSIONS: Children with chronic conditions are more likely than those without to receive nirmatrelvir/ritonavir prescriptions. However, nirmatrelvir/ritonavir prescribing to children with chronic conditions remains infrequent. Pediatric data concerning nirmatrelvir/ritonavir safety and effectiveness in preventing severe disease and hospitalization are critical optimizing clinical decision-making and use among children.

View details for DOI 10.1542/hpeds.2023-007132

View details for PubMedID 39076115

Pediatricians and primary care providers serve an important role in building trust with families and communities. To support the critical role of front-line providers, this perspective seeks to reflect on the work of the Centers for Disease Control and Prevention's (CDC) Advisory Committee on Immunization Practices to support COVID-19 pandemic response efforts. Although ACIP recommends vaccines for all age groups, this perspective focuses on the pediatric lens and is tailored to Academic Pediatrics. ACIP adapted from in-person meetings 3 times yearly to virtual meetings on an emergency basis to ensure a thorough review and presentation of all the components of Evidence to Recommendation framework, including explicit consideration of equity in the decision-making process. The need for diverse enrollment in clinical trials was highlighted as critical for supporting recommendations and enhancing trust. Near real-time vaccine safety surveillance was implemented at scale and emphasized the importance of collaboration between federal partners engaged in vaccine safety in the U.S. and extended to other countries with similar safety surveillance systems to enable early recognition and response to safety concerns. A key equity opportunity for future pandemics is to shorten the time between vaccine was available for adults and young children.

View details for DOI 10.1016/j.acap.2024.06.019

View details for PubMedID 38972350

During the COVID-19 pandemic, candidate COVID-19 vaccines were being developed for potential use in the United States on an unprecedented, accelerated schedule. It was anticipated that once available, under U.S. Food and Drug Administration (FDA) Emergency Use Authorization (EUA) or FDA approval, COVID-19 vaccines would be broadly used and potentially administered to millions of individuals in a short period of time. Intensive monitoring in the post-EUA/licensure period would be necessary for timely detection and assessment of potential safety concerns. To address this, the Centers for Disease Control and Prevention (CDC) convened an Advisory Committee on Immunization Practices (ACIP) work group focused solely on COVID-19 vaccine safety, consisting of independent vaccine safety experts and representatives from federal agencies - the ACIP COVID-19 Vaccine Safety Technical Work Group (VaST). This report provides an overview of the organization and activities of VaST, summarizes data reviewed as part of the comprehensive effort to monitor vaccine safety during the COVID-19 pandemic, and highlights selected actions taken by CDC, ACIP, and FDA in response to accumulating post-authorization safety data. VaST convened regular meetings over the course of 29 months, from November 2020 through April 2023; through March 2023 FDA issued EUAs for six COVID-19 vaccines from four different manufacturers and subsequently licensed two of these COVID-19 vaccines. The independent vaccine safety experts collaborated with federal agencies to ensure timely assessment of vaccine safety data during this time. VaST worked closely with the ACIP COVID-19 Vaccines Work Group; that work group used safety data and VaST's assessments for benefit-risk assessments and guidance for COVID-19 vaccination policy. Safety topics reviewed by VaST included those identified in safety monitoring systems and other topics of scientific or public interest. VaST provided guidance to CDC's COVID-19 vaccine safety monitoring efforts, provided a forum for review of data from several U.S. government vaccine safety systems, and assured that a diverse group of scientists and clinicians, external to the federal government, promptly reviewed vaccine safety data. In the event of a future pandemic or other biological public health emergency, the VaST model could be used to strengthen vaccine safety monitoring, enhance public confidence, and increase transparency through incorporation of independent, non-government safety experts into the monitoring process, and through strong collaboration among federal and other partners.

View details for DOI 10.1016/j.vaccine.2023.12.059

View details for PubMedID 38341293

Multisystem inflammatory syndrome in children (MIS-C) is a severe post-acute sequela of SARS-CoV-2 infection. The highly diverse clinical features of MIS-C necessities characterizing its features by subphenotypes for improved recognition and treatment. However, jointly identifying subphenotypes in multi-site settings can be challenging. We propose a distributed multi-site latent class analysis (dMLCA) approach to jointly learn MIS-C subphenotypes using data across multiple institutions.We used data from the electronic health records (EHR) systems across nine U.S. children's hospitals. Among the 3,549,894 patients, we extracted 864 patients < 21 years of age who had received a diagnosis of MIS-C during an inpatient stay or up to one day before admission. Using MIS-C conditions, laboratory results, and procedure information as input features for the patients, we applied our dMLCA algorithm and identified three MIS-C subphenotypes. As validation, we characterized and compared more granular features across subphenotypes. To evaluate the specificity of the identified subphenotypes, we further compared them with the general subphenotypes identified in the COVID-19 infected patients.Subphenotype 1 (46.1%) represents patients with a mild manifestation of MIS-C not requiring intensive care, with minimal cardiac involvement. Subphenotype 2 (25.3%) is associated with a high risk of shock, cardiac and renal involvement, and an intermediate risk of respiratory symptoms. Subphenotype 3 (28.6%) represents patients requiring intensive care, with a high risk of shock and cardiac involvement, accompanied by a high risk of >4 organ system being impacted. Importantly, for hospital-specific clinical decision-making, our algorithm also revealed a substantial heterogeneity in relative proportions of these three subtypes across hospitals. Properly accounting for such heterogeneity can lead to accurate characterization of the subphenotypes at the patient-level.Our identified three MIS-C subphenotypes have profound implications for personalized treatment strategies, potentially influencing clinical outcomes. Further, the proposed algorithm facilitates federated subphenotyping while accounting for the heterogeneity across hospitals.

View details for DOI 10.1101/2024.01.26.24301827

View details for PubMedID 38343837

View details for PubMedCentralID PMC10854314

COVID-19 vaccines represent a great scientific and public health achievement in the face of overwhelming pressures from a global pandemic, preventing millions of hospitalizations and deaths due to COVID-19 vaccines in the United States. Over 675 million doses of COVID-19 vaccines have been administered in the United States, and over 80% of the U.S. population has had at least 1 dose of a COVID-19 vaccine. Over the course of the COVID-19 pandemic in the United States, over one million people died from COVID-19, and over six million were hospitalized. It has been estimated that COVID-19 vaccines prevented more than 18 million additional hospitalizations and more than 3 million additional deaths due to COVID-19 in the United States. From the beginning of the COVID-19 pandemic in 2020 through June 2023, ACIP had 35 COVID-19 focused meetings and 24 votes for COVID-19 vaccine recommendations. ACIP had the critical task of rapidly and thoroughly reviewing emerging and evolving data on COVID-19 epidemiology and vaccines, as well as making comprehensive population-based recommendations for vaccine policy and considerations for implementation through a transparent and evidence-based framework. Safe and effective COVID-19 vaccines, recommended through transparent policy discussions with ACIP, remain the best tool we have to prevent serious illness, hospitalization and death from COVID-19.

View details for DOI 10.1016/j.vaccine.2023.12.022

View details for PubMedID 38158297

Importance: Pediatric ventilator-associated events (PedVAEs, defined as a sustained worsening in oxygenation after a baseline period of stability or improvement) are useful for surveillance of complications from mechanical ventilation. It is unclear whether interventions to mitigate known risk factors can reduce PedVAE rates.Objective: To assess whether adherence to 1 or more test factors in a quality improvement bundle was associated with a reduction in PedVAE rates.Design, Setting, and Participants: This multicenter quality improvement study obtained data from 2017 to 2020 for patients who were mechanically ventilated and cared for in neonatal, pediatric, and cardiac intensive care units (ICUs). These ICUs were located in 95 hospitals participating in the Children's Hospitals' Solutions for Patient Safety (SPS) network in North America. Data analyses were performed between September 2021 and April 2023.Intervention: A quality improvement bundle consisted of 3 test factors: multidisciplinary apparent cause analysis, daily discussion of extubation readiness, and daily discussion of fluid balance goals. This bundle was distributed to a subgroup of hospitals that volunteered to participate in a collaborative PedVAE prevention initiative under the SPS network guidance in July 2018.Main Outcomes and Measures: Each SPS network hospital submitted monthly PedVAE rates from January 1, 2017, to May 31, 2020, and test factor data were submitted from July 1, 2018, to May 31, 2020. Analyses focused on hospitals that reliably submitted PedVAE rate data, defined as outcomes data submission through May 31, 2020, for at least 80% of the baseline and postbaseline periods.Results: Of the 95 hospitals in the SPS network that reported PedVAE data, 21 were grouped in the Pioneer cohort and 74 in the non-Pioneer cohort. Only 12 hospitals (57%) from the 21 Pioneer hospitals and 33 (45%) from the 74 non-Pioneer hospitals were considered to be reliable reporters of outcome data. Among the 12 hospitals, the PedVAE rate decreased from 1.9 to 1.4 events per 1000 ventilator days (absolute rate difference, -0.6; 95% CI, -0.5 to -0.7; P<.001). No significant change in the PedVAE rate was seen among the 33 hospitals that reliably submitted PedVAE rates but did not implement the bundle. Of the 12 hospitals, 3 that reliably performed daily discussion of extubation readiness had a decrease in PedVAE rate from 2.6 to 1.2 events per 1000 ventilator days (absolute rate difference, -1.4; 95% CI, -1.0 to -1.7; P<.001), whereas the other 9 hospitals that did not implement this discussion did not have a decrease.Conclusions and Relevance: This study found that a multicenter quality improvement intervention targeting PedVAE risk factors was associated with a substantial reduction in the rate of PedVAEs in hospital ICUs. The findings suggest that ICU teams seeking to reduce PedVAEs incorporate daily discussion of extubation readiness during morning rounds.

View details for DOI 10.1001/jamanetworkopen.2023.46545

View details for PubMedID 38060226

Multi-system inflammatory syndrome in children (MIS-C) is a severe post-acute sequela of SARS-CoV-2 infection in children, and there is a critical need to unfold its highly heterogeneous disease patterns. Our objective was to characterize the illness spectrum of MIS-C for improved recognition and management. We conducted a retrospective cohort study using data from March 1, 2020-September 30, 2022, in 8 pediatric medical centers from PEDSnet. We included 1139 children hospitalized with MIS-C and used their demographics, symptoms, conditions, laboratory values, and medications for analyses. We applied heterogeneity-adaptive latent class analyses and identified three latent classes. We further characterized the sociodemographic and clinical characteristics of the latent classes and evaluated their temporal patterns. Class 1 (47.9%) represented children with the most severe presentation, with more admission to the ICU, higher inflammatory markers, hypotension/shock/dehydration, cardiac involvement, acute kidney injury and respiratory involvement. Class 2 (23.3%) represented a moderate presentation, with 4-6 organ systems involved, and some overlapping features with acute COVID-19. Class 3 (28.8%) represented a mild presentation. Our results indicated that MIS-C has a spectrum of clinical severity ranging from mild to severe and the proportion of severe or critical MIS-C decreased over time.

View details for DOI 10.1038/s41598-023-47655-y

View details for PubMedID 38017007

View details for PubMedCentralID 7220177

COVID-19 vaccines protect against severe COVID-19-associated outcomes, including hospitalization and death. As SARS-CoV-2 has evolved, and waning vaccine effectiveness has been noted, vaccine formulations and policies have been updated to provide continued protection against severe illness and death from COVID-19. Since September 2022, bivalent mRNA COVID-19 vaccines have been recommended in the United States, but the variants these vaccines protect against are no longer circulating widely. On September 11, 2023, the Food and Drug Administration (FDA) approved the updated (2023-2024 Formula) COVID-19 mRNA vaccines by Moderna and Pfizer-BioNTech for persons aged ≥12 years and authorized these vaccines for persons aged 6 months-11 years under Emergency Use Authorization (EUA). On October 3, 2023, FDA authorized the updated COVID-19 vaccine by Novavax for use in persons aged ≥12 years under EUA. The updated COVID-19 vaccines include a monovalent XBB.1.5 component, which is meant to broaden vaccine-induced immunity and provide protection against currently circulating SARS-CoV-2 XBB-sublineage variants including against severe COVID-19-associated illness and death. On September 12, 2023, the Advisory Committee on Immunization Practices recommended vaccination with updated COVID-19 vaccines for all persons aged ≥6 months. These recommendations will be reviewed as new evidence becomes available or new vaccines are approved and might be updated.

View details for Web of Science ID 001100985200001

View details for PubMedID 37856366

View details for PubMedCentralID PMC10602621

View details for DOI 10.1155/2023/8798997

View details for Web of Science ID 001051843300001