About

Laura Murphy received her BSN and MSN in maternity nursing from The University of Texas School of Nursing and her PhD in nursing science from Texas Woman's University. She has been teaching perinatal nursing at UT Austin for 11 years. Laura has experience working in labor and delivery, postpartum, and newborn nursery. She has also taught practicum courses at various hospitals in the area.

Research topics

• Medicine
• Internal medicine
• Natural Language Processing
• Computer Science
• Surgery
• Medical physics
• Anesthesia
• World Wide Web
• Oncology
• Pathology
• Psychology
• Medical emergency
• Nuclear medicine
• Environmental health

Selected publications

• De novo colorectal cancer after kidney transplantation: a systematic review and meta-analysis

  British Journal of Cancer · 2025-04-05 · 2 citations

  reviewOpen access
  PublisherOA PDFDOI

• Postoperative Activity Restrictions After Reconstructive Pelvic Surgery

  Obstetrical & Gynecological Survey · 2025-06-01

  article

  (Abstracted from Urogynecology (Phila) 2025;31(3):266–275 Physical activity restrictions after reconstructive pelvic surgery are not standardized. Traditional recommendations for abstaining from physical exertion seem to be founded on the belief that limiting physical stress can positively influence anatomical and functional recovery, although there is little evidence this is true.

  PublisherDOI

• Clinical Presentation, Imaging Features, and Management of Morel-Lavallee Lesions: A Scoping Review

  The American Surgeon · 2025-10-23 · 2 citations

  articleOpen access

  BACKGROUND: Morel-Lavallee lesions (MLLs) are rare closed soft tissue degloving injuries caused by shearing forces that separate skin and subcutaneous tissue from underlying fascia. These injuries may be underrecognized in polytrauma patients, increasing the risk of delayed diagnosis, infection, prolonged healing, and long-term morbidity. This scoping review aims to provide a comprehensive summary of clinical presentation, imaging features, and evolving management for MLLs, with the goal of identifying gaps in the literature and guiding clinical decision making. METHODS: We systematically searched PubMed, Scopus, Web of Science, Cochrane CENTRAL, and ClinicalTrials.gov between 1992 and 2025 for studies pertaining to adult patients with MLLs. Four independent reviewers screened full texts for inclusion. Data were extracted and summarized using narrative synthesis. Of 633 unique articles, 102 underwent full-text review, and 69 studies met inclusion/exclusion criteria. RESULTS: MLLs commonly present with fluctuance, ecchymosis, swelling, and skin hypermobility. Diagnosis is clinical, with imaging playing a supportive role. Computed tomography and ultrasound can aid diagnosis in acute trauma settings, while magnetic resonance imaging is best for evaluating composition, chronicity, and extent. Management strategies vary based on size, chronicity, and anatomic locations. Acute lesions may be managed with compression or aspiration. Chronic or recurrent MLLs often require sclerodesis, minimally invasive debridement, or open surgical excision, often with subsequent soft tissue reconstruction. CONCLUSIONS: MLLs remain a diagnostic and therapeutic challenge in trauma. Given the heterogeneity of existing evidence, high-quality prospective studies are needed to establish evidence-based optimal management guidelines and improve long-term outcomes.

  PublisherDOI

• Accuracy of Prospective Assessments of 4 Large Language Model Chatbot Responses to Patient Questions About Emergency Care: Experimental Comparative Study

  Journal of Medical Internet Research · 2024 · 48 citations

  • Computer Science
  • Natural Language Processing
  • Psychology

  BACKGROUND: Recent surveys indicate that 48% of consumers actively use generative artificial intelligence (AI) for health-related inquiries. Despite widespread adoption and the potential to improve health care access, scant research examines the performance of AI chatbot responses regarding emergency care advice. OBJECTIVE: We assessed the quality of AI chatbot responses to common emergency care questions. We sought to determine qualitative differences in responses from 4 free-access AI chatbots, for 10 different serious and benign emergency conditions. METHODS: We created 10 emergency care questions that we fed into the free-access versions of ChatGPT 3.5 (OpenAI), Google Bard, Bing AI Chat (Microsoft), and Claude AI (Anthropic) on November 26, 2023. Each response was graded by 5 board-certified emergency medicine (EM) faculty for 8 domains of percentage accuracy, presence of dangerous information, factual accuracy, clarity, completeness, understandability, source reliability, and source relevancy. We determined the correct, complete response to the 10 questions from reputable and scholarly emergency medical references. These were compiled by an EM resident physician. For the readability of the chatbot responses, we used the Flesch-Kincaid Grade Level of each response from readability statistics embedded in Microsoft Word. Differences between chatbots were determined by the chi-square test. RESULTS: Each of the 4 chatbots' responses to the 10 clinical questions were scored across 8 domains by 5 EM faculty, for 400 assessments for each chatbot. Together, the 4 chatbots had the best performance in clarity and understandability (both 85%), intermediate performance in accuracy and completeness (both 50%), and poor performance (10%) for source relevance and reliability (mostly unreported). Chatbots contained dangerous information in 5% to 35% of responses, with no statistical difference between chatbots on this metric (P=.24). ChatGPT, Google Bard, and Claud AI had similar performances across 6 out of 8 domains. Only Bing AI performed better with more identified or relevant sources (40%; the others had 0%-10%). Flesch-Kincaid Reading level was 7.7-8.9 grade for all chatbots, except ChatGPT at 10.8, which were all too advanced for average emergency patients. Responses included both dangerous (eg, starting cardiopulmonary resuscitation with no pulse check) and generally inappropriate advice (eg, loosening the collar to improve breathing without evidence of airway compromise). CONCLUSIONS: AI chatbots, though ubiquitous, have significant deficiencies in EM patient advice, despite relatively consistent performance. Information for when to seek urgent or emergent care is frequently incomplete and inaccurate, and patients may be unaware of misinformation. Sources are not generally provided. Patients who use AI to guide health care decisions assume potential risks. AI chatbots for health should be subject to further research, refinement, and regulation. We strongly recommend proper medical consultation to prevent potential adverse outcomes.

  PublisherDOI

• Accuracy of Prospective Assessments of 4 Large Language Model Chatbot Responses to Patient Questions About Emergency Care: Experimental Comparative Study (Preprint)

  2024-05-07

  preprintOpen access

  <sec> <title>BACKGROUND</title> Recent surveys indicate that 48% of consumers actively use generative artificial intelligence (AI) for health-related inquiries. Despite widespread adoption and the potential to improve health care access, scant research examines the performance of AI chatbot responses regarding emergency care advice. </sec> <sec> <title>OBJECTIVE</title> We assessed the quality of AI chatbot responses to common emergency care questions. We sought to determine qualitative differences in responses from 4 free-access AI chatbots, for 10 different serious and benign emergency conditions. </sec> <sec> <title>METHODS</title> We created 10 emergency care questions that we fed into the free-access versions of ChatGPT 3.5 (OpenAI), Google Bard, Bing AI Chat (Microsoft), and Claude AI (Anthropic) on November 26, 2023. Each response was graded by 5 board-certified emergency medicine (EM) faculty for 8 domains of percentage accuracy, presence of dangerous information, factual accuracy, clarity, completeness, understandability, source reliability, and source relevancy. We determined the correct, complete response to the 10 questions from reputable and scholarly emergency medical references. These were compiled by an EM resident physician. For the readability of the chatbot responses, we used the Flesch-Kincaid Grade Level of each response from readability statistics embedded in Microsoft Word. Differences between chatbots were determined by the chi-square test. </sec> <sec> <title>RESULTS</title> Each of the 4 chatbots’ responses to the 10 clinical questions were scored across 8 domains by 5 EM faculty, for 400 assessments for each chatbot. Together, the 4 chatbots had the best performance in clarity and understandability (both 85%), intermediate performance in accuracy and completeness (both 50%), and poor performance (10%) for source relevance and reliability (mostly unreported). Chatbots contained dangerous information in 5% to 35% of responses, with no statistical difference between chatbots on this metric (&lt;i&gt;P&lt;/i&gt;=.24). ChatGPT, Google Bard, and Claud AI had similar performances across 6 out of 8 domains. Only Bing AI performed better with more identified or relevant sources (40%; the others had 0%-10%). Flesch-Kincaid Reading level was 7.7-8.9 grade for all chatbots, except ChatGPT at 10.8, which were all too advanced for average emergency patients. Responses included both dangerous (eg, starting cardiopulmonary resuscitation with no pulse check) and generally inappropriate advice (eg, loosening the collar to improve breathing without evidence of airway compromise). </sec> <sec> <title>CONCLUSIONS</title> AI chatbots, though ubiquitous, have significant deficiencies in EM patient advice, despite relatively consistent performance. Information for when to seek urgent or emergent care is frequently incomplete and inaccurate, and patients may be unaware of misinformation. Sources are not generally provided. Patients who use AI to guide health care decisions assume potential risks. AI chatbots for health should be subject to further research, refinement, and regulation. We strongly recommend proper medical consultation to prevent potential adverse outcomes. </sec> <sec> <title>CLINICALTRIAL</title> <p/> </sec>

  PublisherOA PDFDOI

• Postoperative Activity Restrictions After Reconstructive Pelvic Surgery

  Urogynecology · 2024-12-17 · 2 citations

  review

  IMPORTANCE: Restrictive physical activity after pelvic reconstructive surgery is recommended, although the optimal duration and intensity are not standardized. OBJECTIVE: This systematic review and meta-analysis aimed to evaluate the existing literature comparing clinical outcomes for liberal postoperative physical activity versus standard of care, defined as restricted postoperative physical activity, after pelvic reconstructive surgery. STUDY DESIGN: PubMed, CENTRAL, Scopus, Web of Science, and CINAHL databases were searched for observational and randomized studies comparing liberal postoperative physical activity and standard of care in women undergoing pelvic reconstructive surgery, reporting anatomic and functional outcomes. Statistical analysis was performed using RevMan software, presenting results as mean difference (MD) or odds ratio in a random-effects model, with 95% confidence intervals (CIs). RESULTS: Five randomized trials, representing total n = 434, were included, with 2 studies on sling surgical procedures and 3 on prolapse surgical procedures. Data from 3 studies suggest no significant difference between liberal and standard postoperative instructions in surgical outcomes, measured by Point Ba from POP-Q assessment tool, up to 3 months follow-up (MD, -0.04; 95% CI, -0.16 to 0.07; P = 0.49). Disease-specific symptom distress, measured by Pelvic Floor Distress Inventory-20, favored the liberal approach (MD, -10.09; 95% CI, -18.33 to -1.86; P = 0.02). Other domains, including Urinary Distress Inventory-6, also showed significant improvements with liberal postoperative activities (MD, -4.29; 95% CI, -7.84 to -0.74; P = 0.02). CONCLUSIONS: Patients with liberal postoperative physical activity recommendations in prolapse repair surgical procedures had similar short-term anatomic outcomes compared with standard restrictions, with more favorable outcomes in disease-specific symptom relief and quality of life.

  PublisherDOI

• Duloxetine for Postoperative Pain Control Following Knee or Hip Replacement: A Systematic Review and Meta-Analysis

  Arthroplasty Today · 2023 · 7 citations

  • Medicine
  • Anesthesia
  • Surgery

  Background: Duloxetine is a Food and Drug Administration-approved selective norepinephrine reuptake inhibitor for treating depression, anxiety, fibromyalgia, and neuropathic and chronic musculoskeletal pain. This meta-analysis aims to evaluate the efficacy of duloxetine in reducing pain and postoperative opioid use following lower extremity total joint arthroplasty. Methods: A literature search was performed, identifying randomized controlled trials investigating duloxetine for pain management after total hip and total knee arthroplasty. Data from the visual analog scale (VAS) for pain during movement and at rest were extracted for postoperative days (PODs) 1, 3, 7, and 14, as well as postoperative week 6 and postoperative month 3. Opioid use data were obtained at 24, 48 and 72 hours. All data were analyzed using inverse variance with random effects and presented as weighted mean difference. Results: Eight unique studies were identified and included, 7 of which were analyzed quantitatively. Duloxetine decreased postoperative opioid consumption at 48 and 72 hours. For VAS for pain at rest, significantly reduced pain was reported by duloxetine-treated patients at POD 3, POD 7, and postoperative week 6. For VAS for pain at movement, significantly reduced pain was reported by duloxetine-treated patients at POD1, POD 3, POD 7, POD 14, postoperative week 6, and postoperative month 3. Conclusions: Duloxetine appears to decrease postoperative pain and opioid consumption following total joint arthroplasty. However, definitive conclusions are limited by small sample size and study heterogeneity. While there is a need for follow-up studies to determine the optimal dose, duration, and patient population, strong preliminary data provide robust support for future large-scale efficacy studies.

  PublisherDOI

• Prognostic Factors and Outcomes of De Novo Sinonasal Squamous Cell Carcinoma: A Systematic Review and Meta‐analysis

  Otolaryngology · 2021 · 20 citations

  • Medicine
  • Internal medicine
  • Oncology

  OBJECTIVE: To review overall survival (OS), recurrence patterns, and prognostic factors of de novo sinonasal squamous cell carcinoma (DN-SCC). DATA SOURCES: PubMed, Scopus, OVID Medline, and Cochrane databases from 2006 to December 23, 2020. REVIEW METHODS: The study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Articles were required to report either recurrence patterns or survival outcomes of adults with DN-SCC. Case reports, books, reviews, meta-analyses, and database studies were all excluded. RESULTS: < .001). Recurrence was reported in 628 of 1471 patients from 26 studies (42.7%) with an aggregate 5-year locoregional control rate of 67.1% (range, 50.4%-93.3%). CONCLUSION: This systematic review and meta-analysis suggests that the 5-year OS rate for DN-SCC may approach 54.5% and recurrence rate approaches 42.7%. In addition, various tumor characteristics including advanced T stage, positive nodal status, maxillary sinus origin, and negative HPV status are all associated with decreased survival.

  PublisherOA PDFDOI

• Research Guides: Evidence-Based Neonatology Subject Guide: EBN Resources & Tools

  2021-02-23

  libguides1st authorCorresponding

  This LibGuide is specifically designed for the UCI School of Medicine Neonatal Medicine Fellowship Program. This page includes resources and tools in Evidence-Based Neonatal Medicine.

  Publisher

• The Risk of Cancer from CT Scans and Other Sources of Low-Dose Radiation: A Critical Appraisal of Methodologic Quality

  Prehospital and Disaster Medicine · 2020 · 81 citations

  • Medicine
  • Medical physics
  • Nuclear medicine

  INTRODUCTION: Concern exists that radiation exposure from computerized tomography (CT) will cause thousands of malignancies. Other experts share the same perspective regarding the risk from additional sources of low-dose ionizing radiation, such as the releases from Three Mile Island (1979; Pennsylvania USA) and Fukushima (2011; Okuma, Fukushima Prefecture, Japan) nuclear power plant disasters. If this premise is false, the fear of cancer leading patients and physicians to avoid CT scans and disaster responders to initiate forced evacuations is unfounded. STUDY OBJECTIVE: This investigation provides a quantitative evaluation of the methodologic quality of studies to determine the evidentiary strength supporting or refuting a causal relationship between low-dose radiation and cancer. It will assess the number of higher quality studies that support or question the role of low-dose radiation in oncogenesis. METHODS: This investigation is a systematic, methodologic review of articles published from 1975-2017 examining cancer risk from external low-dose x-ray and gamma radiation, defined as less than 200 millisievert (mSv). Following the PRISMA guidelines, the authors performed a search of the PubMed, Cochrane, Scopus, and Web of Science databases. Methodologies of selected articles were scored using the Newcastle Ottawa Scale (NOS) and a tool identifying 11 lower quality indicators. Manuscript methodologies were ranked as higher quality if they scored no lower than seven out of nine on the NOS and contained no more than two lower quality indicators. Investigators then characterized articles as supporting or not supporting a causal relationship between low-dose radiation and cancer. RESULTS: Investigators identified 4,382 articles for initial review. A total of 62 articles met all inclusion/exclusion criteria and were evaluated in this study. Quantitative evaluation of the manuscripts' methodologic strengths found 25 studies met higher quality criteria while 37 studies met lower quality criteria. Of the 25 studies with higher quality methods, 21 out of 25 did not support cancer induction by low-dose radiation (P = .0003). CONCLUSIONS: A clear preponderance of articles with higher quality methods found no increased risk of cancer from low-dose radiation. The evidence suggests that exposure to multiple CT scans and other sources of low-dose radiation with a cumulative dose up to 100 mSv (approximately 10 scans), and possibly as high as 200 mSv (approximately 20 scans), does not increase cancer risk.

  PublisherOA PDFDOI

Frequent coauthors

• Philip S. Low

  Purdue University West Lafayette

  68 shared

• Philip T. Cagle

  36 shared

• Qihui Zhai

  36 shared

• Nikki Parker

  Endocyte (United States)

  26 shared

• Patrick Klein

  St. Antonius Ziekenhuis

  26 shared

• James F. Naughton

  Purdue University West Lafayette

  26 shared

• Deepika Dhawan

  Purdue University West Lafayette

  26 shared

• Christopher P. Leamon

  26 shared

Education

• MLIS - Master of Library & Information Studies , School of Library Information Studies

  University of California Berkeley

  1986

• BS - Bachelor of Science in Community Health Education, Health Sciences

  San Jose State University

  1984

• LMTA - Library Media Technical Assistance, Library

  Foothill College

  1980