Day 1: Platforms, Abstracts 001-004

001. Postpartum Maternal Opioid Therapy and the Risk of Adverse Neonatal Outcomes

Jonathan Zipursky1,2, Tara Gomes3,2, Karl Everett2, Andrew Calzavara2, Michael Paterson2, Peter C Austin2, Muhammad M Mamdani3, David N Juurlink1,2

1Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada. 2ICES, Toronto, Ontario, Canada. 3Unity Health, Toronto, Ontario, Canada

Background: It is widely taught that maternal use of opioids while breastfeeding, and use of codeine in particular, can predispose neonates to opioid toxicity.

Research Question: Is postpartum maternal opioid therapy associated with adverse neonatal outcomes?

Methods: Population-based retrospective cohort study using propensity score matching in Ontario, Canada, from September 1, 2012 to March 31, 2020. We identified all mother-infant pairs discharged from hospital alive within seven days of delivery. From this group we identified women who filled an opioid prescription within seven days of discharge. We matched each to one woman who did not, matching on age, calendar year, mode of delivery, and logit of the propensity score. The primary outcome was neonatal readmission to hospital for any reason within 30 days of the maternal opioid prescription. Secondary outcomes included any emergency department visit, hospitalization for any injury, admission to an intensive care unit, hospitalization with resuscitation or assisted ventilation, and neonatal death.

Results: We studied 85,675 women who filled an opioid prescription within seven days of hospital discharge after delivery and 85,675 women who did not. The median age was 32 years, 60.3% were primigravid, and 80.8% underwent cesarean delivery. In the primary analysis, 2,962 (3.46%) children born to mothers who filled an opioid prescription were hospitalized, compared to 3,038 (3.55%) children born to mothers who did not receive an opioid. Children of mothers who filled an opioid prescription were no more likely to be hospitalized for any reason compared to children born to mothers who did not (hazard ratio 0.98; 95% CI 0.93 to 1.03). We found no differences in any other adverse neonatal outcome, and similar results in a sensitivity analysis involving 17,037 women prescribed codeine.

Conclusions: Maternal receipt of an opioid prescription following delivery is not associated with several measures of neonatal harm.

002. Exposure to the Endocrine-Disrupting Metal Lead and Serum Estrogen Levels in Women

Anna Gerald1, Srividya Ganapathy1, Jianmin Zhu2, Yudan Wei1

1Mercer University School of Medicine, Macon, GA, USA. 2Fort Valley State University, Fort Valley, GA, USA

Background: Multiple factors affect estrogen levels in the body; however, the impact of exposure to environmental chemicals, such as heavy metals, on estrogen levels in humans remains inconclusive. Hypothesis: Is there a relationship between exposure to endocrine-disrupting metal lead and serum estrogen levels in adult women?

Methods: This cross-sectional study assesses the association between blood lead levels (BLLs) and serum estradiol in women (aged ≥ 20 years) who participated in the National Health and Nutrition Examination Survey between 2013-2016. Due to the impact of pregnancy and female hormone use on estrogen levels, we excluded participants who were pregnant or who used female hormones, resulting in a final sample of 1618 for the analysis. Using multiple general linear models, we estimated the changes (regression coefficient β) of serum estradiol levels (pg/mL) in association with BLLs (μg/dL), adjusting for potential confounders. Age-specific analysis was further conducted.

Results: The median of BLLs was 0.76 μg/dL (range: 0.11-12.80) and the median level of serum estradiol was 31.10 pg/mL (range: 2.12-523.00) among women aged 20-80 years. A statistically significant inverse association was found between blood lead and estrogen levels (β = -0.121; p = 0.0301), after adjusting for age, race/ethnicity, poverty status, education level, BMI, physical activity, cigarette smoking, alcohol consumption, and use of birth control pills. When stratified by age, a statistically significant inverse association was only seen among women aged 50-80 years (n = 600) (β= -0.363; p < 0.0001).

Conclusion: Our study demonstrated that increases in lead exposure were associated with decreased estrogen levels in older women, which might be implicated in the earlier menopause associated with increased levels of lead exposure that was observed in previous studies. Further studies are warranted to address limitations, including the cross-sectional nature of the study, and to characterize the interaction between lead exposure, reproductive hormones, and its associated reproductive toxicity.

003. Pediatric Risk of Mortality-III (PRISM III) and Pediatric Index of Mortality (PIM3) Scores among Pediatric Poisonings

Rachel E Culbreth1, Paul Wax1, Kim Aldy1,2, Christina Hantsch1, Theresa Mikhailov3, Nancy Brundage,4,5Jeffrey Brent5, In Collaboration with the Toxicology Investigators Consortium (ToxIC)

1American College of Medical Toxicology, Phoenix, Arizona, USA. 2Baylor University Medical Center, Dallas, Texas, USA. 3Medical College of Wisconsin, Milwaukee, Wisconsin, USA. 4Virtual Pediatric Systems (VPS, LLC), Los Angeles, California, USA. 5University of Colorado School of Medicine, Aurora, Colorado, USA

Background: The Pediatric Risk of Mortality-III (PRISM III) and the Pediatric Index of Mortality (PIM3) scores are commonly used in pediatric critical care as a measure of severity of illness as well as risk of mortality. However, it’s unclear if these scores adequately predict outcomes specifically for poisonings.

Research Question: What is the association of PRISM III and PIM3 scores with mortality and length of stay (LOS) among pediatric patients in the Pediatric Intensive Care Unit (PICU)?

Methods: Data consisted of all poisonings from the Virtual Pediatric System (VPS, LLC) from years 2019-2021. VPS is a multi-institutional database supporting comparative analysis of PICU management and outcomes. PRISM III and PIM3 probability of mortality scores were used. Multivariable logistic and log-linear regression analyses were separately conducted for each outcome (PICU mortality, PICU LOS, hospital LOS) for overall poisonings and by poisoning type (opioids, acetaminophen, antidepressants, stimulants, cardiovascular agents). Area under the curve (AUC) values were used to assess discrimination for PRISM III and PIM3.

Results: Among a sample of all poisonings (N=29,595), each additional percentage increase in PRISM III was associated with a 7% higher odds of PICU mortality (OR 1.07; 95% CI: 1.07, 1.08). Each additional percentage increase in PIM3 was also associated with a 8% higher odds of PICU mortality (OR 1.08; 95% CI: 1.01, 1.16). The AUC was 0.876 for PIM and 0.886 for PRISM III, reflecting adequate discrimination for the two measures. Higher PRISM III and PIM3 probabilities were associated with longer PICU and hospital LOS for overall poisonings. This also held true within certain types of poisonings (opioid, acetaminophen, antidepressant, and cardiovascular), but not for stimulants.

Conclusion: Higher PRISM III and PIM3 scores are predictive of higher mortality and longer LOS among pediatric poisonings in PICUs, with the exception of stimulant poisonings. 

*ToxIC: This research was performed in collaboration with the ACMT Toxicology Investigators Consortium

004. Can We Predict the Next “Benadryl Challenge?” A Side-by-Side Comparison of RADARS® Web Monitoring and NPDS Data

Hannah S St. Francis1, Christopher O Hoyte2,3, Anastasia N Cornell3, Faith E Lyons3, Christopher Pitotti4

1New York University School of Medicine, New York, NY, USA. 2University of Colorado School of Medicine, Aurora, CO, USA. 3Rocky Mountain Poison and Drug Safety, Denver, CO, USA. 4Uniformed Services University of the Health Sciences, Bethesda, MD, USA

Background: The “Benadryl Challenge” propagated by social media in 2020 resulted in adolescents ingesting toxic doses of diphenhydramine, and at least one associated fatality. Dangerous internet fads such as this continue. The Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS®) Web Monitoring Program surveys the internet for mentions of xenobiotics and observes frequency spikes which may uncover new trends in medication misuse or abuse.

Research Question: The purpose of this study is to determine temporally how detection of diphenhydramine ingestion trends in the RADARS® Web Monitoring Program compares to the National Poison Data System (NPDS).

Methods: This is a retrospective study analyzing the RADARS® Web Monitoring Program database of publicly available internet posts on Twitter, blogs, forums, and news articles. Search terms included “benadryl challenge,” “benadryl challenge,” “benadryl trip,” and/or “benadryl trip” from July 1 to December 31, 2020. Descriptive statistics were used to compare this to 2020 NPDS primary diphenhydramine ingestions by age group.

Results: We observed a peak frequency of 1,904 term mentions from 8/31 to 9/6/20 compared to an average of nine mentions per week from 8/1 to 8/30/20. NPDS data demonstrate that there were 2,171 reported diphenhydramine ingestions among 13-19 year-olds from 9/1 to 10/31/20. This represents a 16% increase over the mean (1,873) of the three preceding two-month periods.

Conclusion: The RADARS® Web Monitoring Program was able to identify a significant spike in internet mentions of the “Benadryl Challenge” in early September, 2020, which preceded an observable increase in adolescent diphenhydramine ingestions reported to poison centers in September and October, 2020. While not definitive, web monitoring may be a useful tool for predicting trends in medication misuse and could help guide public health messaging, education, and resource allocation.

Day 1: Moderated Posters, Abstracts 005-011

005. Medical Toxicology Consultations for Poisoned Children Managed in Pediatric Intensive Care Units

Kim Aldy1,2, Rachel E Culbreth1, Christina Hantsch1, Theresa Mikhailov3, Nancy Brundage4, Paul Wax1, Jeffrey Brent5, In Collaboration with the Toxicology Investigators Consortium (ToxIC)

1American College of Medical Toxicology, Phoenix, Arizona, USA. 2Baylor University Medical Center, Dallas, Texas, USA. 3Medical College of Wisconsin, Milwaukee, Wisconsin, USA. 4Virtual Pediatric System (VPS, LLC), Los Angeles, California, USA. 5University of Colorado School of Medicine, Aurora, Colorado, USA

Background: Drug overdose and poisoning (“poisoning”) increased by 83.6% from 2019 to 2020 among children and adolescents, becoming the third leading cause of death in that age group. Medical toxicologists (MTs) are specially trained to treat children with poisoning, commonly managed in Pediatric Intensive Care Units (PICUs). MT consultation is not available in all PICUs.

Research Question: What is the association between MT consultation and outcomes for poisoned PICU patients?

Methods: Virtual Pediatric System (VPS) is a multi-institutional database supporting comparative analysis of PICU management and outcomes. Data from years 2019-2021 on poisonings were included in this analysis. MT consultations were operationalized as a binary variable (yes/no). The comparison group consisted of cases without MT access at their hospital. Multivariable logistic regression analyses were used to determine the association between MT consultations and mortality, adjusting for age, sex at birth, race/ethnicity, PRISM III scores, and trauma (yes or no). Log-linear regression analyses were used to examine the association between MT consultations and PICU length of stay (LOS). All analyses were conducted in R 4.1.2.

Results: The analytic sample included 1,792 (11.2%) poisoned patients who received MT consultations and 14,205 (88.8%) poisoned patients without MT access. Patients who received MT consultations had a slightly higher PRISM III probability of mortality (median=0.43) compared to patients without MT access (median=0.34) (p < 0.001). In the multivariable model, MT consultations were associated with a 66% lower odds of mortality compared to those without MT access (OR: 0.34; 95% CI: 0.15, 0.77), after adjusting for covariates. Additionally, MT consultations were associated with a 18% shorter PICU LOS compared to patients without MT access (Estimate: -0.20; 95% CI: -0.26, -0.14).

Conclusion: VPS data analyses determined that MT consultation for poisoned PICU patients is associated with a reduction in mortality and shorter LOS.

*ToxIC: This research was performed in collaboration with the ACMT Toxicology Investigators Consortium

006. Analysis of Interventions Required in Pediatric Patients with Acute Intoxications Admitted to the Phoenix Children's PICU

Michael D Wade1, Diane E Hindman1,2, M'Hamed H Temkit1

1Phoenix Children's, Phoenix, AZ, USA. 2University of Arizona, Tucson, AZ, USA

Background: Acute intoxications in children account for ~5% of annual Pediatric Intensive Care Unit (PICU) admissions in the USA, while only about 30% of these admissions require any advanced interventions. In a healthcare environment where there is a shortage of critical care pediatric beds, or where responsible fiscal management is essential, having a better understanding of appropriate allocation of patients with acute intoxications is beneficial.

Hypothesis: Most pediatric patients with acute intoxications admitted to the Phoenix Children’s (PC) PICU require PICU-level interventions.

Methods: This was a retrospective chart review of all PICU admissions for patients less than or equal to 18 years of age with a diagnosis of acute intoxication, poisoning, ingestion, or overdose during the period July 1st, 2014, to June 30th, 2020. Patient charts were evaluated to determine patient demographics, toxicants, toxicant effects, interventions, necessary monitoring, and outcomes. Data were summarized using frequencies and proportions for categorical variables and mean, standard deviation, median, first/third quartiles, and range for continuous data.

Results: In total, 497 patient admissions (0.4%) met study criteria, 56% being female. Median age was 3.00 years (interquartile range, 1.50 – 14.00). Median PICU length of stay was 2.00 days (range, 1.00 – 18.00). Of these patients, 25.8% had intentional intoxications; the remainder were classified as accidental. The most frequently encountered toxicants were unknown substances (11.1%), detergents/degreasers/cleaning substances (9.5%), opioid analgesics (7.8%), and antihistamines (6.2%). In all, one patient died and only 10% of total cases required PICU-level interventions.

Conclusion: After review of all eligible PICU cases with acute intoxication, 90% did not require an ICU-level intervention. Most of these cases could seemingly have been managed on the medical floor, liberalizing PICU space and improving fiscal management. Future research should consider the impact of such a change on patient safety and hospital staffing.

007. Frequency of Critical Care Interventions by Mechanism of Action in Pediatric Anti-Epileptic Drug Overdose

Andrew B Troger1,2, Michael S Toce1,2, On Behalf of the Toxicology Investigators Consortium (ToxIC)

1Boston Children's Hospital, Boston, MA, USA. 2Harvard Medical Toxicology, Boston, MA, USA

Background: Anti-epileptic drugs (AEDs) are commonly prescribed to pediatric patients for both primary seizure prophylaxis as well as psychiatric indications. AED overdoses are associated with a wide range of toxicity, which is dependent on the mechanism of action (MOA) of the specific drug. Hypothesis: Pediatric exposures to AEDs with sodium-channel blockade require critical care interventions (CCIs) more frequently than exposures to other AEDs.

Methods: We performed a cross-sectional analysis of single-agent pediatric (0-19 years) AED exposures reported to the Toxicologic Investigators Consortium (ToxIC) database between 2010-2020. AEDs were sorted by MOA and compared to a composite measure of CCIs. This measure included vasopressors, endotracheal intubation, renal replacement therapy, gastric lavage, whole bowel irrigation, and hemodynamic measures such as mechanical cardiac pacing or extracorporeal circulatory support. Descriptive statistics were used to describe the cohort. Chi-squared test was used to compare drug-classes and CCIs.

Results: 497 AED exposures were identified, with 41 (8.2%) requiring CCIs. The majority of exposures (n = 298; 60.0%) were female. The median exposures per year was 49 (IQR 33-56.5), with a peak incidence of 71 exposures in 2014. Lamotrigine (n = 101), clonazepam (n = 86), and carbamazepine (n = 84) were the most common exposures; lamotrigine (14/101) and carbamazepine (14/84) required the most CCIs. AEDs with sodium-channel blockade (n = 240) had the highest proportion (33/240; 13.8%) of CCIs and were more likely to require CCIs compared to agents with all other primary mechanisms of action (8/257, 3.1%), χ2 (1) = 18.6, p < 0.001. Valproic acid (n = 68) accounted for the largest proportion (5/8; 62.5%) of CCIs among non-sodium channel blocking agents.

Conclusion: Exposures to AEDs with sodium-channel blockade carry a higher probability of requiring CCIs than exposures to agents with no sodium-channel blockade. Clinicians should consider these risks in the determination of monitoring recommendations and disposition for patients poisoned with AEDs.

*ToxIC: This research was performed in collaboration with the ACMT Toxicology Investigators Consortium

008. Cannabis Edible Toxicity in Children: Defining a Toxic Dose

Lesley Pepin1, Mark Simon1, Shireen Banerji1, Christopher Hoyte1,2, George S Wang3,1

1Rocky Mountain Poison & Drug Safety, Denver, CO, USA. 2University of Colorado Hospital, Aurora, CO, USA. 3Children's Hospital Colorado, Aurora, CO, USA

Background: Cannabis edible exposures have increased with legalization. Understanding the tetrahydrocannabinol (THC) dose leading to severe symptoms has implications for medical management and regulations.

Hypothesis: THC mg/kg dose can predict duration and symptom severity in children.

Methods: A retrospective review within a children’s hospital network where recreational cannabis is legal (01/01/2015 - 06/15/2022) was conducted. Edible ingestions in patients < 6 years were identified (ICD codes F12, T40.7X) and included if dose was known. Severe signs and symptoms were defined as: dysrhythmia, bradycardia, hypotension/sinus tachycardia requiring interventions, receiving sedatives, seizure/myoclonic jerks, significant sedation, intubation/respiratory failure/apnea, or oxygen support. Logistic regression and receiver operative curve (ROC) were performed to predict severe signs and symptoms and duration > 8 hours.

Results: Sixty-six cases were analyzed. Median age was 2.95 years (IQR 1.6). THC dosing ranged 0.2 to 69 mg/kg (median 2.083 mg/kg, IQR 4.33). Thirty patients had severe signs and symptoms (45%, 95% CI: 34%, 57%): 29 involved a dose >/= 2 mg/kg. Dose (p < 0.001) and age (p < 0.05) were significant predictors of signs and symptoms > 8 hours with odds increase of 2.03 per 1 mg/kg THC increase (95%CI: 1.34, 3.1). Dose was significant (p < 0.0001) in predicting severe signs and symptoms with odds increase of 2.8 per 1 mg/kg THC increase (95% CI: 1.72, 4.62). ROC demonstrated 2.33 mg/kg had sensitivity/specificity of 90%/73% for severe signs and symptoms (AUC 91.9%); for duration > 8 hours (AUC 84.7%), 1.72 mg/kg had a sensitivity/specificity of 86%/63%.

Conclusion: Edible ingestions exceeding 1.72 and 2.33 mg/kg were associated with prolonged and severe signs and symptoms in children < 6 years, respectively. These findings support healthcare providers and poison centers with management decisions. These thresholds help inform regulators on dose limitations for cannabis edibles in legal markets.

009. Cannabis Legalization and Unintentional Pediatric Poisonings: a Population-based, Repeated Cross-sectional Study

Daniel T Myran1, Peter Tanuseputro1, Nathalie Auger1, Lauren Konikoff1, Robert Talarico2, Yaron Finkelstein3

1University of Ottawa, ON, Canada. 2ICES, Toronto, Ontario, Canada. 3Hospital for Sick Children, University of Toronto, Toronto, ON, Canada

Background: Canada legalized cannabis in October 2018 but initially prohibited the sale of edibles. Starting in January 2020 some provinces permitted the sale of commercial cannabis edibles. Question: What are the effects of cannabis legalization edibles on pediatric poisoning?

Methods: We examined changes in hospitalizations for cannabis poisoning and all-cause poisoning, in children <10 years over three legalization policy periods in Canada’s four most populous (total 3.4 million children) provinces. We conducted a repeated cross-sectional population-level study of all pediatric hospitalizations in Ontario, Alberta, British Columbia, and Quebec. Exposure periods: Pre-legalization (January 2015 – September 2018), period one in which dried flower only was legalized in all provinces (October 2018 – December 2019), and period two, in which edibles were legalized in three provinces (exposed provinces) and restricted in one province (control province) (January 2020 – September 2021).

Results: During the seven-year study period, there were 581 hospitalizations for cannabis poisoning (53.9% male; mean age 3.6 years) and 4,406 hospitalizations for all-cause poisonings. Overall, pediatric hospitalizations for cannabis poisoning increased 6.3-fold. Out of all poisonings, the proportion of cannabis poisoning hospitalizations pre-legalization was 5.7% in the exposed provinces and 3.8% in the control provinces. During period one, the proportion increased to 15.0% in the exposed provinces (incidence rate ratio [IRR] 2.55; 95% CI 1.88-3.46) and to 11.8% in the control province (IRR 3.05; 95% CI 1.82-5.11). During period two, the proportion of poisoning hospitalizations due to cannabis more than doubled to 31.8% in the exposed provinces (IRR 2.16; 95% CI 1.68-2.80) but remained similar at 13.8% in the control province (IRR 1.18; 95% CI 0.71-1.97).

Conclusion: Following legalization, cannabis edible sales were associated with increases in pediatric hospitalizations for poisonings. In jurisdictions with legalized edibles, almost one in three pediatric hospitalizations for poisoning are due to recreational cannabis.

010. Pediatric Rattlesnake Envenomations Treated with Crotalidae Equine Immune F(ab') 2 Antivenom: A 3-Year Retrospective Observational Analysis

Justin A Seltzer1,2, Garret Winkler3, Nathan Friedman1,4, Henrik Galust1,4, Jeremy Hardin1,4, Priya Srihari5, Sam Ontiveros6, Bryan Corbett1, Richard Clark1

1UC San Diego Health, San Diego, CA, USA. 2Rady Children's Hospital-San Diego, San Diego, CA, USA. 3UTHealth Houston, Houston, TX, USA. 4VA San Diego Health Care, San Diego, CA, USA. 5Keck Medicine of USC, Los Angeles, CA, USA. 6Medical University of South Carolina, Charleston, SC, USA

Background: Rattlesnake envenomations are uncommon and the majority occur in adults. Though Crotalidae equine immune F(ab')2 antivenom (F(ab')2AV; trade name ANAVIP®) was introduced in 2018, no pediatric specific studies of F(ab')2AV have been reported to date.

Hypothesis: F(ab')2AV is well tolerated in pediatric patients.

Methods: A single center, retrospective chart review was performed on patients with rattlesnake envenomations presenting to a children's hospital between October 2018 and August 2022. Inclusion criteria were age less than 18 years and F(ab')2AV use. Exclusion criteria were other antivenom use at any time and presentation beyond 24 hours. Demographic characteristics, hemoglobin, platelet count, fibrinogen, and INR, number of F(ab')2AV vials used, infusion-related complications, and clinical outcomes were collected.

Results: Twenty-six patients, 19 males and seven females, with a mean age of 7.7 years (0.67-16 years) met inclusion criteria. 14 (54%) required only the initial 10 vial F(ab')2AV dose. 12 patients required additional doses with a mean additional vials given of 10.5 (4-34 vials). The mean total vials given for all patients was 15.7 (10-44 vials). Two patients developed acute infusion reactions. Both were treated by slowing the infusion rate and with medications (diphenhydramine, corticosteroids). No delayed reactions were noted. No patients required blood products or surgical interventions. Following discharge, no complications, recurrent symptoms, return visits, or readmissions were reported. Post-discharge follow up by chart review or phone was obtained for 18 patients, and no complications were noted. Seven patients had post-discharge hematologic laboratory evaluations, and all were normal.

Conclusions: Though limited by small sample size and post-discharge follow-up, F(ab')2AV was well tolerated in our series of pediatric patients, consistent with prior studies of all age groups.

011. Analysis of New York City Heavy Metal Testing of Consumer Products

Emma DeLoughery, Rob Hendrickson

Oregon Health and Science University, Portland, OR, USA

Background: Though some heavy metals are necessary for life, others are harmful in exposure or ingestion. This project was designed to analyze the findings of the New York City Department of Health and Mental Hygiene’s heavy metal testing of consumer products.

Research question: Do consumer products contain excessive heavy metals?

Methods: The dataset “Metal Content of Consumer Products Tested by the NYC Health Department” was analyzed using Microsoft Excel. The dataset contained testing of 6073 products. Data regarding guidance for maximum acceptable concentration of heavy metals were obtained from the FDA.

Results: Of the 6073 products tested from 2011-2021, 2878 (47.4%) contained measurable amounts of heavy metals (lead, arsenic, mercury, cadmium, and chromium). Cadmium and chromium were detected in only one product each and were excluded from further analysis, as were 246 products with data not recorded in ppm. Arsenic: 112 products; 97.3% dietary supplements, 2.6% spices. All products were over the 100 ppb limit. Lead: 2293 products; 50.5% spices, 12.4% dietary supplements, 10.2% toys. 100% of candy and other foods, 99.8% of spices, and 99.6% of dietary supplements were above the 0.1ppm limit for ingested substances; 68.4% of cosmetics were above the limit of 10 ppm. Mercury: 225 products; 70.7% dietary supplements, 28.4% cosmetics. 60.9% of cosmetics and 60.0% of dietary supplements were above the 1 ppm limit.

Conclusion: Nearly half of tested products contained measurable amounts of heavy metals, and of these many contained a greater than recommended amount of the metal in relation to an approximate reference standard. Dietary supplements and spices represent a common source of contamination, and dietary supplements in particular are often used by patients without discussion with a physician. This data highlights another reason it is important to discuss dietary supplement use with patients and to caution patients about the use of unregulated products.

Day 1: Posters, Abstracts 012-071

012. Massive Isolated Topiramate Ingestion with Marked Improvement after Hemodialysis

Trevor T Thiss1, Thom S Maciulewicz2,3, Jaiva B Larsen4,2, Farshad M Shirazi5,2

1Des Moines University, Des Moines, Iowa, USA. 2Arizona Poison and Drug Information Center, Tucson, AZ, USA. 3R. Ken Coit College of Pharmacy, University of Arizona, Tucson, AZ, USA. 4University of Arizona, Tucson, AZ, USA. 5University Of Arizona, Tucson, AZ, USA

Background: Though topiramate is dialyzable, there are no clear consensus guidelines on when hemodialysis (HD) is indicated. A recent review of the literature identifies 29 topiramate overdoses, however dialysis was rarely performed, co-ingestions were common, and many levels were only modestly elevated or not available. What parameters should trigger the decision to dialyze a patient in the setting of massive ingestion of topiramate?

Methods: This is a case study of a 32-year-old female with an intentional ingestion of 36 g of topiramate. The patient was evaluated by the toxicology service 24 hours after presentation to the ED. She had progressive non-anion gap metabolic acidosis, decreased level of consciousness with minimal response to painful stimuli, and abnormal myoclonic jerking of the extremities and facial muscles. In addition to toxicology, neurology and nephrology were also consulted. Neurology recommenced transfer for electroencephalography (EEG) to rule out focal seizures because EEG was not available at the treating facility. Unfortunately, timely transfer was not possible because the receiving hospital was at capacity. The decision to dialyze the patient at the current facility was made due to the significant ongoing encephalopathy.

Results: After receiving HD, the patient had marked improvement in encephalopathic features; She was alert and oriented to person, place, and time and myoclonic jerking ceased. Topiramate levels before and after HD were 220.0 mcg/mL and 97.7 mcg/mL, respectively. GC/MS showed topiramate only.

Conclusion: Dialysis effectively lowered the serum levels of topiramate, which correlated with significant clinical improvement. Dialysis likely prevented an unnecessary patient transfer to another facility. Further data is needed to establish reliable guidelines regarding HD in the setting of topiramate intoxication.

013. Patience is Prudent: a Case Report of Unusual Phenytoin Toxicokinetics from Fosphenytoin

Samantha S Klein1, Mary Ann Howland2, Mark K Su3

1Division of Medical Toxicology Ronald O. Perelman Department of Emergency Medicine, New York, NY, USA. 2St. John’s University College of Pharmacy and Health Sciences, New York, NY, USA. 3New York City Poison Control Center, Department of Health and Mental Hygiene, New York, NY, USA

Background: Phenytoin is well known to exhibit Michaelis-Menten pharmacokinetics, switching from first-order to zero-order kinetics at concentrations greater than 10 mg/L. However, phenytoin has been reported to display even slower metabolism at the onset of toxic concentrations often depicted as a plateau, followed by expected zero-order metabolism. This has often been attributed to ongoing absorption from oral ingestions. We present a case report of this phenomenon following intravenous fosphenytoin administration one day earlier.

Case Report: A 37-year-old, 79 kg man presented to the Emergency Department (ED) for a supratherapeutic phenytoin concentration after receiving an intravenous loading dose of fosphenytoin 1000 mg PE the day before. Fosphenytoin is a prodrug that is metabolized to phenytoin with a conversion half-life of 15 min after IV administration. Prior to this loading dose, his phenytoin concentration was undetectable. A phenytoin concentration two hours after the fosphenytoin dose was 45.8 mg/L (therapeutic range: 10 – 20 mg/L). A second concentration 21 hours later was 44 mg/L. The poison center (PC) was contacted for help with management. The patient was asymptomatic with an unremarkable physical examination. His triage vitals were: BP, 131/86; HR, 76 beats/minute; RR, 17 breaths/minute; T, 97 degrees Fahrenheit; O2 Sat, 98% on room air. The PC recommended monitoring serial phenytoin concentrations as an outpatient. Over the next five days, the phenytoin concentrations decreased in zero-order fashion: 38.5 mg/L at 29 hours, 28.3 mg/L at 57 hours, 21.7 mg/L at 82 hours and 17.6 at 132 hours.

Conclusion: Phenytoin demonstrates Michaelis-Menten pharmacokinetics above serum concentrations of ten mg/L. At concentrations above 40 mg/L phenytoin, metabolism seems to be stalled initially. Although often attributed to ongoing oral absorption, this case suggests another explanation and may be due to changes in tissue binding or adaptation.

014. Treatment of Phenytoin Poisoning by Medical Toxicologists

Jessica L Winkels, David B Liss, On Behalf of the Toxicology Investigators Consortium (ToxIC)

Washington University, St Louis, Missouri, USA

Background: In 2015, the Extracorporeal Treatments in Poisoning (EXTRIP) workgroup published guidelines indicating hemodialysis can be used in certain patients with severe phenytoin poisoning. It is unknown how medical toxicologists have since implemented these recommendations.

Research Question: We sought to characterize the key features of phenytoin poisoning cases submitted to the Toxicology Investigators Consortium (ToxIC) database and to identify factors associated with hemodialysis use.

Methods: This is a descriptive study of phenytoin poisoned patients and their treatment by medical toxicologists from January 2016 to October 2022 using the ToxIC database. Cases were excluded if there were no signs/symptoms or if symptoms were deemed unlikely to be related to a toxicologic exposure. Pearson Chi-Square and Fisher’s Exact tests were used where appropriate to assess for significant associations.

Results: We analyzed 178 cases. Phenytoin was the primary agent in 92.1% of patients. Most ingestions involved therapeutic use (41.0%) or were unintentional (27.0%), as compared to attempted self-harm (18.5%). Common symptoms were coma/central nervous system (CNS) depression (33.1%), ataxia (30.3%), nystagmus (14.0%), and seizures (9.6%). Toxicological treatment such as decontamination, elimination therapy, or nonpharmacologic support (including intubation) was given in 47.8% of cases. In total, five patients (2.8%) received hemodialysis for toxin removal. Overall mortality was low at 1.1% (n = two).

Patients with CNS depression (p = 0.004), who were intubated (p = <0.001), and/or required ICU admission (p = <0.001) were all significantly more likely to be treated with hemodialysis. Report of phenytoin as a primary agent was associated with higher rates of ICU admission (p = 0.014) but not with more frequent CNS depression, intubation, or hemodialysis use.

Conclusion: Phenytoin-poisoned patients are rarely treated with hemodialysis. Factors suggestive of more clinically severe toxicity such as CNS depression, intubation, and ICU admission are associated with hemodialysis. 

*ToxIC: This research was performed by the ACMT Toxicology Investigators Consortium

015. Severe Phenytoin Toxicity Treated with Intermittent Hemodialysis

Johana Gallerani, Michael Semple, Ari Filip, Michael Mullins, David Liss

Washington University/Barnes Jewish Hospital, St. Louis, MO, USA

Background: Although phenytoin has a volume distribution of 0.5 to 0.8 L/kg in the serum, it is highly protein bound. Therefore, the use of hemodialysis to treat phenytoin toxicity has been controversial as its utility and benefit in improving phenytoin removal is unclear.

Research Question: What is the rate of phenytoin elimination before, during, and after hemodialysis in a patient with phenytoin toxicity?

Methods: This was a single patient chart review of a patient with phenytoin toxicity. Phenytoin concentrations, both bound and free, were serially collected during the patient’s hospitalization. Rates of elimination for the period before, during, and after hemodialysis were calculated using a simple regression.

Results: A 56-year-old male with epilepsy presented to the emergency department (ED) with ataxia, confusion, and falls. His sister reported he had been taking his phenytoin as directed. Vital signs were normal upon presentation to the ED, but he was disoriented and confused with ataxic movements and bilateral lateral-gaze nystagmus. Frontal bone bossing and gingival hyperplasia were present. Phenytoin levels, both bound and free, were gathered at several time points during the patient’s hospital stay. Initial total phenytoin level was found to be 76.0 mcg/mL. At hour seventeen of admission, after IV fluid hydration, total phenytoin remained significantly elevated at 65.0 mcg/mL. After two sessions of hemodialysis, the patient’s mental status improved and his phenytoin concentration decreased to 32.6 mcg/mL. Average elimination rate by simple regression analysis of the time periods before hemodialysis, during hemodialysis, and after dialysis were 0.68, 0.88, and 0.22 mcg/ml/hr, respectively.

Conclusion: Despite phenytoin being highly protein bound, phenytoin clearance appeared to be higher during the two day time period while the patient was undergoing hemodialysis than in the periods before or after dialysis.

016. National Trends in Bupropion Exposures Reported to the United States Poison Centers

Abigail F Kerns1, Saumitra V Rege1, Rita Farah1, Benjamin J Holstege2, Christopher P Holstege1

1University of Virginia, Charlottesville, VA, USA. 2Eastern Virginia Medical School, Norfolk, VA, USA

Background: Bupropion sales remain robust in the United States (U.S.), with ~23 million prescriptions in 2011 increasing to ~29 million in 2020. Medical toxicologists often are called to care for bupropion overdoses.

Research Question: Have bupropion exposures reported to U.S. poison centers (PCs) increased proportionally to prescriptions and are those exposure outcomes significant?

Methods: The National Poison Data System (NPDS) was queried for bupropion exposures reported to U.S. PCs from 2011 to 2021. We identified and descriptively assessed the relevant demographic and clinical characteristics. Poisson regression models were used to evaluate the trends in the number and rates (per 100,000 human exposure calls) of bupropion exposures.

Results: There were 148,210 bupropion exposures (48.1% were single substance) reported during the study period; 53.1% from acute care hospitals. Ages between 20-29 years (29.1%) constituted the most common age group. Teenage exposures (13 – 19 years) increased during the study period (15.7% to 18.7%). Females accounted for 62.8%. Ingestion was the most common route of exposure, followed by inhalational misuse. The most frequently co-occurring substances were other antidepressants (35.1%) and atypical antipsychotics (16.2%). Suspected suicides (48.3%) and therapeutic errors (30.1%) were the most common call reasons. Major effects were encountered in 9.6% with 0.6% case fatality rate; 27% of cases were admitted to the critical care unit. Intravenous fluids and benzodiazepines were the most frequently used therapy. Agitation and drowsiness/lethargy were the most common clinical effects. During the study period the frequency and rate of bupropion exposures increased by approximately 87% and 91%, respectively.

Conclusions: Bupropion exposures increased significantly over the study period. This is greater than can be attributed to the increase in number of prescriptions. Intentional exposure through suicide attempt was common. Many required admission to critical care units. Risk reduction strategies should be formulated to address this growing problem.

017. Bilateral Lower Limb Compartment Syndromes Resulting from Neuroleptic Malignant Syndrome

Daniel Menza, Jasmeen Kaur, Alisha Sherwani, Daria Falkowitz

Hackensack University Medical Center, Hackensack, NJ, USA

Background: Neuroleptic malignant syndrome (NMS) is an uncommon but potentially fatal syndrome secondary to the use of antipsychotics. It is characterized by muscular rigidity, rhabdomyolysis, fever, and autonomic dysfunction.Severe rhabdomyolysis can lead to rising pressures in peripheral fascial compartments. We describe an unusual case of a patient with neuroleptic malignant syndrome who developed bilateral lower extremity compartment syndrome.

Methods: A 24-year-old male on olanzapine, fluphenazine, clozapine, and quetiapine presented to the emergency department with undifferentiated altered mental status and complained of bilateral leg pain. During ED course, he developed fever with CK levels rising up to > 42,670 consistent with rhabdomyolysis, and findings suspicious for tense lower limb anterolateral compartments.

Results: He was treated for NMS with propofol and benzodiazepines. He underwent four compartment bilateral fasciotomies for his bilateral lower extremity compartment syndrome. He required several sessions of hemodialysis but recovered and was eventually trialed on quetiapine and discharged to rehab.

Conclusion: Compartment syndrome, although rare, is an important complication of neuroleptic malignant syndrome and should be considered in all patients presenting with this rare condition. A high index of suspicion can aid in early recognition and treatment in order to prevent long-term complications.

018. Prolonged Serotonin Toxicity after Massive Duloxetine Overdose

Patrick Fann, Andrew Taylor, Andrew New

Henry Ford Jackson, Jackson, Michigan, USA

Background: Duloxetine is a second-generation serotonin-norepinephrine reuptake inhibitor prescribed for treatment of both psychiatric disorders and neuropathic pain. Duloxetine overdose overall is uncommon, with a total of 6281 cases involving the drug reported in 2020 by the American Association of Poison Control Centers. Survival after massive overdose has previously been reported, however prior case reviews have typically described resolution of symptoms within 24 hours.

Methods: This is a case report obtained via retrospective chart review.

Results: A 74-year-old female who presented to the emergency department after intentional ingestion of an unknown quantity of duloxetine and cephalexin. After an unremarkable initial workup, the patient was admitted medically for observation. She was noted by the hospitalist team 13 hours after ingestion to have mild generalized tremors and nystagmus. The next day the patient developed severe agitation, jerking body movements, and worsening horizontal nystagmus. After not responding to initial therapy, psychology and neurology were consulted. Ultimately, the patient was diagnosed with opsoclonus-myoclonus syndrome secondary to serotonin toxicity 64 hours after her initial ingestion. A duloxetine level was drawn 81 hours post-ingestion and returned a serum level of 3300 ng/mL. The majority of the patient’s neurologic symptoms resolved five days after her admission.

Conclusions: This case highlights the potential for both delayed onset as well as prolonged serotonin toxicity in cases of massive duloxetine overdose. Few case reports outside of postmortem studies have identified serum concentrations above 2000 ng/mL. This patient’s level returned at 3300 ng/mL more than three days after her initial ingestion. Several mechanisms may have delayed clearance and prolonged toxicity in this case including delayed absorption secondary to diabetic gastroparesis, inhibition of CYP2D6 by duloxetine, and reversal of CYP1A2 induction by smoking cessation upon admission.

019. Single Substance Ingestions of Citalopram Versus Escitalopram Reported to U.S Poison Centers from 2010-2022

Will R Goodrich, David H Schaffer, Rita Farah, Nathan P Charlton

University of Virginia Health, Charlottesville, Virginia, USA

Background: Citalopram and escitalopram are among the most commonly prescribed selective serotonin reuptake inhibitors worldwide. Previous research presents conflicting data regarding toxicity between these two enantiomers.

Research Question: Are there meaningful differences in toxicity after single substance ingestions of citalopram versus escitalopram?

Methods: We conducted a retrospective review of the National Poison Data System (NPDS) database from 2010-2022 for intentional single substance ingestions of citalopram or escitalopram among adults (age ≥ 20 years). We compared medical outcomes and the incidences of seizure, tachycardia, and ECG changes. We used Pearson’s Chi-square test to assess differences in clinical effects and medical outcomes between citalopram and escitalopram.

Results: There were 12,338 citalopram and 10,385 escitalopram intentional single substance ingestions reported to the NPDS. Statistical significance in outcome severity and incidences of seizure, tachycardia, and ECG changes were observed. Death or major effect resulted in 4.4% (n = 431) of citalopram exposures compared to 2.2% (n = 170) with escitalopram (p <0.001). Seizure developed in 6.1% (n = 749) of citalopram ingestions compared to 1% (n = 105) from escitalopram (p <0.001). Tachycardia occurred in 19.6% (n = 2422) of citalopram ingestions compared to 17.2% (n = 1790) with escitalopram (p <0.001). Before 2019, QRS, QT and PR abnormalities were coded collectively as conduction disturbances. Between 2010 and 2018, 7.4% (n =741) of citalopram ingestions resulted in conduction disturbances compared to 6.5% (n = 396) with escitalopram (p = 0.024). QTc interval data in the NPDS database became available in 2019. Since 2019, 13.4% (n = 317) of citalopram exposures had QT prolongation compared to 11.3% (n = 503) from escitalopram (p = 0.009).

Conclusion: Death or major effect were more than twice as common after intentional single substance ingestions of citalopram compared to escitalopram. Citalopram was associated with statistically significant higher rates of seizure, tachycardia, and ECG changes. This data suggests a higher risk of clinically significant outcomes after isolated citalopram overdose as compared to escitalopram.

020. Mobitz I as Manifestation of Acute Lithium Cardiotoxicity

Henrik Galust1,2, Jeremy Hardin3,2, Nathan Friedman3,2, Justin Seltzer3,2, Alicia Minns3,2

1UCSD, San Diego, California, USA. 2VA San Diego Health Care, San Diego, CA, USA. 3UCSD, San Diego, CA, USA

Background: A 13-year-old female presented one hour following an acute overdose of 10 unknown tablets. The tablets were likely obtained from her grandmother’s medications. The 12-lead ECG demonstrated the Mobitz I block, a novel manifestation of lithium cardiotoxicity.

Methods: This is a single patient case report.

Results: Physical examination, vital signs, CBC, and CMP including serum K, Ca, and Mg were unremarkable. Acetaminophen concentration four hours post ingestion was mildly elevated (28 mcg/ml) and well below the treatment threshold. ECG revealed a Mobitz Type I atrioventricular block but was otherwise morphologically normal for the patient’s age. No prior ECGs were available for comparison. Serum lithium concentration was noted to be 1.7 mEq/L (0.6-1.2 mEq/L); digoxin and salicylate were not detected in the serum. While admitted on telemetry monitoring the patient continued to have occasional recurrent episodes of Mobitz type I block which lasted for minutes and spontaneously resolved. During each episode the patient remained asymptomatic and normotensive. The patient was treated only with IV fluids and repeat lithium concentration at 14 hours post-ingestion was undetectable. Repeat ECG at 20 hours post-ingestion showed normal sinus rhythm with normal intervals. She was discharged 36 hours after ingestion with an ambulatory cardiac monitor per pediatric cardiology recommendations.

Conclusion: Mobitz I atrioventricular block has not, to our knowledge, been described in the setting of lithium exposure. The most common cardiotoxic effects of lithium include QT prolongation, T-wave abnormalities, and to lesser extent SA node dysfunction and ventricular arrhythmias. This case highlights the widely variable potential presentations of lithium cardiotoxicity.

021. Massive Verapamil Ingestion With A Second Peak Of Toxicity

Alexander M Sidlak, Lani Kroese

Inova Fairfax, Falls Church, VA, USA

Background: Verapamil is a phenylalkylamine calcium-channel blocker that leads to cardiogenic and vasoplegic shock in overdose. Massive ingestions of sustained- or extended-release formulations can lead to bezoar formation and delayed toxicity.

Hypothesis: Delayed, acute toxicity can occur with massive ingestions of verapamil despite initial stabilization.

Methods: This is a single patient case report managed with bedside toxicology consultation.

Results: A 16-year-old girl ingested eighty-eight 240 mg verapamil SR tablets (21.12 g). She presented to a local hospital and was initially hypotensive (76/40 mmHg). She was given an insulin bolus (1U/kg) and started on norepinephrine (0.1 mcg/kg/min) and insulin infusions (1 U/kg/h). A nasogastric tube was placed and polyethylene glycol 3350 was started (1 L/h). She was transferred to a tertiary care hospital and was cannulated onto venoarterial extracorporeal membrane oxygenation (ECMO) 6 h after ingestion. Ten hours after ECMO initiation, her lactate, which had peaked at 4 mmol/L, improved to 1.6 mmol/L. A verapamil concentration obtained 24 h after ingestion was 3100 ng/mL. Despite whole bowel irrigation, the patient had no bowel movements until HD 2 at which point she began to pass pill fragments. Thirty hours after ingestion, the patient's glucose began to rise (peak: 581 mg/dL). During this time the patient decompensated with worsening bradycardia. She lost native cardiac function and was restarted on insulin (titrated to 5 U/kg/h). Her lactate increased to 15 mmol/L, 40 h post-ingestion. Additional vasopressors were started (epinephrine and vasopressin) and the patient’s hemodynamics stabilized.

Conclusion: With massive ingestions of calcium channel blockers, not only is a protracted course likely, but acute decompensation with a second peak of toxicity is possible. Restoration of stool output may be a leading indicator of impending decompensation.

022. Fatal Thrombosis Event Following ECMO Cannulation in Beta Blocker/Calcium Channel Blocker Overdose: a Case Report