Meetings & Events > Conferences & Courses > Annual Scientific Meeting > 2014 Annual Scientific Meeting > Research Abstracts > 2014 Abstracts & Posters, 1-20

2014 ACMT Annual Scientific Meeting Research Abstracts and Posters

Abstract submissions were reviewed and scored in blinded fashion by members of the ACMT Research Committee

Select abstract number to go to text, click abstract title to access poster (PDF).

 1,  2,  3,  4,  5,  6,  7,  8,  9,  10,  11,  12,  13,  14,  15,  16,  17,  18,  19,  20

Go to Abstracts 21-40      Go to Abstracts 41-60        Go to Abstracts 61-80        Go to Abstracts 81-96

Abstracts are published in the March 2014 edition of the Journal of Medical Toxicology. To access an online version of the abstract reprint, click here

Original Research Platform Sessions

1. Snake Venom Binding Activity of Expired Antivenoms

Rentmeester LL1, 2, Burton DW1,2, Fitzgerald RL1,2, Clark RF1,3
1University of California-San Diego, San Diego, CA, USA; 2San Diego Veterans Affairs Healthcare System, San Diego, CA, USA; 3California Poison Control System, San Diego, CA, USA

Background: Antivenoms are believed to have limited shelf lives. Research indicates antivenom activity may surpass expiration dates. 

Research Question: Does binding of antivenom to venom persist after expiration date? 

Methods: Fluorescent immunoassays compared binding of antivenom to venom.  Ninety-six well plates were coated with venom, incubated with antivenom, and then incubated with biotinylated anti-horse IgG. Streptavidin conjugated to β-galactosidase was added and hydrolysis of substrate generated fluorescence. Plates were analyzed using a fluorescent reader. 

Results: Faboterapico Polivalente Antivipmyn Tri antivenoms with expiration dates of 1/00, 3/09, and 10/10 bound Bothrops jararaca venom at respective rates of 95%, 68% and 73%, relative to unexpired control. Soro Antibotiopico Laquetico antivenom with an expiration date of 01/00, bound Borthrops jararaca venom at a rate of 141% relative to unexpired control. Faboterapico Polivalente Antivipmyn Tri antivenoms with expiration dates of 1/00, 3/09, and 10/10 bound Crotalus atrox venom at respective rates of 115%, 73%, and 92%, relative to unexpired control. Soro Antibotiopico Laquetico antivenom with an expiration date of 01/00 bound Crotalus atrox venom at a rate of 147% relative to unexpired control. SAIMR Polyvalent antivenoms with expiration dates of 04/98, 04/03, 12/05, and 10/10 bound Naja naja venom at respective rates of 81%, 81%, 93%, and 103% relative to unexpired control. SAIMR Polyvalent antivenom with expiration dates of 04/98, 04/03, 12/05, and 10/10 bound Naja nivea venom at respective rates of 88%, 88%, 100%, and 106%, relative to unexpired control. 

Discussion: SAIMR antivenom binds Naja venom for at least 9 years after expiration. Soro Antibotiopico Laquetico antivenom binds Bothrops and Crotalus venom for at least 14 years past expiration, demonstrating in some cases, higher venom binding than unexpired antivenom. Faboterapico Polivalente Antivipmyn Tri retained some binding to Bothrops and Crotalus venoms, but not to the same degree as Soro Antibotiopico Laquetico. However, binding did not differ significantly among all lots of Faboterapico Polivalente Antivipmyn Tri. This study is limited by its in vitro  nature and suitability for patient safety was not addressed. 

Conclusion: Antivenoms from three manufacturers demonstrated equivalent binding to venom in vitro  despite surpassing expiration dates.


2. Trypsin and Rosmarinic Acid Reduce the Toxicity of Eastern Coral Snake (Micrurus fulvius) Venom in Mice 

Parker-Cote JL, O'Rourke D, Rosenbaum M, Brewer KL, Miller SN, Meggs WJ
Brody School of Medicine at East Carolina University, Greenville, NC, USA 

Objective: Since antivenom is expensive and not always available, alternative treatments for toxic bites and stings are needed. The efficacy of trypsin and rosmarinic acid (RA) in treating Eastern Coral Snake (M. fulvius) envenomation in a murine model is determined in an in vitro  model. 

Hypothesis: Both trypsin and RA will reduce the toxicity of Eastern Coral Snake venom. 

Methods: Design: randomized controlled blinded study. Subjects: Fifty mice (20-30g). Study groups: Intraperitoneal injections of: 1) 2mg/kg M. fulvius venom (approximately twice the LD50 for mice, n=10), 2) 2mg/kg M. fulvius venom incubated in vitro  for 1 hour prior to injection with RA at a 1:10 ratio (n=17), 3) 2mg/kg M. fulvius venom incubated in vitro  for 1 hour prior to injection with 1 mg of trypsin (n=17), 3)1 mg trypsin IP without venom (n=3), and 4) RA IP without venom (n=3). Mice were observed for 12 hours for signs of toxicity. Main outcome: time to toxicity (respiratory distress (respiratory rate < 25 breaths/min.), loss of spontaneous locomotor activity, or inability to upright self). Statistical analysis: Time to toxicity using Tukey-Kramer HSD and survival to 4, 6, and 12 hours using Chi-square analysis. 

Results: Onset of toxicity: group 1, 120.3 minutes, Group 2, 238.1 minutes (p=0.15 relative to Group 1), Group 3, 319.7 minutes (p=0.007 relative to Group 1). Group 3 but not Group 2 survival to 4 hours was significant compared to Group 1 (p=0.023). Two mice in the trypsin group and 1 mouse in the RA group survived to twelve hours. Mice receiving trypsin alone or RA alone survived to 12 hours. 

Conclusion: In vitro neutralization of M. fulvius venom by trypsin justifies progressing to an in vivo model in future studies.


3. Point of Care Testing in Setting of Nitromethane and Methanol Co-ingestion Will Not Mask True Creatinine, Anion, or Osmolar Gap

Cao D1, Maynard S2, Mitchell AM3, Kerns WP4, Beuhler MB5
1Rocky Mountain Poison and Drug Center, Denver, CO, USA; 2Carolinas Pathology Group, Charlotte, NC, USA; 3Indiana University School of Medicine, Indianapolis, IN, USA; 4Carolinas Medical Center, Charlotte, NC, USA; 5Carolinas Poison Center, Carolinas HealthCare System, Charlotte, NC, USA

Background: Nitromethane interferes with the Jaffé colorimetric reaction used to measure serum creatinine, potentially mimicking acute kidney injury. This lab interference may confound the clinical management of nitromethane exposure, especially when co-ingested with a toxic alcohol. Bedside point-of-care (POC) testing platforms measure creatinine by an enzymatic method, which may result more accurate measurements. We further hypothesize that the anion and osmolar gaps remain unchanged in the presence of nitromethane. 

Methods: Nitromethane was added to whole blood from healthy volunteers to achieve 5 concentrations (0, 0.25, 0.5, 1, and 2 mmol/L), and the following tests were performed: creatinine (Jaffé and POC), electrolytes (associated with Jaffé and POC), plasma osmolality, and nitromethane concentration (gas chromatography [GC]). Remaining samples were refrigerated for 7 days and reanalyzed by GC. Anion and osmolar gaps were calculated at each concentration. The proportional recovery of nitromethane and degradation of nitromethane were measured by GC. Data were analyzed for agreement with single factor ANOVA. The magnitude of nitromethane interference on creatinine was assessed by linear regression. 

Results: Mean creatinine for enzymatic and Jaffé methods were 0.93 vs. 0.76 mg/dL, respectively. Using the Jaffé method, the creatinine concentrations increased linearly with increasing nitromethane concentrations (R2=1, p<0.01): creatinine (mg/dL) = 7.1 * nitromethane (mmol/L) + 0.79. With the enzymatic method, creatinine remained unchanged across the range of nitromethane concentrations (p=0.99). Anion and osmolar gaps also remained consistent. Nitromethane was reliably identified in all sample concentrations using GC on day 0. Detection of nitromethane at the 0.25 mmol/L concentration was not recovered in all samples on day 7. Nitromethane degradation was most pronounced at the 2 mmol/L concentrations with a mean 81% recovery. 

Conclusions: Nitromethane alters measures of creatinine using the Jaffé reaction in a linear fashion but not when using the enzymatic reaction. In the clinical setting of toxic alcohol exposure, a measured difference between Jaffé and enzymatic creatinine may identify the presence of nitromethane and may reliably estimate the nitromethane concentration. The presence of nitromethane did not result in alterations in the anion or osmolar gap. Retrospective nitromethane detection may be feasible after 7 days using GC.

Supported by the 2012 – 2013 Emergency Medicine Foundation / Medical Toxicology Foundation Resident Research Grant

Previously published as: Cao D, Maynard SM, Mitchell-Smith AM, Kerns WP, Beuhler MC. Point-of-care testing in setting of nitromethane and methanol co-ingestion will not mask true creatinine, anion gap, or osmolar gap. Ann Emerg Med 2013; 62(4 suppl):S42-3. Reprinted with permission of the journal.


4. Internet Training Resulted in Improved Intern Performance in a Simulated Poisoned Patient as Measured by Checklist, But Not Global Assessment

Stolbach A1, Harry H1, Cordeiro M1, Vasquez V1, Kim H2, Barrueto F2
1Johns Hopkins University, Baltimore, MD, USA; 2University of Maryland School of Medicine, Baltimore, MD, USA

Background: Internet-based didactics can be used as a substitute for live teaching. However, there is little research on the effectiveness of internet-based teaching on outcomes relevant to clinical competence in medical toxicology. Medical simulation can measure competency in the management of poisoned patients. Simulation may be an ideal assessment tool for clinical competence, because it allows standardization of scenarios and no potential for patient harm.

Objective: In a population of emergency medicine (EM) interns, we measured the effect of brief internet-based training on resuscitation competence in a simulated opioid-poisoned patient.

Methods: We enrolled all eligible EM interns from two consecutive classes. The two groups were similar in background characteristics, including prior toxicology training, simulation experience, and EM training. All participants completed identical simulated poisoned patient scenarios during the first week of residency. The case, presented as an unknown, was a comatose opioid-poisoned patient. The case was adapted from a medical simulation textbook. One group of participants performed a brief internet-based training on management of toxicologic coma 1 week to 1 day prior to the patient simulation. The other group received no training. An unblinded rater scored participants’ simulation performance with 3 validated tools- a simple checklist, a time-weighted checklist (which gives higher scores for faster completion of items), and a “global” scoring tool, which uses the rater’s general impression of the participant’s performance in various domains.Results: Scores were compared using unpaired t-tests. Using a simple checklist, internet-trained participants (n=12) had a mean score of 71.67%, while untrained participants (n=11) had a mean score of 49.09%. The difference in scores was 22.58% (95% CI: 1.59%-43.56%.) Using a time-weighted checklist, internet-trained participants had a mean score of 65.00%, while untrained participants had a mean score of 38.18%. The difference in scores was 26.81% (95% CI: 4.75-48.88). Using a global assessment score, the internet-trained participants had a mean score of 23.17, while untrained participants had a mean score of 22.45. The difference was 0.7122 (P>0.05).

Conclusion: A brief internet-based teaching module resulted in improved EM intern performance using checklist scoring, but not global weighted scoring, in a standardized simulated toxicology patient scenario. 

5. Trends in Opioid Prescribing for Acute Headache in U.S. Emergency Departments

Mazer-Amirshahi M1, Mullins PM1, Pines JM1, Perrone J2, Nelson LS3
1The George Washington University, Washington, DC, USA; 2Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA; 3New York University School of Medicine, New York, NY, USA

Background: Prescribing of opioid analgesics has risen dramatically over the past decade. Although not recommended by the American Academy of Neurology and the American College of Emergency Physicians’ consensus guidelines as first-line therapy for acute headache, opioids are commonly used to treat this condition. Research Question: To assess the frequency of use and trends in opioid prescribing for headaches in U.S. emergency departments (EDs) from 2001-2010.

Methods: A retrospective review of data from the CDC’s National Hospital Ambulatory Medical Care Survey (NHAMCS) 2001-2010 was performed. Visits for headache and head pain were identified. Medications used for the treatment of acute headache were identified and categorized based on medication class. Trends in ED prescribing of five common opioids (codeine, hydrocodone, hydromorphone, morphine and oxycodone) were specifically explored. The proportion of visits for which each medication was prescribed or administered was tabulated and trends were analyzed using survey-weighted logistic regression.

Results: Visits for headache during which any of the five designated opioids were prescribed increased significantly between 2001 and 2010 (20.8% vs. 34.3%, p<0.001). Prescribing of the individual drugs hydromorphone, morphine and oxycodone increased significantly, with the largest relative increase (454.1%) noted with hydromorphone (1.7% to 9.5% p<0.001). Codeine utilization decreased from 2.3% to 2.0% and hydrocodone use did not significantly change during the study period. Utilization of opioid alternatives, including acetaminophen, butalbital and non-steroidal anti-inflammatory drugs (NSAIDS) did not change significantly. Prescribing of antiemetic agents decreased modestly whereas the use of intravenous fluids increased significantly. The estimated visits for acute headache increased from 5,539,206 in 2001 to 7,747,342 in 2010.

Discussion: Our analysis was limited because we could not account for the use of multiple medications during the same visit and the order in which medications were given. Future initiatives should focus on promoting the use of guideline-concordant agents for the treatment of acute headache in the ED and to promote rational prescribing of opioids.

Conclusion: Despite guideline recommendations, there was increased utilization of opioid analgesics for acute headache in U.S. EDs over time. Antiemetic use declined, despite consensus recommendations endorsing their use.  


6. Trends in Opioid Prescribing in U.S. Emergency Departments Based on Provider Level of Training

Mazer-Amirshahi M1, Mullins PM1, Pines JM1, Nelson LS2, Perrone J3
1The George Washington University, Washington, DC, USA; 2New York University School of Medicine, New York, NY, USA; 3Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, USA

Background: Over the past decade, there have been significant increases in opioid analgesic prescribing. Variation in opioid prescribing based on provider level of training and specific opioids is less well characterized.

Research Question: To assess trends in opioid prescribing based on provider level of training and specific agents in U.S. EDs.

Methods: We reviewed all ED visits from the CDC’s National Hospital Ambulatory Medical Care Survey (NHAMCS) 2001-2010 involving patients ≥18 years. Trends in ED prescribing of all opioids stratified by Drug Enforcement Agency (DEA) schedule and provider level of training were analyzed. Prescribing trends for five common opioids (codeine, hydrocodone, hydromorphone, morphine and oxycodone) were individually explored. Visits were stratified into three categories by provider level of training: those that involved an attending only, a resident or mid-level provider (nurse practitioners, physician assistants). The proportion of visits for which each medication was prescribed was tabulated and trends were analyzed using survey-weighted logistic regression.

Results: The weighted estimate of adult ED visits increased from 81,251,195 in 2001 to 100,027,879 in 2010 and the proportion of visits during which an opioid was prescribed by any provider increased from 10.0% to 19.0%, p<0.001. Overall opioid prescribing increased for all visit types but was most pronounced in visits involving residents. Prescribing of schedule II agents increased more than schedule III-V agents for all groups (Table). Hydromorphone use increased the most for all provider groups (433.4%-529.9%, p<0.001), followed by morphine (139.7%-219.9%, p<0.001) and oxycodone (94.7%-142.3%, p<0.001). A variable effect was noted on codeine and hydrocodone prescribing among providers.

Discussion: Opioid prescribing increased dramatically in the past decade in the ED, and visits involving residents had the greatest growth. Although attendings may influence resident prescribing, these trends could reflect increased emphasis on pain management in training programs. Other forces such as patient experience and accreditation requirements could be driving overall prescribing rates higher.

Conclusion: There were significant increases in opioid prescribing among visits involving all providers over time, with hydromorphone demonstrating the greatest increase among the opioids studied. Visits involving residents showed the largest change among providers.


7. A 3-year Analysis and Comparison of Opioid Prescribing Practices by Emergency Care Providers for Chronic Pain

Ganem VJ1, Bebarta VS1, Mora AG2, Varney SM1
1San Antonio Military Medical Center, Fort Sam Houston, TX, USA; 2US Army Institute of Surgical Research, Fort Sam Houston, TX, USA

Background: Chronic pain is a common reason for emergency department (ED) visits. Emergency medicine providers commonly prescribe opioids, however, rates of opioid misuse have been high in the last several years. Previous studies have not described the variability in prescribing habits of emergency medicine providers.

Objective: To describe opioid prescribing practices for providers in an emergency department.

Methods: In our IRB-approved, retrospective study we evaluated opioid prescriptions from an emergency department at two military facilities between June 2009 and June 2012. We queried the outpatient record database to obtain a list of opioid medications prescribed and ICD-9 codes associated with ED visits for chronic pain. We also collected the number of pills prescribed, type of medication, and medication refill status. For statistical analysis we compared the incidence and proportions with chi-square or Fisher’s exact tests where appropriate. Wilcoxon test was used for non-parametric continuous variables. Data were reported as mean±SD (median [IQR]). A p<0.05 was considered significant.

Results: Over a 3-year period, emergency care providers wrote 28,298 opioid prescriptions. A total of 3,936 (13.9%) prescriptions were associated with a visit attributed to chronic pain. Providers were 49% emergency physicians, 48% physician assistants (PAs), and <1% nurse practitioners (NPs). Medications prescribed were 40% hydrocodone-acetaminophen, 35% oxycodone-acetaminophen, 18% codeine-acetaminophen and 7% other (tramadol, propoxyphene-acetaminophen, oxycodone, hydromorphone and codeine). The average number of pills prescribed was 23±26 (20 [15-20]). Oxycodone-acetaminophen was more likely to have been prescribed by a physician than by a PA (65% vs 35%) as was hydrocodone-acetaminophen (59% vs 41%) (p<.0001). Codeine-acetaminophen was more likely to have been prescribed by a PA than by a physician (90% vs 10%) as was tramadol (56% vs 44%) (p<0.0001). PAs prescribed a larger number of pills compared to physicians (26±23 vs 20±29, p<0.0001).

Conclusions: Physicians were more likely to prescribe opioids that were more potent and had higher risk for misuse (oxycodone-acetaminophen and hydrocodone-acetaminophen) compared to PAs. PAs prescribed more opioid pills per prescription and were more likely to prescribe tramadol and acetaminophen-codeine.   


8. Buprenorphine/Naloxone Pediatric Ingestion: Exposure Rates Differ Between Film and Tablet Formulations

Green JL1, Martinez EM1, Severtson SG1, Dart RC1,2, Lavonas EJ1,2
1Rocky Mountain Poison and Drug Center, Denver, CO, USA; 2University of Colorado School of Medicine, Aurora, CO, USA

Background: Buprenorphine ingestion can cause life-threatening poisoning in young children. Previous reports have found that film formulations are associated with lower pediatric exposure rates than tablet formulations.

Research Question: The purpose of this study is to determine whether differences in pediatric exposure rates to different buprenorphine products are stable over time.

Methods: Data from Researched Abuse, Diversion, and Addiction-Related Surveillance (RADARS®) System Poison Center Program, January 2011 – March 2013, involving unintentional exposure to buprenorphine sublingual tablets or film by children aged < 6 years were analyzed. To adjust for medication availability, event ratios (rates) were based on the number of patients filling prescriptions for each formulation (“Unique Recipients of a Dispensed Drug”, URDD). Negative binomial regression was used to produce quarterly rates, average rates, and 95% confidence intervals (CIs).

Results: 1,695 reports were analyzed. Exposure rates for buprenorphine/naloxone combination tablets (7.0 exposures per 10,000 URDD (CI: 6.6 – 7.3)) exceeded those for buprenorphine monoingredient tablets (2.8 (CI: 2.4 – 3.2)) and combination film (0.9 (CI: 0.8 – 1.0)). The combination tablet and monoingredient tablet rates were significantly greater than film rates (Rate Ratios (RR): 7.6 (CI: 6.7 – 8.6, p<0.0001) for combination tablets and RR: 3.1 (CI: 2.6 – 3.7, p<0.0001) for monoingredient tablets compared with film). Relationships were consistent over time except for slight decreases in the monoingredient tablet rate.

Discussion: This study cannot determine whether the differences are caused by packaging or formulation. This analysis did not include generic buprenorphine/naloxone tablets, introduced in February 2013.

Conclusion: The rate of unintentional exposures to buprenorphine/naloxone sublingual film by young children is significantly less than the rate of exposure to buprenorphine/naloxone or buprenorphine monoingredient tablets.


9. Buprenorphine/Naloxone Abuse and Diversion: Film Rates are Less Than Tablet Rates

Green JL1, Martinez EM1, Servertson SG1, Cicero TJ2, Kurz SP3, Rosenblum A4, Surratt, HL3, Dart RC1, Lavonas EJ1
1Rocky Mountain Poison and Drug Center, Denver Health and Hospital Authority, Denver, CO, USA; 2Washington University School of Medicine, St. Louis, MO, USA; 3Center for Applied Research on Substance Abuse and Health Disparities, Nova Southeastern University, Coral Gables, FL, USA; 4Institute for Treatment and Services Research, National Development and Research Institutes, New York, NY, USA

Background: Buprenorphine is sometimes diverted and abused. Previous reports showed differences in diversion and abuse rates between formulations, but observation periods were short. This study extends the comparison of diversion and abuse rates between buprenorphine sublingual formulations.

Research Question: Are buprenorphine diversion and abuse rates stable over time?

Methods: Data from the Researched Abuse, Diversion, and Addiction-Related Surveillance (RADARS®) System Drug Diversion (DD), Opioid Treatment (OTP), and Survey of Key Informants’ Patients (SKIP) Programs were analyzed. The DD program captures new police investigations. The treatment programs (OTP and SKIP) collect patient reports of using a product “to get high” in the previous 30 days. Quarterly data from 2010Q4-2013Q1 (DD) and 2011Q2-2013Q1 (OTP/SKIP) were analyzed. To account for availability, event ratios (rates) were based on the number of patients filling prescriptions for each formulation (“Unique Recipients of a Dispensed Drug,” URDD). Quarterly rates, average rates, and 95% confidence intervals (CIs) were calculated using negative binomial regression.

Results: 1,505 diversion reports and 5,293 abuse reports were analyzed. Average diversion rates for buprenorphine/naloxone tablets (13.6 reports/10,000 URDD, CI: 12.8-14.5) and monoingredient tablets (8.7, CI: 7.6-9.8) exceeded the combination film rate (1.3, CI: 1.1-1.5) (Rate ratio (RR) c/w film: 10.6 (CI: 9.0-12.4, p<0.0001) for combination tablets and 6.7 (CI: 5.5-8.2, p<0.0001) for monoingredient tablets). Average abuse rates for buprenorphine monoingredient tablets (61.8 reports/10,000 URDD, CI: 59.2-64.6) and buprenorphine/naloxone tablets (21.3, CI: 20.3-22.3) exceeded the combination film rate (9.1, CI: 8.7-9.6) (RR c/w film: 6.8 (CI: 6.3-7.3, p<0.0001) for monoingredient tablets and 2.3 (CI: 2.2-2.5, p<0.0001) for combination tablets).

Discussion: This analysis excludes generic buprenorphine/naloxone tablets, introduced in February, 2013.

Conclusion: Diversion and abuse rates for buprenorphine and buprenorphine/naloxone tablets consistently exceed those of buprenorphine/naloxone sublingual film.


10. Sub-Anesthetic Slow Infusion of Ketamine for Treatment of Depression Produces Rare but Currently Unpredictable and Potentially Serious Psychosomatic Effects 

Rasimas JJ
Penn State College of Medicine, Hershey, PA, USA

Background: Ketamine has been associated with psychotomimesis, dissociation, and addiction. It is also used as a rapid-acting antidepressant at doses and infusion rates lower than those for anesthesia.

Hypothesis: Ketamine for treatment of depression may cause adverse psychosomatic effects, mental illness may confer greater risk.

Methods: Three concurrent prospective studies enrolled patients in open-label (unipolar depressives (UPs)) and placebo-controlled (UPs and bipolar depressives (BPs)) trials to receive single-dose ketamine 0.5 mg/kg IV over 40 minutes. Serious medical conditions were exclusionary. No CNS-active agents were permitted for 18 days before infusion. Brief Psychiatric Rating Scale (BPRS) and Clinician-Administered Dissociative States Scale (CADSS) were employed to measure symptoms. Fisher’s exact test compared groups regarding SAEs. ANOVAs analyzed symptom data with repeated measures from baseline to 24 hours post-infusion.

Results: Forty-one patients (29 UPs, 12 BPs) and 50 controls were treated from September 2011 to September 2013. One UP experienced profound sedation with a Riker score of 2, she recovered fully within 2 hours. Another UP suffered cocaine relapse 22 days post-infusion. One BP had severe vomiting. Four subjects (2 UP, 1 BP, 1 control) experienced acute psychiatric reactions (“K-holes”) with varying degrees of anxiety, fear, violent ideation, dissociation, and aphasia—one UP failed to return to pre-infusion symptom intensity at 190 minutes post-infusion. Patients reported more anxiety than controls at baseline, it decreased following infusion. Dissociation, substantial in all groups, was greater in BPs vs. UPs vs. controls. On measures of agitation, grandiosity, delusions, and disorientation, patients did not differ from controls. Most measures peaked at infusion end or within 40 minutes. Increases in thought disorder, odd behavior, excitement, hallucinosis, and dissociation were greater in open-label vs. blinded subjects. History of substance abuse, PTSD, or trauma did not correlate with study measures. Baseline data did not predict SAEs.

Discussion: Due to thought disorder and aphasia, some patients were unratable when most symptomatic.

Conclusion: Most subjects tolerate 0.5 mg/kg ketamine infusion over 40 minutes, although dissociation, thought disorder, and aphasia are common. Major adverse reactions are more likely (p=0.043) in mood disorder patients, but not predicted by any demographic or clinical measure studied.


11. Survival of Swine with Nifedipine Toxicity Treated with Methylene Blue

Murphy CM1, Williams C1, Quinn M1, Nicholson B1, Shoe T1, Dulaney AR1,2, Buehler MC1,2, Kerns WP1,2
1Carolinas Medical Center, Charlotte, NC, USA; 2Carolinas Poison Center, Charlotte, NC, USA 

Background: Calcium channel antagonist-induced shock remains a significant treatment challenge.  Nifedipine, a dihydropyridine calcium channel antagonist, has a proposed mechanism of vasodilatation through increased nitric oxide (NO) production.  Methylene blue (MB) inhibits NO production by inhibiting the activity of soluble guanylyl cyclase, thus may be useful to reverse hypotension associated with nifedipine toxicity.  Methylene blue has not been studied as an antidote for nifedipine toxicity. 

Hypothesis:  Methylene blue will improve survival following nifedipine intoxication in a swine model.

Methods: This IACUC approved study used 24 swine that were sedated with alpha-chloralose, mechanically ventilated, and instrumented for drug delivery and hemodynamic measures. After stabilization and basal measures, nifedipine (0.01875 mg/kg/min) was infused until toxicity, defined as a reduction in product of cardiac output and mean arterial pressure of 20%, was reached.  Animals received a bolus of 20 mL/kg 0.9% normal saline once toxicity occurred immediately followed by equal volume amounts of either normal saline as a sham treatment, MB (1 mg/kg as a bolus and subsequent infusion), epinephrine (0.1 mcg/kg/min), or MB and epinephrine.  All treatments were continued for 5 hours after the initiation of toxicity or until death occurred. Hemodynamics were monitored throughout the study.  Surviving animals were euthanized.  Survival data was analyzed using the student’s t-test.

Results: Nifedipine toxicity was characterized by vasodilatory hypotension, impaired contractility, and tachycardia with terminal bradycardia.  The mean time to death after reaching toxicity was 232±67.5 min.  There was no statistically significant change in survival in animals treated with MB, epinephrine, or MB plus epinephrine (Table 1).

Discussion: We observed no survival treatment effect with MB even in combination with epinephrine.  Potential limitations of this experiment include: excessive severity of toxicity, insufficient dose of MB, untreated direct cardiac stress from prolonged compensatory tachycardia, and NO/soluble guanylyl cyclase may play a minor role in nifedipine-induced hypotension.

Conclusion: Methylene blue demonstrated no improvement in survival of swine with nifedipine-induced toxicity.  Further studies are needed to elucidate the value of MB in treating calcium channel antagonist-induced shock.


12. Incidence and Outcomes of Adult Cardiac Arrest Associated with Toxic Exposure Treated with Therapeutic Hypothermia (ToxiCool)

Modisett KL1, Gowda IM2, Walsh SJ3, Heffner A1, Pearson D1, Kerns WP1
1Carolinas Medical Center, Charlotte, NC, USA; 2University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, NC, USA; 3Einstein Medical Center, Philadelphia, PA, USA

Background: Therapeutic hypothermia improves neurologic recovery in cardiac arrest patients who regain spontaneous circulation. The incidence and outcome of patients who undergo therapeutic hypothermia after cardiac arrest associated with a toxic exposure is unknown.

Research Question: To describe the incidence, epidemiologic characteristics, and outcomes of patients who suffer cardiac arrest associated with toxic exposure that are treated with therapeutic hypothermia.

Methods: This is a retrospective review of a post cardiac arrest database and medical records of all patients registered to receive therapeutic hypothermia via our institution’s clinical pathway between November 2007 and February 2013. The database includes prospectively collected clinical data utilizing Utstein criteria and the Cerebral Performance Categories (CPC) Scale. All patients were treated in an evidence-based clinical pathway that included therapeutic cooling. The database and each patient’s medical record was systematically reviewed independently by two physician investigators to determine a toxic versus non-toxic cardiac arrest with toxic defined as a xenobiotic that directly and immediately caused the patient’s cardiac arrest. Causality was determined by consensus of two of three investigators. Groups were compared using Fisher’s exact test.

Results: 389 patients underwent treatment during the study period and 48 of 389 (12%) were deemed toxic arrests. Patients who suffered toxic arrests were slightly younger, less likely to have an initial shockable rhythm, and less likely to receive bystander CPR as compared to non-toxic cases (see table). The most common xenobiotics included cocaine (n=16), benzodiazepines (n=13), and opioids (n=9). Within the toxic subset, an initial shockable rhythm was associated with greater survival rate (11/16) than a non-shockable rhythm (9/31) (p=0.01)

Discussion: Toxic patients treated with therapeutic hypothermia had similar survival and neurologic function compared to non-toxic causes. Limitations to our study include accuracy of assigning causality, incomplete confirmation of exposures, relatively small patient population, CPC scores extrapolated from medical records, and inability to control for potentially confounding co-morbid conditions.

Conclusion: Toxin-associated cardiac arrests accounted for a significant proportion of patients in this study. Additional, larger studies may help to elucidate the optimal role for therapeutic hypothermia in toxin-induced cardiac arrest.


13. The Use of Physostigmine by Toxicologists in Anticholinergic Toxicity

Watkins JW, Schwarz EA, Arroyo-Plasencia AM,On behalf of the ACMT Toxicology Investigators Consortium (ToxIC)
Washington University, St. Louis, MO, USA

Background: The anticholinergic toxidrome is well described and relatively common, seen over 350 times by toxicologists reporting to the ToxIC registry in 2012.  Administration of physostigmine is generally regarded as the antidote to anticholinergic toxicity.  While physicians without toxicology training may be reticent to use physostigmine due to their unfamiliarity, we would expect that trained toxicologists would be relatively liberal in its use.

Research Question: How often is physostigmine administered to patients with anticholinergic toxicity that are evaluated by a toxicologist?

Methods: We retrospectively analyzed data in the ToxIC registry, representing data from medical toxicologists in multiple institutions nationwide, searching for patients who exhibited an anticholinergic toxidrome, determining what treatment(s) they received, and classifying the treatments as physostigmine, benzodiazepines, physostigmine and benzodiazepines, antipsychotics or no definitive treatment.  

Results: 352 patients were seen by Toxicologists for anticholinergic toxidromes in 2012, of which 113 (32.1%) were given benzodiazepines alone, 46 (13.1%) were given physostigmine alone, 32 (9.1%) received both physostigmine and benzodiazepines, 12 (3.4%) were given antipsychotics and 149 (42.3%) were given no definitive treatment.  Of the patients who received physostigmine alone or in combination 5 (6.4%) required intubation, 1 (1.3%) developed rhabdomyolysis. Of those who received benzodiazepines alone or in combination 17 (11.7%) required intubation, 4 (2.8%) developed rhabdomyolysis. Of those who did not receive physostigmine 25 (9.1%) required intubation and 8 (2.9%) developed rhabdomyolysis. Those who received physostigmine had a lower rate of intubation (6.4% vs 9.1%) and rhabdomyolysis (1.3% vs 2.9%) than those who did not, but the differences were not significant (OR 0.68, 95% CI 0.25-1.84 p=0.45 and OR 0.56 95%CI 0.07-4.53 p=0.59 respectively).

Discussion: These data suggest that patients with anticholinergic toxicity are more likely to receive benzodiazepines than physostigmine (32.1% vs 13.1%) as monotherapy, and a significant number of these patients did not receive treatment for their toxidrome.  The use of physostigmine was not correlated with intubation rates or rhabdomyolysis though numbers were small.

Conclusion: We find it interesting that physostigmine was infrequently used as treatment by toxicologists, given its recommendation for use in anticholinergic toxicity.


14. Effects of Calcium Channel Blocker Poisoning Interventions: A Systematic Review

St-Onge M1, Dube PA2, Gosselin S3, Guimont C4, Godwin J1, Blais R3, et al.
1University of Toronto, Ontario Poison Centre, Toronto, ON, Canada; 2Direction of Environmental Health and Toxicology, Institut national de santé publique du Québec, Québec, QC, Canada; 3McGill University Health Centre, Centre antipoison du Québec, Montreal, QC, Canada; 4Centre Hospitalier Universitaire de Québec, Quebéc, QC, Canada

Context: No summary of evidence has been published for calcium channel blocker (CCB) overdose.

Objective: This review sought to evaluate effects of proposed treatments on mortality, morbidity, hemodynamics, functional status, hospital length of stay (LOS), intensive care (ICU) LOS, duration of vasopressors use, and serum levels of CCB in poisoned adults.

Methods: We performed a search of both commonly used databases and gray literature to identify studies examing treatment effects on targeted outcomes. Two reviewers independently selected studies and a group of reviewers abstracted all relevant data using a pilot-tested form. A second group analyzed risk of bias and overall quality with standardized tools. Qualitative synthesis was used to summarize evidence.

Results: 15,621 references were identified by our search strategy, and 81 articles and 103 case reports were included. The only observational studies were for high-dose insulin (HDI) and extra-corporeal life-support (ECLS). Only case series and animal studies were found for atropine, vasopressors, calcium, glucagon, lipid emulsion, 4-aminopyridine, and levosimendan. Decontamination, pacemaker, and plasma exchange were studied in the case series. Risk of bias across studies was high for all interventions and moderate to high for ECLS. HDI was beneficial to hemodynamics and lessened mortality at the risks of hypoglycemia and hypokalemia (low quality of evidence - QOE). ECLS showed potential to improve survival in severe shock and cardiac arrest patients at costs of limb ischemia, thrombosis, and bleeding (low QOE). Calcium, dopamine, or norepinephrine, could improve hemodynamic parameters and survival without severe side effects (very low QOE).

Conclusion: HDI and ECLS were the most strongly supported interventions in the literature. However, evidence on CCB poisoning treatment is of low quality and is drawn from a heavily biased heterogenous literature.


Original Research: Poster Presentations


15. Overdoses and Organ Donation: Barriers to Salvaging Good from Tragedy 

Ahari S, Wiegand TJ
University of Rochester Medical Center, Rochester, NY, USA

Background: Several drugs known to mimic brain death in overdose have led to confused prognoses and unintended harm. We present three cases of fatal opioid poisoning that would have delayed pronouncement of death via standard means undermining organ donation but for the addition of a cerebral perfusion study (CPS).

Hypothesis: Relying on prerequisite findings in brain death (coma, absent brainstem reflexes, apnea) have limited predictive value and the potential for catastrophic harm (premature organ donation) in drug overdose while pending drug concentrations cause harm by delay, in these cases CPS can ultimately hasten diagnosis.

Methods: A review of consult service records was performed and all patients on whom a CPS was performed in order to facilitate determination of brain death were included.

Results: Three patients with accidental opioid exposure, an 8 month-old boy, a 16 year-old girl and a 28 year-old male heroin addict, were found comatose and with respiratory failure. Two had arrested prior to hospitalization. All had severe anoxic injury and deteriorated over 24-48 hours becoming increasingly difficult to support. Formal brain-death exams in each demonstrated absent electrical activity on electroencephalogram, lack of response to ventilatory drive (PCO2, hypoxia) and absent brain-stem reflexes, however families were hesitant to allow organ procurement until they were reassured drug intoxication was not blunting response during the death exam. In each patient a CPS demonstrated a lack of perfusion ultimately facilitating brain-death diagnosis and organ donation.

Discussion: The medical literature includes reports where various types of drug overdose mimicked brain death. State criteria for determining brain death typically include a formal neurologic exam, lack of patient response to ventilatory drive off sedation (i.e. 15 mmHg increases in PCO2) and optional EEG. Additionally, in overdoses, it is suggested brain-death not be diagnosed until toxic drug concentrations are absent. Determining drug concentrations however often take significant time potentially undermining continued stability compromising procurement of viable organs for transplantation in applicable cases.

Conclusion: Cerebral perfusion study may provide us with a more efficient way to ascertain patient death in overdose victims when drug toxicity mimics brain death.

16. Use of Multi-Dose Activated Charcoal in the ToxIC Registry

Ahari S, O'Geen, R, Wiegand, TJ, On behalf of the ACMT Toxicology Investigators Consortium (ToxIC)
University of Rochester Medical Center, Rochester, NY, USA

Background: A position paper from AACT and EAPCCT in 1999 provides an evidence-based review of the literature and guidelines on the use of Multi-Dose Activated Charcoal (MDAC) for acute poisonings. The ToxIC registry is a robust data source that offers important epidemiological insight poisoning trends and practices.

Hypothesis: Do current MDAC treatment practices by Toxicologists involved in the Toxicologic Investigators Consortium align with published guidelines?

Methods: We reviewed the ToxIC Case Registry in its entirety (2010-present) for cases involving MDAC treatment. Descriptive statistics were used for analysis.

Results: Out of nearly 25,000 case entries, the Registry contained 66 patients treated with MDAC for an acute poisoning. 2 (3%) were between 7-12 years-old, 20 (30%) were between 13-18 years-old, 41 (62%) were 19-65 years-old, and 3 (5%) were older than 65.44 (66%) were women and 1 was pregnant. Salicylates were involved in 29 of 66 cases (44%), 18 (27%) of which they were the sole toxin. Valproic acid was second most common in 12 cases (18%) (5 polydrug). 8 (12%) were phenytoin ingestions (1 polydrug). 7 (11%) involved acetaminophen (6 polydrug). 4 (6%) involved carbamazepine, 3 (5%) involved theophylline, 2 (3%) involved cyclobenzaprine (1 polydrug), 2 (3%) involved amitriptyline (1 polydrug), 2 involved ibuprofen (1 polydrug), 3 involved phenobarbital (3 polydrug) and several drugs/toxins were involved in a single poisoning: lamotrigine, cyclopeptide-containing mushrooms, colchicine, XR diltiazem, propoxyphene, atenolol and quetiapine.  Secondary agents involved in more than one of the ingestions included: diphenhydramine (3), caffeine (3) and benzodiazepines (3).

Discussion: Previously published practice guidelines suggest MDAC should be considered in life-threating poisonings involving carbamezapine, dapsone, phenobarbital, quinine, or theophylline. While some of these ingestions are not common, carbamezapine, phenobarbital and theophylline were among the ingestions treated with MDAC in the ToxIC Registry. The most common drug ingestion, however, in which MDAC was used in the Registry were salicylates despite consensus guideline specifically not recommending MDAC be used for salicylates (although this recommendation remains controversial).

Conclusions: MDAC is infrequently used to facilitate drug elimination in ToxIC cases and practice contrasts somewhat from published guidelines. An update on MDAC use is warranted.


17. Fluoroamphetamine(4-FA)-Induced Cardiogenic Shock Confirmed with Serum and Urine Levels

Al-Abri SA1, Meier KH1, Colby J2, Benowitz NL1
1California Poison Control System, San Francisco Division, University of California, San Francisco, CA, USA; 2Department of Laboratory Medicine, University of California, San Francisco, CA, USA

Background: 4-Fluoroamphetamine (4-FA) is a para-substituted phenethylamine-type synthetic stimulant with numerous internet blog postings, but no clinical case reports in the medical literature.

Purpose: We report a case of a young man who developed severe toxicity after using 4FA with laboratory confirmation.

Methods: An 18 year-old male presented to the ED with vomiting and chest tightness. He was alert, oriented, diaphoretic, with respiratory rate 16, HR=103, BP=130/52, SpO2=100 % on R.A, T=97.7 and an otherwise unremarkable exam. Two days prior, he received naltrexone IM as part of an opioid addiction treatment and was taking fluoxetine and trazodone. He used a new drug of abuse about 5 hours before presentation. Initial laboratory studies were significant for WBC 38.5, HgB 17.8 and urine drug screen positive for amphetamine, THC and opiates. Initial ECG showed sinus tachycardia with upward sloping ST depression in inferior leads and no ST segment elevation. After few hours, he became extremely diaphoretic and hypoxic. He was intubated and sedated then had a brief PEA arrest. CXR showed pulmonary edema and his cardiac echo showed LV hypokinesia, sparing the apex, with EF =10%, troponin = 8.3 ng/ml. He became hypotensive and was treated with dobutamine and epinephrine infusions and placed on an intra-aortic balloon pump. Cardiac function improved within 48 hours and repeat Echo showed EF =35%.Comprehensive toxicology testing revealed fluoroamphetamine, naproxen, fluoxetine, trazodone, naltrexone, nicotine and cotinine in urine, and fluoroamphetamine, naproxen, trazodone, and cotinine in serum.The 4FA urine level was 64000 ng/ml and serum level was 118 ng/ml.

Discussion: Our patient experienced severe toxicity after taking 4-FA, which included cardiogenic shock requiring invasive management and life support. With no previously reported 4FA clinical poisoning cases to compare with, and relying on its comparative pharmacology to other stimulant drugs, we suspect the toxic mechanism was an acute cardiomyopathy due to a catecholamine-induced myocarditis and/or small vessel myocardial ischemia. Cardiogenic shock due to acute cardiomyopathy has been reported with methamphetamine abuse.

Conclusion:Recreational use of 4-FA may present with life threatening cardiac toxicity including cardiomyopathy and cardiogenic shock and pulmonary edema.


18. An Algorithm for the Recognition and Treatment of Fire-Related Cyanide Exposure

Blohme E1, Babu KM1, Bebarta VS2
1University of Massachusetts, Worcester, MA, USA; 2Wilford Hall Medical Center, San Antonio, TX, USA

Background: The clinical indications for the treatment of hydrogen cyanide toxicity in the critically ill patient have been described. However, minimally symptomatic patients often present to emergency departments in high numbers after potential cyanide exposure in the setting of house fires.  Evidence-based guidelines are needed to support rapid treatment of fire victims (or emergency responders) with a critical cyanide exposure, while limiting unnecessary testing or therapy in patients without evidence of toxicity.

Hypothesis: A single algorithm can be developed for the evaluation and treatment of cyanide toxicity in patients with smoke inhalation.

Methods: In September 2013, we performed a systematic review of the English and German literature indexed in Pubmed. Abstracts that appeared applicable to the diagnosis, management, and complications of inhalational cyanide toxicity were systematically screened. We included animal studies, case reports, systematic reviews, expert panel recommendations, consensus statements and prospective studies. No human randomized controlled trials were found. Studies older than 25 years or those not available in electronic format were excluded. In addition, a non-systematic literature search was performed to answer specific questions such as long-term outcomes of cyanide toxicity and laboratory interference of antidotes. We inspected the reference lists of all relevant articles.

Results: Our search found 1484 reports, of which 207 abstracts and 95 studies were analyzed. We found that inhalational cyanide toxicity associated with CNS and respiratory depression, acidemia, seizure, hypotension, cardiac ischemia, or cardiac arrest, required empiric treatment. Altered mental status, vomiting, hyperpnea, tachypnea, and bradypnea should prompt a search for cyanide toxicity (using serum lactate > 8mmol/ L as a marker) and carbon monoxide exposure. The presence of other, mild symptoms (e.g., headache, nausea, dizziness, weakness) without vital sign abnormalities is unlikely to indicate clinically significant cyanide toxicity, and should not result in antidotal therapy, or specific laboratory evaluation, however, carbon monoxide toxicity should be considered in this patients.

Conclusions: We propose a diagnostic algorithm that delineates evaluation and treatment indications for patients who range from minimally symptomatic to critically ill after potential cyanide exposure via smoke inhalation.


19. Neurotoxicity from Sidewinder Rattlesnake Envenomation in Arizona

Bosak AB, Ruha AM
Banner Good Samaritan Medical Center, Phoenix, AZ, USA 

Background: North American rattlesnake envenomations typically result in local tissue injury and hematotoxicity. Neurotoxicity is uncommon and most often associated with bites by the Mojave rattlesnake, Crotalus scutulatus. Neurotoxicity following bites by the Sidewinder rattlesnake, C. cerastes, has not been reported.

Hypothesis: This is the first reported case of a Sidewinder envenomation resulting in neurotoxicity.

Methods: This is a case report of a 56 y/o man who was bitten on the right foot through a leather boot by a snake he described as having horns and exhibiting sideways movement. Two independent herpetologists confirmed the species to be C. cerastes by photo. The patient developed painful right-sided paresthesias and weakness progressing from the toes to the distal thigh. Physical exam three hours after the bite revealed ecchymosis at the bite site, decreased sensation in the right foot, mild weakness of the leg, and pronounced fasciculations of the anterior thigh musculature. Fasciculations progressed to involuntary contractions of the large muscles of the thigh. Laboratory studies revealed normal platelets, protime, fibrinogen, CPK and electrolytes throughout hospitalization. Antivenom was withheld based on unclear benefit for treatment of neurotoxicity. Over the next 48-72 hours symptoms progressed to include right arm, bilateral ptosis, and respiratory muscle weakness with 3 consecutive worsening negative inspiratory force measurements. Respiratory failure did not occur. He was anorexic and unable to walk independently. On day 5 he had improved enough to ambulate with a walker and was discharged. On 10-day follow-up, he reported continued but improved paresthesias, right sided weakness, atypical chest pain, poor appetite and new abdominal cramping. The patient was lost to follow up.

Results: The patient developed prolonged neurotoxicity without hematotoxicity or significant local tissue injury.

Discussion: Crotalus cerastes has not previously been implicated in the development of neurotoxic venom effects.

Conclusion: Neurotoxicity may develop following bites by the Sidewinder rattlesnake. It is unknown whether such effects would be prevented or reversed with antivenom.


20. Clinical Presentations with Different Glyphosate Containing Herbicides

Bosak AB1, Heise CW1, Levine M1,2, Graeme KA1
1Banner Good Samaritan Medical Center, Phoenix, AZ, USA; 2University of Southern California, Los Angeles, CA, USA

Background: Glyphosate is a commonly used herbicide associated with toxicity and death following large ingestions. Surfactants are implicated as the primary contributor.

Hypothesis: Glyphosate herbicide preparations contain various surfactants and salts which may lead to diverse clinical toxicity.

Methods: (Case-1) A 50 y/o man presented with N/V, abdominal pain, and somnolence 12 hours after ingesting 6.5 ounces of concentrated Roundup® containing 50.2% glyphosate as isopropylamine salt. Labs revealed metabolic acidosis (AG 40), lactate 9.2mmol/L, creatinine 2.8 mg/dL, potassium 4.5 mmol/L, WBC 41.5 K/mm3, and lipase 2,528 IU/L. Electrocardiogram QRS was 142 ms and responsive to bicarbonate. Patient became hypoxic and hypotensive requiring intubation, vasopressors, and CVVHD for anuric renal failure.  A CT scan showed multiple loops of dilated small bowel and possible pneumatosis with negative exploratory surgery on day 1. On day 5 he had peritonitis prompting the resection of the ischemic terminal ileum and cecum. Over the next 3 weeks he developed recurrent GI bleeds unrelated to surgery, remained anuric on CVVHD and ventilator dependent. (Case-2) A 48 y/o landscaper ingested 250-350 mL of concentrated glyphosate herbicide and presented with AMS and N/V. Labs showed WBC 21.2 K/mm3, potassium 7.1 mmol/L, metabolic acidosis (AG 10), and creatinine of 1.2 mg/dL (0.60 baseline). Electrocardiogram QRS was 192 ms with wide complex rhythm, bradycardia, and responsive to bicarbonate. Persistent hyperkalemia continued despite medical treatment requiring high flux hemodialysis. He required ventilator support for hypoxia and had a negative EGD.

Results: Case-1 developed anuric renal failure, QRS prolongation, ischemic bowel and recurrent GI bleeds. Care was withdrawn on day 25. Case-2 had persistent hyperkalemia, non-anuric renal failure, QRS prolongation, and bradycardia.

Discussion: We hypothesize Case-2 ingested a potassium salt preparation due to the persistent hyperkalemia that only resolved with hemodialysis. QRS prolongation was present in both cases responsive to bicarbonate.

Conclusion: Glyphosate containing herbicides may have diverse clinical toxicity depending on surfactant and salt preparation.