Faculty Bibliography

BACKGROUND: Prescription drug monitoring programs (PDMPs) have emerged as one tool to combat prescription drug misuse and diversion. New York State mandates that prescribers use its PDMP (called ISTOP) before prescribing controlled substances. We surveyed physicians to assess their experiences with mandatory PDMP use. METHODS: Electronic survey of attending physicians, from multiple clinical specialties, at one large urban academic medical center. RESULTS: Of 207 responding physicians, 89.4% had heard of ISTOP, and of those, 91.1% were registered users. 45.7% of respondents used the system once per week or more. There was significant negative feedback, with 40.4% of respondents describing ISTOP as "rarely" or "never helpful," and 39.4% describing it as "difficult" or "very difficult" to use. Physicians expressed frustration with the login process, the complexity of querying patients, and the lack of integration with electronic medical records. Only 83.1% knew that ISTOP use is mandated in almost all situations. A minority agreed with this mandate (44.2%); surgeons, males, and those who prescribe controlled substances at least once per week had significantly lower rates of agreement (22.6%, 36.2%, and 33.0%, respectively). The most common reasons for disagreement were: time burden, concerns about helpfulness, potential for under-treatment, and erosion of physician autonomy. Emergency physicians, who are largely exempt from the mandate, were the most likely to believe that ISTOP was helpful, yet the least likely to be registered users. 48.4% of non-emergency physicians reported perfect compliance with the mandate; surgeons and males reported significantly lower rates of perfect compliance (18.2% and 36.8%, respectively). CONCLUSIONS: This study offers a unique window into how one academic medical faculty has experienced New York's mandatory PDMP. Many respondents believe that ISTOP is cumbersome and generally unhelpful. Furthermore, many disagree with, and don't comply with, its mandatory use.

BACKGROUND: Although intravenous lipid emulsion (ILE) was first used to treat life-threatening local anesthetic (LA) toxicity, its use has expanded to include both non-local anesthetic (non-LA) poisoning and less severe manifestations of toxicity. A collaborative workgroup appraised the literature and provides evidence-based recommendations for the use of ILE in poisoning. METHODS: Following a systematic review of the literature, data were summarized in four publications: LA and non-LA poisoning efficacy, adverse effects, and analytical interferences. Twenty-two toxins or toxin categories and three clinical situations were selected for voting. Voting statements were proposed using a predetermined format. A two-round modified Delphi method was used to reach consensus on the voting statements. Disagreement was quantified using RAND/UCLA Appropriateness Method. RESULTS: For the management of cardiac arrest, we recommend using ILE with bupivacaine toxicity, while our recommendations are neutral regarding its use for all other toxins. For the management of life-threatening toxicity, (1) as first line therapy, we suggest not to use ILE with toxicity from amitriptyline, non-lipid soluble beta receptor antagonists, bupropion, calcium channel blockers, cocaine, diphenhydramine, lamotrigine, malathion but are neutral for other toxins, (2) as part of treatment modalities, we suggest using ILE in bupivacaine toxicity if other therapies fail, but are neutral for other toxins, (3) if other therapies fail, we recommend ILE for bupivacaine toxicity and we suggest using ILE for toxicity due to other LAs, amitriptyline, and bupropion, but our recommendations are neutral for all other toxins. In the treatment of non-life-threatening toxicity, recommendations are variable according to the balance of expected risks and benefits for each toxin. For LA-toxicity we suggest the use of Intralipid(R) 20% as it is the formulation the most often reported. There is no evidence to support a recommendation for the best formulation of ILE for non-LAs. The voting panel is neutral regarding ILE dosing and infusion duration due to insufficient data for non-LAs. All recommendations were based on very low quality of evidence. CONCLUSION: Clinical recommendations regarding the use of ILE in poisoning were only possible in a small number of scenarios and were based mainly on very low quality of evidence, balance of expected risks and benefits, adverse effects, laboratory interferences as well as related costs and resources. The workgroup emphasizes that dose-finding and controlled studies reflecting human poisoning scenarios are required to advance knowledge of limitations, indications, adverse effects, effectiveness, and best regimen for ILE treatment.

A recent analysis of the American Association of Poison Control Centers database, showed that poisonings from toxins not usually considered amenable to extracorporeal purification ("non-classic toxins" such as ethanol and tricyclic antidepressants) continue to be reported. This publication investigates factors that may explain these findings. Our results suggest that: 1) the relatively high absolute number of ECTR performed for non-classic toxins may simply reflect the large number of exposures to these toxins, 2) poisoning from another toxin may have been the reason for ECTR initiation in some exposures to non-classic toxins, 3) poisoning from non-classic toxins may receive ECTR for purposes other than toxin removal, and 4) the decisional threshold to initiate ECTR may be lower for non-classic toxins because of heightened toxicity.

Endogenous digoxin-like factors (EDLF) are compounds that promote natriuresis, inhibit sodium-potasium-ATPase, and cross-react with antibodies raised against cardiac glycosides. Although EDLF are well described in experimental animal models and a variety of clinical conditions in humans, no investigation previously evaluated an association between EDLF and hypothermia. We used two separate models to examine the presence of EDLF in hypothermic humans. The first study consisted of a case-controlled comparison of patients with unintentional environmental hypothermia. All patients with environmental hypothermia were prospectively enrolled if they were not taking cardiac glycosides and had no conditions previously associated with EDLF. A digoxin concentration obtained at presentation was compared to one obtained from presentation on patients from a demographically matched control group of normothermic patients. The second study enrolled a convenience sample of patients undergoing controlled cold cardioplegia for surgery. Once again, patients were excluded for conditions previously associated with EDLF, and were required to have a non-detectable preoperative digoxin concentration. Digoxin concentrations were subsequently compared in the pre- and post-operative periods. In addition to serving as their own controls, a comparison group of normothermic patients were added to help control for confounding factors associated with surgery under general anesthesia. Twenty-two patients with environmental hypothermia (mean temperature 91° Fahrenheit, range, 83.8-93.8° Fahrenheit) and 22 controls were included in the first study. Ten hypothermic patients had measurable digoxin concentrations, 6 of which were above 0.15 ng/mL. All control patients had digoxin concentrations of 0.00 ng/mL (p = 0.03, Fisher's exact test). The second study enrolled ten patients undergoing surgery with cold cardioplegia and ten control patients undergoing normothermic surgery under general anesthesia. None of the ten control patients had measurable digoxin concentrations on either pre or postoperative testing. Although all of the cold cardioplegia hypothermic patients had negative preoperative digoxin concentrations, one developed a postoperative concentration of 0.44 ng/mL. We conclude that hypothermia is associated with the presence of EDLF. An animal model and further human testing will be required to establish causation

Background: Adverse cardiovascular events (ACVE) complicate up to 16.9% of hospitalizations for acute drug overdose. We previously derived a risk prediction rule for ACVE in these patients with 97.1% negative predictive value. Objectives: Our aim was to internally validate the ACVE rule test characteristics. Methods: This prospective cohort study was conducted over 17 months (2012-14) at 2 urban teaching hospitals. Patients were adults with suspected acute drug overdose enrolled from the ED. The study outcome, ACVE, was defined as any of the following: myocardial injury (elevated troponin I), shock (requiring vasopressors), ventricular dysrhythmia (VT/VF/TdP), or cardiac arrest (pulseless requiring CPR). The rule included any of these 3 factors: (1) prior cardiac disease (CAD or CHF); (2) QTc > 500ms; (3) initial serum bicarbonate < 20 mmol/L. Sample size was predetermined in order to produce test characteristics with 95% CI widths <5%; we calculated the need to analyze 900 patients. Results: There were 1,457 patients screened, of whom 552 were excluded (185 non-drug overdose, 145 pediatrics, 111 missing data, 110 alternate diagnosis, 1 chronic), leaving 905 for analysis (mean age, 41 years; female, 44%; suicidal, 40%). ACVE occurred in 65 (7.2%, CI 5.6- 9.1) patients (myocardial injury, 44; shock, 31; dysrhythmia, 16; cardiac arrests, 17) and there were 16 (1.8%, CI 0.9-2.6) deaths. The multivariable model adjusting for the previously derived risk factors, controlling for age, confirmed the following independent predictors of ACVE: QTc >500 msec (OR 5.5, CI 2.8-10.9), bicarbonate <20 mmol/L (OR 2.7, CI 1.5-4.9), and prior cardiac disease (OR 39.5, CI 17.9-87). The validated prediction rule had 75.4% (CI 63.1-85.2) sensitivity, 82.3% specificity (CI 79.9-85.1), and 97.8% negative predictive value (CI 96.4- 98.7). The presence of 2 or more risk factors had 51.5% positive predictive value (CI 34.5-68.6). Conclusion: The risk prediction rule for ACVE in patients with acute drug overdose performed with slightly improved sensitivity and negative predictive value in the validation cohort. External validation in distinct patient populations and clinical settings remains warranted

Background: ASA poisoning remains a significant public health threat with upwards of 20,000 exposures annually in the US and morbidity/mortality rates of up to 25%. Objectives: In order to facilitate targeted treatment to lower these rates, we aimed to establish early predictors of severe in-hospital outcomes in ED patients presenting with ASA poisoning. Methods: This was a secondary data analysis of ASA overdoses from a prospective cohort study of suspected acute drug overdoses at two urban university teaching hospitals from 2009-2013. Patients were enrolled consecutively and were considered eligible for inclusion based on clinical suspicion of ASA ingestion. Children (<18) and alternate diagnoses were excluded. Demographics, clinical parameters, serum ASA concentrations, treatment modalities and death/admission rate were collected from the medical record. Severe outcome was defined as a composite occurrence of any of the following: acidemia (pH<7.3 or bicarbonate<16mEq/L), hemodialysis, or death. Results: 48 patients met inclusion criteria, with 43.8% male, median age 32, mean initial ASA concentration 28.1, and 10 (21%) classified as severe outcome. There were two deaths, neither of whom received hemodialysis. Patients were treated with sodium bicarbonate in onethird of cases, while 54.2% received activated charcoal and 64.6% were admitted. Univariate analysis indicated that age, respiratory rate (RR), creatinine, lactate, coma, and presence of co-ingestions were significantly associated with severe outcome, while ASA alone had no association. However, when adjusted for serum ASA concentration, only age (OR 1.02 per additional year, 95%CI 1.0-1.1), RR (1.09 per additional breath/min, 95%CI 1.03-1.15), creatinine (2.8 per additional mg/dL 95%CI 1.1-7.1), and co-ingestions (OR 6.4, 95%CI 2.3-17.8) were independent predictors of severe outcome. Conclusion: Age, RR, creatinine, and co-ingestions are predictive of severe outcome in ED patients with acute ASA poisoning, while serum ASA concentration alone is not. Despite the severity of these cases, only one-third received sodium bicarbonate, suggesting potential barriers to administration which require further study