Josh J. Wang

Sedating the Agitated Patient: A Moving Target?

1. Peterson DC, Martin-Gill C, Guyette FX, et al. Outcomes of medical emergencies on commercial airline flights. N Engl J Med. 2013;368:2075-2083. 2. Health Canada. Units of measurement for blood glucose meters sold in Canada. Health Canada, 2010. Available at: http://www.hc-sc.gc.ca/ dhp-mps/md-im/activit/announce-annonce/md_notice_glucose meters_im_avis_glycometres-eng.php. Accessed November 4, 2016.

Letter in reply to Levine et al. Re: hypoglycemia and lactic acidosis outperform King’s College criteria for predicting death or transplant in acetaminophen toxic patients

three-bag IV NAC was initially studied [2]. Based on these results the author report that the absence of unexpected hepatotoxicity may be due to the low risk cohort included in this study. It is important to remember that the three-bag method was designed to match the amount of NAPQI formed from APAP with the amount of glutathione provided by NAC. The author’s proposed dosing regimen would not succeed in providing the needed amount of glutathione early on in a patient more severely poisoned with APAP. Second, the authors report a lower rate of adverse reactions with the two-bag protocol compared to the three bad protocol. In Table 2, the authors list a higher incidence of anaphylactoid reactions (14% vs 5%) and higher use of antiallergy medication (11% vs 4%) in the three bag protocol compared to the two bag protocol. However the definition of anaphylactoid reaction is not stated clearly and there is a risk for bias and misreporting as the data for the three-bag control group were collected retrospectively while the data for the two-bag group was collected prospectively. More importantly, even if the rate of adverse reactions is truly lower in the two-bag protocol, this is not a sufficient reason to abandon the three-bag protocol. Anaphylactoid reactions or antiallergy medication administration rarely change care plans, length of stay or clinical outcomes. Finally, the time to reaction in Table 2 was analyzed on a log scale. Log transformation helps normalize widely spread out data. While it serves no purpose with events ranging from 1 to 4 h, the use of log analysis may be concerning for incongruent or unreported results. In conclusion, this paper is underpowered both for safety and efficacy, included patients with low risk ingestions and suffers from methodological limitations. Most importantly this paper should not be used to change practice until it can be proved to be safe for patients with acetaminophen concentrations well above the Rumack Matthew line of 993lmol/L (200mcg/mL) at 4 h post ingestion. Disclosure statement

Balancing the risks and benefits of medication organizers: reply to Wang et al.

We read with interest the article by Wang et al. that states higher odds of unintentional pediatric ingestions in households having a medication organizer [1]. Although we laud the authors’ efforts to investigate factors responsible for unintentional ingestion of pharmaceuticals, a number of methodological issues in their study shed doubt on their conclusions. One concern, as the authors rightly noted, is that the use of medication organizers is a well-proven method for increasing medication adherence, and thus improving patient outcomes [2,3]. Homes and caregivers that require a medication organizer may vary in important aspects that impact the ability of a child to accidentally ingest medication. Although the authors controlled for age, grandparents, and number of caregivers in the home, most families with children under six years of age do not require a medication organizer. Therefore, the case group may have confounding factors such as polypharmacy or cognitive impairments that were not accounted for. Indeed, the greater number of previous encounters with poison control observed in the case group may point to significant factors in overall home safety. Recall bias may have also caused caretakers in the case group to report more frequent use of medication organizers to avoid the perception of being negligent. Furthermore, various medication organizers have different levels of packaging safety. The study did not specify between blister-packs, pill-minders with child locks, or other child-proof methods. If the authors were able to differentiate a safety profile of the different pill organizers, this would provide valuable information on the risk associated with various devices. Finally, we note that only 40 (63%) of the patients in the case group actually ingested pills from the medication organizer. If the authors are inferring that ingestions were due to the medication organizer, then the ingestions from the medication organizer should be 100%, if indeed the medication organizer is the cause. Given this, it does not make sense to keep the ingestions that were not in the medication organizer in the case group. By removing the patients that did not ingest from the medication organizer, the odds ratio becomes non-significant (OR 1⁄41.22, 95%CI: 0.80–1.87). Given these shortcomings, to associate medication organizers with higher odds of unintentional pediatric ingestion is premature, and may tarnish an effective low-cost tool to increase medication compliance. We would thus encourage further exploration of other factors that may be associated with the need for a medication organizer, given the important role medication organizers play in medication adherence.

Extracorporeal Treatments In Poisonings From Four Non-Traditionally Dialyzed Toxins (Acetaminophen, Digoxin, Opioids, and Tricyclic Antidepressants): A Combined Single-Centre and National Study

The use of extracorporeal treatments (ECTRs) for poisonings with four non‐traditionally dialysed toxins (NTDTs) is increasing in the United States. This study evaluated whether ECTRs are prescribed for toxin removal or the treatment of other medical illnesses or complications. We performed a 2‐Phase retrospective analysis evaluating the main indication for ECTRs in patients with poisoning from a NTDT (defined for this study as acetaminophen, opioids, tricyclic antidepressants (TCAs) or digoxin) and ECTR. The first phase assessed all cases from a single site (New York City Poison Control Center) between the years 2000 and 2016, and the second phase surveyed all United States Poison Control Centers (PCCs). In Phase 1, demographics, toxin ingested and main indication for ECTR were extracted. In Phase 2, a query to the National Poison Data System using the a pragmatic subset of inclusion criteria from Phase 1 restricted to single toxin ingestions over a narrower time frame (2014‐2016) provided the cases for study. A structured online questionnaire was sent to all United States PCCs to request their database review regarding the indication for ECTR for their cases. In Phase 1, 92 cases met inclusion criteria. In Phase 2, 519 cases were screened and 425 met inclusion criteria. In Phase 1 91/92 (98.9%) and Phase 2 411/425 (96.7%), of extracorporeal treatments were used to treat underlying medical conditions or poisoning‐related complications rather than accelerate toxin removal. The increasing number of ECTRs reported in patients who ingested one of the four NTDTs thus appears to be for medical indications rather than attempts at toxin removal, a distinction that is important.

Letter in Reply to Arno et al. “Pharmacokinetics in Morbid Obesity: Influence of Two Bariatric Surgery Techniques on Paracetamol and Caffeine Metabolism”

To the Editor: We read with interest the article from Arno and colleagues [1]. In the evaluation of caffeinemetabolism, the authors accounted for various confounders among study participants, such as cytochrome P450 (CYP)-modulating medications, caffeinecontaining or xanthine-related products, and smoking status. However, we believe that by limiting the analysis to a dichotomous variable for smoking status, the authors neglected an important information needed to accurately compare study groups and interpret their results. Tobacco smoke is rich in polycyclic aromatic hydrocarbons, compounds known to be potent CYP1A2 inducers [2]. Caffeine metabolism, as measured by the paraxanthine/ caffeine ratio, is higher among smokers compared to nonsmokers. Furthermore, caffeine metabolism among smokers increases with the quantity of cigarettes consumed per day, indicating that CYP1A2 induction is proportional to the intensity of tobacco exposure. As such, assessing the quantity of cigarettes smoked per day for each study participant would help distinguish the effects of body mass index (BMI) on CYP activity. It is unclear from the reported data whether study participants identified as smokers during study enrolment were actively smoking at the time blood samples were taken and the quantity of tobacco exposure. Also as CYP1A2 induction quickly stops after smoking cessation, with a 20% reduction in caffeine clearance 2 days after interruption of smoking [3]. Any change in smoking pattern in the days prior to blood sampling likely affected the observed paraxanthine/caffeine ratio. Quantifying the smoking history and exposure more precisely might modify the results and the authors’ conclusions on the effect of obesity and bariatric surgery on drug metabolism. Future efforts to characterize CYP metabolism in obese patients may also benefit from examining the contribution of enzyme induction from weight-loss related behaviors such as vigorous physical exercise, cruciferous vegetable ingestion, and dietary grilled meat—all of which are subject to change preand post-bariatric surgery [4, 5].

Letter in response to “Efficacy and effectiveness of anti-digoxin antibodies in chronic digoxin poisonings from the DORA study (ATOM-1)”

We read with interest the study “Efficacy and effectiveness of anti-digoxin antibodies in chronic digoxin poisonings from the DORA study (ATOM-1)” by Chan et al.[1] The study provides further insight into the clinical course of patients with suspected chronic digoxin overdose treated with anti-digoxin Fab. A strength of the study was its emphasis that patients chronically taking digoxin are sick with a high incidence of comorbidities and are on multiple medications that affect their hemodynamic status. This fact fuels one of our concerns that the use of hyperkalemia and renal failure to define digoxin toxicity is problematic. These patients have many reasons for their hyperkalemia, including chronic kidney disease and medications such as spironolactone. While it is known that acute kidney injury or chronic disease can precipitate digoxin toxicity due to impaired elimination, digoxin is not intrinsically renal toxic. The majority of these patients also receive drugs that independently cause bradycardia, such as beta blockers. Thus, assessing the efficacy of anti-digoxin Fab in chronic digoxin overdose is complicated by the authors’ nonspecific definition of toxicity. With bradycardia and hyperkalemia possibly being due to medication effect or other etiologies, using these criteria to define the patient as toxic may be misleading. The lack of response to Fab may be due to the fact that these patients are not digoxin toxic at all. Another concern is the potential use of predistribution digoxin concentrations to define chronic overdose. It is recommended to obtain a serum concentration at least 6–8 h after the last dose.[2] The failure to note the time of last ingestion may have led to the erroneous inclusion of patients with therapeutic concentrations whose serum samples were collected prior to tissue redistribution. In addition, most patients who received anti-digoxin Fab in this study achieved undetectable serum concentrations of free digoxin. We dispute the assertion that free serum digoxin reflects lack of clinical toxicity in this setting. Although digoxin is found in the serum, its primary sites of tissue toxicity are the comparatively vast cardiac and skeletal muscle compartments. As such, free serum digoxin should be treated as a marker – not the cause – of clinical toxicity and any anti-digoxin Fab dosing strategy aimed solely at reducing the free serum concentration risks patient harm by under-dosing the antidote. Despite these issues, the study identified some patients with true chronic digoxin overdose. Gastrointestinal (GI) symptoms figure prominently in digoxin toxicity and we believe that patients whose GI complaints disappeared after antidotal therapy likely had true disease.[3] Additionally, some patients appeared to have dramatic improvements in heart rate following Fab therapy. Analyzing the clinical and laboratory outcomes in this subgroup may provide a more accurate estimate of antidotal efficacy.

Letter in response to “Lithium poisoning in the intensive care unit: predictive factors of severity and indications for extracorporeal toxin removal to improve outcome”

We read with interest the article from Vodovar and colleagues “Lithium poisoning in the intensive care unit: predictive factors of severity and indications for extracorporeal toxin removal to improve outcome.”[1] We commend the authors for using non-parametric tests and acknowledge that prospective validation of their threshold values is needed. However, we would like to raise our concerns with several aspects of their methodology and the reported results. First, the study population consisted of patients admitted to the intensive care unit and may not be representative of the larger spectrum of lithium (Li) poisonings without listing what criteria merit intensive care admission in this particular health care system. The authors themselves note that the use of the Hansen and Amdisen scale to categorize all ingestions is questionable considering that this scale was originally developed from a subject pool that consisted predominantly of patients with chronic lithium poisoning and has previously been shown to lack usefulness in acute ingestions.[2,3] The authors state that their decision to use a threshold value for creatinine of 200 lmol/L and a peak serum [Li]> 5.2mmol/L as indications for ECTR was derived from multiple logistic regression analysis. However, they do not explain why the selection of a single creatinine concentration was chosen instead of using the KDIGO criteria to determine renal function, as was reported in their Table 3. In addition to other shortcomings, use of a single creatinine concentration fails to distinguish whether the renal dysfunction is an evolving process due to the lithium intoxication or from an underlying chronic renal disease. Indeed, serum creatinine does not accurately reflect the GFR in patients in whom it is not in steady state. Use of KDIGO criteria, or at the very least successive creatinine concentrations would have facilitated the distinction. The relationship between persisting neurotoxicity and the length of time of central nervous system (CNS) exposure to toxic lithium concentrations is not discussed even though some authors have previously implied that this was the primary mechanism responsible for syndrome of irreversible lithium-effectuated neurotoxicity (SILENT).[4] While the authors’ study is notable for its detailed record of specific neurological impairments in the study population, the criteria for neurological recovery are neither listed nor discussed. Germane to the assessment of neurological recovery is each patient’s premorbid neurological function – which may be impaired despite therapeutic or subtherapeutic serum [Li] in the context of chronic toxicity and SILENT. Although the authors report in their results and discussion sections that only three chronic poisonings were included and had serum [Li] <4mmol/L and that renal failure should dictate the use of extracorporeal toxin removal (ECTR) in this population, it is not mentioned in the Abstract nor the Conclusion that the threshold value of 5.2mmol/L should only be used for acute and acute-on-chronic poisoning in patients with normal renal function. It would have been interesting to publish the receiver operator curve (ROC) analysis mentioned and see the performance of various threshold values reported in the literature including the latest EXTRIP recommendations.[5] The timing of ECTR with regards to the onset of toxicity and the length of time to return to a Li concentration under 1.0mmol/L is not mentioned. Moreover, as this study recruitment spanned two decades, treatment changes, intensive care unit (ICU) admission criteria and especially technological advances concerning ECTR techniques, membranes or availability, need to be discussed. Intermittent hemodialysis, sustained low-efficiency dialysis, and continuous renal replacement therapy all have unique toxin removal profiles and it would have been interesting if the authors reported the use and outcomes associated with specific ECTR modalities. In conclusion, while we laud the authors’ efforts to define and identify criteria for ECTR use in Li intoxications, the lack of information and precision surrounding the variables discussed make the results difficult to interpret and inform on management strategies with lithium-poisoned patients.

Letter in response to “Idarucizumab for dabigatran overdose”

We read with interest Peetermans et al.’s communication “Idarucizumab for dabigatran overdose”.[1] The report was interesting for the clinical team’s prudent use of gastric lavage, activated charcoal, and intravenous fluids to minimize absorption and maximize excretion of dabigatran and the detailed timeline of laboratory investigations. Our attention was also drawn to the indication for idarucizumab administration: central catheter insertion for emergency dialysis. While the proposed benefit of extracorporeal treatment in dabigatran overdose or bleeding is still anecdotal,[2] the risks of placing a large-bore catheter in an asymptomatic patient with dabigatran-induced coagulopathy and normal renal function – as in this published case – likely outweigh the benefits. Vlad et al. [3] had previously reported deliberate massive dabigatran overdose in two such patients. Both were managed expectantly and experienced no major bleeds, spontaneous normalization of plasma dabigatran to therapeutic concentrations by 32 to 50 h (<92 ng/ml, the median trough serum concentration based on data from the RE-LY trial [4]), and normalization of coagulation parameters shortly thereafter. Further tipping the scale towards increasing risk, five thrombotic events and eighteen deaths were reported in the 62 analyzed patients who received idarucizumab in REVERSE AD,[5] an ongoing prospective cohort study on the safety and efficacy of the drug, and several concerns have been raised regarding its true efficacy.[6] The authors refer to the case patient’s eligibility to receive idarucizumab as part of RE-VERSE AD. In reviewing that study’s published interim analysis, we note that her data were included in the results and informed those authors’ conclusion that idarucizumab safely reverses the anticoagulant activity of dabigatran. Because research into idarucizumab’s treatment effect on patient-oriented outcomes such as major bleeding, disability, and death is still in its early stages, presenting existing RE-VERSE AD data as a case report with a positive outcome may mislead readers to overestimate the evidence-base in favour of idarucizumab administration. Our most striking concern is the lack of transparency in the authors’ “Disclosure statement”. Based on previously released disclosure forms from RE-VERSE AD, five co-authors on this report have or have had financial ties to Boehringer Ingelheim, the manufacturer of both idarucizumab and dabigatran, which include direct employment, grant funding, speaking fees, and positions on the steering committee for idarucizumab. While a partial list of these associations is included under “Funding information”, these relationships constitute real conflicts of interest and potential commercial bias and warrant disclosure in the appropriate section as well. Readers must account for potential industry influence in this report before drawing their own conclusions on the use of idarucizumab in dabigatran overdose.