Dexmedetomidine Overdose in the Perioperative Setting Victor SB Jorden, Robert M Pousman, Mary M Sanford, Per AJ Thorborg, and Michael P Hutchens Author information provided at the end of the text. Dr. Jorden is Associate Medical Director, Abbott Laboratories; Dr. Pousman is a paid investigator and speaker/consultant for Abbott Laboratories; Ms. Sanford and Dr. Thorborg are paid speaker/ consultants for Abbott Laboratories.
OBJECTIVE: To report 3 cases of accidental dexmedetomidine overdose in the perioperative setting and review the pathophysiology of α2-agonist overdose.
CASE SUMMARIES: Three patients accidentally received overdoses of dexmedetomidine, one intraoperatively (192 µg over 20 min) and 2 postoperatively (4 and 2 rather than 0.4 and 0.2 µg/kg/h; 0.5 µg/kg/min rather than 0.5 µg/kg/h). Hemodynamic parameters remained stable for all 3 patients. The most notable sign was oversedation diagnosed either clinically or using a bispectral index monitor; Naranjo criteria suggest possible or probable association of the reactions with dexmedetomidine. In all 3 cases, oversedation resolved within one hour of drug discontinuation. There were no other sequelae, and the remainder of each patient’s hospital course was unremarkable.
DISCUSSION: As of this writing, dexmedetomidine dosing in excess of the label recommendation has been reported, but accidental dexmedetomidine overdose in clinical practice has not been described. Excessive levels of sedation were the only significant finding in all 3 patients. Dexmedetomidine’s short redistribution half-life of 6 minutes should lead to rapid resolution of oversedation induced by overdoses if the overall duration of infusion is short (≤8 h). While the patients reported here were hemodynamically stable, dexmedetomidine may engender significant hemodynamic changes either because of sympatholysis at normal doses or vasoconstriction at higher than recommended doses. The absence of a significant hypertensive response to high dexmedetomidine concentrations suggests that dexmedetomidine-induced hypertension may be multifactorial, not simply related to plasma drug concentrations.
CONCLUSIONS: Practitioners presented with dexmedetomidine overdose should be prepared to manage oversedation. While hemodynamic alterations may be seen with dexmedetomidine use, hypertension from high dexmedetomidine plasma concentrations is not a consistent response. Practitioners using dexmedetomidine should carefully note that dosing for this agent is described by the manufacturer in µg/kg/h, not µg/kg/min.
KEY WORDS: dexmedetomidine, overdose.
Ann Pharmacother 2004;38:803-7. Published Online, 23 Mar 2004, www.theannals.com, DOI 10.1345/aph.1D376 Downloaded from aop.sagepub.com by guest on October 11, 2013 area burns. Hospitalization had been complicated by respiratory failure, pneumonia, sepsis, and renal failure. At the time of this surgery, the patient was still on renal replacement therapy, but was spontaneously ventilating via tracheostomy. While not indicated for use as an adjunct to general anesthesia, dexmedetomidine has been used for this purpose successfully and was incorporated into the anesthetic plan with the desired goal of reducing hypertensive and tachycardic events.5 A total intravenous anesthetic combining dexmedetomidine, propofol, and fentanyl was planned, with the goal of achieving an intraoperative bispectral index (BIS) of <60. To manage any pain associated with transfer, the patient would be administered a loading dose of dexmedetomidine 1 µg/kg before leaving the burn unit. His existing fentanyl infusion of 1.2 µg/kg/h was to be continued throughout the perioperative period. Upon initiation of the dexmedetomidine loading infusion, the patient’s vital signs were BP 122/83 mm Hg, mean arterial pressure (MAP) 96 mm Hg, HR 83 beats/min, arterial oxygen saturation (SpO2) 99%, and spontaneous RR 22 breaths/min. The loading dose was completed without incident, with the patient becoming unresponsive. A dexmedetomidine maintenance infusion of what was believed to be 0.4 µg/kg/h was initiated, and transport to the operating room was accomplished with stable vital signs. Approximately 3 minutes following uneventful transfer to the operating room (6 min after loading dose completion), the anesthesia team recognized that most of the contents of the 60-mL dexmedetomidine syringe (200 µg) had been administered. The actual dose given was 192 µg, representing approximately 2.6 times the prescribed dose of drug for the 20 minutes the patient had been receiving dexmedetomidine (loading dose 1 µg/kg × 68 kg, plus maintenance dose 0.4 µg/kg/h × 68 kg × 1/10 h = 68 mg + 2.7 µg = a prescribed dose of 70.7 µg for the time he was receiving dexmedetomidine). The patient’s vital signs at this time were BP 124/52 mm Hg, MAP 76 mm Hg, and HR 84 beats/min; the BIS was 22–30. The dexmedetomidine infusion was stopped for 20 minutes as the BIS gradually increased to >50; a maintenance infusion was then started at 0.2 µg/kg/h. The decision was made to proceed with surgery since the BIS had returned to an acceptable level and the patient’s hemodynamic status was stable. During the surgery, systolic BP dipped to 90 mm Hg, with HR ranging from 48 to 54 beats/min. Intraoperative management included bolus doses of colloid and crystalloid totaling 1500 mL, coupled with incremental doses of phenylephrine totaling 400 µg. Estimated blood loss was 500 mL. Dexmedetomidine was discontinued approximately 10 minutes prior to the completion of surgery. Vital signs upon return to the ICU were BP 114/84 mm Hg, MAP 94 mm Hg, HR 78 beats/min, RR 16 breaths/min, and SpO2 98%. The patient became completely awake and appropriate within 3 hours of arrival to the ICU, and the duration of his hospital stay was uneventful. CASE 2 A 51-year-old, 95-kg white man with a history of atherosclerotic cardiovascular disease was scheduled for emergency bypass grafting of 4 coronary arteries and mitral valve replacement. Earlier that day, the patient had sustained an acute inferior wall myocardial infarction, and management with tenecteplase had proven unsuccessful. Emergent cardiac catheterization revealed critical triple vessel disease, moderate left-ventricular dysfunction (ejection fraction 36%), and moderate mitral regurgitation. Following an unremarkable intraoperative course, the mechanically ventilated patient was admitted to the cardiothoracic ICU at 1900. Receiving nitroglycerin 33 µg/min and sodium nitroprusside 0.25 µg/kg/min, his vital signs at 1915 were AV paced at 90 beats/min, BP 141/79 mm Hg, MAP 100 mm Hg, pulmonary artery pressure 58/32 mm Hg, right atrial pressure 17 mm Hg, cardiac output 4.47 L/min, and systemic vascular resistance 1145 dyne•sec•cm–5 . A patient-controlled analgesia system was initiated with morphine (basal rate 1 mg/h, bolus 2 mg, lockout interval 10 min). Because of moderate chest tube drainage (150–250 mL/h), the decision was made to delay extubation until morning. At approximately 2300, the patient became agitated, attempting to sit up in bed and remove his endotracheal and chest tubes. A dexmedetomidine infusion was initiated at what was thought to be 0.4 µg/kg/h. The infusion of sodium nitroprusside was rapidly increased to 0.7 µg/kg/min, which, combined with the dexmedetomidine, kept the patient’s MAP just below 100 mm Hg. Thirty minutes later, the patient was calm, cooperative, and denying pain, with an MAP of approximately 95 and HR of 90 beats/min with normal sinus rhythm. At 0100, he was noted to be more difficult to arouse than earlier, and the dexmedetomidine infusion was titrated down to what was believed to be 0.2 µg/kg/h. The patient continued receiving this dose until 0715, with stable vital signs throughout this period. He had received a total of 3 units of packed red blood cells and 2 units of fresh frozen plasma for anemia and coagulopathy. The pulmonary artery catheter was removed at 0400, and the nitroprusside infusion was discontinued at 0530. At 0715, the incoming nurse noted the patient to be unarousable and determined that a dosing error had occurred. Instead of receiving the prescribed doses of dexmedetomidine 0.4 and 0.2 µg/kg/h, a tenfold error had resulted in the patient receiving 4 and 2 µg/kg/h, respectively. Vital signs and other hemodynamic parameters were stable as before, without the need for vasopressors, vasodilators, or pacing. The dexmedetomidine infusion was decreased to 0.7 µg/kg/h, and the basal morphine infusion was discontinued. By 0800, the patient was awake and calm, and the dexmedetomidine infusion was discontinued. The femoral sheath was removed and, following the mandated 4-hour supine period, extubation was performed uneventfully at 1200. The remainder of the patient’s recovery was uneventful.
CASE 3 A 29-year-old, 214-kg white man was scheduled for his third debridement of Fornier’s gangrene. He had been admitted with poorly controlled diabetes mellitus and a suppurating wound of the right thigh 9 days earlier. The patient had remained intubated and mechanically ventilated since admission, but the attending intensivist felt that his overall status had improved sufficiently that extubation may be reasonably considered. The postoperative plan included switching the patient’s sedative from propofol to dexmedetomidine, with the expectation that dexmedetomidine’s respiratory stability and opioid-sparing quality would facilitate weaning from mechanical ventilation for this morbidly obese patient. Following an uneventful surgery, the patient returned to the ICU at 2300 intubated and mechanically ventilated; propofol 22 µg/kg/min was being infused. At 2400, a dexmedetomidine infusion was initiated at what was thought to be 0.5 µg/kg/h without a loading dose. The propofol was titrated to discontinuation over the subsequent 2 hours, and the patient had an uneventful night during which he received a total of 8.0 mg of morphine for 2 episodes of restlessness. At 0700 the following morning, the incoming nurse noted the patient to be unresponsive to sternal rub and other stimulation. No change was noted in his vital signs, which had been stable since 2400 with BP 98–128/35–55 and HR 85–95 beats/min. Inspection of the intravenous infusion pump revealed that the dexmedetomidine infusion was running at 0.5 µg/kg/min rather than the ordered 0.5 µg/kg/h, a 60 times rate increase. An immediate review of the pump contents indicated that the patient had received 1600 µg more dexmedetomidine than would have been expected from the prescribed dose since initiation of the infusion. This overdose translates into a 60-fold increase in the infusion rate for approximately 15 minutes, although the actual time during which the excessive drug had been administered could not be confirmed. The dexmedetomidine infusion rate was immediately reduced to 0.5 µg/kg/h. Within the following hour, the patient became responsive and appropriate, with vital signs remaining within the same range as earlier. Extubation was accomplished uneventfully later that day, and the remainder of the patient’s recovery was unremarkable.
Discussion MEDICATION ERRORS IN CRITICAL CARE Medication errors are a common cause of morbidity and mortality.6 Recent data suggest such errors are commonplace in intensive care venues.7 Even with a high ratio of staff to patients in these areas, the risk of medication errors is similar to that in less acute areas of the hospital, probably because of the high number of interventions per patient 804 ■ The Annals of Pharmacotherapy ■ 2004 May, Volume 38 www.theannals.com VSB Jorden et al. seen in critical care.8 Medication errors in the critical care setting may have a profoundly adverse effect on patients, not only because of the patients’ pathophysiology, but also the low therapeutic indices of some drugs commonly used in critical care. In both the first and third cases presented here, the pump settings for the maintenance infusion were mistakenly predicated on a dexmedetomidine dose of 0.2–0.7 µg/kg/min instead of the correct 0.2–0.7 µg/kg/h.
These errors may be attributable to the vast majority of critical care agents being described by per minute instead of per hour calculations. In case 2, a “naked decimal” (decimal not preceded by a 0) on the available instructional materials led to miscalculating the prescribed dose by a factor of 10 times.9 α2-ADRENORECEPTOR OVERDOSE The class of α2-adrenoreceptor agonists has been described as having a “remarkably wide safety margin.”10 In 2001, a major national database of poisonings indicated that, of 4721 clonidine overdoses, only 7 proved fatal (0.15%) and only 151 led to major morbidity (3.2%).11 The safety of α2-agonists is further supported by a growing literature of pediatric clonidine poisonings, as that medication has attracted recent attention in the management of attention deficit/hyperactivity disorder. A review of >10 000 pediatric poisoning cases included only one death, and 2% of the victims were noted as having “major” toxic effects.12 However, most children lack the cardiovascular comorbidities that make clonidine overdoses especially dangerous. The greatest overdose of clonidine reported has been 50 mg in a 5-year-old child who survived without sequelae.13 Adverse effects of clonidine overdoses have been reported to include hypotension (<7 mg/day), hypertension (>7 mg/day), bradycardia, and mental deterioration; less frequently, myocardial infarction, seizures, tachycardia, and cardiac conduction defects have been observed.14,15 Management has included supportive therapy, administration of α2-adrenoreceptor blocking agents, and hemoperfusion in patients with poor renal function.16-18 Dexmedetomidine notably differs from clonidine in being an order of magnitude more specific for the α2- than the α1-receptor. Because of this difference, toxicities specific to dexmedetomidine overdose cannot necessarily be simply extrapolated from the clonidine literature. Dexmedetomidine administration in excess of the package insert recommendation has been reported previously, but only as part of a study protocol. Loading infusions of up to 2 µg/kg were administered to healthy subjects, double the recommended loading dose under the existing dosing recommendations.19,20 Furthermore, these doses were administered over 2 minutes, one-fifth the minimum time recommended in the package insert; no severe adverse events were noted. Ebert et al.21 administered computercontrolled dexmedetomidine infusions in a study of maximally tolerated doses. Eight of the 10 healthy volunteers tolerated plasma concentrations of at least 4 times greater than those expected with infusions at the recommended maximum dosage. None of these subjects required cardiovascular support, but they did show a biphasic hemodynamic response to dexmedetomidine, with blood pressure reductions at low plasma concentrations and hypertension at high plasma concentrations. In a clinical trial, dexmedetomidine was administered in loading doses of up to 2 µg/kg and maintenance infusion doses of up to 2.5 µg/kg/h to ICU subjects. The investigators found that “modest falls in arterial pressure, heart rate and cardiac output were seen….Adverse cardiovascular events were nearly all confined to the initial loading dose period….”
3 CASE SERIES ANALYSIS This series is the first report in the medical literature of dexmedetomidine overdose attributable to human error. All 3 patients probably had plasma concentrations significantly in excess of those normally seen with dexmedetomidine use. Patient 1 received 60 times the prescribed dose for approximately 6 minutes, patient 2 received 10 times the prescribed dose for approximately 8 hours, and patient 3 received 60 times the prescribed dose for approximately 15 minutes. Dexmedetomidine may have a significant effect on hemodynamics. This effect has been characterized as the combination of 2 competing actions.22 A central sympatholytic action causing vasodilation usually predominates at plasma concentrations associated with the recommended maintenance infusion rates. The resulting mild to moderate hypotension has been well described clinically, occurring with an incidence of 30% compared with 13% in subjects not receiving dexmedetomidine. Dexmedetomidine also appears to have a vasoconstricting effect that has been attributed to direct activity at peripheral vascular α2b-receptors, contrasted with α2a-receptors in the central nervous system that mediate dexmedetomidine’s sedative and analgesic effects. The incidence of hypertension in preapproval clinical studies was 16%, which is similar to the rate of 18% observed in subjects receiving placebo with rescue sedation. However, while the incidence of hypertension was similar between the treatment and control groups, the temporal context for hypertension may have differed significantly. In one large postoperative study, dexmedetomidine subjects became hypertensive almost exclusively during the loading phase.23 This temporal relationship supports the concept of supratherapeutic dexmedetomidine concentrations, such as those associated with loading, causing vasoconstriction that dominates the centrally induced vasodilatation. The lack of hypertensive events outside of the loading phase suggests that subjects receiving dexmedetomidine may have had reduced anxiety, pain, and exaggerated sympathetic nervous system activity commonly seen in critically ill postoperative patients. Earlier studies have documented dexmedetomidine’s ability to reduce plasma norepinephrine concentrations in the immediate postoperative period.24 No patient in this case series suffered from marked alterations in blood pressure. Patient 1 required a small dose of Dexmedetomidine Overdose www.theannals.com The Annals of Pharmacotherapy ■ 2004 May, Volume 38 ■ 805 phenylephrine intraoperatively, but this need did not emerge until surgery was underway and was likely related to the ongoing blood loss. Additionally, the timing of the need for the pressor may have correlated with reduction of dexmedetomidine plasma concentrations into the normal range, making mild hypotension an adverse effect essentially resulting from routine use. Mild hypertension occurred in case 2, requiring a minimal dose of sodium nitroprusside. However, the increase in blood pressure necessitating nitroprusside manifested when the patient was awakening for the first time following surgery, and the nitroprusside became unnecessary more than one hour prior to discontinuation of dexmedetomidine. These 2 circumstances strongly suggest that the mild hypertension in this case was more likely the result of postsurgical emergence from anesthesia than a specific dexmedetomidine effect. The lack of significant hypertensive response to dexmedetomidine overdose is an interesting finding in this case series. While hypertension in association with loading doses of dexmedetomidine has been well documented, this phenomenon occurs in <20% of all patients. We theorize that unrecognized factors may enable or exaggerate the hypertensive response to high plasma concentrations of the drug, and these factors were absent in our patients. Such enabling factors may include drug– drug interactions or unusual baseline levels of sympathetic tone. This theory is supported by the fact that some patients may actually react with hypotension during the loading phase.3 Bradycardia has been documented in patients receiving dexmedetomidine, especially those with high vagal tone.1 Several mechanisms may be causative, including central sympatholysis and a peripheral ganglionic blocking effect.25 In this case series, mild sinus bradycardia was reported only in case 1; this may have been due in part to the use of phenylephrine, an α1-agonist vasoconstrictor that may induce a baroreceptor reflex. Patient 2 was originally paced following cardiac surgery, but was in normal sinus rhythm during the overdose and did not revert to a paced rhythm. In all 3 cases, the high doses of dexmedetomidine appeared to cause deep hypnosis: in case 1, the patient’s BIS decreased to the lower 20s and, in cases 2 and 3, the patients became unarousable. Analysis utilizing Naranjo criteria for adverse drug reactions suggests a probable (case 1) or possible (cases 2 and 3) association to dexmedetomidine.26 In all 3 cases, opioids had been administered concomitantly; it is therefore unclear whether dexmedetomidine could have engendered such a deep level of sedation from high serum concentrations alone. Dexmedetomidine has been demonstrated to not be a general anesthetic unto itself and should not be considered as a sole agent for intraoperative use.27 The deep hypnosis observed in these cases resolved within one hour of drug discontinuation, and this rapid resolution may be at least partially accounted for by dexmedetomidine’s redistribution half-life of 6 minutes.28 This brisk recovery following high doses may be extremely valuable in the differential diagnosis of unresponsiveness among sedated patients. However, the total infusion times in all of these patients were relatively short (<8 h). The rate of recovery following more prolonged infusions of high-dose dexmedetomidine is unknown. Summary This case series represents the first published report of accidental dexmedetomidine overdose. None of the 3 patients developed adverse effects other than deep hypnosis, an effect that rapidly corrected following discontinuation of drug infusion. The patients in this case series had no evidence of a significant hypertensive response to high dexmedetomidine concentrations, suggesting that dexmedetomidine-induced hypertension may be multifactorial and not simply related to plasma concentrations. Practitioners using dexmedetomidine should carefully note that dosing for this agent is described by the manufacturer in µg/kg/h, not µg/kg/min.
Victor SB Jorden MD MPH, Associate Medical Director, Abbott Laboratories, Inc., Abbott Park, IL; Clinical Assistant Professor, Department of Anesthesiology, The Chicago Medical School, Chicago, IL Robert M Pousman DO, Assistant Professor of Anesthesiology, Division of Critical Care and Perioperative Medicine, Vanderbilt University, Nashville, TN Mary M Sanford ARNP MSN ACNP, Nurse Practitioner for Critical Care, Catholic Medical Center, Manchester, NH Per AJ Thorborg MD PhD, Associate Professor, Department of Anesthesiology and Perioperative Medicine, Oregon Health Sciences University, Portland, OR Michael P Hutchens MD MA, Chief Resident, Department of Anesthesiology and Perioperative Medicine, Oregon Health Sciences University Reprints: Victor SB Jorden MD MPH, Department R440, Abbott Laboratories, 200 Abbott Park Dr., Abbott Park, IL 60046-6229, fax 847/935-7633, [email protected] References 1. Package insert. Dexmedetomidine (dexmedetomidine HCl). Abbott Park, IL: Abbott Laboratories, Inc., February 2001. 2. Martin E, Ramsay G, Mantz J, Sum-Ping STJ. The role of the α2- adrenoreceptor agonist dexmedetomidine in postsurgical sedation in the intensive care unit. J Intensive Care Med 2003;18:29-41. 3. Venn RM, Newman PJ, Grounds RM. A Phase II study to evaluate the efficacy of dexmedetomidine for sedation in the medical intensive care unit. Intensive Care Med 2003;29:201-7. 4. Arain SR, Ebert TJ. The efficacy, side effects, and recovery characteristics of dexmedetomidine versus propofol when used for intraoperative sedation. Anesth Analg 2002;95:461-6. 5. Jalonen J, Hynynen M, Kuitunen A, Heikkila H, Perttila J, Salmenpera M, et al. Dexmedetomidine as an anesthetic adjunct in coronary artery bypass grafting. Anesthesiology 1997;86:331-45. 6. Kohn LT, Corrigan JM, Donaldson MS, eds. To err is human, building a safer health system. Committee on Quality of Health Care in America, Institute of Medicine. Washington, DC: National Academies Press, 2000. 7. Van dem Bemt PM, Fijn R, van der Voort PH, Gossen AA, Egberts TC, Brouwers JR. Frequency and determinants of drug administration errors in the intensive care unit. Crit Care Med 2002;30:846-50. 8. Cullen DJ, Sweitzer BJ, Bates DW, Burdick E, Edmondson A, Leape LL. Preventable adverse drug events in hospitalized patients: a comparative study of intensive care and general units. Crit Care Med 1997;25:1289-97. 9. Karch AM, Karch FE. The naked decimal point. And eight other common errors that can be avoided. Am J Nurs 2001;101:22. 10. Kamibayashi T, Maze M. Clinical uses of alpha-2 agonists. Anesthesiology 2000;93:1345-9. 11. Litovitz TL, Klein-Schwartz W, Rodgers GC, Cobaugh DJ, Youniss J, Omslaer JC, et al. 2001 Annual Report of the American Association of Poison Control Centers Toxic E