Sedation in the PICU is vital for patient comfort and to

Transcription

1 Long-Term Dexmedetomidine Use nd Sfety Profile Among Criticlly Ill Children nd Neontes* Lest D. Whlen, MD; Jne L. Di Gennro, MD; Gretchen A. Irby, PhrmD; Ofer Yny, MD; Jerry J. Zimmermn, MD, PhD, FCCM Objective: To determine whether long-term dexmedetomidine dosing is ssocited with lower opioid nd benzodizepine use without risk of significnt hemodynmic chnges nd/or withdrwl. Design: Retrospective, observtionl study. Setting: PICU, crdiovsculr ICU, nd neontl ICU in single, tertiry cre, cdemic children s hospitl. Subjects: We included ll ptients less thn or equl to 21 yers old, who received dexmedetomidine for greter thn or equl to 72 hours from December 2008 to December 2010 resulting in 98-subject cohort. Interventions: None. Mesurement nd Min Results: The medin durtion of dexmedetomidine use ws 141 hours. A decrese in systolic blood pressure nd hert rte ws seen fter initition of dexmedetomidine. After dexmedetomidine ws discontinued, systolic blood pressure ws sttisticlly significntly higher thn bseline. Similrly, hert rte showed significnt increse from bseline following discontinution of dexmedetomidine. Strting dexmedetomidine ws not ssocited with significnt difference in the dosing of opites or benzodizepines. Comfort scores were significntly lower t 2 nd 72 hours of dexmedetomidine infusion. After stopping dexmedetomidine, the comfort score for ptients t 1 hour ws sttisticlly higher thn for ptients t cesstion of the infusion. Thirty percent of ptients who were tken off dexmedetomidine, whether wened or bruptly stopped, hd withdrwl symptoms nd scores recorded with gittion, tremor, nd decresed sleep being most prominent. Conclusions: Hemodynmic effects of dexmedetomidine did not limit long-term use in this diverse popultion. After the ddition of dexmedetomidine, opioid nd benzodizepine doses did *See lso p All uthors: Deprtment of Peditric Criticl Cre, Settle Children s Hospitl, Settle, WA. Dr. Zimmermn received grnt support from the Ntionl Institutes of Helth, roylties from Elsiever Publishing, served s council bord member of the Society of Criticl Cre Medicine, nd is employed by the Settle Children's Hospitl. The remining uthors hve disclosed tht they do not hve ny potentil conflicts of interest. For informtion regrding this rticle, E-mil: lest.whlen@gmil.com Copyright 2014 by the Society of Criticl Cre Medicine nd the World Federtion of Peditric Intensive nd Criticl Cre Societies DOI: /PCC not significntly esclte, nd ptients were more comfortble s evidenced by decresing comfort scores. Withdrwl from dexmedetomidine my be n issue nd mnifests s gittion, tremors, nd decresed sleep. (Peditr Crit Cre Med 2014; 15: ) Key Words: benzodizepines; criticlly ill; dexmedetomidine; opites; peditrics; prolonged infusion; withdrwl Sedtion in the PICU is vitl for ptient comfort nd to fcilitte cre delivery. The use of dexmedetomidine hydrochloride (Precedex; Hospir, Lke Forest, IL), selective α2-drenoreceptor gonist, hs incresed in peditrics s n djunct to trditionl sedtion regimens nd hs become incresingly used s primry sedtive. The therpeutic effects of dexmedetomidine re medited throughout the CNS: with the sedtive nd nxiolytic effects resulting from its ctivity in the locus ceruleus nd the nlgesic effects from the dorsl horn of the spinl cord (1 3). These ctions result in sedtion, nxiolysis, nd nlgesi coupled with miniml concern for respirtory depression (3). Becsue of its symptholytic ctivity, the most significnt dverse rections ssocited with dexmedetomidine use re hypotension nd brdycrdi (2). Despite its populrity in the peditric popultion, dexmedetomidine only hs Food nd Drug Administrtion pprovl for use in dults for up to 24 hours. Most dt concerning peditric use of dexmedetomidine remin centered on the use for less thn 72 hours (4 16). There re few reports nd cse studies with smll subject numbers describing the sfety of dexmedetomidine with verge use longer thn 72 hours (17 23), nd fewer tht highlights possible withdrwl nd rebound symptoms (17, 18, 23). Our originl dexmedetomidine policy t Settle Children s Hospitl, Settle, WA, limited its use to 24 hours. As comfort incresed with the use of the drug nd its known side effects, the use expnded beyond these confines. This shift in clinicl prctice nd experience with gittion fter discontinution of dexmedetomidine infusions prompted us to design this study to describe the hemodynmic vribles, comfort scores, opioid nd benzodizepine use, nd withdrwl symptoms during October 2014 Volume 15 Number 8

2 Feture Articles initition, fter 72 hours of use nd up to 48 hours fter discontinution of dexmedetomidine. MATERIALS AND METHODS Study Design nd Setting The study ws pproved by the Institutionl Review Bord of Settle Children s Hospitl, which wived the need for informed consent. We performed retrospective, observtionl study of ptients in the PICU, crdic ICU, nd neontl ICU t Settle Children s between December 1, 2008, nd December 1, Unit protocols t the time of this study included strting dose of 0.2 μg/kg/hr, mx dose of 0.7 μg/kg/hr, nd use for up to 48 hours unless pproved by the unit medicl director for longer durtion. Study Popultion We included ll ptients less thn 21 yers old who hd received continuous infusion of dexmedetomidine longer thn 72 hours. Exclusion criteri include greter thn or equl to21 yers, inbility to obtin clinicl informtion, or discontinution of dexmedetomidine for ny length of time during the first 72 hours. Ptients were identified using our informtion technology dtbse, discern nlysis, tht enbled us to identify ptients with phrmcy chrges for dexmedetomidine use longer thn 72 hours. Dt Collection Ech ptient s record ws serched for mjor dignosis nd surgicl procedures. Indiction for strting dexmedetomidine, initition nd discontinution times, initil nd mximum dose, durtion of use, nd the utiliztion of dexmedetomidine wen were recorded. Blood pressure nd hert rte (HR) were recorded for ech ptient t initition (considered bseline), 30 minutes nd 1, 2, nd 72 hours of use, s well s t discontinution, 30 minutes nd 1, 2, 4, 24, nd 48 hours fter discontinution. These intervls were chosen given the men distribution hlf-life of pproximtely 10 minutes, durtion of ction of 4 hours, nd the terminl elimintion hlf-life of 2.7 hours (13, 22). Given recent work in our institution on sedtion protocols by Deeter et l (24), we used our estblished Settle PICU Comfort Score to ssess depth of sedtion (Tble 1). This sedtion protocol ws used in both the PICU nd the crdic ICU (CICU), Tble 1. Sedtion Score Used to Assess nd Tret Ptient Level of Sedtion/Comfort Score Signs nd Symptoms 0 Neuromusculr blockde 1 Hevily sedted/unconscious 2 Sleepy/clm/rousble 3 Tering/restless/fussy 4 Anxious/crying/not clm 5 Agitted/crying/tense 6 Dngerous/restrints needed wheres the newborn ICU (NICU) hd n djusted regimen. The typicl progression of sedtion in the PICU nd CICU ws opioid infusion nd s needed intermittent benzodizepine dosing. If incresed sedtion ws needed, our sedtion protocol instructed the ddition of benzodizepine infusion. After this ddition, it ws physicin preference whether to dd dexmedetomidine infusion or to increse both opioid nd benzodizepine infusion doses. The NICU protocol ws similr with the exception tht benzodizepine infusions were not recommended. Dexmedetomidine infusions were noted to either be wened off or stopped bruptly. There ws not dexmedetomidine unit protocol or common prctice t the time of this study. In regrds to opioid nd benzodizepine wening, it ws unit prctice to stop the infusions bruptly if on for less thn 5 dys or wen dily by 20% of the mximum dily dose every dy if on for 5 10 dys. For those on for greter thn 10 dys, the opioid nd benzodizepine infusions were typiclly wened by 10% on lternting dys. However, there ws vribility bsed on provider preference nd ptient tolernce of wen. There ws no protocol in plce directing which gent should be wened or discontinued first. However, dexmedetomidine ws typiclly discontinued first, but there ws vribility mong providers. Withdrwl scores were ssessed using the Opioid nd Benzodizepine Withdrwl Score (25). Withdrwl scores were ssessed on ny ptient receiving ny form of opioid or benzodizepine for greter thn 5 dys or if there ws provider concern of withdrwl symptoms during the wening of sedtion. Dily totl opioid nd benzodizepine dose from the dy before dexmedetomidine initition to 48 hours fter dexmedetomidine discontinution ws recorded nd clculted s intrvenous morphine nd lorzepm equivlents. To ssess significnce of hemodynmic vribility, vsoctive mediction dosing ws recorded t initition, throughout dexmedetomidine dministrtion, nd fter discontinution. The use of tropine for brdycrdi or sline bolus for hypotension within 4 hours of dexmedetomidine initition ws recorded. The ddition of ny ntihypertensive within 48 hours fter dexmedetomidine discontinution nd the ddition of clonidine t ny time during dexmedetomidine use were lso noted. Sttisticl Anlysis Bsic descriptive sttistics for ptient chrcteristics re presented s proportions for ctegoricl vribles nd medin nd rnge for continuous vribles. Pired-smple t tests were used to estimte the men difference in blood pressure, HR, nd comfort score fter initition nd discontinution of dexmedetomidine t the vrious time points noted bove. Wilcoxon mtched-pirs signed-rnk tests were used to compre dily totl cumultive opite nd benzodizepine dosing. Sttisticl nlyses were performed using Excel nd Stt softwre, version 10 (SttCorp LP, College Sttion, TX). RESULTS Ptients who received dexmedetomidine for greter thn or equl to 72 hours numbered 114. After excluding 16 ptients (14 hd dexmedetomidine pused during the 72 hours, nd two ptients Peditric Criticl Cre Medicine 707

3 Whlen et l for whom we could not locte clinicl informtion), 98 remined for nlysis. Of these 98 ptients, 11 died while on dexmedetomidine. These ptients did not contribute dt pertining to discontinution of dexmedetomidine nd withdrwl symptoms. Medin ge ws 3.8 yers with nerly even gender distribution (Tble 2). Forty six of the ptients were treted in the PICU, 13 in the NICU, nd 39 in the CICU. Sixty-five percent hd undergone surgicl procedure, predominntly crdic surgery (Tble 3). Most children (92%) were strted on dexmedetomidine for incresed sedtion need. Medin durtion of dexmedetomidine use ws 141 hours (Tble 4). Eleven of 98 ptients died while on dexmedetomidine, nd upon review it ws felt tht dexmedetomidine ws not contributing fctor to these deths. Of these ptients, the medin length of dexmedetomidine infusion ws 336 hours (Tble 5). Of the remining 87 ptients, 52 (60%) were extubted while on dexmedetomidine, 30 (34%) remined intubted fter dexmedetomidine discontinution, three were never intubted, nd two hd pre-existing trcheostomies. Sixty-three (72%) of the 87 ptients were wened off dexmedetomidine while 24 (28%) hd infusions stopped bruptly. Wening ws not dictted s policy, nd the choice to wen off dexmedetomidine infusion versus stop bruptly ws mde by the ttending physicin, in conjunction with the clinicl phrmcist. Becuse of the lck of wening policy for dexmedetomidine, the wening regimen ws not stndrd nd vried widely. Strting dose of dexmedetomidine for ll ptients ws 0.2 μg/kg/hr nd in 90 ptients (92%) the dose incresed to the unit mximum 0.7 μg/ kg/hr during their course. Eighty-five ptients (87%) incresed to the unit mximum dose within the first 24 hours of dexmedetomidine infusion. Two ptients received infusions beyond 0.7 μg/kg/hr (1 nd 1.5 μg/kg/hr) both of whom were on dexmedetomidine for greter thn 30 dys. A sttisticlly significnt decrese in systolic blood pressure (SBP) from bseline ws seen t 1, 2, nd 72 hours fter initition of dexmedetomidine (p = 0.022, 0.004, 0.027, respectively) (Tble 6). Distolic blood pressure (DBP) did not chnge significntly with initition nd prolonged use. HR dropped significntly from bseline t 1, 2, nd 72 hours fter initition (p for ech comprison). Becuse of hypotension, seven of 98 ptients (7%) received sline bolus within 4 hours of strting the dexmedetomidine infusion. Atropine ws never Tble 2. Demogrphics of All Ptients Who Received Dexmedetomidine for Greter Thn 72 Hours Gender (men), n (%) 48 (49) Age (yr) 3.8 ( ) Weight (kg) 6.85 ( ) Surgery, n (%) 64 (65) Intubted t discontinution, n (%) Withdrwl scores used, n (%) 30 (34) 26 (30) Of the 87 surviving ptients. Dt expressed re medin nd rnge except where specified. Tble 3. Procedures Performed While Ptient Ws on Dexmedetomidine Procedure n (%) None 34 (35) Extrcorporel membrne oxygention cnnultion 14 (14) Crdic 4/14 (29) Noncrdic 10/14 (71) Crdic 27 (28) Generl 11 (11) Congenitl diphrgmtic herni 1 (1) Orthopedic 1 (1) Plstic surgery 2 (2) Solid orgn trnsplnttion 3 (3) Otolryngology 3 (3) Neurosurgery 2 (2) required. No ptients hd the dexmedetomidine infusion stopped becsue of hemodynmic instbility. Vsoctive infusions (including milrinone, epinephrine, dopmine, dobutmine, vsopressin, nitroprusside, nicrdipine, nd esmolol) were recorded during initition of dexmedetomidine, nd there ws no sttisticlly significnt chnge in these infusions. After cesstion of the dexmedetomidine infusion, SBP rose sttisticlly significntly higher t 1, 4, nd 24 hours fter discontinution (p = 0.027, 0.034, 0.028, respectively) (Tble 7). Discontinution of dexmedetomidine ws ssocited with higher DBP t 30 minutes nd 4 hours fter infusion (p = Tble 4. Dexmedetomidine Administrtion Indiction for strting dexmedetomidine Incresed sedtion needs, n (%) 90 (92) Fcilitte extubtion, n (%) 5 (5) Wen opites, n (%) 3 (3) Durtion (hr) 141 (72 2,472) Ptients wened off vs brupt 63 (64) discontinution, n (%) Mximum infusion dose (μg/kg/hr) 0.7 ( ) Durtion of intubtion before 116 (2 4,607) dexmedetomidine (hr) Durtion of intubtion fter 133 (0.3 1,710) dexmedetomidine (hr) Antihypertensive dded within 48 hr 9 (10) fter dexmedetomidine, n (%) Sline bolus within 4 hr, n (%) 7 (7) Atropine used within 72 hr, n (%) 0 (0) Dt expressed re medin nd rnge except where specified October 2014 Volume 15 Number 8

4 Feture Articles Tble 5. Demogrphics of Ptients Who Died on Dexmedetomidine Infusion Age/Rce Primry Dignosis Cuse of Deth Time on Dexmedetomidine (Hr) 5 mo/mle Severe combined immunodeficiency disorder, s/p bone mrrow trnsplnt Respirtory filure humn metpneumovirus 7 mo/femle Algille syndrome Liver filure yr/mle Hypoplstic left hert syndrome s/p Fontn 288 Hert filure d/femle Left congenitl diphrgmtic herni Respirtory filure ECMO 82 2 yr/femle Heptoblstom Respirtory filure virl illness 1 mo/femle Pertussis pneumoni Grm-negtive rod sepsis ECMO 38 wk/mle Unknown genetic disorder Pneumoni/sepsis ECMO d/femle Pulmonry hypoplsi Respirtory filure ECMO yr/mle Hunter syndrome Aortic vlve regurgittion wk/mle History totl nomlous pulmonry venous return, pulmonry hypertension Respirtory filure virl illness 16 d/mle Left congenitl diphrgmtic herni Respirtory filure ECMO 336 s/p = sttus post, ECMO = extrcorporel membrne oxygention nd 0.006). HR ws significntly higher t 30 minutes nd 2, 4, 24, nd 48 hours fter infusion (p = 0.01, 0.006, <0.001, 0.002, 0.03, respectively). No sttisticlly significnt chnge ws mde to vsoctive infusions fter stopping dexmedetomidine. Antihypertensives, excluding clonidine, were strted in nine of 87 ptients (10%) within 48 hours of discontinuing dexmedetomidine; however, six of those ptients hd crdic lesions tht would typiclly be treted with ntihypertensives. Fifteen of 87 ptients (17%) received clonidine t the time of dexmedetomidine discontinution. Five of those ptients (33%) received clonidine before initition of dexmedetomidine s this ws bseline mediction before hospitliztion, wheres the other 10 ptients (67%) hd clonidine dded to fcilitte tpering of dexmedetomidine nd to ttenute withdrwl symptoms. Our dt did not show significntly lower doses of opioids or benzodizepines fter initition of dexmedetomidine (Tble 8). Opioid use on the dy before dexmedetomidine initition ws sttisticlly significntly lower when compred with the dy of dexmedetomidine initition (p = 0.015). Opioid use 24 nd 48 hours fter discontinution of dexmedetomidine ws lso sttisticlly significntly lower when compred with dosing on the dy of discontinution (p = nd ) (Tble 9). The only sttisticlly significnt lowering in benzodizepine dose ws seen t 48 hours fter discontinution of dexmedetomidine (p = 0.022). The comfort scores were significntly lower t 2 nd 72 hours of dexmedetomidine infusion (p = 0.007, 0.001, respectively) (Tble 6). After stopping dexmedetomidine, the comfort score t 1 hour ws sttisticlly higher thn t cesstion of the infusion, indicting less sedted ptient (p = 0.030) (Tble 7). Once strted on dexmedetomidine infusion, 12 of 98 of the ptients (12%) required no increse in opioid dily dosing. Of the 86 ptients who required increse in their opioid dosing fter initition of dexmedetomidinedexmedetomidine, 20 (23%) reched their mximum dose of opioid on dy 1 of DEX Tble 6. Men Chnge From Initition of Dexmedetomidine for Pired Smples (95% CI) Time After Dexmedetomidine Strt Hert Rte (Bets/Min) Systolic Blood Pressure (mm Hg) Distolic Blood Pressure (mm Hg) Crdic Surgery 30 min 1.7 ( 4.7 to 1.2) 0.9 ( 3.1 to 1.3) 0.7 ( 1.5 to 2.8) 0.3 ( 0.6 to 0.004) 1 hr 5.8 ( 8.9 to 2.7) 3.3 ( 6.1 to 0.5) 0.7 ( 2.6 to 1.2) 0.2 ( 0.5 to 0.2) 2 hr 7.6 ( 11.1 to 4.0) 4.5 ( 7.5 to 1.5) 1.0 ( 3.2 to 1.2) 0.4 ( 0.7 to 0.1) 72 hr 16.1 ( 21.4 to 10.7) 3.7 ( 7.1 to 0.4) 1.8 ( 4.4 to 0.8) 0.6 ( 0.9 to 0.2) p < 0.05 (pired t-test). Peditric Criticl Cre Medicine 709

5 Whlen et l Tble 7. Men Chnge From Dexmedetomidine Discontinution in Pired Smples (95% CI, p vlue) Time After Stop of Dexmedetomidine Hert Rte (Bets/Min) Systolic Blood Pressure (mm Hg) Distolic Blood Pressure (mm Hg) Crdic Surgery 30 min 7.0 (1.7 to 12.2) 4.4 ( 0.6 to 9.4) 6.0 (1.9 to 10.0) 0.1 ( 0.04 to 0.3) 1 hr 2.7 ( 2.2 to 7.7) 4.9 (0.6 to 9.2) 1.3 ( 1.5 to 4.1) 0.2 (0.02 to 0.4) 2 hr 6.4 (1.9 to 10.9) 4.0 ( 0.6 to 8.6) 2.1 ( 1.3 to 5.5) 0.01 ( 0.2 to 0.2) 4 hr 8.8 (3.9 to 13.7) 4.4 (0.3 to 8.4) 4.1 (1.2 to 7.0) 0.02 ( 0.1 to 0.2) 24 hr 8.1 (3.1 to 13.1) 4.0 (0.5 to 7.6) 0.4 ( 3.1 to 3.9) 48 hr 5.8 (0.5 to 11.1) 1.8 ( 2.3 to 5.8) 0.2 ( 3.2 to 3.5) p < 0.05 (pired t test). Dshes indicte no dt. infusion, 18 (21%) on dy 2, 14 (16%) on dy 3, seven (8%) on dy 4, nd six (7%) on dy 5 of dexmedetomidine infusion. The medin increse from bseline opioid dose ws 33%. Nineteen (22%) reched their mximum dose of opioid on or fter dy 6 of dexmedetomidine with medin dose increse of 78% from bseline. The medin length of tretment for ptients who hd no increse in opioid fter dexmedetomidine infusion ws strted t 5 dys. For those who hd their mximum opioid dose on dys 2 5 of dexmedetomidine infusion, the medin length of tretment ws 6 dys. For the ptients who reched their mximum opioid dose on or fter dy 6 of dexmedetomidine infusion, the medin length of tretment ws 14 dys. Two ptients were wened off of ll opioids before discontinution of dexmedetomidine infusion, nd the medin length of tretment for this group ws 4.5 dys. Once off of dexmedetomidine infusion, 22 of 89 ptients (25%) required n increse in their dily opioid dose (14% dose increse), wheres 65 (66%) continued to wen off of opioids. Benzodizepine dosing ws ffected similr to opioid dosing for ptients on dexmedetomidine infusions, except tht more ptients were ble to wen off of benzodizepines before discontinution of dexmedetomidine infusions (14 of 89, 16%). Twenty-four percent of ptients (24 of 98) required no increse in benzodizepine dily dose fter initition of dexmedetomidine infusion. Of the 74 ptients who required increse in their benzodizepine dosing fter initition of dexmedetomidine, nine (12%) reched their mximum dose of benzodizepine on dy 1 of dexmedetomidine infusion, 12 (16%) on dy 2, six (8%) on dy 3, 13 (18%) on dy 4, nd nine (12%) on dy 5 of dexmedetomidine infusion. The medin increse from bseline benzodizepine dose ws 42%. Twenty-five (34%) reched their mximum dose of benzodizepine on or fter dy 6 of dexmedetomidine with medin increse of 61% from bseline. The medin length of tretment for ptients who hd no increse in benzodizepine fter dexmedetomidine infusion ws strted t 6 dys. For those Tble 8. Dily Morphine nd Lorzepm Equivlents Around the Time of Initition of Dexmedetomidine Dily Dose (mg/kg/d), Medin (IQR) Comprison With Dy Strt Dexmedetomidine, p Morphine equivlents Dy before dexmedetomidine 1.60 (0.80, 2.67) Dy strt dexmedetomidine 1.74 (1.01, 2.91) Dy 1 dexmedetomidine 1.73 (1.05, 2.65) Dy 3 dexmedetomidine 1.88 (1.10, 2.76) Lorzepm equivlents Dy before dexmedetomidine 0.65 (0.23, 1.44) Dy strt dexmedetomidine 0.75 (0.31, 1.26) Dy 1 dexmedetomidine 0.71 (0.29, 1.30) Dy 3 dexmedetomidine 0.89 (0.28, 1.52) IQR = interqurtile rnge. Wilcoxon mtched-pirs signed-rnk test. Dshes indicte it is the dy strted with no comprison October 2014 Volume 15 Number 8

6 Feture Articles Tble 9. Dily Morphine nd Lorzepm Equivlents Around the Time of Discontinution of Dexmedetomidine Dily Dose (mg/kg/d) Medin (IQR) Comprison With Dy Stop Dexmedetomidine p Morphine equivlents Dy stop dexmedetomidine 1.15 (0.60, 2.30) Dy 1 off dexmedetomidine 0.96 (0.30, 2.40) Dy 2 off dexmedetomidine 0.83 (0.30, 2.15) Lorzepm equivlents Dy stop dexmedetomidine 0.34 (0.10, 1.13) Dy 1 off dexmedetomidine 0.34 (0, 1.0) Dy 2 off dexmedetomidine 0.37 (0, 0.90) IQR = interqurtile rnge. Wilcoxon mtched-pirs signed-rnk test. Dshes indicte it is the dy strted with no comprison. who hd their mximum benzodizepine dose on dys 2 5 of dexmedetomidine infusion, the medin length of tretment ws 5 dys. For the ptients who reched their mximum opioid dose on or fter dy 6 of dexmedetomidine infusion, the medin length of tretment ws 13 dys. After discontinution of dexmedetomidine infusion 25 of 89 ptients (28%) required n increse in the dily benzodizepine dose (21% increse), wheres 49 (56%) continued to wen. Fourteen ptients (16%) Tble 10. Recorded on 26 Ptients Who Were Tken Off of Dexmedetomidine (Withdrwl Score Ws Bsed on the Opioid nd Benzodizepine Withdrwl Score) Withdrwl Signs nd Symptoms Number of Ptients Withdrwl scores, medin (rnge) 24 hr 11 (2 14) 48 hr 11 (0 16) Agittion 17 (65) Tremors 10 (38) Incresed secretions 7 (27) Emesis 3 (12) Dirrhe 5 (19) Decresed sleep 14 (54) Incresed sleep 4 (15) Incresed tone 4 (15) Diphoresis 2 (8) Incresed temperture 4 (15) Hllucintion 1 (4) Dt expressed re totl number nd percentge except when specified. were wened off of benzodizepine before dexmedetomidine infusion ws discontinued, nd the medin length of tretment for these ptients ws 5 dys. Of the 87 ptients who did not die, 26 (30%) hd withdrwl scores nd symptoms recorded, the other 61 ptients were not formlly ssessed for withdrwl symptoms (Tble 10). These ptients hd medin durtion of use of 178 hours when compred with tht of 123 hours for those without withdrwl scores recorded. Of the 26 ptients, 18 (69%) were wened off of dexmedetomidine. The most common withdrwl score t 24 nd 48 hours fter discontinution ws 11 (rnge, 0 16). Of ptients wened off dexmedetomidine, the most common withdrwl score t 24 hours ws 3 (2 14) nd t 48 hours ws 4 (0 12). Seventeen of 26 ptients (65%) hd gittion recorded s withdrwl symptom, 10 of 26 (38%) hd tremors, nd 14 of 26 (54%) exhibited decresed sleep. Ptients who exhibited withdrwl symptoms were treted with the increse in opioid or benzodizepine dosing or hd clonidine dded if dexmedetomidine withdrwl ws concern to the provider. DISCUSSION We report the use of dexmedetomidine infusion in criticlly ill infnts nd children t doses of μg/kg/hr for medin durtion of 141 hours. The mjor findings from our study include 1) prolonged infusions of dexmedetomidine cn be tolerted sfely nd re ssocited with miniml cliniclly significnt hemodynmic effects, 2) dexmedetomidine infusions my decrese the overll opioid nd benzodizepine burden, nd 3) prolonged infusions of dexmedetomidine my be ssocited with rebound tchycrdi, hypertension, nd withdrwl symptoms with rpid discontinution, suggesting need for wening of the infusion. Mny dult studies hve demonstrted the sfety of dexmedetomidine use longer thn 24 hours (26 29), nd limited peditric studies show similr sfety profile of long-term Peditric Criticl Cre Medicine 711

7 Whlen et l dexmedetomidine to tht of dults (8, 10, 12, 20, 22). Our study mirrors the findings of both the dult nd the peditric dt. Known hemodynmic effects, such s decrese in HR nd blood pressure, were noted on initition of dexmedetomidine. However, only seven ptients (7%) received norml sline bolus nd none needed djustment in their crdiovsculr medictions, suggesting tht the clinicl significnce of the hemodynmic chnges is miniml nd long-term infusion is well tolerted. This could be becuse of our voidnce of loding dose or the overll low infusion dose when compred with current prctice. Unlike other peditric studies (26, 30, 31), the only sttisticlly significnt decrese in opioid nd benzodizepine use ws fter discontinution of dexmedetomidine, suggesting n overll wen off of sedtive medictions. The mjority of our ptients did require n increse in opioid (88%) nd benzodizepine (76%) dosing, despite the initition of dexmedetomidine infusion. We rgue tht this could be nticipted on the dy of dexmedetomidine initition becuse most of our ptients were treted with dexmedetomidine becuse of incresed sedtion needs. Of the ptients who reched their mximum opioid nd benzodizepine doses fter dy 1 of dexmedetomidine infusion, 24 (of totl 41, 59%) hd n initil decrese in opioid nd benzodizepine need. We speculte tht the initil decline, then subsequent increse, could be becuse of tchyphylxis to either the trditionl sedtion regimen or to the dexmedetomidine infusion. Further suggesting this phenomenon re those ptients who reched their mximum opioid nd benzodizepine dose fter 6 dys of dexmedetomidine infusion s they were on dexmedetomidine twice s long s the ptients nd hd higher increse opioid nd benzodizepine dosing. With one qurter of our ptients requiring incresed opioid nd benzodizepine doses fter dexmedetomidine discontinution, this could suggest tht this is becuse of the lck of dexmedetomidine effect. This must be noted cutiously, however, s it my indicte tretment for withdrwl symptoms. Comfort scores were significntly lower with initition of dexmedetomidine, suggesting tht dexmedetomidine did id in incresing ptient comfort. It should be noted tht during our study, the verge mximum infusion dose ws 0.7 μg/kg/hr, which, when compred with more recent prctice, would be considered low dose. This my explin our lck of sttisticlly significnt chnges in opioid nd benzodizepine dily dosing, given tht there re dt supporting high-dose dexmedetomidine infusion when compred with low-dose when ttempting to decrese opioid burden (32). Of the previous peditric studies noted, the one by Czj nd Zimmermn (12) most closely resembles our study in number nd unit use of dexmedetomidine infusion. Interestingly, this study showed higher rte of hemodynmic side effects nd sttisticlly significnt decrese in opioid nd benzodizepine use with the initition of dexmedetomidine infusions when compred with our work. Possible resons for this difference could be the erlier dte of their study in generl, which my hve been ssocited with less experience with the drug nd less tolernce of the hemodynmic side effects, s noted in their discussion. Dexmedetomidine use nd the hemodynmic effects hd been well described by 2008, nd this could hve incresed comfort mong our providers to wit out the decrese in HR nd blood pressure. It is lso possible tht by design of our study, we selected ptients who better tolerted the drug, s they hd to be on the dexmedetomidine infusion for greter thn or equl to 72 hours, nd did not include study ptients who my hve hd more hemodynmic side effects cusing cesstion of the dexmedetomidine infusion erlier. Their study lso shows n impressive decrese of opioid nd benzodizepine burden for their ptients while on dexmedetomidine infusion (36% nd 42%, respectively), which ws not found in our study. There re some notble differences in our study tht my hve contributed to this discrepncy. First, during the period of time of the study of Czj nd Zimmermn (12), the use of dexmedetomidine in their unit ws specificlly used primrily with the gol to decrese opioid nd benzodizepine doses in ptients requiring high doses of benzodizepines nd/or opioids, cycling on nd off dexmedetomidine typiclly every 24 to 48 hours (J. J. Zimmermn, personl communiction, 2012). In ddition, 52% of the ptients in the study of Czj nd Zimmermn (12) were on dexmedetomidine to decrese longer cting sedtive gents to fcilitte extubtion s opposed to 5% in this study. Thus, in the study of Czj nd Zimmermn (12), the gol of using dexmedetomidine ws to decrese infusions of benzodizepines nd/or opioids, wheres during the timefrme of our study dexmedetomidine ws primrily dded erlier to the sedtion regimen to prevent significnt escltion of opioids nd benzodizepines without n explicit gend to wen these ctively. However, s noted bove, there is suggestion to support tht dexmedetomidine infusion cn ct synergisticlly nd decrese the overll burden of both of these drugs s there ws not sttisticlly significnt escltion of doses fter strting dexmedetomidine in our study. We did show n increse in opioid nd benzodizepine dily dose in pproximtely one qurter of our ptients once off dexmedetomidine infusion, similr to the study of Czj. Limited dt bout withdrwl nd rebound hypertension or tchycrdi fter dexmedetomidine use longer thn 72 hours. Peditric dt suggest trend towrd rebound nd withdrwl symptoms with incresed use (17, 19, 23, 25). Of the studies tht report no concerns for withdrwl symptoms (5, 7, 20 22, 31), two wened the ptients off dexmedetomidine (21, 22) nd one strted 30% of their ptients on clonidine (31), which could explin the lck of withdrwl symptoms. Our study rises concern for withdrwl nd rebound. Both SBP nd DBP incresed significntly fter discontinution of dexmedetomidinedexmedetomidine nd this increse persisted over the first 24 hours. HR lso significntly incresed over the first 48 hours fter discontinution. Yet, despite the sttisticlly significnt increse, tretment with either new ntihypertensive or clonidine ws prescribed for only 21% of our ptients. The ddition of clonidine ws specificlly for meliortion of withdrwl symptoms nd ws strted t the discretion of the provider. It is possible tht these ptients would hve experienced withdrwl or rebound symptoms if October 2014 Volume 15 Number 8

8 Feture Articles not strted on clonidine. Antihypertensive mediction, excluding clonidine, ws strted in nine ptients (10%) fter 36 hours of hypertension. In ll of these ptients, incresed sedtion ws triled for concern of discomfort cusing hypertension. Agin, it is worth noting tht six of these nine ptients would hve been strted on n ntihypertensive mediction given their crdic lesion nd not necessrily in response to hypertension. We cknowledge tht t the cesstion of dexmedetomidine infusion, ptients were presumbly improved from clinicl stndpoint leding to higher HR nd blood pressure reflected by less sedted ptient. We did find tht ptients hd incresed opioid use in 25% of the cses nd benzodizepine in 28% of the ptients. Whether this is becuse of incresed discomfort, withdrwl not specificlly noted or incresing BP nd HR from dexmedetomidine discontinution cn only be speculted. Despite this question, our findings suggest tht lthough rebound hypertension nd tchycrdi re seen with cesstion of long-term dexmedetomidine infusion, tretment with clonidine or other ntihypertensive is rrely required, further illustrting the sfety of long-term infusions. Our study confirmed the risk of withdrwl symptoms following long-term dexmedetomidine infusion. Of our ptients, 30% hd recorded withdrwl scores nd symptoms tht suggested withdrwl phenomenon from dexmedetomidine infusions. These ptients hd higher medin infusion durtion (178 vs 123 hr). Symptoms recorded were similr to those seen in previous cse reports nd included gittion, tremors, nd decresed sleep. The symptoms most commonly recorded re similr to delirium seen in ICU ptients. It ws our unit prctice to normlize the ptients schedule s much s possible when concerns for delirium were noted. This included schedule set by nursing nd child life, norml dytime nd nighttime ctivities s possible, including light djustment, music therpy, nd ctivity in room. Despite these efforts, it is impossible to sy with certinty tht these ptients were not experiencing degree of delirium. However, we speculte tht this question cn be rised with ny sedtion withdrwl evlution, not just dexmedetomidine withdrwl. We cknowledge tht we cnnot completely conclude tht the withdrwl symptoms were not from opioid or benzodizepine becuse the mjority of the ptients were on these infusions s well t time of dexmedetomidine discontinution. Our study suggests tht wening the dexmedetomidine infusion, which hppened in 69% of the ptients with withdrwl scores recorded, ws ssocited with lower withdrwl scores nd less use of rescue tretment. The retrospective nture of the study is susceptible to design flws nd bis. We re unble to differentite between the use of dexmedetomidine or other resons (eg, other sedtion nd ptient condition) s the cuse of hemodynmic nd sedtive effects or withdrwl symptoms. Despite being one of the lrger peditric studies, which ws not limited to just one subset of PICU ptients, it my not be generlizble to ll PICU ptients given other unit sedtion nd nlgesi prctices. Specificlly, this study ws done while our unit policies nd prctice were to keep the mximl dose of dexmedetomidine infusion t 0.7 μg/kg/hr. We cknowledge tht this is not the most recent prctice in either our unit or mny others round the country. We cn only speculte tht our findings of withdrwl phenomenon would be ugmented by higher doses of dexmedetomidine infusion, but this study cnnot spek specificlly to this. It is lso possible tht concerning hemodynmic chnges my be ugmented on higher dose of dexmedetomidine infusion. Our study recorded chnge from bseline hemodynmics s the prmeter of significnce. This is not idel becuse it did not recognize the ge pproprite hemodynmic prmeters, which ptients my hve devited from. Ptients were lso used s their own controls, limiting our bility to mke ny firm conclusions bout dexmedetomidine use s comprison with no dexmedetomidine use. We cknowledge tht there ws self-selection of ptients who tolerted dexmedetomidine well enough to be treted for greter thn or equl to 72 hours given tht our records show tht 459 ptients were treted with ny mount of dexmedetomidine during this 2-yer time period. Finlly, it ws not our routine unit prctice to record withdrwl scores during dexmedetomidine discontinution. Scores could lso reflect opioid or benzodizepine withdrwl. CONCLUSIONS We hve evluted the use of long-term dexmedetomidine dministrtion in diverse popultion of criticlly ill infnts nd children 0 21 yers old. On the bsis of our findings, we conclude tht dexmedetomidine is effective nd sfe for longterm use in the PICU. Attention to known dverse hemodynmic effects while on dexmedetomidine is necessry, but need for tretment is rre. Comfort nd sedtion is improved with dexmedetomidine, nd decrese in the overll opioid or benzodizepine burden is probbly given tht there ws not significnt escltion of benzodizepine or opioid dosing fter strting dexmedetomidine. Vigilnce for tchycrdi nd hypertension t dexmedetomidine discontinution is pproprite; however, these rrely require tretment. The most prominent withdrwl symptoms include gittion, tremors, nd decresed sleep nd my be meliorted by wening dexmedetomidine. REFERENCES 1. Tobis JD: Dexmedetomidine: Applictions in peditric criticl cre nd peditric nesthesiology. Peditr Crit Cre Med 2007; 8: Buck ML: Dexmedetomidine use in peditric intensive cre nd procedurl sedtion. J Peditr Phrmcol Ther 2010; 15: Guinter JR, Kristeller JL: Prolonged infusions of dexmedetomidine in criticlly ill ptients. Am J Helth Syst Phrm 2010; 67: Brton KP, Munoz R, Morell VO, et l: Dexmedetomidine s the primry sedtive during invsive procedures in infnts nd toddlers with congenitl hert disese. Peditr Crit Cre Med 2008; 9: Bejin S, Vlsek C, Nigro JJ, et l: Prolonged use of dexmedetomidine in the peditric crdiothorcic intensive cre unit. Crdiology in the Young 2009; 19: Berkenbosch JW: Dexmedetomidine nd peditric (crdic) criticl cre re we there yet? Peditr Crit Cre Med 2010; 11: Buck ML, Willson DF: Use of dexmedetomidine in the peditric intensive cre unit. Phrmcotherpy 2008; 28:51 57 Peditric Criticl Cre Medicine 713

9 Whlen et l 8. Crroll CL, Krieger D, Cmpbell M, et l: Use of dexmedetomidine for sedtion of children hospitlized in the intensive cre unit. J Hosp Med 2008; 3: Chrysostomou C, Zebllos T: Use of dexmedetomidine in peditric hert trnsplnt ptient. Peditr Crdiol 2005; 26: Chrysostomou C, Di Filippo S, Mnrique A, et l: Use of dexmedetomidine in children fter crdic nd thorcic surgery. Peditr Crit Cre Med 2006; 7: Chrysostomou C, De Toledo JS, Avolio T, et l: Dexmedetomidine use in peditric crdic intensive cre unit: Cn we use it in infnts fter crdic surgery? Peditr Crit Cre Med 2009; 10: Czj AS, Zimmermn JJ: The use of dexmedetomidine in criticlly ill children. Peditr Crit Cre Med 2009; 10: Diz SM, Rodrte A, Foley J, et l: Phrmcokinetics of dexmedetomidine in postsurgicl peditric intensive cre unit ptients: Preliminry study. Peditr Crit Cre Med 2007; 8: Hosokw K, Shime N, Kto Y, et l: Dexmedetomidine sedtion in children fter crdic surgery. Peditr Crit Cre Med 2010; 11: Tobis JD, Berkenbosch JW: Initil experience with dexmedetomidine in peditric-ged ptients. Peditr Anesth 2002; 12: Tobis JD, Berkenbosch JW, Russo P: Additionl experience with dexmedetomidine in peditric ptients. South Med J 2003; 96: Burbno NH, Otero AV, Berry DE, et l: Discontinution of prolonged infusions of dexmedetomidine in criticlly ill children with hert disese. Intensive Cre Med 2012; 38: Drnell C, Steiner J, Szmuk P, et l: Withdrwl from multiple sedtive gent therpy in n infnt: Is dexmedetomidine the cuse or the cure? Peditr Crit Cre Med 2010; 11:e1 e Enomoto Y, Kudo T, Sito T, et l: Prolonged use of dexmedetomidine in n infnt with respirtory filure following living donor liver trnsplnt. Peditr Anesth 2006; 16: Hmmer GB, Philip BM, Schroeder AR, et l: Prolonged infusion of dexmedetomidine for sedtion following trchel resection. Peditr Anesth 2005; 15: Reiter PD, Pietrs M, Dobyns EL: Prolonged dexmedetomidine infusions in criticlly ill infnts nd children. Indin Peditr 2009; 46: Wlker J, McCllum M, Fischer C, et l: Sedtion using dexmedetomidine in peditric burn ptients. J Burn Cre Res 2006; 27: Weber MD, Thmmsitboon S, Rosen DA: Acute discontinution syndrome from dexmedetomidine fter protrcted use in peditric ptient. Peditr Anesth 2008; 18: Deeter KH, King MA, Ridling D, et l: Successful implementtion of peditric sedtion protocol for mechniclly ventilted ptients. Peditr Crit Cre Med 2011; 39: Frnck LS, Nughton I, Winter I: Opioid nd benzodizepine withdrwl symptoms in peditric intensive cre ptients. Intensive Crit Cre Nurs 2004; 20: Riker RR, Shehbi Y, Bokech PM, et l: Dexmedetomidine vs midzolm for sedtion of criticlly ill ptients: A rndomized tril. JAMA 2009; 301: Shehbi Y, Ruettimnn U, Admson H, et l: Dexmedetomidine infusion for more thn 24 hours in criticlly ill ptients: Sedtive nd crdiovsculr effects. Intensive Cre Med 2004; 30: Ruokonen E, Prviinen I, Jkob SM, et l: Dexmedetomidine versus propofol/midzolm for long-term sedtion during mechnicl ventiltion. Intensive Cre Med 2009; 35: Abuhsn S, Al Jundi A, Abdeltty W, et l: Evlution of long-term infusion of dexmedetomidine in criticlly ill ptients: A retrospective nlysis. Int J Crit Illn Inj Sci 2012; 2: Pndhripnde PP, Pun BT, Herr DL, et l: Effect of sedtion with dexmedetomidine vs lorzepm on cute brin dysfunction in mechniclly ventilted ptients: The MENDS rndomized controlled tril. JAMA 2007; 298: Gupt P, Whiteside W, Sbti A, et l: Sfety nd efficcy of prolonged dexmedetomidine use in criticlly ill children with hert disese. Peditr Crit Cre Med 2012; 13: Tobis JD, Berkenbosch JW: Sedtion during mechnicl ventiltion in infnts nd children: Dexmedetomidine versus midzolm. South Med J 2004; 97: October 2014 Volume 15 Number 8