Opioid and benzodiazepine withdrawal symptoms in paediatric intensive care patients

Transcription

1 Intensive and Critical Care Nursing (2004) 20, ORIGINAL ARTICLE Opioid and benzodiazepine withdrawal symptoms in paediatric intensive care patients Linda S. Franck, Ita Naughton, Ira Winter Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust, Great Ormond Street, Level 7 Old Building, London WC1N 3JH, UK Accepted 13 July 2004 KEYWORDS Opioid; Benzodiazepine; Withdrawal; PICU; Children Summary The purposes of this prospective repeated measures study were to: (a) describe the occurrence of withdrawal symptoms with the use of a standardised protocol to slowly taper opioids and benzodiazepines; and (b) to test the predictive validity of an opioid and benzodiazepine withdrawal assessment scoring tool in critically ill infants and young children after prolonged opioid and benzodiazepine therapy. Fifteen children (6 weeks 28 months of age) with complex congenital heart disease and/or respiratory failure who received opioids and benzodiazepines for 4 days or greater were evaluated for withdrawal symptoms using a standardized assessment tool. Thirteen children showed moderate to severe withdrawal symptoms a median 3 days after commencement of tapering. Symptom intensity was not related to prior opioid or benzodiazepine exposure, extracorporeal membrane oxygenation (ECMO) therapy or length of tapering. Children who received fentanyl in addition to morphine more often exhibited signs of withdrawal. This study demonstrated that significant withdrawal symptoms occur in critically ill children even with the use of a standardised assessment tool and tapering management protocol. The predictive validity and utility of the Opioid and Benzodiazepine Withdrawal Score (OBWS) was adequate for clinical use, but areas for further improvement of the tool were identified. Problems with the clinical withdrawal prevention and management guidelines were also identified. More research is needed to establish the optimal methods for prevention and management of iatrogenic opioid and benzodiazepine withdrawal in paediatric critical care Elsevier Ltd. All rights reserved. Introduction * Corresponding author. Tel.: ; fax: address: [email protected] (L.S. Franck). The use of opioids and benzodiazepines to provide children with effective relief of pain and anxiety is standard practice in the paediatric intensive care setting (Matthews, 1993). Analgesia and sedation /$ see front matter 2004 Elsevier Ltd. All rights reserved. doi: /j.iccn

2 Opioid and benzodiazepine withdrawal symptoms 345 are often necessary for comfort during mechanical ventilation and may be essential to blunt physiological stress responses and reduce mortality and morbidity risk in the very critically ill patient (Carr and Todres, 1994). Unfortunately, the use of these drugs for more than a few days may lead to tolerance, in which larger doses are needed to maintain the same effects, and physical dependence, in which continued administration of the drug is required to prevent withdrawal syndrome (Anand and Ingraham, 1996). Withdrawal syndrome is the term used for a characteristic pattern of unpleasant signs and symptoms that typically follows abrupt cessation of drugs with central nervous system depressant effects. The predominant manifestations include central nervous system hyperirritability and autonomic nervous system dysregulation (sneezing, yawning, sweating, tachycardia, mydriasis), gastrointestinal dysfunction (diarrhoea, nausea, vomiting), respiratory distress (tachypnoea), and abnormal motor movements (tremors, hypertonia, hyper-reflexivity, repetitive movements). Benzodiazepine withdrawal differs from opioid withdrawal in that it does not cause gastrointestinal dysfunction and abnormal motor movements may be more pronounced (Anand and Ingraham, 1996). Withdrawal symptoms range from mild to severe, and seizures can occur in severe untreated cases (Franck and Vilardi, 1995). Since the early 1980s, retrospective studies and case reports have documented withdrawal syndrome in the infants and children following long term exposure and abrupt discontinuation of opioid and benzodiazepine therapy in the neonatal and paediatric critical care settings (Hasday and Weintraub, 1983; Miser et al., 1986; Norton, 1988; Sury et al., 1989; Tobias et al., 1990; Bergman et al., 1991; Lane et al., 1991; Carnevale and Ducharme, 1997; Fonsmark et al., 1999; Arnold et al., 1990). Iatrogenic withdrawal syndrome caused by too rapid tapering or abrupt discontinuation of opioid or benzodiazepine therapy in critically ill patients complicates medical treatment and may prolong recovery and hospital stay. Current prevalence is unknown, although anecdotal and survey evidence suggests it remains problematic (Franck and Vilardi, 1995; Brennan et al., 2003; Matthews, 1993). Only five studies prospectively assessed withdrawal symptoms (Katz et al., 1994; French and Nocera, 1994; Franck, 1998, unpublished data; Lugo et al., 2003; Franck et al., 1998; Bicudo et al., 1999) and associated factors. Only four prospective or retrospective (Franck et al., 1998; Tobias, 2003; Tobias, 1999; Lugo et al., 2003) used a consistent approach to tapering. Variability in drugs, doses, length of therapy, length of taper, assessment of symptoms and treatment of symptoms make it difficult to draw inferences from these studies. However, the findings suggest that use of fentanyl, longer duration of therapy and more rapid discontinuation increase the risk and severity of withdrawal syndrome. Nevertheless, at present there is no consensus amongst clinicians as to how analgesia and sedation should be managed and discontinued to prevent iatrogenic withdrawal, or on how withdrawal syndrome should be assessed and treated (Carr and Todres, 1994; Franck and Vilardi, 1995; Anand and Ingraham, 1996; Brennan et al., 2003; Lugo et al., 2001; Tobias et al., 1994; Siddappa et al., 2003). The clinical assessment of withdrawal syndrome in infants and children is based on a few studies conducted in infants with prenatal opioid exposure (Lipsitz, 1975; Green and Suffet, 1981; Zahorodny et al., 1998) and on case reports of iatrogenic benzodiazepine withdrawal in children (Bergman et al., 1991; Sury et al., 1989). There are no published data on reliability and validity of assessment tools for neonatal or paediatric intensive care patients. The lack of valid measures hampers research and clinical practice to determine the most effective method of terminating long-term analgesia and sedation for critically ill infants and children. Therefore, the purposes of this study were to: (a) describe the occurrence of withdrawal symptoms with the use of a standardised protocol to slowly taper opioids and benzodiazepines; and (b) to test the predictive validity of an opioid and benzodiazepine withdrawal assessment scoring tool in critically ill infants and young children after prolonged opioid and benzodiazepine therapy. Methods Design A prospective repeated measures study design was used to describe the occurrence and severity of withdrawal symptoms using a standardised assessment tool and clinical management guidelines from commencement to completion of tapering from opioids and benzodiazepines. Settingand sample The study was initiated in a 10-bed paediatric Cardiac Intensive Care Unit (CICU) that provided care for children pre- and post-cardiac surgery. Because of the severity of illness (i.e., severe congenital

3 346 L.S. Franck et al. heart defects and respiratory failure) and invasiveness of medical treatment (i.e., multiple chest drains, mechanical ventilation, central venous and arterial lines), all children receive opioid analgesia with or without benzodiazepine sedation as a routine part of their medical care. Prior to the study, clinical guidelines for initiating and discontinuing analgesia and sedation therapy (available from authors on request) were introduced with the aim of reducing the variation in practice and the frequency and severity of withdrawal syndrome. For patients with opioid and benzodiazepine administration for greater than 5 days, a 20% daily taper (from the pre-taper dose) was recommended. For patients who had received treatment for greater than 14 days, a 10% daily taper was recommended. If withdrawal symptoms occurred, recommendations were given regarding slowing or suspending the taper and reinitiating opioid treatment, depending on severity (Carr and Todres, 1994; Franck et al., 1998; Anand and Arnold, 1994). All patients regardless of age, diagnosis, or analgesia/sedation regimen were eligible for enrolment in this study, once they received opioid and/or benzodiazepine therapy for greater than 5 days. The inclusion criteria were modified to include children received opioid and benzodiazepine therapy for 4 days, after withdrawal symptoms were observed in one child after commencement of the taper. However, this resulted in the inclusion of only the one additional child in the sample. Children with a significant neurological insult (documented by EEG or head ultrasound) or seizure disorder were excluded from the study. The Local Research Ethics Committee approved the study but did not require written consent, as this study did not alter patients medical treatment in any way. The study was carried out in accordance to the Department of Health Research Governance Framework for Health and Social Care, with particular attention to data protection (Department of Health, 2001), and parents were provided with the usual information about pain and sedation management and withdrawal symptom monitoring. Instruments Opioid and Benzodiazepine Withdrawal Score (OBWS) A 21-item checklist that rates the frequency and severity of withdrawal symptoms (available from authors on request). The OBWS was adapted from the Children s Hospital Oakland Opioid Withdrawal Flowsheet (Franck and Vilardi, 1995). Preliminary face validity, content validity, and reliability were established in previous pilot work (Franck, 1998, unpublished data). Nurses clinical judgment of withdrawal syndrome Nurses were asked to give their clinical judgment as to whether the patient was experiencing withdrawal by indicating a yes or no response on the bottom of the OBWS form. Patient medical record review form A 21-item data sheet to record patient demographic and clinical variables that may influence withdrawal risk and or symptom presentation (Katz et al., 1994; Franck et al., 1998; Fonsmark et al., 1999; Arnold et al., 1990). All opioid and benzodiazepine exposure during the episode of illness were recorded. The rate of tapering and discontinuing analgesia and/or sedation therapy were also recorded until all opioids and benzodiazepines were discontinued. Procedures As part of the implementation of the opioid and benzodiazepine clinical management guideline, the OBWS was incorporated into the standard bedside documentation in the CICU. Patients who received opioid and/or benzodiazepine therapy for greater than 5 days were assessed every four hours using the OBWS until approximately 2 days after the drugs had been discontinued. Two investigators (I.N. and I.W.) made daily bedside rounds to collect data and discuss the use of the OBWS with the nursing staff. Statistical analyses Descriptive statistics were used to summarise the frequency, severity, and duration of withdrawal symptoms for each patient and for the whole sample. Plots were produced showing the change in OBWS over time for each patient and highlighting peak OBWS. Non-parametric statistics were used to explore associations (Spearman rank correlations, r s ) and differences (Mann Whitney U, chi-square) between the variables of interest, including peak infusion rates, total dose, duration of infusion, peak OBWS, nurses judgment of withdrawal and length of tapering. Sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operator characteristic (ROC) curve were calculated to test the predictive validity of the OBWS.

4 Opioid and benzodiazepine withdrawal symptoms 347 Results Sample characteristics Fifteen children (6 weeks 28 months of age) met the inclusion criteria over an 11-month period. Fourteen children had complex congenital heart disease and underwent cardiac surgery with cardiopulmonary bypass, and five of these children required extracorporeal membrane oxygenation (ECMO) following failure to wean from bypass. One child received medical treatment only for cardiac failure. Thirteen of the children received both opioids and benzodiazepines and two children received opioids alone. Four children received fentanyl in addition to morphine. Three children were converted from intravenous to oral morphine before tapering was completed. Most of the children received at least one other drug during the tapering period such as chloral hydrate or triclofos (n = 10), clonidine (n = 9), or lorazepam (n = 3). Additional characteristics of opioid and benzodiazepine therapy and tapering are shown in Table 1. Peak and total doses and length of treatment with opioids and benzodiazepines were all moderately to highly correlated (r s =.58.81, p <.05). Children who received ECMO therapy received larger cumulative and peak opioid doses than children who did not (U = 7.0, p =.03; U = 4.0, p =.01, respectively). Withdrawal symptoms Complete data were obtained for 693 OWS assessments, of which 151 indicated symptoms of withdrawal by nurses clinical judgment. Table 2 shows the occurrence of symptoms when nurses judged the children to be experiencing withdrawal compared to assessments when nurses judged no withdrawal. The most common symptoms when patients were experiencing withdrawal, occurring in over one-third of the assessments, were tempera- ture greater than 37.2 C, sleeplessness, diarrhoea, pupils larger than 4 mm, and tremors. However, temperature elevation was also the most common symptom noted when withdrawal was judged not to be present, occurring in two-thirds of the observations. The remainder of the most common signs of withdrawal occurred in 20% or less of nonwithdrawal assessments. Median OBWS was strongly correlated with nurses judgment of withdrawal symptoms (r s =.66, p =.007). Using the predetermined cut-off score of greater than 8, the sensitivity of the OBWS was 50% and specificity was 87% and the area under the ROC curve was The positive likelihood ratio was 4 and the negative likelihood ratio was With pre-test odds of 28%, the use of the OWS increased positive diagnostic probabilities only slightly (Sackett et al., 2000). Thirteen children showed moderate to severe withdrawal symptoms (OBWS greater than 8) on at least three assessments. Clinically significant symptoms developed a median 2 days after commencement of tapering and OBWS remained over 8 for a median of 3 days. OBWS intensity was not related to prior opioid or benzodiazepine exposure, ECMO therapy or length of tapering. However, the length of tapering was strongly correlated with nurses judgment of withdrawal (r s =.70, p =.004). Children who received fentanyl in addition to morphine more often exhibited signs of withdrawal as judged by nurses (U =5,p =.03) and received higher doses of opioids and benzodiazepines prior to tapering (U =5,p =.03; Fig. 1). Compliance and usability Seventy percent of the nursing staff received training in the assessment of withdrawal symptoms and the use of the OBWS and achieved inter-rater reliability of 80% or greater prior to implementation of the withdrawal clinical management guideline. Nurses daily compliance with completion of the Table 1 Characteristics of opioids and benzodiazepine treatment (n = 15). Median Range Length of opioid treatment prior to tapering (days) Peak opioid dose during treatment (mcg/kg/h) Opioid treatment cumulative dose (mg/kg) Peak midazolam dose during treatment (mcg/kg/min) Midazolam treatment cumulative dose (mg/kg) Tapering period (days) Number of days with OBWS > Number of days nurses judged withdrawal symptoms Day first symptoms appear after commencement of taper 2 0 6

5 348 L.S. Franck et al. Table 2 Sign or symptom Frequency of withdrawal symptoms. Nurse s assessment of withdrawal Present (%) Absent (%) Crying/agitated >75% of interval Crying/agitated 26 75% of interval Sleeping <25% of the interval Hyperactive Moro reflex Pupils >4mm Tremors Movement disorder Hallucinations Temperature >37.2 C Respiratory rate high for age Frequent suction required Sweating Yawning Sneezing Nasal stuffiness Vomiting Diarrhoea OBWS was 85%, but within each day, on average 55% of the 4-hourly assessments were made. Overall, nurses were enthusiastic about the use of the OBWS and felt that it improved their assessment of withdrawal symptoms. However, several issues about the use of the OBWS were consistently raised by nurses during the investigators daily rounds, including the need for ongoing training in the use of the tool because of staff turnover, the difficulty in maintaining inter-rater reliability among nurses, nurses preference for yes/no items than rating scales for symptoms, nurses preference for fewer items on the tool, and nurses preference for longer interval between assessments. All patients in the study were to undergo scheduled tapering of opioids and benzodiazepine according to the withdrawal management guideline introduced on the unit prior to the study. Eleven children received drug therapy for less than 14 days and should have been tapered on the 20% per day Figure 1. Selected characteristics of children who received morphine only or morphine and fentanyl (median values).

6 Opioid and benzodiazepine withdrawal symptoms 349 protocol. Four children received drug therapy for greater than 14 days and should have been tapered on the 10% per day protocol. However, in all but two cases, the clinical management guideline was not followed. Five children were tapered too fast, four were tapered too slowly and four were tapered both too fast and too slowly. Fig. 2 shows a child s initial taper that was too rapid, resulting in withdrawal symptoms and subsequent slowing of the taper. These actions likely resulted in a prolonged taper compared to the expected duration if the protocol had been followed. Supplemental drug use was not in accordance with the guideline in any of the cases. For example, for one child clonidine administration was commenced at the start of tapering at the consultant s request. Discussion In this study, we observed the frequency and presentation of withdrawal symptoms following tapering of opioid and benzodiazepines according to a clinical guideline developed by the critical care unit staff. Signs and symptoms of withdrawal in this study were similar to those reported by others (Katz et al., 1994; Fonsmark et al., 1999; Franck et al., 1998; Lugo et al., 2003; Arnold et al., 1990; Anand and Arnold, 1994). None of the above studies used a clinical guideline for tapering and in some studies the reduction in dose was dramatic or discontinued abruptly (Carnevale and Ducharme, 1997; Katz et al., 1994). However, even when guidelines were used, 41% of children experienced severe symptoms requiring treatment and the most common signs observed were: hyper-reflexia, tremors, tacypnoea, agitation/irritability, dilated pupils, diarrhoea, and sleeplessness (Lugo et al., 2003). In this study, the four most common signs of withdrawal were: sleeplessness, diarrhoea, mydriasis, and tremors. Previous studies suggest that withdrawal symptoms appear 1 hour 3 days after a decrease in analgesia or sedation dosage (Norton, 1988; French and Nocera, 1994; Carnevale and Ducharme, 1997). In this study, symptoms occurred as late as 6 days after commencement of tapering greater than 10%. Symptom onset appears related to increment or speed of the taper. Children in this study received lower doses of opioids and benzodiazepines than in previous studies but nevertheless exhibited clinically significant withdrawal symptoms. A larger study sample is needed to determine the influence of risk factors, such as choice of drug, amount, or frequency of delivery, on withdrawal symptoms when therapy is tapered rather than abruptly discontinued. Compliance with assessment using a standardized scoring tool was reasonably good. However, Figure 2. Case example of rapid initial taper.

7 350 L.S. Franck et al. suggestions to improve ease of use and decrease the frequency should be considered, if the improvements can be made without sacrificing validity. The OBWS yielded fewer false positives (good specificity), but missed almost 50% of cases (low sensitivity). Although the ideal assessment tool would be both sensitive and specific, this is rarely achieved and there is often a trade-off between sensitivity and specificity in clinical diagnostics. The prognostic value of the OBWS would improve if the prevalence of withdrawal was higher, as reported in other studies (Katz et al., 1994). However, further research should examine modifications to the tool that will improve sensitivity and clinical utility. Protocols for analgesia and sedation management and tapering have demonstrated usefulness in the optimising patient care (Alexander et al., 2002; Parran and Pederson, 2002; Siddappa et al., 2003). However, compliance with the tapering guideline in this study was extremely poor. Although the multidisciplinary team on the clinical unit had collaborated on the guideline development, approved its use, and received training in its use, practical and attitudinal obstacles emerged only after implementation. For example, practical difficulties included the number of other drugs including inotropes, vasodilators, multiple antibiotics, muscle relaxants, and heparin that children received concurrently. They were frequently fluid restricted and therefore drug solutions would be made up quite concentrated. This would sometimes affect tapering of opioids and benzodiazepines by small increments (e.g., 12% rather than 10%), as it would not be possible to be more exact unless the drug was made up in a less concentrated form. Attitudinal obstacles included the persistent belief that gradual tapering of analgesia and sedation would prolong ventilation and intensive care length of stay. Some clinicians also felt that opioids and benzodiazepines must be tapered quickly or discontinued abruptly to enable extubation. There is little objective evidence to support these concerns. For example, most patients develop rapid tolerance to the respiratory depressant effects of opioids whereas they rarely develop tolerance to other opioid side effects such as miosis and slowed gastrointestinal motility (Franck and Vilardi, 1995). The findings from this study must be interpreted with caution because of the limitations of the methods. First, variability in the tapering of withdrawal symptoms may have confounded detection of symptoms. Second, nurses judgment of withdrawal have been influenced by their use of the OBWS (observer bias) because the same person undertook the scoring using the withdrawal tool as made the assessment of whether withdrawal was present or not. For this small study it was not possible to create independence between total score and the nurses judgment, but it should be considered in a further study. Conclusions This study demonstrated that significant opioid and benzodiazepine symptoms occur in critically ill children even with the use of a standardised assessment tool and tapering management protocol. The predictive validity and utility of the OBWS was adequate for clinical use, but areas for further improvement of the tool were identified. Problems with the clinical withdrawal prevention and management guidelines were also identified. More research is needed to establish the optimal methods for prevention and management of iatrogenic opioid and benzodiazepine withdrawal in paediatric critical care. References Alexander E, Franco A, Razack S. Evaluation of a sedation protocol for intubated critcally ill children. Intensive Crit Care Nurs 2002;18: Anand KJ, Arnold JH. Opioid tolerance and dependence in infants and children. Crit Care Med 1994;22: Anand KJS, Ingraham J. Tolerance, dependence, and strategies for compassionate withdrawal of analgesics and anxiolytics in the pediatric ICU. Crit Care Nurs 1996;16: Arnold JH, Truog RD, Orav EJ, Scavone JM, Hershenson MB. Tolerance and dependence in neonates sedated with fentanyl during exracorporeal membrane oxygenation. Anesthesiology 1990;73: Bergman I, Steeves M, Burckart G, Thompson A. Reversible neurologic abnormalities associated with prolonged intravenous midazolam and fentanyl administration. J Pediatr 1991;119: Bicudo JN, de Souza N, Mangia CM, de Carvalho WB. Withdrawal syndrome associated with cessation of fentanyl and midazolam in pediatrics. Rev Assoc Med Brasileira 1999;45:15 8. Brennan NM, Crossland DS, O Sullivan JJ. Drug withdrawal after PICU: are we doing enough? Eur Soc Paediatr Carnevale FA, Ducharme C. Adverse reactions to the withdrawal of opioids and benzodiazepines in paediatric intensive care. Intensive Crit Care Nurs 1997;13: Carr DB, Todres ID. Fentanyl infusion and weaning in the pediatric intensive care unit: towards science-based practice. Crit Care Med 1994;22: Department of Health. Research governance framework for health and social care (England). research/rd3/nhsrandd/researchgovernance.htm July Fonsmark L, Rasmussen YH, Peder C. Occurrence of withdrawal in critically ill sedated children. Crit Care Med 1999;27:196 9.

8 Opioid and benzodiazepine withdrawal symptoms 351 Franck LS, Vilardi J. Assessment and management of opioid withdrawal in ill neonates. Neonatal Netw 1995;14(2): Franck LS, Vilardi J, Durand D, Powers R. Opioid withdrawal in neonates after continuous infusions of morphine or fentanyl during extracorporeal membrane oxygenation. Am J Crit Care 1998;7(5): Franck LS. Preliminary validity and reliability of the Opioid Weaning Flowsheet. Unpublished (available from author on request) French JP, Nocera M. Drug withdrawal symptoms in children after continuous infusions of fentanyl. J Pediatr Nurs 1994;9: Green M, Suffet F. The neonatal withdrawal index: a device for the improvement of care in abstinence syndrome. Addict Dis 1981;8: Hasday JD, Weintraub M. Propoxyphene in children with iatrongenic morphine dependence. Am J Dis Child 1983;137: Katz R, Kelly HW, Hsi A. Prospective study on the occurrence of withdrawal in critically ill children who receive fentanyl by continuous infusion. Crit Care Med 1994;22: Lane JC, Tennison MB, Lawless ST, Greenwood RS, Zaritsky AL. Movement disorder after withdrawal of fentanyl infusion. J Pediatr 1991;119: Lipsitz PJ. A proposed narcotic withdrawal score for use with newborn infants. A pragmatic evaluation of its efficacy. Clin Pediatr 1975;14: Lugo RA, Lee WE, Cash J, Vernon DD, Ward RM. Opioid abstinence syndrome in the PICU: frequency of signs. J Pediatr Pharmacol Ther 2003;8:232. Lugo RA, MacLaren R, Cash J, Pribble CG, Vernon DD. Enteral methadone to expidite fentanyl discontinuation and prevent opioid abstinence syndrome in the PICU. Pharmacotherapy 2001;21: Matthews AJ. An audit of sedation, analgesia and muscle relaxation in paediatric intensive care in the United Kingdom. Paediatr Anaesth 1993;3: Miser AW, Chayt KJ, Sandlund JT, Cohen PS, Dothage JA, Miser JS. Narcotic withdrawal syndrome in young adults after the therapeutic use of opiates. Am J Dis Child 1986;140: Norton SJ. Aftereffects of morphine and fentanyl analgesia: a retrospective study. Neonatal Netw 1988;7:25 8. Parran L, Pederson C. Effects of an opioid taper algorithm in hematopoietic progenitor cell transplant recipients. Oncol Nurs Forum 2002;29: Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. London: Churchill Livingstone; Siddappa R, Fletcher JE, Heard AMB, Kielma D, Cimino M, Heard CMB. Methadone dosage for prevention of opioid withdrawal in children. Paediatr Anaesth 2003;13: Sury MR, Billingham I, Russell GN, Hopkins CS, Thornington R, Vivori E. Acute benzodiazepine withdrawal syndrome after midazolam infusions in children. Crit Care Med 1989;17: Tobias JD. Subcutaneous administration of fentanyl and midazolam to prevent withdrawal after prolonged sedation in children. Crit Care Med 1999;27: Tobias JD. Pain management for the critically ill child in the pediatric intensive care unit. In: Pain in infants, children and adolescents. Philadelphia: Lippincott, Williams and Wilkins; p Tobias JD, Deshpande JK, Gregory DF. Outpatient therapy of iatrogenic drug dependency following prolonged sedation in the pediatric intensive care unit. Intensive Care Med 1994;20: Tobias JD, Schleien CL, Haun SE. Methadone as treatment for iatrogenic narcotic dependency in pediatric intensive care unit patients. Crit Care Med 1990;18: Zahorodny W, Rom C, Whitney W, Giddens S, Samuel M, Maichuk G, Marshall R. The neonatal withdrawal inventory: a simplified score of newborn withdrawal. J Dev Behav Pediatr 1998;19:89 93.