Development of a Rapid Observational Assessment Tool

Intensive Care Med (2012) 38:1025–1031 DOI 10.1007/s00134-012-2518-z PEDIATRIC ORIGINAL

Gabrielle Silver Detecting pediatric delirium: development Chani Traube Julia Kearney of a rapid observational assessment tool Daniel Kelly Margaret J. Yoon Wendy Nash Moyal Maalobeeka Gangopadhyay Huibo Shao Mary Jo Ward

W. Nash Moyal Á M. Gangopadhyay performed in a PICU at a university Received: 13 July 2011 Psychiatry, NY Presbyterian Hospital, Accepted: 23 January 2012 hospital. Fifty patients, ages 3 months New York, NY, USA Published online: 10 March 2012 to 21 years, admitted to the PICU Ó Copyright jointly held by Springer and over a 6-week period were included. ESICM 2012 H. Shao Public Health, Weill Cornell Medical No interventions were performed. College, New York, NY, USA Results: After informed consent was obtained, two study teams inde- M. J. Ward pendently assessed for delirium by Child Development, Weill Cornell Medical completing the CAP-D and by con- G. Silver Á M. J. Yoon College, NY Presbyterian Hospital, ducting psychiatric evaluation based Child Psychiatry, Weill Cornell Medical New York, NY, USA College, NY Presbyterian Hospital, on the DSM-IV criteria. Concordance New York, NY, USA between the CAP-D and DSM-IV Abstract Objective: Develop- criteria was excellent, at 97%. Prev- C. Traube ()) ment of a novel screening tool for the alence of delirium in this sample was Pediatric Critical Care, Weill Cornell detection of delirium in pediatric 29%. Conclusion: The CAP-D may Medical College, NY Presbyterian Hospital, be a valid screen for identiﬁcation of 525 East 68th Street M-508, intensive care unit (PICU) patients of New York, NY 10065, USA all ages by comparison with psychi- delirium in PICU patients of all ages. e-mail: [email protected] atric assessment based on the Further studies are required to explore reference standard Diagnostic and its validity, inter-rater reliability, and J. Kearney Statistical Manual of Mental Disor- feasibility of use as a nursing screen. Child Psychiatry, Memorial Sloan Kettering ders (DSM-IV) criteria. Cancer Center, New York, NY, USA Methods: This was a prospective Keywords Delirium Á Pediatrics Á Intensive care units Á Critical care Á D. Kelly blinded pilot study investigating the Pediatrics, NY Presbyterian Hospital, feasibility of the Cornell Assessment Detection Á Screening New York, NY, USA of Pediatric Delirium (CAP-D)

Introduction illness itself (e.g., serious infection or neoplasm), by- products of the illness (e.g., metabolic or endocrine dys- Delirium in critically ill children represents acute brain function), or side-effects of the treatment (e.g., sedative dysfunction and is characterized by a range of altered drugs or drug withdrawal) [1–8]. mental states and behaviors. The Diagnostic and Statis- Evidence-based assessments of outcomes and inter- tical Manual of Mental Disorders (DSM-IV) definition of ventions for pediatric delirium are lacking, largely due to delirium involves acute onset, fluctuating course, and two the absence of a simple and reliable screening tool [9]. To major categories: disturbance of consciousness (including date, delirium has not been considered a meaningful attention and awareness) and cognition (including mem- sequela of severe pediatric illness, but there is an emerging ory, orientation, language, and perception). This must be literature challenging this assumption [9–12]. The unique the result of a medical condition and can be caused by the combination of vulnerability and resiliency of children’s 1026 developing brains makes it imperative to conduct research The Delirium Rating Scale (DRS) [23], the most to improve recognition, establish efficient, reliable, and comprehensive scale used to date in adults, has been valid methods of assessment, and develop strategies to shown to have comparable cutoff scores in a pediatric prevent, treat, and understand long-term implications cohort in a retrospective chart review [5]. However, the [1–3, 10, 11, 13]. DRS requires considerable time and a trained psychiatrist. Pediatric delirium is associated with prolonged hos- It includes questions which are not applicable to younger pital stay [1, 3, 13] and residual perceptual-motor and children, including self-report of symptoms such as hal- behavior problems [2, 10, 14]. Children have reported lucinations and delusions. No prospective validation delusional memories of intensive care unit (ICU) expe- studies in pediatrics have been conducted. riences, which can become the basis for post-traumatic In the adult population the Confusion Assessment stress disorder (PTSD) [5, 11, 15, 16]. Small sample sizes, Method for the ICU (CAM-ICU) has been developed for lack of a common diagnostic tool, lack of prospective non-psychiatric trained ICU medical professionals to design, and other issues in these studies have contributed screen for delirium [24, 25]. The Vanderbilt Delirium to the difficulty in interpreting the impact of delirium on Group has developed the Pediatric Confusion Assess- critically ill children. ment Method for the ICU (pCAM), the first screening instrument developed specifically for critically ill children [26, 27]. Through the use of yes/no questions, Materials and methods pictures, and hand signals, the pCAM-ICU allows for the evaluation of children’s ‘‘inattention’’ and ‘‘disorganized Background thinking.’’ Sensitivity and specificity are currently being evaluated in pilot studies in the USA and Europe. A trial Screening seriously ill children for delirium is challenging of the pCAM-ICU in our unit found its neurocognitive due to the developmental differences in presentation [2]. approach to be impractical in very sick, often pharma- Agitated or combative behavior is readily recognized cologically sedated, and non-verbal patients. While a [17], but delirium is more often associated with hypoac- well-designed and elegant tool for diagnosing delirium in tive behavior, e.g., withdrawn affect and sparse responses cooperative older children, the pCAM-ICU requires [3, 18]. The great range and subtlety of delirium in chil- intensive training of the nursing staff and is not appli- dren have led some to suggest this condition is under- cable to patients who are developmentally impaired or recognized, under-studied, and likely common in pediat- less than 5 years old [28]. ric critical care [9, 18–21]. The Pediatric Anesthesia Emergence Delirium scale In clinical practice the diagnosis of pediatric delirium (PAED) was developed in studies of ‘‘emergence delir- is dependent on recognition by ICU staff who then consult ium’’, a proposed delirium subtype in anesthesiology, a child psychiatrist [5]. The psychiatrist’s assessment occurring when children are awakening from anesthesia, includes an interview of the family and medical staff and often with agitated or hyperactive symptoms. This a psychiatric examination of the patient. The reference observational scale evaluates the child’s eye contact with standard for diagnosing delirium is this psychiatric eval- the caregiver, awareness of surroundings, purposeful uation which assesses for the signs and symptoms nature of movements, ability to be consoled, and rest- delineated by the American Psychiatric Association’s lessness [29] (see footnote in Table 1). It is a quick, (APA) DSM [7]. This time-consuming and labor-inten- flexible tool which can be used by non-physician staff. It sive approach may work for managing delirium in the has been validated in children from age 19 months to agitated child, but it is not suitable for broad-based 6 years in postoperative settings and is being studied for screening or research on prevalence. As diagnosis use in pediatric intensive care unit (PICU) settings as part depends upon referral to psychiatry, many children with of a ‘‘diagnostic algorithm,’’ which includes caregiver and hypoactive or mixed delirium may go unrecognized [21, multidisciplinary team consensus [13]. The strength of the 22]. Identifying the prevalence of delirium in all its forms PAED as a screening tool is its focus on the observation and improving functional outcomes for children with of neurobehavioral symptoms, which makes it uniquely delirium has been hampered by the lack of an effective, suited for younger children (and potentially for develop- user-friendly tool. mentally delayed adults), in whom neurocognitive An ideal screen would be simple, quick, applicable to symptoms are difficult or impossible to assess. In a recent children of all ages and cognitive levels, and could be article, Jan Schieveld’s [30] group compared the PAED to administered by non-psychiatrists multiple times a day. It other delirium screening tools in a PICU population and would be flexible for use in intubated and non-intubated found it to be the most promising of the available tools, children, and would not require child cooperation. It but identified the need for improving upon it. The PAED would detect all delirium subtypes and take advantage of lacks any items to account for observation of hypoactive the ICU staff’s ongoing observation of the patient. delirious symptoms and has not been validated in infants Existing scales have not met these criteria. under age 19 months. 1027

Table 1 Cornell Assessment of Pediatric Delirium (CAP-D)

RASS score ______(if -4or-5 do not proceed) Not at all 4 Just a little 3 Quite a bit 2 Very much 1 Extremely 0 Score

1. Does the child make eye contact with the caregiver?a 2. Are the child’s actions purposeful?a 3. Is the child aware of his/her surroundings?a Not at all 0 Just a little 1 Quite a bit 2 Very much 3 Extremely 4 4. Is the child restless?a 5. Is the child inconsolable?a 6. Is the child underactive: very little movement and interaction?b 7. Are the child’s responses sparse and/or delayed?b Total (C10 delirium present)c

RASS Richmond Agitation-Sedation Scale c a The original PAED was scored: 0–6, no delirium and no further Elements of the original PAED [29], modiﬁed from statement to evaluation needed; 7–9, subsyndromal delirium and reevaluate question form soon; C10, delirium present b Questions added to improve detection of hypoactive and mixed delirium

To address the limitations of existing scales and the Evaluators need for a tool that could be easily applied to the general PICU population, we modiﬁed the PAED to capture Two independent teams of investigators were formed: the hypoactive as well as hyperactive and mixed delirium in ‘‘screen team’’ and the ‘‘psychiatry team.’’ The screen team all pediatric age groups by adding questions about evaluators included an attending pediatric intensivist and a underactive behaviors and sparse or delayed responses. pediatric chief resident. The psychiatry team evaluators We changed the format to questions rather than state- included two attending consultation-liaison (CL) child psy- ments in order to make it more acceptable to our nursing chiatrists and two CL child psychiatry fellows. A complete staff, as it is ultimately intended to be a nursing tool. We assessment included both a CAP-D screen and a formal named our adapted PAED scale the Cornell Assessment psychiatric evaluation. Each assessment was completed of Pediatric Delirium (CAP-D) (Table 1). We report here without knowledge of the outcome of the other assessment. a comparison of the new CAP-D with psychiatric evaluation in a consecutive sample of children (n = 50) ages 3 months to 21 years. Setting

This study was executed in a 20-bed, academic PICU at a Design university-based medical center. The PICU admits critically ill children with medical, surgical, cardiac, and This pilot study was designed to assess the feasibility of neurologic diagnoses. Patients range in age from 2 days the CAP-D as a screening tool for delirium, and to to 21 years. compare it to the DSM-IV-based psychiatric interview. The study included every child hospitalized in our PICU during a predeﬁned period of time, regardless of age, Assessments developmental level, or diagnosis. Two exclusion criteria were applied: (1) subjects whose parents were unavailable On consecutive weekdays for 6 weeks, parents or guard- to provide consent; (2) subjects who were deeply sedated ians of every patient in the PICU were approached by the or unarousable (score less than -3 on the Richmond screen team and offered the opportunity to participate in Agitation-Sedation Scale (RASS) [31], a standard mea- this pilot study. When informed consent was obtained, the sure widely used in PICUs around the world. On the evaluators conducted the CAP-D using a paper checklist. RASS, score ranges from -5 for unarousable to ?4 for The CAP-D, a Likert-type scale, consists of seven ques- combative; a score of zero is alert and calm. tions: Does the child make eye contact? Is the child The study was approved by the institutional review purposeful? Is the child aware of his/her surroundings? Is board. Informed consent was obtained from each subject’s the child restless? Inconsolable? Underactive? Are the parent or legal guardian; child assent was obtained from child’s responses sparse and/or delayed? Each question is alert subjects ages 7–17. rated on a 5-point scale (not at all, just a little, quite a bit, 1028

very much, and extremely). The first three questions are Table 2 Subject characteristics (n = 50) scored 0–4, and the last four in reverse, 4–0. A total of at least 10 was used as a cutoff for delirium, as in the PAED. Age (years) N (% total) (We have kept the cutoff score of 10 because the additional 0–2 11 (22) questions are specific for hypoactive delirium. In hyper- 2–5 12 (24) active delirium, they will not add to the score. In 5–10 10 (20) hypoactive delirium, they will increase detection. In mixed 10–15 10 (20) 15–21 7 (14) delirium, they will definitely add to the score, and this Gender higher score may be a way to differentiate mixed delirium.) Male 30 (60) A list of enrolled subjects was provided to the psychi- Female 20 (40) atry team, who completed a psychiatric exam and interview Developmental delay Yes 12 (24) later that same day. The psychiatric interview and exam No 38 (76) were based on the DSM-IV criteria for diagnosing delirium [7]. This includes a fluctuating disturbance of consciousness with an alteration in cognition or perceptual Table 3 Primary diagnoses (n = 50) disturbance and is caused by an underlying physiological condition. A checklist was used to document eight symp- Primary diagnoses Number tom categories: language and thought; mood/affect; Neurologic 5 perception; cognition; attention; orientation; sleep/wake Neurosurgical 8 cycle; and motor activity. Child psychiatrists based their Oncologic 13 assessments of these symptom categories on knowledge of CNS 6 Non-CNS 7 appropriate cognition and behavior for each age and Other surgical 3 developmental stage. They took into consideration the Cardiac 8 multiple possible causes of altered consciousness and were Infectious 5 careful to discriminate between delirium and other causes Metabolic 3 Other 5 of acute CNS dysfunction. At the close of the assessment, the psychiatrist made a clinical assessment as to whether delirium was present or absent, and when present, a clas- to compare the detection of delirium using the adapted sification as to subtype (hypoactive, hyperactive, or CAP-D screen with the diagnosis of delirium by the mixed). Delirium subtypes were determined by the psy- DSM-IV criteria. Data from the DSM-IV diagnosis and chiatric evaluation, as defined by DSM-IV criteria, which the CAP-D were coded as binary variables (positive include psychomotor and behavioral disturbances (agita- versus negative delirium) and cross-tabulated. Cohen’s tion for hyperactive and withdrawal for hypoactive). kappa was used to calculate agreement. Sensitivity and Patients who remained in the PICU on multiple study days specificity were calculated. were re-assessed, as delirium can fluctuate from day to day. As a result of incomplete data, six subjects were excluded from analyses. Included subjects had at least one complete set of data (CAP-D and psychiatric interview); some subjects had more than five separate assessments Results made on separate study days. So as not to bias the data by overrepresentation of individual subjects, analysis was Subject characteristics limited to the first two subject encounters. For the first assessment of all 50 subjects, the preva- Over a 6-week period, 61 patients were identified as eli- lence of delirium diagnosed by psychiatric evaluation was gible and approached for inclusion in this study. Consent 28% (n = 14). The hyperactive subtype was the least rate was excellent at 92%. Subjects ranged in age from common with n = 2 (14%). The hypoactive and mixed 3 months to 21 years; gender distribution was 60% male. subtypes each constituted six cases (43% each). There The ages, genders, and admitting diagnoses of the subjects were 27 subjects with a second assessment, in which the are detailed in Tables 2 and 3. Of note 46% of our subjects prevalence of delirium diagnosed by psychiatric evalua- were less than 5 years old, and 24% were designated as tion was 33% (n = 9). Medical diagnoses spanned the developmentally delayed by prior cognitive testing. range of typical pediatric critical care, with delirium present in all types of illness.

Outcomes CAP-D agreement with psychiatric interview

Fifty-six subjects were enrolled, and more than 100 paired For primary analyses, to avoid repeat-observer bias, only assessments completed. The assessments were reviewed the ﬁrst set of assessments for each subject was used. The 1029

DSM-IV diagnosis and CAP-D classification agreed in 50 the requirement of a collective minimum score of 10, of 50 cases (kappa = 1.0, p \ 0.001). Sensitivity and successfully selects for delirium, as opposed to other causes specificity were 100%. For the second set of assessments, of acute CNS depression. Our blinded study showed this to DSM-IV diagnosis and CAP-D classification agreed in 25 be the case—a substantial portion of our subjects were of 27 cases (92.6%; kappa = 0.82, p \ 0.001). Both sedated, and the tool was able to discriminate and eliminate disagreements were false negatives from the CAP-D, the patients who were underactive due to sedation, from yielding sensitivity of 78% and specificity of 100%. those who were delirious. (The gold standard to which Overall, concordance between DSM diagnosis and the these assessments were compared was the psychiatric CAP-D screen was 97%, with a sensitivity of 91% and a screen by CL child psychiatrists who are trained to dis- specificity of 100%. Our study sample had a rate of criminate between delirium and other causes of acute CNS delirium of 29%. dysfunction in children.)

PAED agreement with psychiatric interview Improved sensitivity as compared to PAED

When study subject scores on the first set of assessments Although items 1–3 on the original PAED address were tabulated using the original PAED, rather than the awareness and orientation, they miss the cognitive and CAP-D, agreement with psychiatric diagnosis of delirium motor findings most prominent in pediatric hypoactive was 86%. The PAED would have missed 7 of the 14 delirium. In our particular subjects, those with hypoactive delirious patients diagnosed by DSM-IV criteria (5 out of delirium as determined by the psychiatric evaluation 6 children with hypoactive delirium, 1 out of 6 children scored 0 or 1 on items four and five of the original PAED; with mixed delirium, and 1 out of 2 children with this lowered their total score to less than the cutoff of 10 hyperactive delirium). The PAED thus had a sensitivity of to screen positive by PAED. By adding items six and only 50% when applied in a PICU setting. seven, we were able to capture these patients. The internal consistency expressed as Cronbach’s alpha for the five items adapted from the PAED is 0.79. The 7-element CAP-D, with the addition of the two novel Subjectivity items, had an internal consistency of 0.87. The increase of Cronbach’s alpha with this new tool reveals that the As this tool is designed to be administered by a large additional items demonstrate reliable internal consistency. group of pediatric critical care nurses, it is important to The item-total score correlations for the two added ele- eliminate (to as large an extent as possible) subjectivity. ments were 0.71 and 0.82, demonstrating good fit with the We have added explicit anchor points to the next version scale construct (these analyses were performed in STATA of the CAP-D that will be used both for pre-education of version 11). the screeners and as a reference tool readily available at the bedside for point-of-care use during the twice-daily screen. (In this pilot, the screen team was composed of Discussion two highly trained pediatricians, so the anchor points were implicit.) Ease of use and feasibility

On average, trained staff required less than 2 min to Point-in-time screens complete the CAP-D screen. Although we were con- cerned that difficulties might arise in applying the The pilot data demonstrated that, as with any screening instrument to the youngest children, with minimal guid- tool, it is possible to miss cases of delirium with single ance as to appropriate developmental expectations, the point-in-time screens as delirium is a fluctuating syn- screen team quickly became adept at using the tool in this drome. There were two cases of hypoactive delirium that population. The CAP-D was easily administered in our were not detected by the screen. In one case, the subject preverbal and developmentally delayed subjects. had been delirious and then was sedated for a procedure; the screener missed the hypoactive delirium present in the post-sedation period. The second was a case in which the Specificity in diagnosing delirium CAP-D score was 8 at the time of the screen; when the psychiatric evaluation was performed several hours later, Although individually, each item in the tool can correlate the subject had hypoactive delirium. In both of these with other causes of acute CNS depression (such as seda- cases, it is possible that on repeat assessment, the CAP-D tion, agitation, pain, anxiety, and myriad other causes), we would have detected the delirium. In a larger validation posit that the combination of items in the revised tool, with study, we propose that this tool should be administered 1030 once a shift to facilitate identification of delirium in a rates of delirium across institutions, pediatric-specific risk timely manner in these rapidly evolving critically ill factors for development of delirium, safety, and efficacy of patients. To highlight the fluctuating course of delirium, therapeutic interventions, and the long-term neurocogni- we will preface the screen with the instruction: ‘‘Please tive and neuropsychiatric outcomes of pediatric delirium. answer the following questions based on your interactions with the patient over the course of your shift.’’ Conclusion Summary The proposed CAP-D is a feasible screening tool for the This study confirms the feasibility of administering the detection of delirium in children of all ages admitted to a CAP-D and comparing it to the gold standard for diag- PICU. If validated, the CAP-D may allow for accurate nosing delirium, the expert psychiatric evaluation based identification of delirious patients in a timely manner, on the DSM-IV criteria. This novel screening tool is easily facilitating appropriate interventions that may hasten res- administered, even in our challenging preverbal and olution of delirium and improve long-term functional developmentally delayed children. Although pilot studies, outcomes. As an easily administered screen, it avoids the by their very nature, are imprecise [32], the 29% preva- problem of referral bias and may enable the pediatric crit- lence rate of delirium in our study group suggests that ical care community to assess the true prevalence of pediatric delirium is a common and under-recognized delirium. This pilot study is a vital next step in the devel- problem that requires the PICU community’s attention. opment of a usable screening tool for pediatric delirium. Larger clinical studies are necessary to validate this tool, assess inter-rater reliability, stratify by age group and Acknowledgments This project is an ongoing collaboration developmental disability, and establish feasibility for use between the PICU, headed by Dr. Chani Traube, and the pediatric CL psychiatry service, headed by Dr. Gabrielle Silver. Drs. Traube by pediatric nurses for routine screening. Once validated, and Silver want to thank their division chiefs and mentors, Drs. the CAP-D may allow pediatric delirium research to move Bruce Greenwald and John Walkup, for their guidance and support. forward, with follow-up studies evaluating prevalence

References

1. Turkel SB, Tavare CJ (2003) Delirium 7. American Psychiatric Association 11. Rennick JE, Morin I, Kim D, Johnston in children and adolescents. (2000) Diagnostic and statistical CC, Dougherty G, Platt R (2004) J Neuropsychiatry Clin Neurosci manual of mental disorders, Text Identifying children at high risk for 15:431–435 Revised (DSM-IV-TR), 4th edn. psychological sequelae after pediatric 2. Prugh DG, Wagonﬁeld S, Metcalf D, American Psychiatric Association, intensive care unit hospitalization. Jordan K (1980) A clinical study of Washington, DC Pediatr Crit Care Med 4:411–412 delirium in children and adolescents. 8. Antonelli M, Azoulay E, Bonten M, 12. Leentjens AF, Schieveld JN, Leonard Psychosom Med 42:177–195 Chastre J, Citerio G, Conti G, De M, Lousberg R, Verhey FR, Meagher 3. Schieveld JN, Leroy PL, van Os J, Backer D, Gerlach H, Hedenstierna G, DJ (2008) A comparison of the Nicolai J, Vos GD, Leentjens AF Joannidis M, Macrae D, Mancebo J, phenomenology of pediatric, adult, and (2007) Pediatric delirium in critical Maggiore SM, Mebazaa A, Preiser JC, geriatric delirium. J Psychosom Res illness: phenomenology, clinical Pugin J, Wernerman J, Zhang H (2011) 64:219–223 correlates and treatment response in 40 Year in review in intensive care 13. Schieveld JNM, van der Valk JA, cases in the pediatric intensive care medicine 2010: I. Acute renal failure, Smeets I, Berghmans E, Wassenberg R, unit. Intensive Care Med 33:1033–1040 outcome, risk assessment and ICU Leroy PLMN, Vos GD, van Os J (2009) 4. Heatherhill S, Flisher AJ (2010) performance, sepsis, neuro intensive Diagnostic considerations regarding Delirium in children and adolescents: a care and experimentals. Intensive Care pediatric delirium: a review and a systemic review of the literature. Med 37:19–34 proposal for an algorithm for pediatric J Psychosom Res 68:337–344 9. DeCarvalho W, Fonseca M (2008) intensive care units. Intensive Care Med 5. Turkel S, Trzepacz P, Tavare C (2006) Pediatric delirium: a new diagnostic 35:1843–1849 Comparing symptoms of delirium in challenge of which to be aware. Crit 14. Kain ZN, Caldwell-Andrews AA, adults and children. Psychosomatics Care Med 36:1986–1987 Maranets I, McClain B, Gaal D, Mayes 47:320–324 10. Silver GS, Kearney JA, Kutko M, LC, Feng R, Zhang H (2004) 6. Schieveld J, Leentjens A (2005) Bartell A (2010) Infant delirium in Preoperative anxiety and emergence Delirium in severely ill young children pediatric critical care settings. Am J delirium and postoperative maladaptive in the pediatric intensive care unit. Psychiatry 167:1172–1177 behaviors. Anaesth Analg J Am Acad Child Adolesc Psychiatry 99:1648–1654 44:392–394 1031

15. Colville G, Kerry S, Pierce C (2008) 22. Heatherhill S, Fisher AJ, Nassen R 28. Schieveld JN (2011) On pediatric Children’s factual and delusional (2010) Delirium among children in an delirium and the use of the pediatric memories of intensive care. Am J urban sub-Saharan African setting. confusion assessment method for the Respir Care Med 177:976–982 J Psychosom Res 69:187–192 intensive care unit. Crit Care Med 16. Myhren H, Tøien K, Ekeberg O, 23. Trzepacz PT, Mittal D, Torres R, 39:220–221 Karlsson S, Sandvik L, Stokland O Kanary K, Norton J, Jimerson N (2001) 29. Sikich N, Lerman J (2004) (2009) Patients’ memory and Validation of the delirium rating scale Development and psychometric psychological distress after ICU stay revised-98: comparison with the evaluation of the pediatric anesthesia compared with expectations of the delirium rating scale and the cognitive emergence delirium scale. relatives. Intensive Care Med test for delirium. J Neuropsychiatry Anesthesiology 100:1138–1145 35:2078–2086 Clin Neurosci 3:433 30. Janssen N, Tan E, Staal M, Janssen E, 17. Schieveld JN, Staal M, Voogd L, 24. Ely EW, Margolin R, Francis J, May L, Leroy P, Lousberg R, van Os J, Fincken J, Vos G, van Os J (2010) Turman B, Dittus B, Speroff T, Gautam Schieveld J (2011) On the utility of Refractory agitation as a marker for S, Bernard G, Inouye S (2001) diagnostic instruments for pediatric pediatric delirium in very young infants Evaluation of delirium in critically ill delirium in critical illness: an at a pediatric intensive care unit. patients: validation of the confusion evaluation of the pediatric anesthesia Intensive Care Med 36:1982–1983 assessment method for the intensive emergence delirium scale, the delirium 18. Karnik NS, Joshi SV, Paterno C, Shaw care unit (CAM-ICU). Crit Care Med rating scale (DRS) 88, and the DRS- R (2007) Subtypes of pediatric 29:1370–1379 revised R-98. Intensive Care Med delirium: a treatment algorithm. 25. Ely EW, Inouye SK, Bernard GR, 37:1331–1337 Psychosomatics 48:253–257 Gordon S, Francis J, May L, Truman B, 31. Ely EW, Truman B, Shintani A, 19. Pandharipande P, Jackson J, Ely EW Speroff T, Gautam S, Margolin R, Hart Thomason JW, Wheeler AP, Gordon S, (2005) Delirium: acute cognitive RP, Dittus R (2001) Delirium in Francis J, Speroff T, Gautam S, dysfunction in the critically ill. Curr mechanically ventilated patients: Margolin R, Sessler CN, Dittus RS, Opin Crit Care 11:360–368 validity and reliability of the confusion Bernard GR (2003) Monitoring 20. Spronk PE, Riekerk B, Hofhuis J, assessment method for the intensive sedation status over time in ICU Rommes JH (2009) Occurrence of care unit (CAM-ICU). JAMA patients: reliability and validity of the delirium is severely underestimated in 286:2703–2710 Richmond agitation-sedation scale the ICU during daily care. Intensive 26. Smith HA, Fuchs DC, Pandharipande (RASS). JAMA 289:2983–2991 Care Med 35:1276–1280 PP, Barr FE, Ely EW (2009) Delirium: 32. Leon A, Davis L, Kraemer H (2011) 21. Antonelli M, Azoulay E, Bonten M, an emerging frontier in the management The role and interpretation of pilot Chastre J, Citerio G, Conti G, De of critically ill children. Crit Care Clin studies in clinical research. J Psychiatr Backer D, Lemaire F, Gerlach H, 25:593–614 Res 45:626–629 Hedenstierna G, Joannidis M, Macrae 27. Smith HAB, Boyd J, Fuchs C, Melvin D, Mancebo J, Maggiore SM, Mebazaa K, Berry P, Shintani A, Eden SK, A, Preiser JC, Pugin J, Wernerman J, Terrerr MK, Boswell T, Wolfram K, Zhang H (2010) Year in review in Sopfe J, Barr F, Pandharipande PP, Ely intensive care medicine 2009: II. EW (2011) Diagnosing delirium in Neurology, cardiovascular, critically ill children: validity and experimental, pharmacology and reliability of the pediatric confusion sedation, communication and teaching. assessment method for the intensive Intensive Care Med 36:412–427 care unit. Crit Care Med 39:150–157