Jan N. M. Schieveld

Delirium phenomenology: What can we learn from the symptoms of delirium?

OBJECTIVES This review focuses on phenomenological studies of delirium, including subsyndromal and prodromal concepts, and their relevance to other elements of clinical profile. METHODS A Medline search using the keywords delirium, phenomenology, and symptoms for new data articles published in English between 1998 and 2008 was utilized. The search was supplemented by additional material not identified by Medline but known to the authors. RESULTS Understanding of prodromal and subsyndromal concepts is still in its infancy. The characteristic profile can differentiate delirium from other neuropsychiatric disorders. Clinical (motoric) subtyping holds potential but more consistent methods are needed. Studies are almost entirely cross-sectional in design and generally lack comprehensive symptom assessment. Multiple assessment tools are available but are oriented towards hyperactive features and few have demonstrated ability to distinguish delirium from dementia. There is insufficient evidence linking specific phenomenology with etiology, pathophysiology, management, course, and outcome. CONCLUSIONS Despite the major advancements of the past decade in many aspects of delirium research, further phenomenological work is crucial to targeting studies of causation, pathophysiology, treatment, and prognosis. We identified eight key areas for future studies.

Late onset autism and anti-NMDA-receptor encephalitis

In December, 2009, a 9-year-old boy was admitted to our hospital with an acute onset of secondary generalised seizures. He had no medical or psychiatric history and functioned very well socially and academically. He presented with speech and swallowing diffi culties, which after 10 days developed into a severely agitated catatonic state with opisthotonic posturing, tonic posturing of limbs, insomnia, and dyskinesia. Initially the electroencephalogram showed a normal background pattern with epileptic discharges, and oligoclonal bands were present in cerebrospinal fl uid (CSF). Brain MRI and extensive blood tests were normal. The neurological diagnosis was atypical childhood epilepsy with centrotemporal spikes, for which oral corticosteroids and antiepileptic drugs were prescribed. His catatonia was treated with benzodiazepines. In January, 2010, our patient presented in a robotic state with complete mutism and negativism, and he did not respond to any form of contact. We provisionally diagnosed acute late onset autism with a diff erential diagnosis of childhood dis integrative disorder or early onset schizophrenia. Childhood disintegrative disorder, early onset schizophrenia, and late onset autism often share a fi nal common pathway: previous normal development, followed by sudden neuropsychiatric regression of social interaction and communication skills, and a decline in intelligence and daily activities. The disorders are sometimes misrecognised and collectively called as autistic disorder. Although judged to be functional psychiatric diagnoses, the marked deterioration and poor prognosis suggest an organic cause, especially in children with catatonia, a normal development up to at least 5 years of age, or both. In our patient, late onset autism was considered because: it is associated with neurological disorders; it is a known end stage of acquired brain injury; progression of symptoms was fast and severe, unlike in early onset schizophrenia; the absence of positive symptoms made schizophrenia less plausible; the age of onset and rare prevalence made chronic disintegrative disorder unlikely; and accompanying catatonic features were present. After extensive diagnostic assessments, our patient was fi nally diagnosed with anti-NMDA-receptor encephalitis on the basis of slightly raised anti-NMDA-receptor antibody titres in serum and highly raised titres in CSF. Clinical characteristics of this condition are acute major neuropsychiatric symptoms including anxiety, aggres sion, agitation, behavioural changes and catatonia, delusional thoughts, progressive speech deterioration, and hallucinations. Neurological symptoms such as dyskinesia, abnormal seizure-like movements, and diff use and profound autonomic instability have also been reported. Anti-NMDA-receptor encephalitis can occur in the context of malignant disease; however for our patient extensive oncological investigations were negative. Electroconvulsive therapy was given for the severe catatonic state, and monoclonal antibody treatment (rituximab) was started because of the unsatisfactory response to the initial treatment with benzodiazepines. The acquired autism gradually subsided, he spoke fl uently and was able to draw a happy picture (fi gure). In June, 2011, he only had some mild cognitive dysfunction. Childhood disintegrative disorder, early onset schizophrenia, late onset autism and all stages of anti-NMDAreceptor encephalitis share core symptoms, as in our patient. We suggest that anti-NMDA-receptor encephalitis might be a possible organic cause underlying these three disorders. Patients previously diagnosed with these diagnoses might need to be re-examined for anti-NMDAreceptor encephalitis. We suggest that forthcoming editions of DSM-5 and ICD-11 exclude and defi ne cases of regressive autism spectrum disorders due to anti-NMDAreceptor encephalitis.

Diagnostic considerations regarding pediatric delirium: a review and a proposal for an algorithm for pediatric intensive care units

ContextIf delirium is not diagnosed, it is unlikely that any effort will be made to reverse it. Given evidence for under-diagnosis, tools that aid recognition are required.ObjectiveRelating three presentations of pediatric delirium (PD) to standard criteria and developing a diagnostic algorithm.ResultsDelirium-inducing factors, disturbance of consciousness and inattention are common in PICU patients: a pre-delirious state is present in most. An algorithm is introduced, containing (1) evaluation of the sedation-agitation level, (2) psychometric assessment of behavior and (3) opinion of the caregivers.DiscussionIt may be argued that the behavioral focus of the algorithm would benefit from the inclusion of neurocognitive measures.LimitationsNo sufficiently validated diagnostic instrument covering the entire algorithm is available yet.ConclusionThis is the first proposal for a PD diagnostic algorithm. Given the high prevalence of pre-delirious states at the PICU, daily evaluation is mandatory. Future algorithmic refinement is urgently required.

A comparison of the phenomenology of pediatric, adult, and geriatric delirium

BACKGROUND The phenomenology of delirium in childhood is understudied. OBJECTIVE The objective of the study is to compare the phenomenology of delirium in children, adults and geriatric patients. POPULATION AND METHODS Forty-six children [mean age 8.3, S.D. 5.6, range 0-17 years (inclusive)], admitted to the pediatric intensive care unit of Maastricht University Hospital, with Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) delirium, underwent assessment with the Delirium Rating Scale (DRS). The scores are compared with those of 49 adult (mean age 55.4, S.D. 7.9, range 18-65 years) and 70 geriatric patients (mean age 76.2, S.D. 6.1, range 66-91 years) with DSM-IV delirium, occurring in a palliative care unit. Score profiles across groups, as well as differences in individual item scores across groups are analysed with multiple analysis of variance, applying a Bonferroni correction. RESULTS Although the range of symptoms occurring in all three groups was similar, DRS score profiles differed significantly across the three groups (Wilks lambda=0.019, F=804.206, P<.001). On item level, childhood delirium is characterized by a more acute onset, more severe perceptual disturbances, more frequent visual hallucinations, more severe delusions, more severe lability of mood, greater agitation, less severe cognitive deficits, less severe sleep-wake cycle disturbance, and less variability of symptoms over time. Adult and geriatric delirium do not differ in their presentations, except for the presence of more severe cognitive symptoms in geriatric delirium (P=.001). CONCLUSION Childhood delirium has a different course and symptom profile than adult and geriatric delirium. Adult and geriatric delirium differ only in the severity of cognitive symptoms.

Delirium in the Pediatric Patient: On the Growing Awareness of Its Clinical Interdisciplinary Importance

“Delirium in the elderly patient” was the headline of the important article Lipowski wrote in 1989.2 It was the first landmark article in the midst of a slowly growing number of publications, dating back to 1968. The response to these first, largely conceptual, articles ignited a fast-growing number of research studies regarding all the positive correlations between the onset of delirium on the one hand and all kinds of negative patient outcomes—for example, regarding quality of life in the cardiac intensive care unit, morbidity, and mortality—on the other. With the passing of time, the knowledge developed that these findings not only apply to cardiac patients, but to all adult and elderly patients with critical illness. Owing to critical illness, the brain can react with several neuropsychiatric presentations of which the most important in this context are sickness behavior, fever, epilepsy, catatonia, delirium, refractory agitation, and coma. In general, delirium is defined as a neuropsychiatric condition secondary to a general medical condition and/or its treatments. Unfortunately, the lack of a proper description of all the neuropsychiatric aspects of critical illness in general, and pediatric delirium (PD) in particular, still remains a serious drawback in the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition), the International Statistical Classification of Diseases, 10th Revision, and Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition).3 This is surprising because recent data suggest that PD also is a similarly serious condition, a finding that is in contradiction to the famous remark of Bleuler (1857-1939), “Children get delirious so often and quickly that this is of no importance to us.”4 However, this statement was made before World War II as opposed to both the concept of critical illness and the first intensive care unit, which both date back to around 1952 at the time of the polio epidemic in Copenhagen, Denmark. Bearing in mind that the first pediatric intensive care unit was set up in 1955 by Goran Haglund, Bleuler probably meant to discuss only the very common daily life occurrence of delirium febrile because the awareness of delirium in critical illness did not exist in those days. Then why is PD so important? Although the prevalence of delirium in adults varies between 40% and 80%, PD’s prevalence in critical illness is estimated to be at least 20%.5,6 Furthermore, research in the medical field has shown that there exists more overlap than differences between children and adults, not only in the reactions to critical illness, but in the presentation of delirium as well.3 Thus, it is remarkable that for a long time almost nobody realized that there is no scientific justification for not trying to generalize the knowledge available in adults to pediatric patients. There are at least 5 reasons for reevaluating PD: (1) According to Lipowski,2 delirium is “acute brain failure in man” and as the brain is the director of the autonomic nervous and endocrine systems the consequences of such a failure may be very severe; (2) the neurometabolic stress of residing in a state of delirium probably has a negative influence on the outcome and recovery from critical illness; (3) the hyperactive phase is accompanied by various other risks, such as pulling out of catheters and autodetubation; (4) being confronted with delirium can be very stressful for the patient who can experience frightening hallucinations and/or delusions; and (5) this holds for the family and interdisciplinary staff as well. The impact of a pediatric intensive care unit stay is best illustrated by the fact that one-third of pediatric patients are diagnosed as having serious posttraumatic stress disorder 3 months after hospital discharge. Finally, growing evidence suggests there is a positive association between pediatric disease severity measure scores and the onset of PD, implicating that disease severity is a risk factor for this condition.5 Taken together, we conclude that it is about time to change the way of thinking about PD. One of the most challenging questions remains how to properly diagnose PD in critically ill children.7 In adults, tools that rely on neurocognitive aspects can be used. However, because most critically ill pediatric patients are either younger than 3 years old, and make up an important group, and/or have intellectual disabilities, we have to rely on behavioral symptoms. In the last few years, an important stride forward had been made in the development of screening tools for PD. Most recently, the Cornell Assessment of Pediatric Delirium (CAP-D) was introduced, which is an adaptation of the Pediatric Anesthesia Emergence Delirium scale.5 Two major adjustments were made: (1) 3 items were added to enhance detection of hypoactive delirium and (2) developmental anchor points were formulated to be suitable for the screening of children who are very young and/or have a developmental delay. However, if these anchor points would not only be used to assess current functioning, but also premorbid baseline functioning, it would even become possible to assess person-specific changes over time. One way to further increase PD detection would be continuously monitoring the patient (eg, after each nurse shift). In this way, one can evaluate the well-known exclamation of parents and caregivers more VIEWPOINT

Pediatric illness severity measures predict delirium in a pediatric intensive care unit.

Context:Delirium in children is a serious but understudied neuropsychiatric disorder. So there is little to guide the clinician in terms of identifying those at risk. Objective:To study, in a pediatric intensive care unit (PICU), the predictive power of widely used generic pediatric mortality scoring systems in relation to the occurrence of pediatric delirium (PD). Design and Methods:Four-year prospective observational study, 2002–2005. Predictors used were the Pediatric Index of Mortality (PIM) and Pediatric Risk of Mortality (PRISM II). Setting:A tertiary 8-bed PICU in the Netherlands. Patients:877 critically ill children who were acutely, nonelectively, and consecutively admitted. Main Outcome Measure:Pediatric delirium. Main Results:Out of 877 children with mean age 4.4 yrs, 40 were diagnosed with PD (Cumulative incidence: 4.5%), 85% of whom (versus 40% with nondelirium) were mechanically ventilated. The area under the curve was 0.74 for PRISM II and 0.71 for the PIM, with optimal cut-off points at the 60th centile (PRISM: sensitivity: 76%; specificity: 62%; PIM: sensitivity: 82%; specificity: 62%). A PRISM II or PIM score above the 60th centile was strongly associated with later PD in terms of relative risk (PRISM II: risk ratio = 4.9; 95% confidence interval: 2.3–10.1; PIM: RR = 6.7; 95% confidence interval: 3.0–15.0). Given the low incidence of PD, values for positive predictive value were lower (PRISM II: 8.3%; PIM: 8.9%, rising to, respectively, 10.1% and 10.6% in mechanically ventilated patients) and values for negative predictive value were higher (PRISM II: 98.3%; PIM: 98.7%). Limitations:Given the relatively low incidence of delirium, a low detection rate biased toward the most severe cases cannot be excluded. Conclusions:Given the fact that PIM and PRISM II are widely used mortality scoring instruments, prospective associations with PD suggest additional value for ruling in, or out, patients at risk of PD.

Anti-NMDAR encephalitis: a new, severe and challenging enduring entity

Two girls, 15- and 17-year old, were consecutively and involuntarily admitted to the local child and adolescent psychiatric hospital with severe first onset psychosis. Due to refractory agitation, ongoing psychosis and insomnia, catatonic features, autonomic instability and the need for one-on-one guidance, the first girl was transferred to the PICU of an academic tertiary hospital and anti-NMDA receptor encephalitis was diagnosed. Given this experience nursing staff suspected, due to similarities in the clinical presentation and course, anti-NMDA receptor encephalitis in the second girl also and this proved to be true. The main clinical features, pharmacological and non-pharmacological treatment strategies and outcomes are presented and discussed. Perhaps, one ought to suspect anti-NMDA receptor encephalitis in every case of severe first onset psychosis with catatonic features.

Catatonia and refractory agitation in an updated flow chart for the evaluation of emotional-behavioral disturbances in severely ill children

Dear Editor, A flow chart regarding the evaluation and management of emotional– behavioral disturbances in severely ill children admitted to a pediatric intensive care unit (PICU) was presented to the scientific community of intensive care medicine in 2009 [1]. At that time we concluded that future refinements were expected to be unveiled. Despite the overall satisfactory experiences with the initial flow chart we increasingly experienced that it needs to be expanded with at least two additional items, i.e., (1) ‘‘Catatonic features’’ and (2) ‘‘Refractory agitation’’ (Fig. 1). Regarding catatonia: adding this item is a necessary adaptation due to the fact that there is an increasing amount of scientific literature, coming from both neurology and psychiatry, regarding a re-appraisal of catatonic features in general and of catatonia in pediatric neuro-psychiatry in particular [2]. These new developments have resulted in a proposed whole new chapter exclusively devoted to catatonia in the forthcoming Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [3]. Catatonia (meaning pervasive (motor) tension) is defined as a severe psychomotor disturbance with a primary presentation due to a mood disorder or schizophrenia. Besides, and in our context much more importantly, it has a secondary presentation attributable to any of the ‘acronymic’ I WATCH DEATH causes of delirium. The most characteristic features of catatonia are: catalepsy (bizarre postural fixity, especially of the limbs and head), opisthotonus (severe backwards hyperextension and rigidity of the whole body) and/or (a)typical motor agitation. Other symptoms are negativism (apparently motiveless resistance), stupor (extreme hypoactivity, minimally responsive to stimuli), and stereotypy (repetitive, non-goal directed motor activity such as finger-play, repeatedly touching, patting or rubbing self). According to a growing consensus, two or more symptoms suffice to diagnose catatonia. The above mentioned features deserve in their own right—especially in critical illness—a thorough evaluation and so we have added them to our updated flow chart. Treatment consists of treating the I WATCH DEATH causes and/or administering benzodiazepines in high dosages and/ or giving electroconvulsive therapy. Refractory agitation is a common and important issue in everyday critical care medicine. It can be defined as: lasting agitation after repeatedly checking the aforementioned flow chart items and its treatment results. If agitation still persists, this leads to the new and last box: ‘‘Refractory agitation’’. This extra item is of clinical importance: for it can be a sign of (1) ongoing agitation in the course of delirium—which is also a very known and common problem in acute adult psychosis [4]; (2) paradoxal agitation due to the administration of possibly any drug, but especially

First-onset psychosis, anti-NMDAR encephalitis, schizophrenia and Consultation–Liaison psychiatry

Anti-N-methyl-D-aspartate receptor (anti-NMDAR) encephalitis is a new, severe and possibly life-threatening disorder as illustrated by a 25% mortality or severe disability rate. Its neuropsychiatric presentation closely resembles schizophrenia. In this paper, the relationship of anti-NMDAR encephalitis to schizophrenia and its consequences in Consultation-Liaison service for diagnostic workup in patients with first-episode psychosis are addressed.

On pediatric delirium and the use of the Pediatric Confusion Assessment Method for the Intensive Care Unit.

Extracorporeal membrane oxygenation has been successfully used to support patients with cardiac arrest failing to respond to conventional cardiopulmonary resuscitation (CPR) (1, 2). Extracorporeal membrane oxygenation is generally deployed in these patients when all conventional therapies have been exhausted and death without other methods of cardiopulmonary support is imminent. Survival for patients using extracorporeal membrane oxygenation to support failed conventional CPR (external CPR [ECPR]) is variable and is reported to be approximately 37% in children and 27% in adults. The American Heart Association in their 2005 CPR guidelines recommended consideration of ECPR at institutions capable of rapidly deploying ECPR for patients with inhospital cardiac arrest receiving good-quality CPR and thought to have a reversible reason for their cardiac arrest (3). The use of ECPR is increasing both in children and adults, although survival rates have not improved despite increasing experience (1, 2). Given the unpredictable nature of the need for ECPR, institutions providing ECPR services maintain extracorporeal membrane oxygenation teams and equipment to capable of rapid ECPR deployment at all times and is thus expensive. Before ECPR becomes more widely adopted, it is critically important to evaluate the efficacy of ECPR in promoting survival and good long-term outcomes compared with conventional CPR therapies. It is equally important to know who may benefit from such therapy and thus it is also crucial to refine patient selection for ECPR so that ECPR is used effectively and for the right patients. The emergent nature at which ECPR is deployed makes conducting a randomized trial of ECPR vs. conventional CPR challenging and very difficult. In this issue of Critical Care Medicine, Shin et al (4) add to the existing knowledge that ECPR may indeed help in some patients failing to respond to conventional CPR therapies. They compare outcomes of patients with cardiac arrest requiring CPR for 10 mins managed with conventional CPR and with ECPR in their institution during January 2003 to June 2009 using a novel case–control study design. Cases (ECPR patients) and control subjects (conventional CPR patients) for this study were selected after matching based on a propensity score predicting the probability to being a candidate for ECPR. In doing so, they attempt to balance important covariates between cases and control subjects to minimize bias resulting from confounding variables much similar to a randomization. Unlike randomization, propensity score matching can only balance measured confounders so residual confounding from unmeasured covariates could still be present (5). However, when randomization is not possible like in the case of ECPR, using propensity score-based matching may be an excellent alternative to minimize bias resulting from confounding. In this article, a cohort of patients receiving CPR for 10 mins for cardiac arrest treated with or without ECPR was used to select a propensity score-based matched cohort of cases (ECPR cases) and control subjects (conventional CPR patients) for analysis (4). Using this cohort of matched patients, the authors show that the odds of mortality and having more than minimal neurologic disability was lower (odds ratio, 0.17; 95% confidence interval, 0.04–0.71) for patients supported with ECPR compared with those managed with conventional CPR. These improved outcomes for ECPR patients persisted even at 6 months after cardiac arrest. Improved survival was also seen in a subgroup of patients with cardiac disease, the largest diagnostic group in their cohort. The authors conclude that ECPR improved survival in patients with refractory inhospital cardiac arrest. These results shown here by Shin et al are similar to those published recently by Chen et al in Lancet in 2009 (6). So is there sufficient evidence to promote the use of ECPR? Not yet. There are several issues to consider when evaluating results from this article. First, in this cohort of patients, 70% of patients had primary cardiac disease and thus generalization to patients with other diagnoses is difficult. Second, the decision to deploy ECPR in 85 (21%) of the 406 patients during the study period was at the discretion of the primary physician conducting CPR. Thus, the majority of patients in this cohort were deemed not suitable to be ECPR candidates by the primary physician managing the event, yet they form the chosen comparison group. Third, the median predicted probability of being an ECPR candidate in the cohort was low (0.32 0.18 for both cohorts) indicating that many patients both in the ECPR and control groups had a low probability of being an ECPR candidate, or in other words, many were not good candidates for ECPR. Thus, the authors are comparing outcomes in a cohort of patients, many of whom may not have been good candidates for ECPR. A stratified analysis showing differences in survival outcomes between the control and ECPR groups based on high or low probability of being an ECPR candidate would have been useful to better understand efficacy of ECPR in promoting survival. Finally, although survival is improved for ECPR patients compared with those who were not, survival to hospital discharge in the ECPR cohort was only 34% raising the question of who should be a candidate for ECPR. *See also p. 1.