Pediatric risk of mortality (PRISM) score

The Pediatric Risk of Mortality (PRISM) score was developed from the Physiologic Stability Index (PSI) to reduce the number of physiologic variables required for pediatric ICU (PICU) mortality risk assessment and to obtain an objective weighting of the remaining variables. Univariate and multivariate statistical techniques were applied to admission day PSI data (1,415 patients, 116 deaths) from four PICUs. The resulting PRISM score consists of 14 routinely measured, physiologic variables, and 23 variable ranges. The performance of a logistic function estimating PICU mortality risk from the PRISM score, age, and operative status was tested in a different sample from six PICUs (1,227 patients, 105 deaths), each PICU separately, and in diagnostic groups using chi-square goodness-of-fit tests and receiver operating characteristic (ROC) analysis. In all groups, the number and distribution of survivors and nonsurvivors in adjacent mortality risk intervals were accurately predicted: total validation group (chi 2(5) = 0.80; p greater than .95), each PICU separately (chi 2(5) range 0.83 to 7.38; all p greater than .10), operative patients (chi 2(5) = 2.03; p greater than .75), nonoperative patients (chi 2(5) = 2.80, p greater than .50), cardiovascular disease patients (chi 2(5) = 4.72; p greater than .25), respiratory disease patients (chi 2(5) = 5.82; p greater than .25), and neurologic disease patients (chi 2(5) = 7.15; p greater than .10). ROC analysis also demonstrated excellent predictor performance (area index = 0.92 +/- 0.02).     

Comparison of the performance of two general and three specific scoring systems for meningococcal septic shock in children

OBJECTIVE: To evaluate the performance at admission to the pediatric intensive care unit (PICU) of five severity scores, two general (the Pediatric Risk of Mortality [PRISM] II and III scores) and three specific for meningococcal septic shock (Leclerc, Glasgow Meningococcal Septicemia Prognostic Score [GMSPS], and Gedde-Dahl's MOC score) in children with this condition. DESIGN: Multicenter, retrospective, cohort study. SETTING: The PICUs from four tertiary centers. PATIENTS: Patients were 192 children ranging in age from 1 month to 14 yrs consecutively admitted to the participating PICUs during a period of 12 yrs and 6 months (January 1983 to June 1995), who were diagnosed with presumed or confirmed meningococcal septic shock. Patients with a length of stay <2 hrs were excluded from the study. INTERVENTIONS: Clinical and laboratory data gathered during the first 2 hrs after admission were used to compute the scoring systems tested. MEASUREMENTS AND MAIN RESULTS: There were 66 deaths (34%). Neisseria meningitidis was cultured from 142 (74%) children. GMSPS and PRISM II provided the best discriminative capability, as measured by the area under the receiver operating characteristic curve (SEM): 0.816 (0.036) and 0.803 (0.041), respectively. The other three scores gave lower receiver operating characteristic areas: PRISM III = 0.777 (0.043), MOC = 0.775 (0.037), and Leclerc = 0.661 (0.045). There was a statistically significant difference between the areas under the receiver operating characteristic curve of GMSPS and Leclerc (p < .01) but not between the GMSPS and the remaining three scores. All five scores presented good calibration with no significant differences between observed and predicted mortality (Hosmer-Lemeshow goodness-of-fit test). CONCLUSIONS: The specific GMSPS and the general pediatric severity system PRISM II performed better than the other three scores, being appropriate tools to assess severity of illness at admission to the PICU in children with presumed meningococcal septic shock.     

Evaluating the prognosis of multiple, severely traumatized children in the intensive care unit

OBJECTIVE: To assess the Pediatric Risk of Mortality (PRISM) score and to identify other prognosis factors in severe, multiple trauma in children. DESIGN: Retrospective study over a 9-year period. SETTING: A Pediatric Intensive Care Unit (PICU) in a University Hospital. PATIENTS AND PARTICIPANTS: One hundred and thirty-three traumatized children, 8.6 years (8 months-16 years), were reviewed. INTERVENTIONS: None. MEASUREMENTS AND RESULTS: Pediatric Trauma Score (PTS), Injury Severity Score (ISS), New ISS (NISS), Glasgow Coma Scale (GCS) score, and PRISM were calculated. The areas under the Receiver Operating Characteristic (ROC Az) curves, were compared. Univariate and multivariate analyses were performed. The mortality rate was 25.6%. PRISM performed well for discrimination between survivors and non-survivors. Az PRISM 0.9387 (0.029) was not different from Az GCS score 0.9451 (0.027) (P=0.568), but was significantly different from Az ISS 0.756 (0.052) (P<0.001), Az NISS 0.7606 (0.051) (P<0.001), and Az PTS 0.8244 (0.047) (P=0.016). Death was significantly associated with head trauma (P=0.014), PRISM >35, PTS <5, GCS <7, and ISS or NISS >32 (P<0.00001). PRISM >35 (P=0.001) and GCS <7 (P=0.003) were independent risk factors of death. CONCLUSIONS: PRISM is a reliable tool for evaluating the prognosis of multiple, severely traumatized children. Its relative simplicity and the fact that it is extremely widespread as a general prognosis score in PICUs represent other arguments for its use. Due to the leading influence of head trauma on mortality, GCS, a score even simpler than PRISM, showed identical accuracy regarding survival prediction.     

The Pediatric Risk of Mortality (PRISM) Score and Injury Severity Score (ISS) for predicting resource utilization and outcome of intensive care in pediatric trauma.

OBJECTIVE: Mortality prediction in trauma is assessed using the Injury Severity Score (ISS) and Revised Trauma Score using Trauma Injury Severity Score (TRISS) methodology. The Pediatric Risk of Mortality (PRISM) score assesses mortality risk in critically ill children. We compared the ability of PRISM and ISS (using TRISS methodology) to predict resource utilization and outcome in pediatric trauma. DESIGN: Retrospective chart and database review. SETTING: Pediatric intensive care unit (PICU). PATIENTS: Consecutive admissions to a PICU over a 2-yr period. MEASUREMENTS AND MAIN RESULTS: Demographic data including PICU resource utilization and outcome were recorded. Data were recorded on 1,052 admissions (31 deaths), including 125 pediatric trauma patients (11 deaths). Patients were stratified into low- and high-risk categories based on PRISM and ISS scores. Patients with PRISM scores <6 and ISS scores <10 were classified as low risk. While both low-risk PRISM and ISS scores readily identified survivors, PRISM was the more sensitive indicator of resource utilization. PRISM, however, performed less well in determining risk-adjusted mortality as compared with ISS. CONCLUSION: PRISM readily stratifies pediatric trauma patients for resource utilization. PRISM appears to underestimate mortality in pediatric trauma as compared with ISS using TRISS methodology.     

Mortality prediction models for pediatric intensive care: comparison of overall and subgroup specific performance

Aim

To validate paediatric index of mortality (PIM) and pediatric risk of mortality (PRISM) models within the overall population as well as in specific subgroups in pediatric intensive care units (PICUs).

Methods

Variants of PIM and PRISM prediction models were compared with respect to calibration (agreement between predicted risks and observed mortality) and discrimination (area under the receiver operating characteristic curve, AUC). We considered performance in the overall study population and in subgroups, defined by diagnoses, age and urgency at admission, and length of stay (LoS) at the PICU. We analyzed data from consecutive patients younger than 16 years admitted to the eight PICUs in the Netherlands between February 2006 and October 2009. Patients referred to another ICU or deceased within 2 h after admission were excluded.

Results

A total of 12,040 admissions were included, with 412 deaths. Variants of PIM2 were best calibrated. All models discriminated well, also in patients <28 days of age (neonates), with overall higher AUC for PRISM variants (PIM = 0.83, PIM2 = 0.85, PIM2-ANZ06 = 0.86, PIM2-ANZ08 = 0.85, PRISM = 0.88, PRISM3-24 = 0.90). Best discrimination for PRISM3-24 was confirmed in 13 out of 14 subgroup categories. After recalibration PRISM3-24 predicted accurately in most (12 out of 14) categories. Discrimination was poorer for all models (AUC < 0.73) after LoS of >6 days at the PICU.

Conclusion

All models discriminated well, also in most subgroups including neonates, but had difficulties predicting mortality for patients >6 days at the PICU. In a western European setting both the PIM2(-ANZ06) or a recalibrated version of PRISM3-24 are suited for overall individualized risk prediction.     

Efficacy and outcome of intensive care in pediatric oncologic patients.

OBJECTIVE: Because the long-term survival of children with cancer has dramatically improved because of multimodal treatment strategies, intensive care medicine has become more relevant for these patients. This study was performed to assess the efficacy of intensive care medicine in newly diagnosed pediatric oncologic patients and in patients under ongoing oncologic treatment. DESIGN: A retrospective analysis of children admitted to the pediatric intensive care unit (PICU) of the University Hospital Duesseldorf for life-threatening conditions between 1995 and 1999 was performed to identify those patients with an oncologic condition. SETTING: University hospital. PATIENTS: A total of 123 patients were identified. Children admitted for uncomplicated postoperative care and children admitted after bone marrow transplantation were excluded from this analysis. Forty-eight patients could be divided into two groups. Group A contained children admitted to the PICU at the time of cancer diagnosis and group B children receiving ongoing oncologic treatment. INTERVENTIONS: The evaluation included diagnosis, risk factors, complications leading to PICU admission, PICU therapy, and outcome. Statistical analysis included evaluation of Pediatric Risk of Mortality (PRISM) and Therapeutic Intervention Scoring System (TISS) scores. MEASUREMENTS AND MAIN RESULTS: Respiratory insufficiency was the leading diagnosis for PICU admission, whereas in the remaining children cardiovascular insufficiency, renal failure, neurologic impairment, ileus, and tumor-associated complications led to PICU admission. The number of organ failures was correlated to outcome. All children but one of group A could be discharged from the PICU, whereas 12 of 35 children in group B died, despite intensive care treatment attempts. The PRISM and TISS scores at admission to the PICU were significantly higher in children who did not survive the period of intensive care treatment in group B. However, all patients with a PRISM score of >20 died. CONCLUSIONS: Diagnosis of cancer does not exclude potential benefit from intensive care medicine in these children, although severe complications might affect the prognosis.     

The prognosis of oncologic patients in the pediatric intensive care unit

Objective To evaluate the predicted mortality rate of oncologic patients in the PICU using the PRISM score and factors that might influence short-term outcomes.Design Retrospective study.Setting: Pedriatic ICU in a university hospitalPatients and Methods The medical charts of all oncologic patients admitted to the PICU during the period from January 1983 to December 1992 were reviewed.Main Results Over a period of 10 years, 51 oncologic patients were admitted on 57 occasions to the PICU. The mortality was 32%. This is significantly higher than the overall mortality in the PICU (8%). Comparison of observed and predicted mortality, derived from the PRISM score, using chi square goodness-of-fit tests showed a significantly higher observed mortality (x ²(5)=20.1,P<0.01). Patients admitted for circulatory failure and the highest mortality (47%), followed by those with respiratory failure due to tachypnea/cyanosis (36%), central nervous system deterioration (27%), respiratory failure due to ariway obstruction (25%), and metabolic disorders (20%). Of the 31 patients who needed mechanical ventilation, 17 died (55%), and when they needed inotropic support as well, the mortality increased to 69%. The mortality rose to 100% when the patient was admitted with a septic shock, necessitating mechanical ventilation and inotropic support. The median PRISM score was 5 in the survivor group and 18.5 in the non-survivor group; this difference was found to be significant using the Wilcoxon test (P=0.01). However, some patients with high scores were found in the survivor group, as well as some with low scores in the non-survivor group.Conclusion The decision to treat opcologic patients in a PICU remains difficult and has to be considered on an individual basis. However, oncologic patients do benefit from admission to the PICU. The PRISM score is not suitable for oncologic patients in the PICU, because it underestimates the observed mortality. Other factors like neutropenia, septic shock, the need for mechanical ventilation, and inotropic support should be taken into consideration.     

The core characteristics and nursing care activities in psychiatric intensive care units in Sweden

Internationally, research on psychiatric intensive care units (PICUs) commonly reports results from demographic studies such as criteria for admission, need for involuntary treatment, and the occurrence of violent behaviour. A few international studies describe the caring aspect of the PICUs based specifically on caregivers' experiences. The concept of PICU in Sweden is not clearly defined. The aim of this study is to describe the core characteristics of a PICU in Sweden and to describe the care activities provided for patients admitted to the PICUs. Critical incident technique was used as the research method. Eighteen caregivers at a PICU participated in the study by completing a semistructured questionnaire. In-depth interviews with three nurses and two assistant nurses also constitute the data. An analysis of the content identified four categories that characterize the core of PICU: the dramatic admission, protests and refusal of treatment, escalating behaviours, and temporarily coercive measure. Care activities for PICUs were also analysed and identified as controlling - establishing boundaries, protecting - warding off, supporting - giving intensive assistance, and structuring the environment. Finally, the discussion put focus on determining the intensive aspect of psychiatric care which has not been done in a Swedish perspective before. PICUs were interpreted as a level of care as it is composed by limited structures and closeness in care.     

脓毒症/严重脓毒症患儿维生素D水平与预后

目的 了解儿童重症监护病房(pediatric intensive care unit,PICU)脓毒症/严重脓毒症患儿维生素D缺乏程度;分析患者入院时维生素D水平与预后之间的关系.方法 本研究为多中心、前瞻性、观察性研究.以2013年3月1日至2014年3月30日期间入住3家儿童医院PICU符合危重症标准的脓毒症/严重脓毒症患儿作为研究对象.入组患者入院24h内留取静脉血2 mL,采用酶联免疫法检测血清25 (OH)维生素D水平.根据维生素D水平,将患者分为维生素D充足、轻度维生素D缺乏和重度维生素D缺乏三组.分析入院时维生素D水平与PICU住院时间、总住院时间、出院病死率、28d病死率和经济费用之间的关系.结果 共纳入病例194例.其中男117例(60.3％),女77例(39.7％).脓毒症/严重脓毒症患分别为96例和98例.出院病死率和28 d病死率分别为6.7％和24.2％.入院时中位维生素D水平9.79 ng/mL (5.32,18.46) ng/mL,维生素D缺乏率77.8％ (151/194),其中重度维生素D缺乏率达50.5％ (98/194).入院时存在重度维生素D缺乏的患者出院病死率显著增加(P=0.011).维生素D水平与PICU住院时间、总住院时间和经济花费间没有进行相关分析.结论 PICU脓毒症/严重脓毒症儿童中维生素D缺乏率高达77.8％;重度维生素D缺乏患者,出院病死率更高.     

Observational study of children admitted to United Kingdom and Republic of Ireland Paediatric Intensive Care Units after out-of-hospital cardiac arrest

AimsTo estimate the prevalence of children admitted after out-of-hospital cardiac arrest (OHCA) to UK and Republic of Ireland (RoI) Paediatric Intensive Care Units (PICUs) and factors associated with mortality to inform future clinical trial feasibility.MethodObservational study using a prospectively collected dataset of the Paediatric Intensive Care Audit Network (PICANet) of 33 UK and RoI PICUs (January 2003 to June 2010). Cases (0 to <16 years), with documented OHCA surviving to PICU admission and requiring mechanical ventilation were included. Main outcomes were prevalence for admission and death within PICU. Factors associated with mortality were examined with multiple logistic regression analysis.Results827 of 111,170 admissions (0.73%; 95% CI [0.48 to 0.98%]) were identified as children admitted following OHCA. PICU mortality for OHCA was 50.5% (418/827). Recruitment into an adequately sized clinical trial would not be feasible with the current prevalence rate. Characteristics at PICU admission associated with increased risk of death included; bilateral unreactive pupils, genetically inherited condition, inter-hospital transfer to PICU, requirement for vasoactive drugs and greater base deficit. Factors associated with reduced risk of death were submersion or a respiratory aetiology and pre-existing respiratory or cardiac conditions.ConclusionsLess than 120 children a year are admitted to PICUs in the UK and RoI after OHCA, limiting options for conducting UK intervention trials. The risk factors associated with mortality identified in this study will allow risk stratification in future studies.     

Pediatric cancer type predicts infection rate,need for critical care intervention,and mortality in the pediatric intensive care unit

Purpose

Up to 38 % of children with cancer require pediatric intensive care unit (PICU) admission within 3 years of diagnosis, with reported PICU mortality of 13–27 % far exceeding that of the general PICU population. PICU outcomes data for individual cancer types are lacking and may help identify patients at risk for poor clinical outcomes.

Methods

We performed a retrospective multicenter analysis of 10,365 PICU admissions of cancer patients no greater than 21 years old among 112 PICUs between 1 January 2009 and 30 June 2012. We evaluated the effect of cancer type, age, gender, genetic syndrome, stem cell transplantation, PRISM3 score, infections, and critical care interventions on PICU mortality.

Results

After excluding scheduled perioperative admissions, cancer patients represented 4.2 % of all PICU admissions (10,365/246,346), had overall mortality of 6.8 % (708/10,365) vs. 2.4 % (5,485/230,548) in the general PICU population (RR = 2.9, 95 % CI 2.7–3.1, p < 0.001), and accounted for 11.4 % of all PICU deaths (708/6,215). Hematologic cancer patients had greater median PRISM3 score (8 vs 2, p < 0.001), rates of sepsis (27 vs 9 %, RR = 2.9, 95 % CI 2.6–3.1, p < 0.001), and mortality (9.6 vs 4.5 %, RR = 2.1, 95 % CI 1.8–2.5, p < 0.001) compared to solid cancer patients. Among hematologic cancer patients, stem cell transplantation, diagnosis of acute myeloid leukemia, PRISM3 score, and infection were all independently associated with PICU mortality.

Conclusions

Children with cancer account for 4.2 % of PICU admissions and 11.4 % of PICU deaths. Hematologic cancer patients have significantly higher admission illness severity, rates of infections, and PICU mortality than solid cancer patients. These data may be useful in risk stratification for closer monitoring and patient counseling.     

Relationship of illness severity and length of stay to functional outcomes in the pediatric intensive care unit: a multi-institutional study

OBJECTIVE: The purpose of this study was to establish relationships between illness severity, length of stay, and functional outcomes in the pediatric intensive care unit (PICU) by using multi-institutional data. We hypothesized that a positive relationship exists between functional outcome scores, severity of illness, and length of stay. DESIGN: The study used a prospective multicentered inception cohort design. SETTING: The study was conducted in 16 PICUs across the United States that were member institutions of the Pediatric Critical Care Study Group of the Society of Critical Care Medicine. PATIENTS: In total, 11,106 patients were assessed, representing all admissions to these intensive care units for 12 consecutive months. MEASUREMENTS: Functional outcomes were measured by the Pediatric Overall Performance Category (POPC) and Pediatric Cerebral Performance Category (PCPC) scales. Both scales were assessed at baseline and discharge from the PICU. Delta scores were formed by subtracting baseline scores from discharge scores. Other measurements included admission Pediatric Risk of Mortality scores, age, operative status, length of stay in the PICU, and diagnoses. Interrater reliability was assessed by using a set of ten standardized cases on two occasions 6 months apart. MAIN RESULTS: Baseline, discharge, and delta POPC and PCPC outcome scores were associated with length of stay in the PICU and with predicted risk of mortality (p < .01). Incorporation of baseline functional status in multivariate length of stay analyses improved measured fit. Mild baseline cerebral deficits in children were associated with 18% longer PICU stays after controlling for other patient and institutional characteristics. Moderate and severe baseline deficits for both the POPC and PCPC score predict increased length of stay of between 30% and 40%. On the standardized cases, interrater consensus was achieved on 82% of scores with agreement to within one neighboring class for 99.7% of scores. CONCLUSIONS: These data establish current relationships for the POPC and PCPC outcome scales based on multi-institutional data. The reported relationships can be used as reference values for evaluating clinical programs or for clinical outcomes research.     

Long-stay patients in the pediatric intensive care unit

OBJECTIVE: Length of stay in the pediatric intensive care unit (PICU) is a reflection of patient severity of illness and health status, as well as PICU quality and performance. We determined the clinical profiles and relative resource use of long-stay patients (LSPs) and developed a prediction model to identify LSPs for early quality and cost saving interventions. DESIGN: Nonconcurrent cohort study. SETTING: A total of 16 randomly selected PICUs and 16 volunteer PICUs. PATIENTS: A total of 11,165 consecutive admissions to the 32 PICUs. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: LSPs were defined as patients having a length of stay greater than the 95th percentile (>12 days). Logistic regression analysis was used to determine which clinical characteristics, available within the first 24 hrs after admission, were associated with LSPs and to create a predictive algorithm. Overall, LSPs were 4.7% of the population but represented 36.1% of the days of care. Multivariate analysis indicated that the following factors are predictive of long stays: age <12 months, previous ICU admission, emergency admission, no CPR before admission, admission from another ICU or intermediate care unit, chronic care requirements (total parenteral nutrition and tracheostomy), specific diagnoses including acquired cardiac disease, pneumonia, and other respiratory disorders, having never been discharged from the hospital, need for ventilatory support or an intracranial catheter, and a Pediatric Risk of Mortality III score between 10 and 33. The performance of the prediction algorithm in both the training and validation samples for identifying LSPs was good for both discrimination (area under the receiver operating characteristics curve of 0.83 and 0.85, respectively), and calibration (goodness of fit, p = .33 and p = .16, respectively). LSPs comprised from 2.1% to 8.1% of individual ICU patients and occupied from 15.2% to 57.8% of individual ICU bed days. CONCLUSIONS: LSPs have less favorable outcomes and use more resources than non-LSPs. The clinical profile of LSPs includes those who are younger and those that require chronic care devices. A predictive algorithm could help identify patients at high risk of prolonged stays appropriate for specific interventions.     

Meningococcal disease: A comparison of eight severity scores in 125 children

Objective To investigate the accuracy of eight different prognostic scores (Stiehm, Niklasson, Leclerc, Garlund, the MOC score, Tesero, the Glasgow Meningococcal Septicaemia Prognostic Score (GMSPS) and Tüyzüs) in the prediction of fatal outcome in meningococcal disease.Design Combined prospective and retrospective study.Setting A 175-bed pediatric department of a university hospital providing secondary care to ±180,000 inhabitants and serving as a referral center. The Pediatric Intensive Care (14 beds) is one of the six PICUs in the Netherlands and provides tertiary care for children under 18 years.Patients During an 8-year period (1986–1994) 125 children (mean age 4 years, 10 months) with culture-proven meningococcal disease were studied: 34 patients presenting with meningitis, 33 patients with septic shock and 58 patients with meningitis and septic shock.Main results All eight scores discriminated above average between survivors and non-survivors, as expressed by the corresponding Receiver Operator Characteristic (ROC) curves. The area under the ROC curve (AUC) ranged from 0.74 for the Garlund score to 0.93 for the GMSPS. The GMSPS performed significantly better than its competitors, even after exclusion of the base deficit as one of the score components (AUC=0.92). It showed above average calibration when logistically transformed into a probability of mortality, and accurately identified a subgroup of patients with no mortality. None of the scores correctly identified non-survivors.Conclusion The GMSPS is a simple score that can be reliably used for risk classification and the identification of low-risk patients.     

Les scores de gravité généraux et de dysfonctions d’organes en réanimation pédiatrique : quoi de neuf en 2016 ?

Severity of illness and organ dysfunction scoring systems are increasingly used in adult and pediatric intensive care unit (PICU) to characterize disease severity and degree of organ dysfunction. In PICU, severity of illness scores are developed independently of diagnosis, taking into account the age of the different populations. Severity of illness scores were developed to better describe the severity of illness at baseline of groups of critically ill patients. Two systems are available: the pediatric index of mortality (PIM)—latest version PIM3 published in 2013, and the Pediatric RISk of Mortality (PRISM) system, which should be used in critically ill neonates, infants, children, or adolescents, (excluding premature infants)—latest version PRISM IV published in 2016. Organ dysfunction scores were developed to better describe the severity of illness during stay in the PICU. The Pediatric Logistic Organ Dysfunction score is the most commonly used pediatric organ dysfunction score—latest version PELOD-2 published in 2013. The purpose of this review is to provide an update about severity of illness and organ dysfunction scoring systems in PICU, to describe tools to evaluate performance and customization, and then discuss utility, limits and perspectives for scoring systems in PICU.     

Outcome of oncology patients in the pediatric intensive care unit

We evaluated the outcome of oncology patients in the Pediatric Intensive Care Unit (PICU) from a total of 72 consecutive admissions. Severity of illness and quantity of care were measured by the Physiologic Stability Index (PSI) and the Therapeutic Intervention Scoring System (TISS), respectively. The overall mortality was 51% and was especially high in patients admitted for acute organ system failure (OSF)-66%. Acute respiratory failure was the most frequent OSF (73%) and the most common cause for PICU admission. A poor outcome was associated with severe leucopenia (<1000 WBC/mm³, 91% mortality), acute renal failure (94% mortality) and central nervous system deterioration (83% mortality). When the outcome was predicted using a quantitative algorithm the observed mortality was significantly higher than the predicted for all admissions with a PSI higher than 5. Improved scoring systems are required to enable characterization of pediatric cancer patients admitted to the PICU.