Cumulative Influence of Organ Dysfunctions and Septic State on Mortality of Critically Ill Children
http://www.100md.com
美国呼吸和危急护理医学 2005年第2期
Pediatric Intensive Care Unit, Jeanne de Flandre University Hospital
Department of Biostatistics, CERIM, Faculty of Medicine, Universitee de Lille
Department of Epidemiology and Public Health, Calmette University Hospital, Lille
Pediatric Intensive Care Unit, Necker-Enfants Malades University Hospital, Paris, France
Pediatric Intensive Care Unit, Sainte-Justine Hospital, Universitee de Montreeal, Montreeal, Quebec, Canada
ABSTRACT
The interaction between sepsis and multiple organ dysfunction syndrome is poorly defined in children. We analyzed by Cox regression models the cumulative influence of organ dysfunctions, using the pediatric logistic organ dysfunction (PELOD) score, and septic state (systemic inflammatory response syndrome or sepsis, severe sepsis, and septic shock) on mortality of critically ill children. We included 593 children (mortality rate: 8.6%) from three pediatric intensive care units; 514 patients had at least a systemic inflammatory response syndrome and 269 had two or more organ dysfunctions. Hazard ratio of death significantly increased with the severity of organ dysfunction, as estimated by the PELOD score, and the worst diagnostic category of septic state. Each increase of one unit in the PELOD score multiplied the hazard ratio by 1.096 (p < 0.0001); hazard ratio of diagnostic category was 9.039 (p = 0.031) for systemic inflammatory response syndrome or sepsis, 18.797 (p = 0.007) for severe sepsis and 32.572 (p < 0.001) for septic shock. Cumulative hazard ratio of death = (hazard ratio of PELOD score) x (hazard ratio of diagnostic category). We conclude that there is a cumulative accrual of the risk of death both with an increasing severity of organ dysfunction and an increasing severity of the diagnostic category of septic state.
Key Words: child critical care multiple organ dysfunction syndrome sepsis septic shock
Sepsis remains an important health problem in children, as it is in adults (1). A recent survey in the United States reported an annual incidence of severe sepsis of 0.56 cases/1,000 children with an overall mortality of 10.3% (2). Systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis, and septic shock, as defined by the American College of Chest Physicians/Society of Critical Care Medicine expert panel (3) with adaptation to children by Hayden (4, 5), are frequently observed in the pediatric intensive care unit (PICU). SIRS has been reported with an incidence up to 82% in the PICU (6eC8). Moreover, infection may lead to multiple organ dysfunction syndrome (MODS; defined as more than two organ dysfunctions); this is particularly true in children with septic shock who receive delayed treatment or have primary or acquired immunodeficiency (9). Only a few pediatric studies have reported that mortality of sepsis was linked to MODS (2, 10, 11), whereas Wilkinson and colleagues failed to demonstrate that sepsis significantly increased mortality in children with multiple organ failure (12).
In previous pediatric studies on sepsis and MODS, each organ failure or dysfunction was defined by its presence or absence (2, 6, 7, 10eC17). However, organ dysfunction is an ordinal process, and thus a score based on scaled and weighted rather than dichotomous variables should be more informative (3, 18eC21). Two scaled MODS scores have been proposed for critically ill children: the pediatric logistic organ dysfunction (PELOD) score (18, 22) and the pediatric multiple organ dysfunction score (published as an abstract) (23). Also, Johnston and colleagues used the International Classification of Diseases, ninth revision, clinical modification diagnostic and procedure codes to evaluate organ dysfunction in pediatric hospitalizations (24).
The purpose of this study was to analyze the cumulative influence of organ dysfunctions and septic state (SIRS, sepsis, severe sepsis, and septic shock) on mortality of critically ill children using the PELOD scoring system to estimate the severity of cases of MODS (18, 22). Some of the results of this study have been previously reported in the form of abstracts (25, 26).
METHODS
Patients
The study was run in three multidisciplinary tertiary care PICUs of university-affiliated hospitals (two French, one Canadian) (Table 1). The study was approved by the Ethics Committee of Ste-Justine Hospital and by the Comitee Consultatif de Protection des Personnes dans la Recherche Biomeedicale de Lille. The number of patients needed to develop the PELOD score was estimated to be 494 (18).
Between January 1997 and May 1997, all consecutive patients admitted to the participating PICUs were prospectively included, unless they met the following exclusion criteria: (1) age 18 years or older; (2) premature at entry into PICU; (3) pregnant; (4) length of stay in the PICU less than 4 hours; (5) admission in a state of continuous cardiopulmonary resuscitation without achieving stable vital signs for at least 2 hours; (6) transfer to another PICU; or (7) admission for scheduled procedures.
Data
Data were prospectively collected by research assistants or site investigators on a standardized case report form: age; sex; emergency admission to PICU; Pediatric Risk of Mortality III score (27); immunodeficiency; operative status; length of PICU stay; variables included in the PELOD score (18, 22) (see Table E1 in the online supplement); diagnostic criteria of SIRS, sepsis, severe sepsis, and septic shock as proposed by American College of Chest Physicians/Society of Critical Care Medicine consensus conference (3) and adapted to children by Hayden (4, 5) (Table E2); and PICU outcome (survival or death). Physiologic data from the preterminal period (the last 2 hours of life) were excluded. Patients were monitored daily until they died or were discharged from the PICU. Variables of the PELOD score were measured and collected as indicated in Table E1. Diagnostic criteria for SIRS, sepsis, severe sepsis, and septic shock were daily recorded during PICU stay, allowing identification of the worst diagnostic category reached by each patient.
Statistical Analysis
The validity of the database was checked by A.D. and B.G. before statistical analyses were undertaken. Interobserver reliability was assessed only during the validation step of the PELOD score, which included the three participating centers of the development step; kappa coefficients for organ dysfunction values ranged from 0.73 to 1 (22). All statistical analyses were done with SAS software (SAS Institute, Cary, NC). p Values < 0.05 were considered statistically significant. Comparisons of frequencies were made using the chi-square or Fisher exact probability test when appropriate, and multiple comparisons were performed using the method described by Fleiss (28). Density of incidence was defined as the number of new cases per unit person-time (1,000 patient-days). Density of incidence was used to compute the cumulative risk (R(t)) of acquiring SIRS, sepsis, severe sepsis, and septic shock for each day after entry into PICU according the following formula: , where t is the current day after admission in PICU. All patients were classified into the following independent categories: no SIRS, SIRS, sepsis, severe sepsis, or septic shock according to the worst septic state observed during their PICU stay. Because no significant difference in mortality was found between the SIRS and sepsis categories, these two groups were pooled in a same category for further analysis. Cumulative influence of organ dysfunctions and septic state on mortality was investigated by multivariate survival analysis using the Cox proportional hazard model. Patients may have had different lengths of stay. which is why the end point was survival from the day of admission in PICU to discharge. The validity of the proportional hazard assumption was assessed using time-dependent coefficients as suggested by Cox (29). Because it was expected that the delay reaching the worst septic state would be different across patients, which might modulate survival, we performed a multivariate Cox regression model using the PELOD score as fixed covariate and the worst septic state as time-dependent covariate. Then, we investigated separately by Cox regression models the influence of each organ dysfunction score (and also the number of organ dysfunctions) as fixed covariate and the worst septic state as time dependent categorical covariate (see the online supplement). Center effect was tested in each model and hazard ratios were adjusted when necessary.
RESULTS
Population Characteristics
There were 594 patients who fulfilled the inclusion criteria. One patient was excluded a posteriori because data were lacking with respect to the septic state. The final sample involved 593 patients with 51 deaths (8.6%). The readmission rate was 8%. The characteristics of the population are given in Table 1. The sample included 50 neonates (8%), 165 infants (28%), 300 children (51%), and 78 adolescents (13%). The male:female ratio was 1:4, and the median age was 30 months (Q1-Q3: 5eC93). There were 275 (46%) surgical patients and 289 (48%) ventilated patients. Sixty-three patients (11%) were diagnosed as immunodeficient. The median Pediatric Risk of Mortality III score was 3 (Q1-Q3: 0eC8), the median length of PICU stay was 3 days (Q1-Q3: 2eC7), and the median PELOD score was 10 (Q1-Q3: 1eC11). Two hundred and sixty nine patients (45%) had MODS defined as two or more organ dysfunction: 12 without SIRS and 257 with SIRS, sepsis, severe sepsis, or septic shock (Table 2).
Incidence of SIRS, Sepsis, Severe Sepsis, and Septic Shock
Among the 593 children, 514 (87%) was in a SIRS or a septic state: there were 376 cases of SIRS (63%), 103 cases of sepsis (17%), 17 cases of severe sepsis (3%), and 18 cases of septic shock (3%). The proportion of patients who reached their worst diagnostic category of the septic state during the first day in the PICU was 75% and the proportion was 88% within 48 hours after admission.
Cumulative Risk Over Time of Acquiring SIRS, Sepsis, Severe Sepsis, and Septic Shock
The risk over time of acquiring SIRS, sepsis, severe sepsis, or septic shock is shown in Figure 1.
Mortality Rate of Children with SIRS, Sepsis, Severe Sepsis, and Septic Shock
The observed mortality rates were 1% when SIRS was absent, 6% for SIRS, 8% for sepsis, 35% for severe sepsis, and 67% for septic shock (SIRS versus sepsis: p = 0.34; sepsis versus severe sepsis: p = 0.007; severe sepsis versus septic shock: p < 0.0001; SIRS and sepsis versus severe sepsis: p < 0.0001). The hazard ratio of death of each diagnostic category, without adjustment with the PELOD score, was 7.43 (95% confidence interval [CI]: 1.01eC54.81; p = 0.049) for SIRS or sepsis, 27.40 (3.26eC230.44; p = 0.002) for severe sepsis and 61.74 (7.84eC486.11; p < 0.0001) for septic shock.
Cumulative Influence of MODS and Septic State on Mortality (Cox Models)
No center effect was found for the different Cox models apart from the neurologic organ dysfunction (Wald test, p = 0.03).
PELOD score.
Each increase of one unit in the PELOD score multiplied the hazard ratio by 1.096 (95% CI: 1.077eC1.116; p < 0.0001), and, thus the hazard ratio of a given PELOD score was 1.096PELOD score. Adjusted hazard ratio of each diagnostic category, taking into account the PELOD score, was 9.039 (95% CI: 1.227eC66.620; p = 0.031) for SIRS or sepsis, 18.797 (2.241eC157.693; p = 0.007) for severe sepsis, and 32.572 (4.179eC253.890; p < 0.001) for septic shock. The cumulative influence of the PELOD score and diagnostic categories in our population is depicted in Figure 2. As an example, a child with a PELOD score of 24 and severe sepsis had a hazard ratio of death = (1.09624) x (18.797) = 169.64.
Score of each organ dysfunction.
Each increase of one unit in organ dysfunction scores multiplied the hazard ratio by a value i, ranging from 1.113 to 2.692 (all p < 0.05), depending on the considered organ dysfunction (Table 3); thus, the hazard ratio of a given organ dysfunction score was value iorgan dysfunction score. Adjusted hazard ratios of each diagnostic category of septic state for the corresponding organ dysfunction are given with their 95% CIs in Table 3. All theoretical values of cumulative hazard ratio of death of each organ dysfunction score and diagnostic categories are given in Table E3. As an example, a child with a respiratory organ dysfunction score of 10 and septic shock had a hazard ratio of death = (1.11310) x (53.022) = 154.67.
Number of organ dysfunctions.
Each organ dysfunction multiplied the hazard ratio by 2.374 (95% CI: 1.916eC2.940; p < 0.0001) and, thus the hazard ratio of organ dysfunction number was 2.374organ dysfunction number. Adjusted hazard ratio of diagnostic category was 8.916 (95% CI: 1.207eC65.872; p = 0.032) for SIRS or sepsis, 19.637 (2.345eC164.433; p = 0.006) for severe sepsis, and 47.039 (6.077eC364.093; p = 0.0002) for septic shock. As an example, a child with two organ dysfunctions had a hazard ratio of death = (2.3742) x (8.916) = 50.25 in case of SIRS or sepsis and a hazard ratio of death = (2.3742) x (47.039) = 265.11 in case of septic shock.
DISCUSSION
This prospective multicenter study is the first that used a validated pediatric MODS score based on scaled and weighted variables (PELOD score) in children with sepsis. It showed that the hazard ratio of death increased with the severity of MODS, as estimated by the PELOD score (or each organ dysfunction score) and the worst diagnostic category (from SIRS to septic shock). These hazard ratios are multiplicative and, thus the final hazard ratio is obtained by multiplying the hazard ratio of each one unit increase in the PELOD (or organ dysfunction score) by the hazard ratio of the diagnostic category of septic state. Translated in practical terms, this means that an increase in the PELOD score of 10 points was associated with a hazard ratio of death of 2.50 (1.09610 x 1) in children without SIRS and of 81.46 (2.501 x 32.572) in children with septic shock. Furthermore, each increase of one unit in the total PELOD score (or each organ dysfunction score) and the number of organ dysfunction significantly affected hazard ratio of death in the three considered diagnostic categories. Finally, it can be seen in Table 3 that neurologic organ dysfunction demonstrated the lowest hazard ratios of death for all septic states (sepsis, severe sepsis and septic shock) as compared with all the other organ dysfunctions. This could mean that death is less frequently attributable or associated with an infectious process in patients with neurologic dysfunction than in those with other dysfunctions.
In this study, we used the database that was prospectively collected for the development phase of the PELOD score (18) rather than the larger database collected while validating the PELOD score (22) because the presence or absence of septic states were annotated prospectively in the former, whereas it was not at all in the latter. However, both collections of data were quite similar with respect to mortality rate (6.4%), distribution of ages and median age (24 months), median length of stay (2 days), percentage of patients with MODS (53%),and median PELOD score (10) (22). These findings also compare with those of previous studies among children admitted in PICUs as reviewed by Johnston and colleagues (24).
We pooled patients with SIRS and those with sepsis, because mortality rates were similar, as previously reported both in children (11, 30) and adults (31, 32). Also, in statistical analysis, we used the PELOD score value rather than probability of death because MODS scoring systems were developed to be used as outcome measures rather than predictive indexes (18, 20).
Two Main Questions of the Study
The first question is: does sepsis influence the risk of death among children with MODS In the series of Wilkinson and colleagues, sepsis, as defined in the article (both bacteremia and clinical sepsis syndrome), did not increase mortality rates in the groups of children with organ system failure (46% versus 47% in those without sepsis) (12). Three studies gave different results (Table 4). Proulx and colleagues reported an incidence of sepsis, severe sepsis, and septic shock of 23%, 4%, and 2% respectively (7). In children having primary (n = 168) or secondary (n = 23) MODS, a trend toward distinct mortality rates according to the diagnostic categories (from no SIRS to septic shock: p = 0.057) was observed (7). Among the 84 children with MODS reported by Goh and colleagues, 11% had sepsis, 24% severe sepsis, and 18% septic shock; the MODS index (total number of organ dysfunction divided by 6 and expressed as percentage) varied among the diagnostic categories (sepsis 36%; severe sepsis 46%; septic shock 58%; p = 0.007) and the observed mortality also varied according to the different categories (sepsis 22%, severe sepsis 65%, septic shock 80%; p = 0.03) (15). In the series from Tantalean and colleagues, the mortality rate of children with MODS plus sepsis was greater than that of those without sepsis (52% versus 29%; p < 0.001) (17). As underlined by Cengiz and Zimmerman, the high prevalence of severe sepsis and septic shock in the latter series of critically ill children cared for in Lima might be explained by their nutrition status (malnutrition was observed in 33%) (34). Our results also demonstrated that mortality of children with MODS (more than two organ dysfunctions) was modulated by the severity of the septic state (Table 2), even though the number of patients in certain categories was small. In fact, in our patients with MODS, the death rate significantly increased with the increasing severity of the septic state (from 8% in children without SIRS to 71% in those with septic shock (Fisher exact test: p < 0.0001).
The second question is: does MODS increase the risk of death among children with sepsis Three series are to be considered (Table 4). In a retrospective study, Saez-Llorens and colleagues reported that the incidence of sepsis, severe sepsis, and septic shock was 4%, 11%, and 3%, respectively; the incidence of MODS in patients with sepsis was 24% (14). Children with septic states and MODS had the worst prognosis with 66% (versus 32% in children without MODS; p < 0.0001) and 84% (versus 37%; p < 0.0001) of them dying from severe sepsis and septic shock, respectively (14). In the series from Duke and colleagues, 64% of children with sepsis and septic shock and MODS at 48 hours after admission died, whereas 94% of those without MODS survived (10). Recently, Kutko and colleagues reviewed a subgroup of 80 children with 96 episodes of septic shock, of whom 71% had an oncologic disease (11). Although the mortality rate was not different between oncologic and nononcologic patients, it was higher in patients with MODS (19%; n = 70 episodes of shock) than in those without MODS (0%; n = 26 episodes of shock; p < 0.05). MODS was present in 100% of the patients who died (11). These data, which suggest a link between MODS and death in septic children, are in line with those from an American survey that included 9,675 cases of severe sepsis in which mortality ranged from 7% for children with one organ system failure (definitions not given) to 53% for those with more than four organ systems failure (2).
Furthermore, such a link between MODS and death has been reported in many conditions in which children are at risk of sepsis: children treated with high frequency oscillatory ventilation (36), after cardiac surgery (37, 38), with malignancies (39, 40), and after liver (41) and bone marrow (42) transplantation. Our study confirmed that children had a worse prognosis when MODS was present, whatever the diagnostic category of septic state.
What are the limitations and strengths of our study One limitation may be that only three PICUs participated in the data collection. However, incidence of SIRS, sepsis, severe sepsis, septic shock, and MODS were in accordance to previous reported studies in children, and mortality rates observed in the three participating PICUs were in the range of that reported in United States (27) and other countries (43). Another limitation is the use of the PELOD score calculated from the most abnormal values of variables measured during the entire PICU stay. Such an approach is advocated by experts to create and validate MODS scores against mortality (44), and maximum multiple organ dysfunction scores accurately predicted outcome in critically ill adults (45, 46). In future studies, the PELOD score (total and component scores) should be calculated daily or repeatedly to describe the variation of organ dysfunction with time, and obtain additional prognostic information (20, 47).
The representativity of the sample of patients collected in this study is probably good for the following reasons. The sites where the study was done are quite typical multidisciplinary European and North American universityeCaffiliated multidisciplinary tertiary care PICUs. The very high cumulative risk to contract a septic state during PICU stay was comparable in the collected sample to the risk reported by Proulx and colleagues (7), with 87 and 82% of patients at least having SIRS, respectively. Another strength of this study is the larger number of patients as compared with other pediatric studies that focused on the modulatory effect of MODS on mortality associated with SIRS, sepsis, severe sepsis, and septic shock (10, 11, 14). Moreover, we collected prospectively and daily the criteria of SIRS, sepsis, severe sepsis, and septic shock during the entire PICU stay. Finally, this is the first study that used a clinimetric MODS scale in children and analyzed the cumulative influence of organ dysfunctions and septic state on mortality. Even though the increase in the number of organ dysfunction was associated with an increase in the hazard ratio of death, we think that the PELOD score should be preferred to the number of organ dysfunction, which does not take into account the severity level of each organ dysfunction. Moreover, the 2001 Society of Critical Care Medicine/European Society of Intensive Care Medicine/American College of Chest Physicians/American Thoracic Society/Surgical Infection Society international sepsis definitions conference suggested to use the PELOD score to measure the severity of organ dysfunction developing over the course of critical illness of children (48, 49). This study reinforces this suggestion by demonstrating that PELOD quantification of MODS is an appropriate surrogate marker for mortality in pediatric septic states.
Conclusions
This study showed that there is a cumulative accrual of the risk of death both with an increasing severity of MODS—as estimated by the PELOD score, organ dysfunction scores, or the number of organ dysfunctions—and an increasing severity of the worst septic state (SIRS-sepsis, severe sepsis, septic shock) in critically ill children. This study also suggests that the PELOD score is a valid scale to measure organ dysfunctions.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
REFERENCES
Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001;29:1303eC1310.
Watson RS, Carcillo JA, Linde-Zwirble WT, Clermont G, Lidicker J, Angus DC. The epidemiology of severe sepsis in children in the United States. Am J Respir Crit Care Med 2003;167:695eC701.
Bone RC, Sibbald WJ, Sprung CL. The ACCP-SCCM consensus conference on sepsis and organ failure. Chest 1992;101:1481eC1483.
Hayden WR. Sepsis and organ failure definitions and guidelines. Crit Care Med 1993;21:1612eC1613. (Letter).
Hayden WR. Sepsis terminology in pediatrics. J Pediatr 1994;124:657eC658.
Proulx F, Gauthier M, Nadeau D, Lacroix J, Farrell CA. Timing and predictors of death in pediatric patients with multiple organ system failure. Crit Care Med 1994;22:1025eC1031.
Proulx F, Fayon M, Farrell CA, Lacroix J, Gauthier M. Epidemiology of sepsis and multiple organ dysfunction syndrome in children. Chest 1996;109:1033eC1037.
Despond O, Proulx F, Carcillo JA, Lacroix J. Pediatric sepsis and multiple organ dysfunction syndrome. Curr Opin Pediatr 2001;13:247eC253.
Carcillo JA. Pediatric septic shock and multiple organ failure. Crit Care Clin 2003;19:413eC440.
Duke TD, Butt W, South M. Predictors of mortality and multiple organ failure in children with sepsis. Intensive Care Med 1997;23:684eC692.
Kutko MC, Calarco MP, Flaherty MB, Helmrich RF, Ushay HM, Pon S, Greenwald BM. Mortality rates in pediatric septic shock with and without multiple organ system failure. Pediatr Crit Care Med 2003;4:333eC337.
Wilkinson JD, Pollack MM, Glass NL, Kanter RK, Katz RW, Steinhart CM. Mortality associated with multiple organ system failure and sepsis in pediatric intensive care unit. J Pediatr 1987;111:324eC328.
Wilkinson JD, Pollack MM, Ruttimann UE, Glass NL, Yeh TS. Outcome of pediatric patients with multiple organ system failure. Crit Care Med 1986;14:271eC274.
Saez-Llorens X, Vargas S, Guerra F, Coronado L. Application of new sepsis definitions to evaluate outcome of pediatric patients with severe systemic infections. Pediatr Infect Dis J 1995;14:557eC561.
Goh A, Lum L. Sepsis, severe sepsis and septic shock in paediatric multiple organ dysfunction syndrome. J Paediatr Child Health 1999;35:488eC492.
Montgomery VL, Strotman JM, Ross MP. Impact of multiple organ system dysfunction and nosocomial infections on survival of children treated with extracorporeal membrane oxygenation after heart surgery. Crit Care Med 2000;28:526eC531.
Tantalean JA, Leon RJ, Santos AA, Sanchez E. Multiple organ dysfunction syndrome in children. Pediatr Crit Care Med 2003;4:181eC185.
Leteurtre S, Martinot A, Duhamel A, Gauvin F, Grandbastien B, Nam TV, Proulx F, Lacroix J, Leclerc F. Development of a pediatric multiple organ dysfunction score: use of two strategies. Med Decis Making 1999;19:399eC410.
Joffe AR. Critical care medicine: major changes in dogma of the past decade. J Intensive Care Med 2001;16:177eC192.
Vincent JL, Ferreira F, Moreno R. Scoring systems for assessing organ dysfunction and survival. Crit Care Clin 2000;16:353eC366.
Vincent JL, Wendon J, Groeneveld J, Marshall JC, Streat S, Carlet J. The PIRO concept: O is for organ dysfunction. Crit Care 2003;7:260eC264.
Leteurtre S, Martinot A, Duhamel A, Proulx F, Grandbastien B, Cotting J, Gottesman R, Joffe A, Pfenninger J, Hubert P, et al. Validation of the paediatric logistic organ dysfunction (PELOD) score: prospective, observational, multicentre study. Lancet 2003;362:192eC197.
Graciano AL, Balko JA, Rahn DS, Giroir BP. Development and validation of a pediatric multiple organ dysfunction score (P-MODS) . Crit Care Med 2001;29:S260.
Johnston JA, Yi MS, Britto AT, Mrus JM. Importance of organ dysfunction in determining hospital outcomes in children. J Pediatr 2004;144:595eC601.
Leteurtre S, Grandbastien B, Duhamel A, Gauvin F, Proulx F, Vu Nam T, Martinot A, Lacroix J, Leclerc F. Etude de l'influence du syndrome de deefaillance multivisceerale, du SRIS, sepsis, sepsis seeveere et choc septique sur la mortalitee en reeanimation peediatrique . Reeanimation 2003;12(Suppl 3):153s.
Leteurtre S, Duhamel A, Proulx F, Grandbastien B, Gauvin F, Martinot A, Tu Vam T, Hubert P, Lacroix J, Leclerc F. Epideemiologie et risque de mortalitee des eetats septiques (SRIS, sepsis, sepsis seeveere et choc septique) en reeanimation peediatrique . Arch Pediatr 2004;11:732.
Pollack MM, Patel KM, Ruttimann UE. PRISM III: an updated pediatric risk of mortality score. Crit Care Med 1996;24:743eC752.
Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley and Sons; 1981.
Marubini E, Valsechi MG. Analyzing survival data from clinical trials and observational studies. New York: John Wiley and Sons; 1972.
Fischer J, Fanconi S. Systemic inflammatory response syndrome (SIRS) in pediatric patients. In: Tibboel D, van der Voort E, editors. Intensive care in childhood. A challenge to the future. Berlin: Springer-Verlag; 1996, pp. 239eC254.
Rangel-Frausto MS, Pittet D, Costigan M, Hwang T, Davis CS, Wenzel RP. The natural history of the systemic inflammatory response syndrome (SIRS). A prospective study. JAMA 1995;273:117eC123.
Muckart DJ, Bhagwanjee S. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions of the systemic inflammatory response syndrome and allied disorders in relation to critically injured patients. Crit Care Med 1997;25:1789eC1795.
Jafari HS, McCracken GH Jr. Sepsis and septic shock: a review for clinicians. Pediatr Infect Dis J 1992;11:739eC748.
Cengiz P, Zimmerman JJ. Prelude to pediatric multiple organ dysfunction syndrome: the golden hours concept revisited. Pediatr Crit Care Med 2003;4:263eC264.
Carcillo JA, Fields AI. American College of Critical Care Medicine Task Force Committee Members. Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock. Crit Care Med 2002;30:1365eC1378.
Brogan TV, Bratton SL, Meyer RJ, O'Rourke PP, Jardine DS. Nonpulmonary organ failure and outcome in children treated with high-frequency oscillatory ventilation. J Crit Care 2000;15:5eC11.
Shime N, Kageyama K, Ashida H, Tanaka Y. Application of modified sequential organ failure assessment score in children after cardiac surgery. J Cardiothorac Vasc Anesth 2001;15:463eC468.
Ben-Abraham R, Efrati O, Mishali D, Yulia F, Vardi A, Barzilay Z, Paret G. Predictors for mortality after prolonged mechanical ventilation after cardiac surgery in children. J Crit Care 2002;17:235eC239.
Heying R, Schneider DT, Korholz D, Stannigel H, Lemburg P, Gobel U. Efficacy and outcome of intensive care in pediatric oncologic patients. Crit Care Med 2001;29:2276eC2280.
Ben Abraham R, Toren A, Ono N, Weinbroum AA, Vardi A, Barzilay Z, Paret G. Predictors of outcome in the pediatric intensive care units of children with malignancies. J Pediatr Hematol Oncol 2002;24:23eC26.
Feickert HJ, Schepers AK, Rodeck B, Geerlings H, Hoyer PF. Incidence, impact on survival, and factors for multi-organ system failure in children following liver transplantation. Pediatr Transplant 2001;5:266eC273.
Lamas A, Otheo E, Ros P, Vazquez JL, Maldonado MS, Munoz A, Martos I. Prognosis of child recipients of hematopoietic stem cell transplantation requiring intensive care. Intensive Care Med 2003;29:91eC96.
Slater A, Shann F, Pearson G. Paediatric Index of Mortality (PIM) Study Group. PIM2: a revised version of the paediatric index of mortality. Intensive Care Med 2003;29:278eC285.
Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995;23:1638eC1652.
Pettila V, Pettila M, Sarna S, Voutilainen P, Takkunen O. Comparison of multiple organ dysfunction scores in the prediction of hospital mortality in the critically ill. Crit Care Med 2002;30:1705eC1711.
Jacobs S, Zuleika M, Mphansa T. The Multiple Organ Dysfunction Score as a descriptor of patient outcome in septic shock compared with two other scoring systems. Crit Care Med 1999;27:741eC744.
Cook R, Cook D, Tilley J, Lee K, Marshall J, for the Canadian Critical Care Trials Group. Multiple organ dysfunction: baseline and serial component scores. Crit Care Med 2001;29:2046eC2050.
Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G, for the International Sepsis Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med 2003;29:530eC538.
Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G, for the International Sepsis Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003;31:1250eC1256.(Francis Leclerc, Steephan)
Department of Biostatistics, CERIM, Faculty of Medicine, Universitee de Lille
Department of Epidemiology and Public Health, Calmette University Hospital, Lille
Pediatric Intensive Care Unit, Necker-Enfants Malades University Hospital, Paris, France
Pediatric Intensive Care Unit, Sainte-Justine Hospital, Universitee de Montreeal, Montreeal, Quebec, Canada
ABSTRACT
The interaction between sepsis and multiple organ dysfunction syndrome is poorly defined in children. We analyzed by Cox regression models the cumulative influence of organ dysfunctions, using the pediatric logistic organ dysfunction (PELOD) score, and septic state (systemic inflammatory response syndrome or sepsis, severe sepsis, and septic shock) on mortality of critically ill children. We included 593 children (mortality rate: 8.6%) from three pediatric intensive care units; 514 patients had at least a systemic inflammatory response syndrome and 269 had two or more organ dysfunctions. Hazard ratio of death significantly increased with the severity of organ dysfunction, as estimated by the PELOD score, and the worst diagnostic category of septic state. Each increase of one unit in the PELOD score multiplied the hazard ratio by 1.096 (p < 0.0001); hazard ratio of diagnostic category was 9.039 (p = 0.031) for systemic inflammatory response syndrome or sepsis, 18.797 (p = 0.007) for severe sepsis and 32.572 (p < 0.001) for septic shock. Cumulative hazard ratio of death = (hazard ratio of PELOD score) x (hazard ratio of diagnostic category). We conclude that there is a cumulative accrual of the risk of death both with an increasing severity of organ dysfunction and an increasing severity of the diagnostic category of septic state.
Key Words: child critical care multiple organ dysfunction syndrome sepsis septic shock
Sepsis remains an important health problem in children, as it is in adults (1). A recent survey in the United States reported an annual incidence of severe sepsis of 0.56 cases/1,000 children with an overall mortality of 10.3% (2). Systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis, and septic shock, as defined by the American College of Chest Physicians/Society of Critical Care Medicine expert panel (3) with adaptation to children by Hayden (4, 5), are frequently observed in the pediatric intensive care unit (PICU). SIRS has been reported with an incidence up to 82% in the PICU (6eC8). Moreover, infection may lead to multiple organ dysfunction syndrome (MODS; defined as more than two organ dysfunctions); this is particularly true in children with septic shock who receive delayed treatment or have primary or acquired immunodeficiency (9). Only a few pediatric studies have reported that mortality of sepsis was linked to MODS (2, 10, 11), whereas Wilkinson and colleagues failed to demonstrate that sepsis significantly increased mortality in children with multiple organ failure (12).
In previous pediatric studies on sepsis and MODS, each organ failure or dysfunction was defined by its presence or absence (2, 6, 7, 10eC17). However, organ dysfunction is an ordinal process, and thus a score based on scaled and weighted rather than dichotomous variables should be more informative (3, 18eC21). Two scaled MODS scores have been proposed for critically ill children: the pediatric logistic organ dysfunction (PELOD) score (18, 22) and the pediatric multiple organ dysfunction score (published as an abstract) (23). Also, Johnston and colleagues used the International Classification of Diseases, ninth revision, clinical modification diagnostic and procedure codes to evaluate organ dysfunction in pediatric hospitalizations (24).
The purpose of this study was to analyze the cumulative influence of organ dysfunctions and septic state (SIRS, sepsis, severe sepsis, and septic shock) on mortality of critically ill children using the PELOD scoring system to estimate the severity of cases of MODS (18, 22). Some of the results of this study have been previously reported in the form of abstracts (25, 26).
METHODS
Patients
The study was run in three multidisciplinary tertiary care PICUs of university-affiliated hospitals (two French, one Canadian) (Table 1). The study was approved by the Ethics Committee of Ste-Justine Hospital and by the Comitee Consultatif de Protection des Personnes dans la Recherche Biomeedicale de Lille. The number of patients needed to develop the PELOD score was estimated to be 494 (18).
Between January 1997 and May 1997, all consecutive patients admitted to the participating PICUs were prospectively included, unless they met the following exclusion criteria: (1) age 18 years or older; (2) premature at entry into PICU; (3) pregnant; (4) length of stay in the PICU less than 4 hours; (5) admission in a state of continuous cardiopulmonary resuscitation without achieving stable vital signs for at least 2 hours; (6) transfer to another PICU; or (7) admission for scheduled procedures.
Data
Data were prospectively collected by research assistants or site investigators on a standardized case report form: age; sex; emergency admission to PICU; Pediatric Risk of Mortality III score (27); immunodeficiency; operative status; length of PICU stay; variables included in the PELOD score (18, 22) (see Table E1 in the online supplement); diagnostic criteria of SIRS, sepsis, severe sepsis, and septic shock as proposed by American College of Chest Physicians/Society of Critical Care Medicine consensus conference (3) and adapted to children by Hayden (4, 5) (Table E2); and PICU outcome (survival or death). Physiologic data from the preterminal period (the last 2 hours of life) were excluded. Patients were monitored daily until they died or were discharged from the PICU. Variables of the PELOD score were measured and collected as indicated in Table E1. Diagnostic criteria for SIRS, sepsis, severe sepsis, and septic shock were daily recorded during PICU stay, allowing identification of the worst diagnostic category reached by each patient.
Statistical Analysis
The validity of the database was checked by A.D. and B.G. before statistical analyses were undertaken. Interobserver reliability was assessed only during the validation step of the PELOD score, which included the three participating centers of the development step; kappa coefficients for organ dysfunction values ranged from 0.73 to 1 (22). All statistical analyses were done with SAS software (SAS Institute, Cary, NC). p Values < 0.05 were considered statistically significant. Comparisons of frequencies were made using the chi-square or Fisher exact probability test when appropriate, and multiple comparisons were performed using the method described by Fleiss (28). Density of incidence was defined as the number of new cases per unit person-time (1,000 patient-days). Density of incidence was used to compute the cumulative risk (R(t)) of acquiring SIRS, sepsis, severe sepsis, and septic shock for each day after entry into PICU according the following formula: , where t is the current day after admission in PICU. All patients were classified into the following independent categories: no SIRS, SIRS, sepsis, severe sepsis, or septic shock according to the worst septic state observed during their PICU stay. Because no significant difference in mortality was found between the SIRS and sepsis categories, these two groups were pooled in a same category for further analysis. Cumulative influence of organ dysfunctions and septic state on mortality was investigated by multivariate survival analysis using the Cox proportional hazard model. Patients may have had different lengths of stay. which is why the end point was survival from the day of admission in PICU to discharge. The validity of the proportional hazard assumption was assessed using time-dependent coefficients as suggested by Cox (29). Because it was expected that the delay reaching the worst septic state would be different across patients, which might modulate survival, we performed a multivariate Cox regression model using the PELOD score as fixed covariate and the worst septic state as time-dependent covariate. Then, we investigated separately by Cox regression models the influence of each organ dysfunction score (and also the number of organ dysfunctions) as fixed covariate and the worst septic state as time dependent categorical covariate (see the online supplement). Center effect was tested in each model and hazard ratios were adjusted when necessary.
RESULTS
Population Characteristics
There were 594 patients who fulfilled the inclusion criteria. One patient was excluded a posteriori because data were lacking with respect to the septic state. The final sample involved 593 patients with 51 deaths (8.6%). The readmission rate was 8%. The characteristics of the population are given in Table 1. The sample included 50 neonates (8%), 165 infants (28%), 300 children (51%), and 78 adolescents (13%). The male:female ratio was 1:4, and the median age was 30 months (Q1-Q3: 5eC93). There were 275 (46%) surgical patients and 289 (48%) ventilated patients. Sixty-three patients (11%) were diagnosed as immunodeficient. The median Pediatric Risk of Mortality III score was 3 (Q1-Q3: 0eC8), the median length of PICU stay was 3 days (Q1-Q3: 2eC7), and the median PELOD score was 10 (Q1-Q3: 1eC11). Two hundred and sixty nine patients (45%) had MODS defined as two or more organ dysfunction: 12 without SIRS and 257 with SIRS, sepsis, severe sepsis, or septic shock (Table 2).
Incidence of SIRS, Sepsis, Severe Sepsis, and Septic Shock
Among the 593 children, 514 (87%) was in a SIRS or a septic state: there were 376 cases of SIRS (63%), 103 cases of sepsis (17%), 17 cases of severe sepsis (3%), and 18 cases of septic shock (3%). The proportion of patients who reached their worst diagnostic category of the septic state during the first day in the PICU was 75% and the proportion was 88% within 48 hours after admission.
Cumulative Risk Over Time of Acquiring SIRS, Sepsis, Severe Sepsis, and Septic Shock
The risk over time of acquiring SIRS, sepsis, severe sepsis, or septic shock is shown in Figure 1.
Mortality Rate of Children with SIRS, Sepsis, Severe Sepsis, and Septic Shock
The observed mortality rates were 1% when SIRS was absent, 6% for SIRS, 8% for sepsis, 35% for severe sepsis, and 67% for septic shock (SIRS versus sepsis: p = 0.34; sepsis versus severe sepsis: p = 0.007; severe sepsis versus septic shock: p < 0.0001; SIRS and sepsis versus severe sepsis: p < 0.0001). The hazard ratio of death of each diagnostic category, without adjustment with the PELOD score, was 7.43 (95% confidence interval [CI]: 1.01eC54.81; p = 0.049) for SIRS or sepsis, 27.40 (3.26eC230.44; p = 0.002) for severe sepsis and 61.74 (7.84eC486.11; p < 0.0001) for septic shock.
Cumulative Influence of MODS and Septic State on Mortality (Cox Models)
No center effect was found for the different Cox models apart from the neurologic organ dysfunction (Wald test, p = 0.03).
PELOD score.
Each increase of one unit in the PELOD score multiplied the hazard ratio by 1.096 (95% CI: 1.077eC1.116; p < 0.0001), and, thus the hazard ratio of a given PELOD score was 1.096PELOD score. Adjusted hazard ratio of each diagnostic category, taking into account the PELOD score, was 9.039 (95% CI: 1.227eC66.620; p = 0.031) for SIRS or sepsis, 18.797 (2.241eC157.693; p = 0.007) for severe sepsis, and 32.572 (4.179eC253.890; p < 0.001) for septic shock. The cumulative influence of the PELOD score and diagnostic categories in our population is depicted in Figure 2. As an example, a child with a PELOD score of 24 and severe sepsis had a hazard ratio of death = (1.09624) x (18.797) = 169.64.
Score of each organ dysfunction.
Each increase of one unit in organ dysfunction scores multiplied the hazard ratio by a value i, ranging from 1.113 to 2.692 (all p < 0.05), depending on the considered organ dysfunction (Table 3); thus, the hazard ratio of a given organ dysfunction score was value iorgan dysfunction score. Adjusted hazard ratios of each diagnostic category of septic state for the corresponding organ dysfunction are given with their 95% CIs in Table 3. All theoretical values of cumulative hazard ratio of death of each organ dysfunction score and diagnostic categories are given in Table E3. As an example, a child with a respiratory organ dysfunction score of 10 and septic shock had a hazard ratio of death = (1.11310) x (53.022) = 154.67.
Number of organ dysfunctions.
Each organ dysfunction multiplied the hazard ratio by 2.374 (95% CI: 1.916eC2.940; p < 0.0001) and, thus the hazard ratio of organ dysfunction number was 2.374organ dysfunction number. Adjusted hazard ratio of diagnostic category was 8.916 (95% CI: 1.207eC65.872; p = 0.032) for SIRS or sepsis, 19.637 (2.345eC164.433; p = 0.006) for severe sepsis, and 47.039 (6.077eC364.093; p = 0.0002) for septic shock. As an example, a child with two organ dysfunctions had a hazard ratio of death = (2.3742) x (8.916) = 50.25 in case of SIRS or sepsis and a hazard ratio of death = (2.3742) x (47.039) = 265.11 in case of septic shock.
DISCUSSION
This prospective multicenter study is the first that used a validated pediatric MODS score based on scaled and weighted variables (PELOD score) in children with sepsis. It showed that the hazard ratio of death increased with the severity of MODS, as estimated by the PELOD score (or each organ dysfunction score) and the worst diagnostic category (from SIRS to septic shock). These hazard ratios are multiplicative and, thus the final hazard ratio is obtained by multiplying the hazard ratio of each one unit increase in the PELOD (or organ dysfunction score) by the hazard ratio of the diagnostic category of septic state. Translated in practical terms, this means that an increase in the PELOD score of 10 points was associated with a hazard ratio of death of 2.50 (1.09610 x 1) in children without SIRS and of 81.46 (2.501 x 32.572) in children with septic shock. Furthermore, each increase of one unit in the total PELOD score (or each organ dysfunction score) and the number of organ dysfunction significantly affected hazard ratio of death in the three considered diagnostic categories. Finally, it can be seen in Table 3 that neurologic organ dysfunction demonstrated the lowest hazard ratios of death for all septic states (sepsis, severe sepsis and septic shock) as compared with all the other organ dysfunctions. This could mean that death is less frequently attributable or associated with an infectious process in patients with neurologic dysfunction than in those with other dysfunctions.
In this study, we used the database that was prospectively collected for the development phase of the PELOD score (18) rather than the larger database collected while validating the PELOD score (22) because the presence or absence of septic states were annotated prospectively in the former, whereas it was not at all in the latter. However, both collections of data were quite similar with respect to mortality rate (6.4%), distribution of ages and median age (24 months), median length of stay (2 days), percentage of patients with MODS (53%),and median PELOD score (10) (22). These findings also compare with those of previous studies among children admitted in PICUs as reviewed by Johnston and colleagues (24).
We pooled patients with SIRS and those with sepsis, because mortality rates were similar, as previously reported both in children (11, 30) and adults (31, 32). Also, in statistical analysis, we used the PELOD score value rather than probability of death because MODS scoring systems were developed to be used as outcome measures rather than predictive indexes (18, 20).
Two Main Questions of the Study
The first question is: does sepsis influence the risk of death among children with MODS In the series of Wilkinson and colleagues, sepsis, as defined in the article (both bacteremia and clinical sepsis syndrome), did not increase mortality rates in the groups of children with organ system failure (46% versus 47% in those without sepsis) (12). Three studies gave different results (Table 4). Proulx and colleagues reported an incidence of sepsis, severe sepsis, and septic shock of 23%, 4%, and 2% respectively (7). In children having primary (n = 168) or secondary (n = 23) MODS, a trend toward distinct mortality rates according to the diagnostic categories (from no SIRS to septic shock: p = 0.057) was observed (7). Among the 84 children with MODS reported by Goh and colleagues, 11% had sepsis, 24% severe sepsis, and 18% septic shock; the MODS index (total number of organ dysfunction divided by 6 and expressed as percentage) varied among the diagnostic categories (sepsis 36%; severe sepsis 46%; septic shock 58%; p = 0.007) and the observed mortality also varied according to the different categories (sepsis 22%, severe sepsis 65%, septic shock 80%; p = 0.03) (15). In the series from Tantalean and colleagues, the mortality rate of children with MODS plus sepsis was greater than that of those without sepsis (52% versus 29%; p < 0.001) (17). As underlined by Cengiz and Zimmerman, the high prevalence of severe sepsis and septic shock in the latter series of critically ill children cared for in Lima might be explained by their nutrition status (malnutrition was observed in 33%) (34). Our results also demonstrated that mortality of children with MODS (more than two organ dysfunctions) was modulated by the severity of the septic state (Table 2), even though the number of patients in certain categories was small. In fact, in our patients with MODS, the death rate significantly increased with the increasing severity of the septic state (from 8% in children without SIRS to 71% in those with septic shock (Fisher exact test: p < 0.0001).
The second question is: does MODS increase the risk of death among children with sepsis Three series are to be considered (Table 4). In a retrospective study, Saez-Llorens and colleagues reported that the incidence of sepsis, severe sepsis, and septic shock was 4%, 11%, and 3%, respectively; the incidence of MODS in patients with sepsis was 24% (14). Children with septic states and MODS had the worst prognosis with 66% (versus 32% in children without MODS; p < 0.0001) and 84% (versus 37%; p < 0.0001) of them dying from severe sepsis and septic shock, respectively (14). In the series from Duke and colleagues, 64% of children with sepsis and septic shock and MODS at 48 hours after admission died, whereas 94% of those without MODS survived (10). Recently, Kutko and colleagues reviewed a subgroup of 80 children with 96 episodes of septic shock, of whom 71% had an oncologic disease (11). Although the mortality rate was not different between oncologic and nononcologic patients, it was higher in patients with MODS (19%; n = 70 episodes of shock) than in those without MODS (0%; n = 26 episodes of shock; p < 0.05). MODS was present in 100% of the patients who died (11). These data, which suggest a link between MODS and death in septic children, are in line with those from an American survey that included 9,675 cases of severe sepsis in which mortality ranged from 7% for children with one organ system failure (definitions not given) to 53% for those with more than four organ systems failure (2).
Furthermore, such a link between MODS and death has been reported in many conditions in which children are at risk of sepsis: children treated with high frequency oscillatory ventilation (36), after cardiac surgery (37, 38), with malignancies (39, 40), and after liver (41) and bone marrow (42) transplantation. Our study confirmed that children had a worse prognosis when MODS was present, whatever the diagnostic category of septic state.
What are the limitations and strengths of our study One limitation may be that only three PICUs participated in the data collection. However, incidence of SIRS, sepsis, severe sepsis, septic shock, and MODS were in accordance to previous reported studies in children, and mortality rates observed in the three participating PICUs were in the range of that reported in United States (27) and other countries (43). Another limitation is the use of the PELOD score calculated from the most abnormal values of variables measured during the entire PICU stay. Such an approach is advocated by experts to create and validate MODS scores against mortality (44), and maximum multiple organ dysfunction scores accurately predicted outcome in critically ill adults (45, 46). In future studies, the PELOD score (total and component scores) should be calculated daily or repeatedly to describe the variation of organ dysfunction with time, and obtain additional prognostic information (20, 47).
The representativity of the sample of patients collected in this study is probably good for the following reasons. The sites where the study was done are quite typical multidisciplinary European and North American universityeCaffiliated multidisciplinary tertiary care PICUs. The very high cumulative risk to contract a septic state during PICU stay was comparable in the collected sample to the risk reported by Proulx and colleagues (7), with 87 and 82% of patients at least having SIRS, respectively. Another strength of this study is the larger number of patients as compared with other pediatric studies that focused on the modulatory effect of MODS on mortality associated with SIRS, sepsis, severe sepsis, and septic shock (10, 11, 14). Moreover, we collected prospectively and daily the criteria of SIRS, sepsis, severe sepsis, and septic shock during the entire PICU stay. Finally, this is the first study that used a clinimetric MODS scale in children and analyzed the cumulative influence of organ dysfunctions and septic state on mortality. Even though the increase in the number of organ dysfunction was associated with an increase in the hazard ratio of death, we think that the PELOD score should be preferred to the number of organ dysfunction, which does not take into account the severity level of each organ dysfunction. Moreover, the 2001 Society of Critical Care Medicine/European Society of Intensive Care Medicine/American College of Chest Physicians/American Thoracic Society/Surgical Infection Society international sepsis definitions conference suggested to use the PELOD score to measure the severity of organ dysfunction developing over the course of critical illness of children (48, 49). This study reinforces this suggestion by demonstrating that PELOD quantification of MODS is an appropriate surrogate marker for mortality in pediatric septic states.
Conclusions
This study showed that there is a cumulative accrual of the risk of death both with an increasing severity of MODS—as estimated by the PELOD score, organ dysfunction scores, or the number of organ dysfunctions—and an increasing severity of the worst septic state (SIRS-sepsis, severe sepsis, septic shock) in critically ill children. This study also suggests that the PELOD score is a valid scale to measure organ dysfunctions.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
REFERENCES
Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001;29:1303eC1310.
Watson RS, Carcillo JA, Linde-Zwirble WT, Clermont G, Lidicker J, Angus DC. The epidemiology of severe sepsis in children in the United States. Am J Respir Crit Care Med 2003;167:695eC701.
Bone RC, Sibbald WJ, Sprung CL. The ACCP-SCCM consensus conference on sepsis and organ failure. Chest 1992;101:1481eC1483.
Hayden WR. Sepsis and organ failure definitions and guidelines. Crit Care Med 1993;21:1612eC1613. (Letter).
Hayden WR. Sepsis terminology in pediatrics. J Pediatr 1994;124:657eC658.
Proulx F, Gauthier M, Nadeau D, Lacroix J, Farrell CA. Timing and predictors of death in pediatric patients with multiple organ system failure. Crit Care Med 1994;22:1025eC1031.
Proulx F, Fayon M, Farrell CA, Lacroix J, Gauthier M. Epidemiology of sepsis and multiple organ dysfunction syndrome in children. Chest 1996;109:1033eC1037.
Despond O, Proulx F, Carcillo JA, Lacroix J. Pediatric sepsis and multiple organ dysfunction syndrome. Curr Opin Pediatr 2001;13:247eC253.
Carcillo JA. Pediatric septic shock and multiple organ failure. Crit Care Clin 2003;19:413eC440.
Duke TD, Butt W, South M. Predictors of mortality and multiple organ failure in children with sepsis. Intensive Care Med 1997;23:684eC692.
Kutko MC, Calarco MP, Flaherty MB, Helmrich RF, Ushay HM, Pon S, Greenwald BM. Mortality rates in pediatric septic shock with and without multiple organ system failure. Pediatr Crit Care Med 2003;4:333eC337.
Wilkinson JD, Pollack MM, Glass NL, Kanter RK, Katz RW, Steinhart CM. Mortality associated with multiple organ system failure and sepsis in pediatric intensive care unit. J Pediatr 1987;111:324eC328.
Wilkinson JD, Pollack MM, Ruttimann UE, Glass NL, Yeh TS. Outcome of pediatric patients with multiple organ system failure. Crit Care Med 1986;14:271eC274.
Saez-Llorens X, Vargas S, Guerra F, Coronado L. Application of new sepsis definitions to evaluate outcome of pediatric patients with severe systemic infections. Pediatr Infect Dis J 1995;14:557eC561.
Goh A, Lum L. Sepsis, severe sepsis and septic shock in paediatric multiple organ dysfunction syndrome. J Paediatr Child Health 1999;35:488eC492.
Montgomery VL, Strotman JM, Ross MP. Impact of multiple organ system dysfunction and nosocomial infections on survival of children treated with extracorporeal membrane oxygenation after heart surgery. Crit Care Med 2000;28:526eC531.
Tantalean JA, Leon RJ, Santos AA, Sanchez E. Multiple organ dysfunction syndrome in children. Pediatr Crit Care Med 2003;4:181eC185.
Leteurtre S, Martinot A, Duhamel A, Gauvin F, Grandbastien B, Nam TV, Proulx F, Lacroix J, Leclerc F. Development of a pediatric multiple organ dysfunction score: use of two strategies. Med Decis Making 1999;19:399eC410.
Joffe AR. Critical care medicine: major changes in dogma of the past decade. J Intensive Care Med 2001;16:177eC192.
Vincent JL, Ferreira F, Moreno R. Scoring systems for assessing organ dysfunction and survival. Crit Care Clin 2000;16:353eC366.
Vincent JL, Wendon J, Groeneveld J, Marshall JC, Streat S, Carlet J. The PIRO concept: O is for organ dysfunction. Crit Care 2003;7:260eC264.
Leteurtre S, Martinot A, Duhamel A, Proulx F, Grandbastien B, Cotting J, Gottesman R, Joffe A, Pfenninger J, Hubert P, et al. Validation of the paediatric logistic organ dysfunction (PELOD) score: prospective, observational, multicentre study. Lancet 2003;362:192eC197.
Graciano AL, Balko JA, Rahn DS, Giroir BP. Development and validation of a pediatric multiple organ dysfunction score (P-MODS) . Crit Care Med 2001;29:S260.
Johnston JA, Yi MS, Britto AT, Mrus JM. Importance of organ dysfunction in determining hospital outcomes in children. J Pediatr 2004;144:595eC601.
Leteurtre S, Grandbastien B, Duhamel A, Gauvin F, Proulx F, Vu Nam T, Martinot A, Lacroix J, Leclerc F. Etude de l'influence du syndrome de deefaillance multivisceerale, du SRIS, sepsis, sepsis seeveere et choc septique sur la mortalitee en reeanimation peediatrique . Reeanimation 2003;12(Suppl 3):153s.
Leteurtre S, Duhamel A, Proulx F, Grandbastien B, Gauvin F, Martinot A, Tu Vam T, Hubert P, Lacroix J, Leclerc F. Epideemiologie et risque de mortalitee des eetats septiques (SRIS, sepsis, sepsis seeveere et choc septique) en reeanimation peediatrique . Arch Pediatr 2004;11:732.
Pollack MM, Patel KM, Ruttimann UE. PRISM III: an updated pediatric risk of mortality score. Crit Care Med 1996;24:743eC752.
Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley and Sons; 1981.
Marubini E, Valsechi MG. Analyzing survival data from clinical trials and observational studies. New York: John Wiley and Sons; 1972.
Fischer J, Fanconi S. Systemic inflammatory response syndrome (SIRS) in pediatric patients. In: Tibboel D, van der Voort E, editors. Intensive care in childhood. A challenge to the future. Berlin: Springer-Verlag; 1996, pp. 239eC254.
Rangel-Frausto MS, Pittet D, Costigan M, Hwang T, Davis CS, Wenzel RP. The natural history of the systemic inflammatory response syndrome (SIRS). A prospective study. JAMA 1995;273:117eC123.
Muckart DJ, Bhagwanjee S. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions of the systemic inflammatory response syndrome and allied disorders in relation to critically injured patients. Crit Care Med 1997;25:1789eC1795.
Jafari HS, McCracken GH Jr. Sepsis and septic shock: a review for clinicians. Pediatr Infect Dis J 1992;11:739eC748.
Cengiz P, Zimmerman JJ. Prelude to pediatric multiple organ dysfunction syndrome: the golden hours concept revisited. Pediatr Crit Care Med 2003;4:263eC264.
Carcillo JA, Fields AI. American College of Critical Care Medicine Task Force Committee Members. Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock. Crit Care Med 2002;30:1365eC1378.
Brogan TV, Bratton SL, Meyer RJ, O'Rourke PP, Jardine DS. Nonpulmonary organ failure and outcome in children treated with high-frequency oscillatory ventilation. J Crit Care 2000;15:5eC11.
Shime N, Kageyama K, Ashida H, Tanaka Y. Application of modified sequential organ failure assessment score in children after cardiac surgery. J Cardiothorac Vasc Anesth 2001;15:463eC468.
Ben-Abraham R, Efrati O, Mishali D, Yulia F, Vardi A, Barzilay Z, Paret G. Predictors for mortality after prolonged mechanical ventilation after cardiac surgery in children. J Crit Care 2002;17:235eC239.
Heying R, Schneider DT, Korholz D, Stannigel H, Lemburg P, Gobel U. Efficacy and outcome of intensive care in pediatric oncologic patients. Crit Care Med 2001;29:2276eC2280.
Ben Abraham R, Toren A, Ono N, Weinbroum AA, Vardi A, Barzilay Z, Paret G. Predictors of outcome in the pediatric intensive care units of children with malignancies. J Pediatr Hematol Oncol 2002;24:23eC26.
Feickert HJ, Schepers AK, Rodeck B, Geerlings H, Hoyer PF. Incidence, impact on survival, and factors for multi-organ system failure in children following liver transplantation. Pediatr Transplant 2001;5:266eC273.
Lamas A, Otheo E, Ros P, Vazquez JL, Maldonado MS, Munoz A, Martos I. Prognosis of child recipients of hematopoietic stem cell transplantation requiring intensive care. Intensive Care Med 2003;29:91eC96.
Slater A, Shann F, Pearson G. Paediatric Index of Mortality (PIM) Study Group. PIM2: a revised version of the paediatric index of mortality. Intensive Care Med 2003;29:278eC285.
Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995;23:1638eC1652.
Pettila V, Pettila M, Sarna S, Voutilainen P, Takkunen O. Comparison of multiple organ dysfunction scores in the prediction of hospital mortality in the critically ill. Crit Care Med 2002;30:1705eC1711.
Jacobs S, Zuleika M, Mphansa T. The Multiple Organ Dysfunction Score as a descriptor of patient outcome in septic shock compared with two other scoring systems. Crit Care Med 1999;27:741eC744.
Cook R, Cook D, Tilley J, Lee K, Marshall J, for the Canadian Critical Care Trials Group. Multiple organ dysfunction: baseline and serial component scores. Crit Care Med 2001;29:2046eC2050.
Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G, for the International Sepsis Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Intensive Care Med 2003;29:530eC538.
Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G, for the International Sepsis Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003;31:1250eC1256.(Francis Leclerc, Steephan)