Q J Med 2004; 97: 507-518
QJM vol. 97 no. 8 © Association of Physicians 2004; all rights reserved.
Acute and chronic paediatric intensive care patients: current trends and perspectives on resource utilization
From the Pediatric Intensive Care Unit, ‘Aghia Sophia’ Children's Hospital, Athens, Greece
Received 22 December 2003 and in revised form 28 April 2004
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Advances in paediatric critical care have resulted in increased survival of critically ill patients, many of whom require long-term ventilation as a means of life support.
Aim: To determine current trends in resource utilization, and problems in the care of acute and chronic paediatric intensive care patients.
Design: Open observational study.
Methods: We evaluated consecutive admissions (n = 1629) to a 10-bed paediatric intensive care unit (PICU) over a 5-year period. Three previously defined criteria for resource utilization were used: mean length of stay (LOS); length of mechanical ventilation (LOMV); and LOMV/LOS ratio.
Results: A total of 10 310 patient bed days and 5223 ventilator days were used. Mean LOS increased from 5.3 ± 12 days in 1998 to 8.7 ± 27 days in 2001 (p < 0.05). Although LOMV/LOS ratio (50.7%) was significantly correlated with Paediatric Risk of Mortality score (p < 0.0001), there was no significant change in mortality rate (12.6% vs. 12%). Patients hospitalized for >2 weeks (n = 320, 20%) used 55% of LOS and 57% of LOMV, in contrast to the 1298 (80%) hospitalized for <7 days, who used only 29% of LOS and 20% of LOMV. Patients hospitalized for >3 months (11, 0.7%) consumed 17% of LOS and 23% of LOMV. Five of these (45%) were eventually discharged home, two on ventilators.
Conclusions: The increasing trend of occupation of PICU bed and ventilator days by critically ill children may be related to the increasing trend for hospitalization of chronic care patients. Severity scoring systems were predictive of resource consumption, but not of the overall trend in mortality rate.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Advances in paediatric critical care and mechanical ventilation (MV) have resulted in an increased salvage of critically ill patients, a number of whom require long-term ventilation as a means of life support. In addition, in recent years, increasing numbers of chronically ill children with acute illnesses or exacerbations with prolonged weaning are occasionally being admitted to PICUs worldwide, with a big impact on staffing. Children receiving long-term ventilation are now treated in many different locations throughout the health care delivery system: the paediatric intensive care unit (PICU), specialized weaning units, and the home.1 These new issues, especially where they concern palliative care patients or patients with chronic (frequently undiagnosed) metabolic or mitochondrial diseases, set a major challenge not only to the staffing needs and cost, but also to the whole intensive care system structure and function.
Ideally, most chronic care patients should be managed at home for at least some of the palliative care period, if appropriate support from a home care team is available.2 Alternatively, intermediate care units might also promote efficient and effective care by increasing the flexibility of patient triage, using personnel efficiently, and providing cost-effective care. Inability to provide reliable step-down resources in a geographic area could reflect a general resource utilization imbalance, but might also clearly reveal current longitudinal trends of patients’ characteristics in a PICU setting.
We hypothesized that there are significant inter-relationships among diagnostic categories or severity of illness and bed and ventilator days, and that there is a trend towards increasing PICU resource consumption by chronic care patients. To determine the relation between temporal trends in basic clinical characteristics and resource utilization in a PICU that lacks an intermediate care unit, we performed an open prospective study. Outcomes among intensive and chronic care patients and ‘after discharge’ health care provision difficulties were evaluated. Resource utilization was expressed using the length of stay (LOS), the length of mechanical ventilation (LOMV), and the LOMV/LOS ratio.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
This prospective observational study was conducted at a university-affiliated teaching hospital in Athens (Aghia Sophia Children's Hospital, 760 beds). Subjects included all admissions to our 10 bed tertiary care PICU over five full calendar years (1 October 1996, the date of PICU establishment, to 1 October 2001). No intermediate care unit is available in our Hospital. As the study did not affect patient care, the institutional review board waived the need for informed parental consent.
Data collections and definitions
Clinical data were collected upon admission and on each consecutive PICU day. PICU length of stay was measured from admittance and discharge times, and coded into days. In patients with a PICU stay <24 h, LOS was defined to be one day. Three indicators of resource utilization in PICU patients were selected: LOS in the PICU (in days), LOMV (in days) and the LOMV/LOS ratio (longitudinal data). The following clinical characteristics were included as explanatory variables for resource utilization: comorbidity (pre-existing chronic disease), age groups, sex, and PICU outcome. Chronic comorbidity was classified as (i) none, independent of care of others for activities of daily life, (ii) chronic disease or handicapped, with genetic factors influencing their admission, partially or fully dependent on care of others, or (iii) cancer patient. Patients who were classified in a prioritization model3 in priorities 1 to 3 (unstable and in need of intensive treatment and monitoring; intensive monitoring and potentially immediate intervention; critically ill unstable patients who receive intensive care to relieve acute illness, but have a reduced likelihood of recovery because of underlying disease or nature of their acute illness) were characterized as chronic care patients if they needed hospitalization for >2 weeks. Meantime, a few chronic care patients, especially a small proportion of patients with untreatable cancer or severe comorbidity, were changed to priority 4 (patients with irreversible illness treated only under unusual circumstances for acute severe medical or surgical illness) representing a potential palliative care group of patients in the future. Because this group, however, could only be established at the discharge period, in our analyses we used the chronic comorbidity classification, which was more representative of the clinical status of patients during PICU hospitalization. We also recorded the primary reason for PICU admission using a modified classification for mutually exclusive disease categories.4
Final diagnostic information was coded by combining modifications of the International Classification of Diseases (appropriate for the age groups and the intensive care settings) and the Guidelines for Developing Admission and Discharge Policies for the Pediatric Intensive Care Unit.5 Severity of illness was assessed by the Pediatric Risk of Mortality (PRISM) Score,6 the Therapeutic Intervention Scoring System (TISS) modified for children,7 and indices of organ failure. Multiple organ system failure (MOSF) was defined using the criteria of Wilkinson et al.8
Statistical analysis
Normally distributed data are expressed as means ± SD, while non-normally distributed data are given as median and range. Statistical analysis was performed with a two-tailed t-test for normally distributed paired data after Levene's correction for equality of variances or by Mann-Whitney U/Wilcoxon rank sum W test for non-normally distributed data. The number of patients was used for analyses including mortality or demographics, whereas admissions were used for analyses including diseases, LOS, LOMV, and PRISM. Analysis of variance (ANOVA) was used when repeated measurements were performed. Scheffe or Bonferroni post hoc tests to detect differences within groups followed ANOVA. Probability values of p < 0.05 with two-tailed tests were considered significant. When a bivariate correlation of measuring linear association was calculated, Pearson's correlation coefficient was used. Fisher's exact test was used for the categorical data. In univariate analysis, we used t tests and one-way analysis of variance (if appropriate, with post hoc multiple comparison tests) to compare continuous variables (LOS and LOMV). Clinical characteristics that showed significant differences in univariate analysis were subsequently examined in stepwise multiple regression analysis with LOS, LOMV, and mortality (logistic regression) as dependent variables, using p = 0.05 for entry and p = 0.1 for removal. In this analysis, the contribution of each determinant, or resource utilization, was assessed in connection with all others. All analyses were done using SPSS for Windows, release 10.0.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patient characteristics
Overall, 1586 patients accounting for 1629 admissions were included. The patients were aged between 7 days and 18 years (mean 5.5 ± 5, median 4 years). The 25th, 50th and 75th age percentiles were 2, 4 and 9.5 years, respectively. The ratio of males (878/1629, 53.9%) to female (750/1629, 46.1%) was 1.43:1. Overall PICU mortality was 196/1629 (12.1%). Expected mortality by PRISM was 13.2 ± 15, PRISM score on admission was 11.9 ± 7 and TISS was 26.8 ± 13. Three hundred and sixty-five patients developed MOSF (23%). Survival did not differ significantly between boys (88.1%) and girls (87.7%). Specific diagnoses, with frequencies, severity of illness and mortality rates, are shown in Table 1.
|
Length of stay
The LOS in PICU was 6.3 ± 16 days (median 2 days, range 1–311 days), totalling 10 310 days. The 25th, 50th and 75th LOS percentiles were 1, 2 and 5 days, respectively. Overall, the LOS differed significantly between survivors (5.8 ± 14 days) and non-survivors (10.3 ± 27 days) (p < 0.03), as well as among patients without (3 ± 4 days) or with MV (10.2 ± 10 days, p < 0.0001). Stepwise linear regression analysis showed that LOS was independently positively associated with MV (p < 0.0001) and comorbidity (p = 0.001), and negatively with age (p = 0.001). Significant differences of LOS were also recorded among survivors and non-survivors of the various diagnostic groups (p < 0.0001). Compared with survivors, LOS was significantly shorter among non-survivors in the infection, trauma, and organic failure groups, and higher in the cancer, respiratory, neurological, and, especially, chronic metabolic disease groups (14.9 vs. 49.4 days).
Mechanical ventilation
Overall the LOMV was 3.2 ± 12 days, (median 0 days, range 0–311 days). Of the 1629 admissions, 751 patients needed MV (46.1%), whereas 878 were not supported with MV (53.9%). Among patients whose ventilation was mechanically supported, the LOMV was 6.9 ± 18 days, (median 2 days, range 1–311 days). The 25th, 50th and 75th LOMV percentiles were 1, 2 and 6 days, totalling 5223 days. The mean of individual patients’ LOMV/LOS ratios was 50.7%. There was a significant positive correlation between LOMV and LOS (r2 = 0.72, p < 0.0001), which was unchanged when repeated separately for survivors (r2 = 0.70) and non-survivors (r2 = 0.73). Overall, the LOMV differed significantly among survivors (2.5 ± 11 days) and non-survivors (8.7 ± 20 days) (p < 0.0001). Significant differences in LOMV were also recorded between the various diagnostic groups (p = 0.002). Compared with survivors, non-survivors of all diagnostic groups had longer LOMV, especially those with chronic metabolic diseases (3.9 vs. 36.3 days). Overall, 8% of admissions, which constituted the chronic care group of patients, mainly with metabolic or mitochondrial diseases, represented 20% and 21% of total LOS and LOMV, respectively, and, among all non-survivors, 9% of patients, which also constituted the chronic care group, represented 38% of total LOS and 44% of LOMV.
Comorbidity
Patients without significant comorbidity were a majority in our sample (1011/1629, 62.1%) (610 boys (60.3%) and 401 girls (39.7%)). Handicapped patients with chronic disease (361/1629, 22.1%) and cancer patients (257/1629, 15.8%) also included more boys (57.5% and 56.8%, respectively, NS). Survival differed significantly among patients without comorbidity (90%), with comorbidity (85.3%), or with cancer (84%) (p = 0.008). The LOS differed significantly among patients without comorbidity (5.6 ± 12 days), with comorbidity (9.3 ± 24 days), or cancer (5.1 ± 11 days), some of whom, however, died shortly after admission (p < 0.0001, Bonferroni post hoc tests for chronic vs. no comorbidity p = 0.001, chronic vs. cancer p = 0.005). ANOVA also showed significant differences among the three comorbidity groups regarding the LOMV (F = 4, p = 0.02, Bonferroni post hoc tests for chronic vs. no comorbidity p = 0.03).
Outcome: severity of illness
Overall mortality was 12%. Logistic regression analysis (forward stepwise conditional) revealed that young age (odds ratio 1.04, p < 0.03), comorbidity (3.4, p = 0.001), PRISM (5.8, p < 0.0001), TISS (1.4, p < 0.05), MOSF (1.7, p < 0.02), need for MV (22, p < 0.0001), and diagnosis (1.9, p = 0.002) were all independently associated with mortality. PRISM correlated significantly with LOS and LOMV/LOS ratio (p < 0.0001). These indices of resource consumption were also significantly correlated with the TISS (p < 0.003) and MOSF (p < 0.05).
Longitudinal data
Longitudinal analysis of resource utilization
The admitted number of patients in our unit varied between 330 and 344 per year, with the exception of 1997 (285 admissions), approaching one admission per day. Such stability, however, was not found in the longitudinal variation of total resource utilizations (sum of the LOS or LOMV of all patients per year), which almost doubled during the 5-year period, leading to a relative increase of LOMV/LOS ratios (Figure 1a), and associated with an increase of PRISM over time (Figure 1b). Mean LOS increased from 5.3 ± 12 days in 1998 to 8.7 ± 27 days in 2001 (post hoc p < 0.05), showing significant longitudinal difference (F = 2.5, p < 0.05). The means of LOMV differed longitudinally, but not significantly, increasing from 2.1 ± 5.3 days in 1998 to 4.6 ± 20 days in 2001 (post hoc p = 0.07). The LOMV/LOS ratio showed a trend towards an increase over time, following a similar trend of frequency of patients with comorbidity (Figure 2a). In contrast, there was no relation of the increasing trend over time of LOS and LOMV to the relatively stable longitudinal number of patients (Figure 2b). Although during the first years of our PICU establishment (1997–1999) the percentage of occupied beds was varied at relatively low levels (56.5%) the patient bed-days increased to 82% in 2001 and is now >90% (data not shown). The mortality rate showed a temporal trend for reduction from 12.6% to 12%, despite a an opposite temporal trend for increased severity of illness (PRISM from 11.3 ± 8.6 in 1997 to 13.2 ± 10.6 in 2001, F 2.1, p < 0.06).
|
|
Characteristics of LOS groups and longitudinal relations
The longitudinal data suggest a rather steady LOS in the years 1997–2000, followed by a marked increase in 2001, partially due to increasing number of hospitalized patients with various chronic diseases. This was in contrast to the 80% of patients who had been hospitalized for <7 days and were responsible for only 9% of LOS and 20.4% of LOMV. More than half of the system resources (55% of LOS and 57% of LOMV) were used by the 19.6% of patients who had been hospitalized for >2 weeks. Finally, 11 patients (0.7%) who had been hospitalized for >3 months (potentially palliative care patients) represented 17% of overall LOMV and 23% of total LOS. Additionally, the longer LOS groups exhibited a steady increase of mortality rates, PRISM, and frequencies of patients with chronic disease, with simultaneous decreases of relative percentages of previously healthy children (Table 2). These sequences and various interrelations of prolonged PICU hospitalization were more pronounced among patients supported with MV (all patients with LOS >3 months), than in patients who did not need MV (Figure 3).
|
|
All eleven patients with LOS >3 months required MV and had one or more episodes of nosocomial pneumonia. The LOMV of these patients occupied 70% of their LOS, while five patients could not be weaned off the ventilator. Seven of these patients (63.6%) were admitted during the last two years, and five of them (45.5%) were admitted in the last year. Two of them (Larsen syndrome, central apnoea syndrome after pneumococcal meningitis, 18.9%) could be discharged home on ventilators, having already occupied 125 and 311 bed and ventilator days in PICU. Three were discharged home on public or private healthcare home-support (metabolic encephalopathy, brain tumour with hydrocephalus, mitochondrial disease, 27.3%), having already spent 166, 105 and 126 PICU days, while repeated efforts to discharge them earlier had failed. Similar difficulties were encountered with the rest of the patients (LOS 92–253, LOMV 43–172), who died after a long course of exacerbations, improvements and relapses in PICU or the wards (mitochondrial disease, glucogoniasis, Joubert's syndrome, metabolic encephalopathy, William syndrome with intracerebral haemorrhage, and genetic progressive myoclonus epilepsy, 54.5%). The patients who had chronic illness of metabolic or congenital origin (72.7%) had a 75% mortality rate, whereas all patients with LOS >3 months suffering from severe complications of acute illness or tumour (27.3%) survived.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
This study provides information about the hospitalization of chronic care patients in a PICU setting and its influence on current trends over time in a PICU population. It also provides evidence concerning the outcome of patients with chronic diseases, and difficulties in discharge and home support policies. The hypothesized relationships among length of stay, length of mechanical ventilation, diagnostic categories, severity of illness, and mortality were confirmed. Ventilator days accounted for 50.7% of total patient-days, compared to 52.5% and 60% reported previously.9 The need for MV and the percentages of patients with chronic diseases affected the characteristics of the patient population, increasing the LOS and the resource usage, as reflected by the LOMV/LOS ratio. Thus, patients requiring MV at any time in the PICU stayed on average 218% longer than patients who did not undergo ventilation. The relationships were further confirmed in regression analyses controlling for sex, age, time, comorbidity, mortality, severity of illness, and disease. Remarkably, the 9% patients who stayed >2 weeks in the PICU, made up 53% of bed days and 62% of ventilator days. Further, the 0.7% of patients staying >3 months made up 17% of bed days and 23% of ventilator days. In a recent multicentre study, long-stay patients comprised from 2.1% to 8.1% of individual PICU patients, and occupied 15.2%–57.8% of bed days.10
Most importantly, our study provides evidence that mitochondrial, metabolic or neurological diseases correlate with significantly greater lengths of stay and increased risk of prolonged ventilatory support. The clinical spectrum of mitochondrial diseases has expanded dramatically in the last decade, setting up new diagnostic and therapeutic dilemmas in a critical care setting.11 In this study, patients with mitochondrial or metabolic disease were more likely on average to contribute to MV patient-days, and the same was true of other chronically ill patients, mainly with diseases of genetic origin. These findings have several implications.
First, the results presented in this study provide current trends for evaluating admission policies or for decisions regarding with the need for organizing intermediate units. Second, the cost of increased admissions and hospitalizations of patients with a poor prognosis for improved outcome or in a severely handicapped state (chronic illness) is greatly increased. Although it has been previously suggested that the patient-days that meet definitions of medical futility are not associated with high resource consumption (6.5% patients representing 2.7% patient-days) compared with non-futile care patient-days,12 the significant influence that chronic metabolic or mitochondrial disease has been exerting on LOS or LOMV/LOS has not been mentioned previously.13 In this study, patients with metabolic disease exhibited high resource consumption (8–9% patients representing a total of 20% total patient-days or 44% ventilator-days among non-survivors).
Most of the eleven chronically ill patients with ventilator dependence had a long LOS for of a variety of reasons, such as recurrent exacerbations of the primary illness and superimposed nosocomial infections with a prolonged critical illness and extremely slow improvement; lack of availability of a room with monitoring and ventilator support facilities; and the inability of health services to provide ventilator-dependent home-care support. Supporting the view that LOS is a positive risk factor for complication risk and mortality,14 these patients had a high mortality rate, and required frequent monitoring of vital signs, prolonged invasive or non-invasive ventilation and frequent nursing interventions. Although they periodically required invasive monitoring, they all needed more care than could be provided on a general ward. In a study of 706 surgical and medical ICU patients, this patient population accounted for approximately 22% of all ICU bed days.15 In a more recent study of 17 440 ICU admissions, 6180 patients were admitted strictly for intensive monitoring, though they had a <10% risk for requiring active treatment based on this monitoring.16 This low percentage contrasts with the 70% need for acute or chronic MV of our chronically ill patients, and the fact that 46% of the long stay patients were never be weaned from the ventilator. As a consequence, intermediate care has been proposed as a more appropriate means of resource utilization for these patients.17,18 The lack of an intermediate care unit in our hospital, may well have restricted flexibility in patient triage, periodically overcrowded the PICU, increased the need for nurse staffing, and deprived patients of both access to a cost-effective alternative to critical care unit admission, and more liberal family visitation policies, particularly for patients with a low risk of major complications.19 There is an escalating need for intensive care beds during the last years: the percentage of occupied beds increased from 56.5% in 1997 to 82% in 2001 and to >90% in 2003 (data not shown). In contrast, cost-containment initiatives focusing on futility in the PICU setting are unlikely to be successful, as only relatively small amounts of resources are used in providing futile care.20
We managed to discharge home (on ventilators or on some healthcare home-support) 50% of the severely handicapped patients (though very low number and small percentage of overall patients), but only after 105–311 bed days. The time of discharge was dictated by medical and social reasons, especially when they were stabilized and families and community were able to accept them as they were with their health problems. In our experience, primary care physicians are becoming increasingly involved in teamwork with paediatric intensivists and physiotherapists in the care of paediatric patients dependent on long-term or home mechanical ventilation. As the number of patients living at home on ventilator support increases, patients and their caregivers need the support of home healthcare nurses for this to be a successful and positive experience.21
There are further problems, however, associated with paediatric home ventilation, such as its negative psychosocial impact on family life, limited home care resources, the financial burden, differences in ethical perception by the local community, and the fact that the equipment is usually designed for adults. Family members, who must be available 24 h a day, especially need psychosocial support, professional home care, and voluntary assistance, which has not been previously been offered.22 Although the combination of monitoring systems and multimedia, which can store real-time data and send it daily to the hospital, could produce effective support system,23 sufficient care networks are essential to facilitate the various interventions necessary to maintain the patient with chronic respiratory failure at home, and should be promptly established.
In this study, if such an effective care network had been established, more families of patients with chronic diseases might had also elected to have their child supported at home, significantly reducing LOS, LOMV and PICU cost. In contrast, most of them expressed their willingness for their loved ones to be fully supported at hospital, whatever the outcome was.
Thus, the major barriers in home ventilation are the lack of multidisciplinary structures able to give regional assistance, and the reluctance of families to take responsibility for care. In some countries there is an increasing use of home ventilation, with a high responsibility placed on parents and families to solve the related ethical and technical problems.24 There are also a number of mechanical ventilators on the market that can be safely used even in very young children. There is also an increasing use of non-invasive ventilation in children, especially in obstructive sleep apnoea or paediatric neuromuscular disease.25
A severe limitation of this study is that it was conducted at a single unit. Presumably, issues such as resource utilization and care of children with chronic medical problems/technological dependence vary from country to country, being dependent on home care facilities, intermediate care units, and availability of long-term acute care hospitals.1 This limitation, however, also offers a unique benefit, since results based on homogeneous data and strategies are fully explored, showing the interrelations and trends over time of basic clinical characteristics and resource utilization in an intensive care setting more clearly, than a multicentre study could. Furthermore, by performing illness severity and therapeutic intervention adjustments in our study, we have the opportunity of comparison within and between PICUs, involving any outcomes or resources analysis. In this study, illness severity scoring systems (PRISM, TISS), and development of MOSF were predictive of LOS and LOMV. In accordance with these findings, previous studies have also shown that PRISM readily stratifies paediatric trauma patients for resource utilization26 and that significant patient-related predictors of LOS included PRISM and first-day use of mechanical ventilation.27 However, in our study, severity scoring systems predicted temporal trends of increased resource consumption, but not the overall trend of mortality rate. This might be explained by the very small proportion of potentially palliative care patients, who have an extremely small impact on the cumulative mortality rate, but a disproportionately high resource consumption.
Undoubtedly, new tensions arise as medical, psychological, and economic forces lead to the increasing use of high technology in the care of children outside of traditional health care institutions.28 Further studies are needed to characterize potentially palliative care services more fully, and to assess the quality of care provided by these services.29 In some areas, long-term acute-care hospitals have evolved their infrastructure to accommodate the growing complex medical patient population, a direct result of the expanded capability in supporting critically ill patients in the PICU setting. A service focusing on the transition from acute illness to recovery might provide an ideal environment for weaning patients from MV, and for patient care for other complex medical illnesses, including populations with oncological, neurological, metabolic, and infectious diseases.30 While promoting post-acute home care or a long-term acute-care hospital system, the availability of intermediate care units might reduce costs, reduce PICU length of stay without increasing hospital length of stay, and improve patient /family satisfaction by providing a physical environment that is quieter and calmer than the ICU.31
|
Footnotes |
|---|
Address correspondence to Dr G. Briassoulis, Pediatric Intensive Care Unit, ‘Aghia Sophia’ Children's Hospital, Thivon & Levadias street, Goudi 11527, Athens, Greece. e-mail: briassg{at}otenet.gr
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Stambouly JJ, Pollack MM, Ruttimann UE. An objective method to evaluate rationing of pediatric intensive care beds. Intensive Care Med 1991; 17:154–8.[CrossRef][Web of Science][Medline]
2. Lubin S. Palliative care–could your patient have been managed at home? J Palliat Care 1992; 8:18–22.[Medline]
3. Guidelines for intensive care unit admission, discharge, and triage. Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Crit Care Med 1999; 27:633–8.[CrossRef][Web of Science][Medline]
4. Knaus WA, Wagner DP, Draper EA, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest 1991; 100:1619–36.
5. American Academy of Pediatrics Committee on Hospital Care and Pediatric Section of the Society of Critical Care Medicine: Guidelines and levels of care for pediatric intensive care units. Crit Care Med 1993; 21:931–7.[Web of Science][Medline]
6. Pollack MM, Ruttimann UE, Getson PR. Pediatric risk of mortality (PRISM) score. Crit Care Med 1988; 16:1110–16.[Web of Science][Medline]
7. Keene AR, Cullen DJ. Therapeutic Intervention Scoring System: Update 1983. Crit Care Med 1983; 11:1–3.[Web of Science][Medline]
8. Wilkinson JD, Pollack MM, Glass NL, Kanter RK, Katz RW, Steihart CM. Mortality associated with multiple organ system failure and sepsis in pediatric intensive care unit. J Pediatr 1987; 111:324–8.[CrossRef][Web of Science][Medline]
9. Mickell JJ, Furgiuele TL. Acute mechanical ventilation: An index of the absolute need for pediatric intensive care unit beds. Crit Care Med 1988; 16:504–9.[Web of Science][Medline]
10. Marcin JP, Slonim AD, Pollack MM, Ruttimann UE. Long-stay patients in the pediatric intensive care unit. Crit Care Med 2001; 29:652–7.[CrossRef][Web of Science][Medline]
11. Clay AS, Behnia M, Brown KK. Mitochondrial disease: a pulmonary and critical-care medicine perspective. Chest 2001; 120:634–48.
12. Sachdeva RC, Jefferson LS, Coss-Bu J, Brody BA. Resource consumption and the extent of futile care among patients in a pediatric intensive care unit setting. J Pediatr 1996; 128:725–7.[CrossRef][Web of Science][Medline]
13. Ruttimann UE, Patel KM, Pollack MM. Length of stay an efficiency in pediatric intensive care units. J Pediatr 1998; 133:79–85.[CrossRef][Web of Science][Medline]
14. Stambouly JJ, McLaughlin LL, Mandel FS, Boxer RA. Complications of care in a pediatric intensive care unit: a prospective study. Intensive Care Med 1996; 22:1098–104.[Web of Science][Medline]
15. Henning RJ, McClish D, Daly B, Nearman H, Franklin C, Jackson D. Clinical characteristics and resource utilization of ICU patients: Implications for organization of intensive care. Crit Care Med 1987; 15:264–9.[Web of Science][Medline]
16. Zimmerman JE, Wagner DP, Knaus WA, Williams JF, Kolakowski D, Draper EA. The use of risk predictions to identify candidates for intermediate care units: Implications for intensive care utilization and cost. Chest 1995; 108:490–9.
17. Kalb PE, Miller DH. Utilization strategies for intensive care units. JAMA 1989; 261:2389–95.
18. Popovich J. Intermediate care units. Graded care options. Chest 1991; 99:4–5.
19. Douglas S, Daly B, Rudy E, Song R, Dyer MA, Montenegro H. The cost-effectiveness of a special care unit to care for the chronically critically ill. J Nursing Administration 1995; 25:47–53.
20. Nowotny ML. Life on a ventilator. Home Healthc Nurse 1999; 17:691–4.[Medline]
21. Goh Ay, Mok Q. Identifying futility in a paediatric critical care setting: a prospective observational study. Arch Dis Child 2001; 84:265–8.
22. van Kesteren RG, Velthuis B, van Leyden LW. Psychosocial problems arising from home ventilation. Am J Phys Med Rehabil 2001; 80:439–46.[CrossRef][Web of Science][Medline]
23. Matsumura T, Saito T, Miyai I, Nozaki S, Kang J. A pilot study to establish a support system using SpO2 monitoring system and videoconference system for patients under home mechanical ventilation. Rinsho Shinkeigaku 2000; 40:683–8.[Medline]
24. Hewitt-Taylor J. Children who require long-term ventilation: staff education and training. Intensive Crit Care Nurs 2004; 20:93–102.[CrossRef][Medline]
25. Katz S, Selvadurai H, Keilty K, Mitchell M, MacLusky I. Outcome of non-invasive positive pressure ventilation in paediatric neuromuscular disease. Arch Dis Child 2004; 89:121–4.
26. Castello FV, Cassano A, Gregory P, Hammond J. The Pediatric Risk of Mortality (PRISM) Score and Injury Severity Score (ISS) for predicting resource utilization and outcome of intensive care in pediatric trauma. Crit Care Med 1999; 27:985–8.[CrossRef][Web of Science][Medline]
27. Ruttimann UE, Pollack MM. Variability in duration of stay in pediatric intensive care units: a multiinstitutional study. J Pediatr 1996; 128:35–44.[CrossRef][Web of Science][Medline]
28. Lantos JD, Kohrman AF. Ethical aspects of pediatric home care. Pediatrics 1992; 89:920–4.
29. Pantilat SZ, Billings JA. Prevalence and structure of palliative care services in California hospitals. Arch Intern Med 2003; 163:1084–8.
30. Hotes LS, Kalman E. The evolution of care for the chronically critically ill patient. Clin Chest Med 2001; 22:1–11.[CrossRef][Web of Science][Medline]
31. Society of critical care medicine. Guidelines on admission and discharge for adult intermediate care units. Crit Care Med 1998; 26:607–10.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||