Performance of six severity-of-illness scores in cancer patients requiring admission to the intensive care unit: a prospective observational study

Introduction  

The aim of this study was to evaluate the performance of five general severity-of-illness scores (Acute Physiology and Chronic Health Evaluation II and III-J, the Simplified Acute Physiology Score II, and the Mortality Probability Models at admission and at 24 hours of intensive care unit [ICU] stay), and to validate a specific score – the ICU Cancer Mortality Model (CMM) – in cancer patients requiring admission to the ICU.     

Determinants of postintensive care unit mortality: a prospective multicenter study

OBJECTIVE: Six to 25 percent of patients discharged alive from the intensive care unit (ICU) die before hospital discharge. Although this post-ICU mortality may indicate premature discharge from a full ICU or suboptimal management in the ICU or ward, another factor may be discharge from the ICU as part of a decision to limit treatment of hopelessly ill patients. We investigated determinants of post-ICU mortality, with special attention to this factor. DESIGN: Prospective, multicenter, database study. SETTING: Seven ICUs in or near Paris, France. PATIENTS: A total of 1,385 patients who were discharged alive from an ICU after a stay of > or = 48 hrs; 150 (10.8%) died before hospital discharge. Decisions to withhold or withdraw life-sustaining treatments were implemented in the ICUs in 80 patients, including 47 (58.7%) who died before hospital discharge. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: In the univariate analysis, post-ICU mortality was associated with advanced age, poor chronic health status, severe comorbidities, severity and organ failure scores (Simplified Acute Physiology Score II, sepsis-related organ failure assessment, and Logistic Organ Dysfunction at admission and at ICU discharge), decisions to withhold or withdraw life-sustaining treatments, and Omega score (reflecting ICU resource utilization and length of ICU stay). Multivariate stepwise logistic regression identified five independent determinants of post-ICU mortality: McCabe class 1 (odds ratio, 0.388 [95% confidence interval, 0.26-0.58]), transfer from a ward (odds ratio, 1.89 [95% confidence interval, 1.27-2.80]), Simplified Acute Physiology Score II score at admission >36 (odds ratio, 1.57 [95% confidence interval, 1.6-2.33]), decisions to withhold or withdraw life-sustaining treatments (odds ratio, 9.64 [95% confidence interval, 5.75-16.6]), and worse sepsis-related organ failure assessment score at discharge (odds ratio, 1.11 [95% confidence interval, 1.03-1.18] per point). CONCLUSIONS: More than 10% of ICU survivors died before hospital discharge. Determinants of post-ICU mortality included variables reflecting patient status before and during the ICU stay. However, the most powerful predictor of post-ICU mortality was the decision to withhold or withdraw life-sustaining treatments in the ICU, suggesting that the decision has been made not to use the unique services of the ICU for these patients.     

Predictors of intensive care unit and hospital length of stay in diabetic ketoacidosis

OBJECTIVE: To determine the predictive value for prolonged intensive care unit (ICU) and hospital length of stay (LOS) in patients with diabetic ketoacidosis (DKA) of the Acute Physiology and Chronic Health Evaluation II (APACHE II) score and Logistic Organ Dysfunction System (LODS), and to identify associated characteristics. DESIGN: Prospective cohort, 18-month observation. SUBJECTS AND SETTING: All admissions to a 12-bed, inner-city, university-affiliated hospital, medical ICU from July 1999 to December 2000. MEASUREMENTS: Data for APACHE II and LODS scoring systems were collected within 24 hours of admission. Lengths of ICU and hospital stay were the primary outcomes. Prolonged ICU and hospital LOS were defined as 3 or more and 6 or more days. RESULTS: A total of 584 patients, mean age 49, 56% men, 82% African American were admitted to the ICU. At admission they had (mean +/-SD) APACHE II (18 +/- 10), LODS (5 +/- 4), and predicted mortality of 32% +/- 29%. DKA was the admitting diagnosis in 42 (7.6%) patients; they had lower APACHE II (12 +/- 6), LODS (2 +/- 1), and predicted mortality 5% +/- 5% than the general ICU population (all, P <.001). Hospital mortality in non-DKA patients was 18%; there were no deaths in patients with DKA. Among DKA patients, those with insulin noncompliance had a shorter hospital stay (2.8 +/- 1 d) than those with an underlying illness as the DKA trigger (4.8 +/- 3, P =.02). Between patients with DKA, regardless of the LOS, there were no significant differences in APACHE II, LODS, or predicted mortality. CONCLUSIONS: ICU-admitted patients with DKA are less ill, and have lower disease severity scores, mortality, and shorter length of ICU and hospital stay than non-DKA patients. Disease severity scores are not, but precipitating cause is, predictor associated with prolonged hospital LOS in patients with DKA.     

Outcome, functional autonomy, and quality of life of elderly patients with a long-term intensive care unit stay

OBJECTIVE: To examine the outcome, functional autonomy, and quality of life of elderly patients (> or = 70 yrs old) hospitalized for >30 days in an intensive care unit (ICU). DESIGN: Prospective cohort study. SETTING: A ten-bed, medical-surgical ICU in a 460-bed, acute care, tertiary, university hospital. PATIENTS: A consecutive cohort of 75 patients, >70 yrs old, admitted to the ICU from January 1, 1993, to August 1, 1998, for >30 days. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Severity at admission and of the underlying disease was estimated according to the Simplified Acute Physiologic Score (SAPS II), the Organ Dysfunction and/or Infection (ODIN) score, the McCabe score, and the Knaus classification. Therapeutic intensity was measured through the French Omega scoring system. All patients were mechanically ventilated during their ICU stay. Outcome measurements were made by two cross-sectional studies using telephone interviews on the first week of September 1996 and 1998 with a questionnaire including measures of functional capacity by Katz's Activities of Daily Living, modified Patrick's Perceived Quality of Life score, and the Nottingham Health Profile. The survival rate was 67% in the ICU and 47% in the hospital. A total of 30 patients were alive and able to participate in at least one of the cross-sectional studies. Independence in activities of daily living was decreased significantly after the ICU stay, except for feeding. However, most of the 30 patients remained independent (class A of the Activities of Daily Living index) with the possibility of going home. Perceived Quality of Life scores remained good, even if the patients estimated a decrease in their quality of life for health and memory. Return to society appeared promising regarding patient self respect and happiness with life. The estimated cost by survivor was of 55,272 EUR ($60,246 US). CONCLUSIONS: This study suggests that persistent high levels of ICU therapeutic intensity were associated with a reasonable hospital survival in elderly patients experiencing prolonged mechanical ventilatory support. These patients presented a moderate disability that influenced somewhat their perceived quality of life. These results are sufficient to justify prolonged ICU stays for elderly patients.     

Post-intensive care unit pediatric hospital stay and estimated costs

OBJECTIVE: For pediatric intensive care unit (ICU) survivors, to determine the proportion of hospital stay and estimated hospital costs accounted for by post-ICU care. DESIGN: Prospective study. SETTING: University teaching hospital. PATIENTS: Pediatric patients who survive an ICU admission. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Estimated relative daily costs were assumed as follows: ICU, with ventilator/ICU, not on ventilator/intermediate care unit/general pediatric hospital day, at 2:1:0.7:0.3, respectively. Estimated costs were expressed in arbitrary units as (number of days at each level of care) x (relative cost per day). The ICU phase was defined as the patient's first ICU admission only, and the post-ICU phase included intermediate care unit and general pediatric hospital days, as well as ICU readmission during the same hospitalization. Pre-ICU hospital activity was excluded from analysis. For 341 ICU survivors, post-ICU days (median, 4 days per patient) accounted for 62% of the total hospital stay. Post-ICU care accounted for one third of the total estimated hospital costs for ICU survivors. Patients with longer post-ICU stays could not be reliably identified at the time of ICU discharge according to their ICU length of stay, duration of mechanical ventilation in the ICU, age, ICU day 1 mortality probability, or diagnostic group (p>.05). CONCLUSIONS: Post-ICU care accounts for a substantial proportion of hospital stay and estimated costs for ICU survivors. These observations suggest that developing strategies to optimize hospital utilization at the time of ICU discharge may be important for controlling costs of recovery from critical illness.     

Postoperative troponin-T predicts prolonged intensive care unit length of stay following cardiac surgery

OBJECTIVE: To evaluate the use of postoperative cardiac troponin T (cTnT) for the prediction of prolonged intensive care unit length of stay following cardiac surgery. DESIGN: Prospective, single-center, observational cohort study of patients following cardiac surgical procedures. The enrollment period was from October through December 2000. Patients were enrolled on admission to the intensive care unit and followed until hospital discharge. SETTING: The cardiac surgical intensive care unit of the Massachusetts General Hospital. PATIENTS: A total of 222 consecutive patients were enrolled. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Perioperative clinical factors and serum concentrations of cTnT measured every 8 hrs after surgery were recorded. These clinical factors and the results of serum cTnT measurement were correlated with the need for prolonged intensive care unit length of stay (defined as >24 hrs). Univariable analysis identified factors predictive of prolonged intensive care unit length of stay. Stepwise logistic regression identified independent predictors of prolonged intensive care unit length of stay. Multiple linear regression was used to explore the direct relationship between cTnT concentrations at several postoperative time points and intensive care unit length of stay. At each time point assessed, cTnT concentrations from patients requiring a prolonged intensive care unit length of stay were significantly higher (all p <.001) than in those individuals with normal length of stay. In contrast, creatine kinase isoenzymes were not significantly different between patients with normal or prolonged intensive care unit length of stay. Multivariable analysis demonstrated that an immediate postoperative cTnT concentration > or =1.58 ng/mL was the strongest predictor of a prolonged intensive care unit length of stay (odds ratio, 5.6; 95% confidence interval, 2.9-10.8). Multiple linear regression analysis revealed that intensive care unit length of stay increased by 0.32 days with each incremental 1.0 ng/mL increase in cTnT measured at 18-24 hrs postprocedure. CONCLUSIONS: Elevated postoperative cTnT concentrations can prospectively identify patients requiring prolonged intensive care unit length of stay after cardiac surgery.     

Recent innovations in intensive care unit risk-prediction models

During the past 20 years, ICU risk-prediction models have undergone significant development, validation, and refinement. Among the general ICU severity of illness scoring systems, the Acute Physiology and Chronic Health Evaluation (APACHE), Mortality Prediction Model (MPM), and the Simplified Acute Physiology Score (SAPS) have become the most accepted and used. To risk-adjust patients with longer, more severe illnesses like sepsis and acute respiratory distress syndrome, several models of organ dysfunction or failure have become available, including the Multiple Organ Dysfunction Score (MODS), the Sequential Organ Failure Assessment (SOFA), and the Logistic Organ Dysfunction Score (LODS). Recent innovations in risk adjustment include automatic physiology and diagnostic variable retrieval and the use of artificial intelligence. These innovations have the potential of extending the uses of case-mix and severity-of-illness adjustment in the areas of clinical research, patient care, and administration. The challenges facing intensivists in the next few years are to further develop these models so that they can be used throughout the IUC stay to assess quality of care and to extend them to more specific patient groups such as the elderly and patients with chronic ICU courses.     

Scoring system for the selection of high-risk patients in the intensive care unit

OBJECTIVE: Patients admitted to the intensive care unit greatly differ in severity and intensity of care. We devised a system for selecting high-risk patients that reduces bias by excluding low-risk patients and patients with an early death irrespective of the treatment. DESIGN: A posteriori analysis of a multiple-center prospective observational trial. SETTING: A total of 89 units from 12 European countries, with 12,615 patients. INTERVENTION: Demographic and clinical data: severity of illness at admission, daily score of nursing workload, length of stay, and hospital mortality. METHODS: We enrolled patients with intensive care unit length of stay of >24 hrs. Three groups of high-risk patients were created: a) Severity group, those with Simplified Acute Physiology Score (SAPS II) over the median; b) Intensity-of-care group, patients with >1 day of high level of care (assessed by logistic analysis); and c) MIX group, patients fulfilling both Severity and Intensity-of-care criteria. The groups were included in a logistic regression model (random split-sample design) to identify the characteristics associated with hospital mortality. We compared the outcome prediction of the SAPS II model (unsplit sample) against our model. MAIN RESULTS: Out of 8,248 patients, the Severity method selected 3,838 patients, Intensity-of-care selected 4,244, and both methods combined selected 2,662 patients. There were 2,828 low-risk patients. Significant associations with hospital mortality were observed for: age, sites of admission, medical/unscheduled surgical admission, acute physiologic score of SAPS II, and the indicator variable "only Severity," "only Intensity-of-care," or MIX (developmental sample: calibration chi-square test, p = .205; area under the receiver operation characteristic curve, 0.814). Calibration and discrimination were better in our model than with the SAPS II model (unsplit sample). CONCLUSION: All three indicator variables select high-risk patients, the Severity/Intensity-of-care MIX being the most robust. These stratification criteria can improve case-mix selection for clinical and organizational studies.     

The effect of prompt physician visits on intensive care unit mortality and cost

OBJECTIVE: To determine the effect on mortality, length of stay, and direct variable cost of physician response time to seeing patients after intensive care unit admission. DESIGN: Retrospective analysis of the intensive care unit database. SETTING: Medical center. PATIENTS: Subjects were 840 patients who had complete direct variable cost data and a subset of 316 patients who were matched by propensity scores. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Median time to first visit by a physician was 6 hrs. One hundred thirty-five patients (16.1%) died in hospital compared with 25.0% predicted by Acute Physiology and Chronic Health Evaluation risk (p < .001). Higher Acute Physiology and Chronic Health Evaluation risk, older age, mechanical ventilation on arrival in the intensive care unit, and longer time until seen by a physician were predictors of hospital mortality. Each 1-hr delay in seeing the patient was associated with a 1.6% increased risk of hospital death, which further increased to 2.1% after including propensity score. However, patients seen more promptly (<6 vs. >6 hrs) had greater hospital direct variable cost ($11,992 +/- $12,043 vs. $10,355 +/- $10,368, p = .04), before controlling for acuity of illness and other factors that may have affected time to evaluation. In the subpopulation of propensity-matched patients, patients seen promptly (<6 vs. >6 hrs) had shorter hospital length of stays (11 +/- 11 vs. 13 +/- 14 days, p = .03) but similar direct variable costs ($10,963 +/- 10,778 vs. $13,016 +/- 13,006, p = .16) and similar mortality rates (24 vs. 30, p = .46). CONCLUSIONS: In the total patient population, delay in seeing patients was associated with an increased risk of death. In the propensity-matched patients, promptly seen patients had shorter hospital stays but similar direct variable costs.     

Five-year survival, quality of life, and individual costs of 303 consecutive medical intensive care patients--a cost-utility analysis

OBJECTIVE: To assess long-term survival, health-related quality of life, and associated costs 5 yrs after discharge from a medical intensive care unit. DESIGN: Prospective cohort study. SETTING: Medical intensive care unit of a German university hospital. PATIENTS: Three hundred and three consecutive patients with predominantly cardiovascular and pulmonary disorders admitted between November 1997 and February 1998 with an intensive care unit length of stay >24 hrs. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Demographic data, Simplified Acute Physiology Score II, Sequential Organ Failure Assessment, simplified Therapeutic Intervention Scoring System, and individual intensive care unit and hospital costs were prospectively recorded. Primary outcomes included 5-yr survival, functional status, health-related quality of life (Medical Outcome Short Form, SF-36), effective costs per survivor, and costs per life year and per quality-adjusted life year gained.Of 303 patients, 44 (14.5%) died in the hospital. Among the remaining 259 patients, 190 (73%) survived the 5-yr follow up and 173 patients (91%) completed the questionnaire. Baseline demographics including gender, age, Simplified Acute Physiology Score II, Sequential Organ Failure Assessment, simplified Therapeutic Intervention Scoring System, and admission diagnosis were similar between hospital and long-term survivors (p > .05 for all). The health status index of those patients surviving the 5-yr follow-up was 0.88, independent of patients' severity of illness. The average effective costs per survivor were 8.827 for intensive care unit costs and 14.130 for intensive care unit and hospital costs. Mean costs per life year and per quality-adjusted life year gained amounted to 19.330 and 21.922 , respectively. Increasing severity of illness was associated with higher costs. CONCLUSIONS: Considering the severity of illness and the patients' outcome, the costs associated with both life year and quality-adjusted life year gained were within generally accepted limits for other potentially life-saving treatments.     

Glomerular permeability and trauma: a correlation between microalbuminuria and Injury Severity Score.

OBJECTIVE: To determine if there is a correlation between an increase in glomerular permeability, the magnitude of trauma, and the severity of illness. DESIGN: Prospective study. SETTING: Two university hospital intensive care units. PATIENTS: Forty consecutive critically ill trauma patients admitted directly to the intensive care unit within 120 mins of their injuries. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: For each patient, urine was collected from the time of admission until 7 am the next day. Within 48 hrs, only one sample of all urine collected (5 mL) was examined for microalbuminuria and urinary creatinine. Results were expressed as the microalbuminuria/urinary creatinine ratio (MACR). The mortality rate in the intensive care unit, Injury Severity Score at the moment of admission, Acute Physiology and Chronic Health Evaluation III score, and Simplified Acute Physiology Score in the first 24 hrs were calculated for each patient. The data were analyzed using the Pearson test for linear regression and Student's t-test. During the first 24 hrs after trauma, there was an increase of MACR (6.9 +/- 0.6 mg/mmol) above normal (reference range, <3 mg/mmol) that was positively correlated with Injury Severity Score (31.4 +/- 1.9; r2 = .73, p < .05). However, there was no correlation between MACR, Acute Physiology and Chronic Health Evaluation III score, Simplified Acute Physiology Score, and mortality rate. CONCLUSIONS: Patients with trauma show an increase in glomerular permeability during the first 24 hrs after injury. The magnitude of this increase is correlated with the extent of trauma but does not seem significant enough to be predictive of severity of illness and/or outcome.     

Mortality rate and length of stay of patients admitted to the intensive care unit in July

OBJECTIVE: At the beginning of each academic year in July, inexperienced residents and fellows begin to care for patients. This inexperience can lead to poor patient outcome, especially in patients admitted to the intensive care unit (ICU). The objective of this study was to determine the impact of July ICU admission on patient outcome. DESIGN: Retrospective, cohort study. SETTING: Academic, tertiary medical center. PATIENTS: Patients admitted to the ICU from October 1994 through September 2002. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Demographics, Acute Physiology and Chronic Health Evaluation (APACHE) III score and predicted mortality, admission source, admission date, intensity of treatment, ICU length of stay (LOS), and hospital mortality of 29,084 patients were obtained. The actual and predicted weighted ICU LOS and their ratio were calculated. Logistic regression analysis was used to compare the hospital mortality rate of patients admitted to the ICU in July with those admitted during the rest of the year, with adjustment for potentially confounding variables. The patients' mean age was 62.3 +/- 17.6 yrs; 57.3% were male and 95.5% white. Both the customized predicted and observed hospital mortality rates of the entire cohort were 8.2%. The majority (76.7%) of the patients were discharged home, and 15.1% were discharged to other facilities. When adjusted for potentially confounding variables, ICU admission in July was not associated with higher hospital mortality rate compared with any other month. There were no significant differences in the discharge location of patients between July and any one of the other months. There were no statistically significant differences in the weighted ICU LOS ratio between July and any of the other months. CONCLUSIONS: ICU admission in July is not associated with increased hospital mortality rate or ICU length of stay.     

Mortality of patients with heatstroke admitted to intensive care units during the 2003 heat wave in France: a national multiple-center risk-factor study

OBJECTIVE: Heatstroke requires active body cooling and organ failure supportive care. Although heat waves are expected to recur over the next decades, little is known about the risk factors for mortality in heatstroke patients. We examined the prognosis and risk factors for hospital mortality in patients with heatstroke admitted to an intensive care unit (ICU) during the heat wave in France in August 2003. DESIGN: A questionnaire was sent to the physicians leading an ICU in France to identify the patients admitted with heatstroke during August 2003. Data included demographics, factors predisposing to heatstroke, severity during the first day in the ICU, air conditioning in the ICU, and hospital mortality. Risk factors for mortality were determined in multivariate Cox proportional hazards analysis. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Data were obtained for 345 patients. Hospital mortality was 62.6%. Occurrence of heatstroke at home or in a healthcare facility rather than in a public location, high Simplified Acute Physiology Score II, high body temperature, prolonged prothrombin time, use of vasoactive drugs within the first day in the ICU, and patient management in an ICU without air conditioning were independently associated with an increased risk of death. CONCLUSIONS: Mortality of patients admitted to the ICU with heatstroke is high. Predictors of mortality are available within the first 24 hrs after admission. Furthermore, in this study, air conditioning in the ICU was associated with improved outcome.     

Impact of standard procedures and clinical standards on cost-effectiveness and intensive care unit performance in adult patients after cardiac surgery

Objective: To investigate the impact of organizational procedures on intensive care unit (ICU) performance and cost-effectiveness after cardiac surgery. Design: Prospective study. Setting: Cardiothoracic ICU at a university hospital. Patients: Thousand five hundred twenty-six consecutive patients over a period of 18 months. Interventions: The first 6 months were used as the control period. Afterwards selected organizational changes were introduced, such as written standard procedures, time schedules and discharge reports. Measurements: Demographic data, surgical procedures, length of ICU and hospital stay and hospital outcome were recorded. Severity of illness was assessed daily using Acute Physiology and Chronic Health Evaluation (APACHE II), Simplified Acute Physiology Score (SAPS II) and Organ Failure Score (OFS). Intensity of treatment and nursing care was monitored by the Therapeutic Intervention Scoring System (TISS). RIYADH ICU Program (RIP 5.0) was used to determine the relationship of observed to predicted mortality (standardized mortality ratio SMR) and the effective costs per survivor. Main results: SMR decreased continuously after the establishment of new management procedures while all other factors all other factors remained unchanged. Comparing outcome according to APACHE II on ICU admission demonstrated a significantly increased ICU performance in high risk patients with an APACHE II of 20–30 points (p < 0.05) while effective costs per survivor decreased significantly from DM 29,988 to DM 13,568 DM (p < 0.05). Conclusions: Organizational changes can improve ICU performance and cost-effectiveness after cardiac surgery. The RIP may be used to monitor the clinical and economical effects of change. Received: 16 December 1998 Accepted: 28 July 1999     

Outcome following prolonged intensive care unit stay in multiple trauma patients

OBJECTIVE: To describe the hospital course and outcomes of trauma patients requiring ICU stays greater than 30 days and the charges they incur. DESIGN: A retrospective case series analysis of data collected from patient charts and trauma registry. SETTING: A Level I regional trauma center that is part of a statewide trauma system. PATIENTS: Over a 3-yr period, 87 patients (3% of all trauma ICU admissions) had prolonged stays (greater than 30 days) in the ICU; they constitute the study group. Blunt trauma was responsible for 90% of injuries, and the mean Injury Severity Score was 34 +/- 16 SD. RESULTS: Mechanical ventilation was required for 78.5% of the time spent in the ICU. The mean time spent on mechanical ventilators was 47 +/- 23 days; in the ICU, 60 +/- 27 days; and in the hospital, 72 +/- 29 days. Infectious complications occurred in 90% and organ dysfunction was seen in 76% of patients. The overall mortality rate was 17.2% (31% for patients greater than 65 yr). Patients less than 40 yr had lower mortality rates despite a significantly higher Injury Severity Score and lower Glasgow Coma Scale score compared with those greater than 65 yr. More patients greater than 65 yr were discharged to chronic care facilities than those younger (23% vs. 5%). The number of patients followed at 3 and 12 months after discharge was 74% and 54%, respectively, with only two deaths. The mean hospital and professional charges to the patients were $101,000 +/- 61,000 and $35,000 +/- 13,000, respectively. CONCLUSION: Length of ICU stay was most closely associated with the need for mechanical ventilation. The presence of premorbid illness, age greater than 65 yr, and organ dysfunction was associated with increased mortality. Although trauma patients requiring prolonged ICU stays utilize many resources, the ultimate outcome may be fairly good.     

Adjustment of intensive care unit outcomes for severity of illness and comorbidity scores

PURPOSE: Comparison of outcomes among intensive care units (ICUs) requires adjustment for patient variables. Severity of illness scores are associated with hospital mortality, but administrative databases rarely include the elements of these scores. However, these databases include the elements of comorbidity scores. The purpose of this study was to compare the value of these scores as adjustment variables in statistical models of hospital mortality and hospital and ICU length of stay after adjustment for other covariates. MATERIALS AND METHODS: We used multivariable regression to study 1808 patients admitted to a 13-bed medical-surgical ICU in a 400-bed tertiary hospital between December 1998 and August 2003. RESULTS: For all patients, after adjusting for age, sex, major clinical category, source of admission, and socioeconomic determinants of health, we found that Acute Physiology and Chronic Health Evaluation (APACHE) II and comorbidity scores were significantly associated with hospital mortality and that comorbidity but not APACHE II was significantly associated with hospital length of stay. Separate analysis of hospital survivors and nonsurvivors showed that both APACHE II and comorbidity scores were significantly associated with hospital length of stay and APACHE II score was associated with ICU length of stay. CONCLUSION: The value of APACHE II and comorbidity scores as adjustment variables depends on the outcome and population of interest.     

Long-stay patients in the pediatric intensive care unit

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

Scoring systems in cancer patients admitted for an acute complication in a medical intensive care unit

OBJECTIVE: To validate and compare two severity scoring systems, the Acute Physiology and Chronic Health Evaluation (APACHE) II and Simplified Acute Physiology Score (SAPS) II and to determine their prognostic value for mortality during the hospital stay and after discharge in a specific group of cancer patients admitted to intensive care unit (ICU) for an acute medical complication. DESIGN: Prospective cohort study. SETTING: The medical ICU of a European cancer hospital. SUBJECTS: A total of 261 consecutive cancer patients admitted to ICU for an acute medical complication. MEASUREMENTS: Variables included into the APACHE II and SAPS II scores, as well as characteristics of the cancer, were collected during the first 24 hrs of the ICU stay. Hospital and in-ICU mortalities, overall survival, and survival after day 30 were measured. RESULTS: Observed hospital and ICU mortalities were 33% and 23%. Median survival time was 94 days and 1-yr survival rate was 23%. The mean predicted risk of death was 26.5% with APACHE II and 26.1% with SAPS II. Correlation between both systems was excellent. Calibration for mortality prediction ability of both scoring systems was similar. Discrimination between survivors and nonsurvivors was superior with SAPS II according to the area under the receiver operating characteristic curve but was better with APACHE II for survivors using thresholds minimizing the overall misclassification rates. Multivariate prognostic analysis showed that the scoring systems were the only significant factors for hospital and in-ICU mortalities, whereas characteristics related to the cancer (extent, phase) were the factors predicting survival after discharge. CONCLUSION: The prognosis of cancer patients admitted to ICU for a medical problem is first determined by the acute physiologic changes induced by the complication, as evaluated by the severity scores. There is no major difference between the two assessed scoring systems. They are, however, not accurate enough to be used in the routine management of these patients. After recovery from complications, characteristics related to the neoplastic disease, however, retrieve their independent influence on the further survival.     

A comparison of methods to predict mortality of intensive care unit patients

This paper presents results of the first study explicitly designed to compare three methods for predicting hospital mortality of ICU patients: the Acute Physiology Score (APS), the Simplified Acute Physiology Score (SAPS), and the Mortality Prediction Model (MPM). With respect to sensitivity, specificity, and total correct classification rates, these methods performed comparably on a cohort of 1,997 consecutive ICU admissions. In these patients from a single hospital, the APS overestimated and the SAPS underestimated the probability of hospital mortality. The MPM probabilities most closely matched the observed outcomes. Each method holds considerable promise for assessing the severity of illness of critically ill patients. The MPM should be particularly useful for comparing ICU performance, since it is independent of ICU treatment and can be calculated at the time a patient is admitted.