Intensive care unit length of stay: recent changes and future challenges.

OBJECTIVE: To compare case-mix adjusted intensive care unit (ICU) length of stay for critically ill patients with a variety of medical and surgical diagnoses during a 5-yr interval. DESIGN: Nonrandomized cohort study. SETTING: A total of 42 ICUs at 40 US hospitals during 1988-1990 and 285 ICUs at 161 US hospitals during 1993-1996. PATIENTS: A total of 17,105 consecutive ICU admissions during 1988-1990 and 38,888 consecutive ICU admissions during 1993-1996. MEASUREMENTS AND MAIN RESULTS: We used patient demographic and clinical characteristics to compare observed and predicted ICU length of stay and hospital mortality. Outcomes for patients studied during 1993-1996 were predicted using multivariable models that were developed and cross-validated using the 1988-1990 database. The mean observed hospital length of stay decreased by 3 days (from 14.8 days during 1988-1990 to 11.8 days during 1993-1996), but the mean observed ICU length of stay remained similar (4.70 vs. 4.53 days). After adjusting for patient and institutional differences, the mean predicted 1993-1996 ICU stay was 4.64 days. Thus, the mean-adjusted ICU stay decreased by 0.11 days during this 5-yr interval (T-statistic, 4.35; p < .001). The adjusted mean ICU length of stay was not changed for patients with 49 (75%) of the 65 ICU admission diagnoses. In contrast, the mean observed hospital length of stay was significantly shorter for 47 (72%) of the 65 admission diagnoses, and no ICU admission diagnosis was associated with a longer hospital stay. Aggregate risk-adjusted hospital mortality during 1993-1996 (12.35%) was not significantly different during 1988-1990 (12.27%, p = .54). CONCLUSIONS: For patients admitted to ICUs, the pressures associated with a decrease in hospital length of stay do not seem to have influenced the duration of ICU stay. Because of the high cost of intensive care, reduction in ICU stay may become a target for future cost-cutting efforts.

Predicting functional status outcomes in hospitalized patients aged 80 years and older.

OBJECTIVE: To develop a model estimating the probability of a patient aged 80 years or older having functional limitations 2 months and 12 months after being hospitalized. DESIGN: A prospective cohort study. SETTING: Four teaching hospitals in the US. PARTICIPANTS: Enrolled patients were nonelective hospital admissions aged 80 years or older who stayed in hospital at least 48 hours. The 804 patients who survived and completed an interview at 2 months and the 450 who completed an interview at 12 months were from the 1266 patients in the Hospitalized Elderly Longitudinal Project (HELP) (76% and 47% of survivors, respectively). Median age of the 2-month survivors was 84.7 years. MEASUREMENTS AND MAIN OUTCOMES: Patient function 2 and 12 months after enrollment was defined by the number of dependencies in Activities of Daily Living (ADLs). Ordinal logistic regression models were constructed to predict functional status. Predictors included demographic characteristics, disease category, geriatric conditions, severity of physiologic imbalance, current quality of life, and exercise capacity and ADLs 2 weeks before study admission. RESULTS: Before admission, 39% of patients were functionally independent in ADLs. Of patients who survived and were interviewed at 2 months, 32% were functionally independent, and at 12 months, 36% were independent. Among patients with no baseline dependencies, 42% had developed one or more limitations 2 months later, and 41 % had limitations 12 months later. The patient's ability to perform activities of daily living at baseline was the most important predictor of functional status at both 2 and 12 months. In a multivariable predictive model, independent predictors of poorer functional status at 2 months included: worse baseline functional status and quality of life; depth of coma, if any; lower serum albumin level; presence of dementia, depression, or incontinence; being bedridden; medical record documentation of need for nursing home; and older age. Model performance, assessed using Somers' D, was 0.61 for 2 months and 0.57 for 12 months (Receiver Operating Characteristic (ROC) area = 0.81 and .79, respectively.) Bootstrap validation of the month 2 model also yielded a Somers' D = 0.60. The models were well calibrated over the entire risk range. The ROC area for prediction of the loss of independence was 0.76 for 2 months and 0.68 for 12 months. CONCLUSIONS: Many older patients are functionally impaired at the time of hospitalization, and many develop new functional limitations. A limited amount of readily available clinical information can yield satisfactory predictions of functional status 2 months after hospitalization. Models like this may prove to be useful in clinical care. This work illuminates a potential method for risk adjustment in research studies and for monitoring quality of care.

Length of stay and therapeutic intervention allow estimation of in-hospital resource use independent of site and inflation.

OBJECTIVES: To develop and evaluate a model to estimate inpatient resource utilization for seriously ill patients using periodic medical record abstraction and hospital length of stay. DESIGN: A prospective cohort study of prognoses and preferences and outcomes of seriously ill hospitalized adults. The predictive model was based on the Therapeutic Intervention Scoring System (TISS), length of stay (LOS), and hospital site, and was developed on Phase I data and validated on Phase II data. SETTING: Five teaching hospitals in the United States PARTICIPANTS: The derivation sample included 4301 patients enrolled in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT). The validation sample included 4804 patients. MEASUREMENTS: Hospital charges were correlated with the product of hospital length of stay and average score from the TISS. Model fit assessment used R2 for total hospital charges. RESULTS: The correlation between hospital charges and the final model, after adjusting for site differences, was 0.86. The correlation was similar when validating the model in an independent data set. CONCLUSIONS: We present a method to estimate hospital resource utilization that is independent of site and inflation and can estimate actual costs in multiyear, multisite research endeavors.

Evaluation of acute physiology and chronic health evaluation III predictions of hospital mortality in an independent database.

OBJECTIVE: To assess the accuracy and validity of Acute Physiology and Chronic Health Evaluation (APACHE) III hospital mortality predictions in an independent sample of U.S. intensive care unit (ICU) admissions. DESIGN: Nonrandomized, observational, cohort study. SETTING: Two hundred eighty-five ICUs in 161 U.S. hospitals, including 65 members of the Council of Teaching Hospitals and 64 nonteaching hospitals. PATIENTS: A consecutive sample of 37,668 ICU admissions during 1993 to 1996; including 25,448 admissions at hospitals with >400 beds and 1,074 admissions at hospitals with <200 beds. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We used demographic, clinical, and physiologic information recorded during ICU day 1 and the APACHE III equation to predict the probability of hospital mortality for each patient. We compared observed and predicted mortality for all admissions and across patient subgroups and assessed predictive accuracy using tests of discrimination and calibration. Aggregate hospital death rate was 12.35% and predicted hospital death rate was 12.27% (p =.541). The model discriminated between survivors and nonsurvivors well (area under receiver operating curve = 0.89). A calibration curve showed that the observed number of hospital deaths was close to the number of deaths predicted by the model, but when tested across deciles of risk, goodness-of-fit (Hosmer-Lemeshow statistic, chi-square = 48.71, 8 degrees of freedom, p< .0001) was not perfect. Observed and predicted hospital mortality rates were not significantly (p < .01) different for 55 (84.6%) of APACHE III's 65 specific ICU admission diagnoses and for 11 (84.6%) of the 13 residual organ system-related categories. The most frequent diagnoses with significant (p < .01) differences between observed and predicted hospital mortality rates included acute myocardial infarction, drug overdose, nonoperative head trauma, and nonoperative multiple trauma. CONCLUSIONS: APACHE III accurately predicted aggregate hospital mortality in an independent sample of U.S. ICU admissions. Further improvements in calibration can be achieved by more precise disease labeling, improved acquisition and weighting of neurologic abnormalities, adjustments that reflect changes in treatment outcomes over time, and a larger national database.

Patient empowerment and feedback did not decrease pain in seriously ill hospitalized adults.

We tested a nurse clinician-mediated intervention to relieve pain in a group of seriously ill hospitalized adults using a randomized controlled trial at five tertiary care academic centers in the US. The study included 4804 patients admitted between January 1992 and January 1994 with one or more of nine high mortality diagnoses; 2652 were allocated to the intervention and 2152 to usual care. Specially-trained nurse clinicians assessed patients' pain, educated them and their families about pain control, empowered patients to expect pain relief, informed patients' nurses and physicians about level of pain and suggested or used other pain management resources. Patients' pain was determined from hospital interviews with patients and surrogates. Pain 2 and 6 months later or after death and satisfaction with its control at all time periods were also assessed. All analyses were adjusted for baseline risk of being in pain and propensity to be in the intervention group. Overall, 50.9% of patients reported some pain. After adjustment for other variables associated with pain, comparing the intervention to the control group, there was not a statistically significant difference in level of pain (OR for higher levels of pain 1.15; CI 1.00-1.32) or satisfaction with control of pain during the hospitalization (OR for higher levels of pain 1.12; CI 0.91-1.39), 2 or 6 months after discharge, or during the last 3 days of life. A multifaceted intervention using information, empowerment, advocacy, counseling and feedback was ineffective in ameliorating pain in seriously ill patients. Control of pain in these patients remains an important problem. More intensive pain treatment strategies addressing the needs of seriously ill hospitalized adults must be evaluated.

The illusion of end-of-life resource savings with advance directives. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment.

OBJECTIVE: Would increasing the documentation of advance directives (ADs) lead to a reduction in resource utilization? We examined this question by conducting three secondary analyses: (1) we tested for a change in resource use among those who died in the hospital at a time before and after an intervention that increased the documentation of ADs in the medical record; (2) we replicated analyses of published studies that reported an association of chart documentation of ADs and hospital resource use; and (3) we examined whether a potential explanation of the observed association is biased documentation of ADs among patients who have completed an AD. DESIGN: Replication of analysis of previous published studies using data from a prospective cohort study and block-randomized controlled trial. SETTING: Five teaching hospitals in the United States. PATIENTS: A total of 9105 seriously ill patients were enrolled in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT), including 4301 patients in the 2 years (1989-91) before the Patient Self-Determination Act (PSDA) and 4804 in the 2 years (1992-94) after the PSDA implementation, with 2652 patients receiving the intervention and 2152 serving as controls. INTERVENTIONS: The SUPPORT intervention provided a nurse to facilitate communication among patients, surrogates, and physicians about preferences for and outcomes of treatments. Documenting existing advance directives was also one of this nurse's tasks. The Patient Self-Determination Act required that health care institutions inquire about and document existing advance directives at the time of hospital admission. MEASUREMENT: Hospital resource use was derived from the Therapeutic Intensity Scoring System and hospital length of stay, converted into 1994 dollars. RESULTS: Chart documentation of existing advance directives at the time of study admission increased with both the PSDA and the SUPPORT intervention. We found that intervention patients were more likely to have pre-existing ADs documented. Despite this increase, there was no corresponding change in hospital resource use for those who died during the enrollment hospitalization. Replication of analyses from published studies using data from the block randomized controlled trial found that ADs documented by the third day of serious illness were associated with a 23% reduction in hospital resource use among control patients ($21,284 with ADs documented compared with $26,127 without, 95% CI 1-48% reduction). However, this association was not observed among intervention patients, who had more pre-existing ADs documented in the medical record. Intervention patients with early documentation of ADs showed a trend toward greater cost ($28,017 compared with $24,178 among those without AD documentation, 95% CI 0-25% increase). The rate of documentation and characteristics of those with documentation differed between control and intervention patients. Intervention patients were more likely (as reported by patient or surrogate interview) to have ADs documented in the medical record by the third day (55% vs 32%, P < .001). In contrast to intervention patients, control patients who were older, less wealthy, less educated, more likely to prefer to forgo CPR, and more likely to want life-sustaining treatment limited had their ADs documented. These associations were not found among intervention patients when comparison was made between those with and those without an AD documented in the medical record. CONCLUSION: Increasing the documentation of pre-existing ADs was not associated with a reduction in hospital resource use. ADs documented without further intervention by the third day of a serious illness were associated with decreased hospital resource use. However, we did not find this association with an intervention that increased AD documentation. One potential explanation of these findings is that classification of those with an AD was based on cha

Pain during hospitalization is associated with continued pain six months later in survivors of serious illness. The SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments.

PURPOSE: To determine the level of pain reported by survivors of serious illness 2 and 6 months after study enrollment and to identify variables associated with later pain. PATIENTS AND METHODS: Observational cohort study of patients with interviews during hospitalization (5,652) and 2 (3,782) and 6 (2,984) months later admitted between June 1989 and January 1994 with 1 or more of 9 high mortality diagnoses admitted to 5 tertiary care academic centers in the United States. Patients' level of pain during the hospitalization and 2 and 6 months later was determined from interviews with patients and surrogates (most often family members). Separate ordinal logistic regressions were constructed with level of pain at months 2 or 6 as the dependent variable and 22 demographic, psychological, chronic, and acute illness measures at the time of hospitalization as independent variables. RESULTS: Of patients reporting level 4 (moderately severe pain occurring most of the time or extremely severe pain occurring half of the time) or 5 (moderately severe pain occurring most or all of the time or extremely severe pain occurring at least half of the time) pain to 5 during hospital interviews, 39.5% and 39.7% reported level 4 or 5 pain 2 and 6 months later, respectively. Level of hospital pain was the variable most strongly associated with later pain. Compared with patients with level 1 hospital pain, those with level 2 (not at all severe pain or moderate, occasional) had a 2.91 (95% confidence interval [CI] 2.50, 3.37) and 1.75 (CI 1.48, 2.07) times greater adjusted odds of increased levels of pain 2 and 6 months later, respectively. Compared with patients with level 1 hospital pain, those with level 5 pain had a 9.20 (CI 7.27, 11.65) and 4.40 (CI 3.39, 5.71) times greater adjusted odds of increased levels of pain 2 and 6 months later, respectively. Age, number of dependencies in activities of daily living, depression, and type of comorbid illnesses were also independently associated with level of pain both 2 and 6 months later. CONCLUSION: Survivors of the serious and common illnesses that we studied have a high level of pain during hospitalization and up to 6 months after hospitalization. Level of hospital pain was most strongly associated with later pain. Better pain control both during hospitalization and after discharge should be given a high priority. Pain during hospitalization should trigger future inquiries about pain and its treatment.

The covariance decomposition of the probability score and its use in evaluating prognostic estimates. SUPPORT Investigators.

The probability score (PS) or Brier score has been used in a large number of studies in which physician judgment performance was assessed. However, the covariance decomposition of the PS has not previously been used to evaluate medical judgment. The authors introduce the technique and demonstrate it by analyzing prognostic estimates of three groups: physicians, their patients, and the patients' decision-making surrogates. The major components of the covariance decomposition--bias, slope, and scatter--are displayed in covariance graphs for each of the three groups. The decomposition reveals that whereas the physicians have the best overall estimation performance, their bias and their scatter are not always superior to those of the other two groups. This is primarily due to two factors. First, the physicians' prognostic estimates are pessimistic. Second, the patients place the large majority of their estimates in the most optimistic category, thereby achieving low scatter. The authors suggest that the calculational simplicity of this decomposition, its informativeness, and the intuitive nature of its components make it a useful tool with which to analyze medical judgment.

Predicting future functional status for seriously ill hospitalized adults. The SUPPORT prognostic model.

OBJECTIVE: To develop a model estimating the probability of an adult patient having severe functional limitations 2 months after being hospitalized with one of nine serious illnesses. DESIGN: Prospective cohort study. SETTING: Five teaching hospitals in the United States. PARTICIPANTS: 1746 patients (model development) who survived 2 months and completed an interview, selected from 4301 patients in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT); independent validation sample of 2478 patients. MEASUREMENTS AND MAIN OUTCOMES: Patient function 2 months after admission categorized as absence or presence of severe functional limitations (defined as Sickness Impact Profile scores > or = 30 or as activities of daily living scores > or = 4 [levels that require near-constant personal assistance]). A logistic regression model was constructed to predict severe functional limitation. RESULTS: One third (n = 590) of patients who were interviewed at 2 months had severe functional limitations. Changes in functional status were common: Of those with no baseline dependencies (not dependent on personal assistance), 21% were severely limited at 2 months; of those with 4 or more baseline limitations, 30% had improved. The patient's ability to do activities of daily living was the most important predictor of functional status. Physiologic abnormalities, diagnosis, days in hospital, age, quality of life, and previous exercise capacity also contributed substantially. Model performance, assessed using receiver-operating characteristic curves, was 0.79 for the development sample and 0.75 for the validation sample. The model was well calibrated for the entire risk range. CONCLUSIONS: Functional outcome varied substantially after hospitalization for a serious illness. A small amount of readily available clinical information can estimate the probability of severe functional limitations.

Carotid atherosclerosis measured by B-mode ultrasound in populations: associations with cardiovascular risk factors in the ARIC study.

To assess whether carotid atherosclerosis measured by B-mode ultrasound is related to cardiovascular risk factors, 386 cases with carotid artery wall thickening and an equal number of controls free of arterial intima-media thickening were drawn from the cohort of the Atherosclerosis Risk in Communities (ARIC) Study examined in four communities in the United States between 1988 and 1990. Cases and controls were individually matched on sex, race, age group, study center, and date of examination. The mean values of total cholesterol, low density lipoprotein (LDL) cholesterol, total triglyceride, blood pressure, and pack-years of cigarette smoking were higher in cases than controls. Mean high density lipoprotein (HDL) cholesterol was lower in cases than controls. Case-control differences were all statistically significant. Multivariable-adjusted odds ratios point to differences of considerable magnitude in the risk of carotid atherosclerosis between groups defined by clinical and public health-oriented risk factor cut-points.

Optimization of the under-agarose assay of porcine neutrophil migration.

Important procedural factors in the under-agarose assay for porcine neutrophil migration were identified, and optimal conditions were established. Three factors were tested: the concentration of zymosan-activated serum inoculated into the outer well; the number of neutrophils inoculated into the center well; and the time of incubation of the agarose plates. All factors had a significant (P less than 0.0001, 0.0001, and 0.01, respectively) effect on the chemotactic index of porcine neutrophils. The optimal combination of these 3 factors was undiluted zymosan-activated serum as the chemoattractant, 8 X 10(5) neutrophils inoculated into the center well, and 5 hours of incubation. The assay was validated, using standard conditions, and the data were used to predict the number of pigs and/or repetitive assays needed to identify differences among experimental groups.