[Epidemiological and prognostic data from 2054 patients of an internal medicine intensive care unit]. / Epidemiologische und prognostische Daten von 2054 Patienten einer internistischen Intensivstation.

BACKGROUND: Computer-based data collection and objective gathering of degree of illness severity and risk of death with a prognostic scoring system make it possible to obtain, in addition to epidemiological and aetiological data, risk-related outcome values for patients in an intensive care unit. PATIENTS AND METHODS: All 2054 patients who during a 2-year period (1995-1996) had stayed in a medical intensive care unit (MICU) for more than 4 hours were studied prospectively. The simplified acute physiology score II (SAPS II), risk of death, duration of stay in the MICU and in the hospital, and death rates during MICU and hospital stay were determined. Mean and median values and histograms of the various parameters as well as the standardized mortality index (SMI: observed/ predicted death rate with 99% confidence limits) were calculated for each of the patients and certain defined subgroups (basic disease, age, risk). Receiver operating characteristics curves (discrimination) and calibration curves were obtained for SAPS II. RESULTS: Mean age for the cohort was 59.8 years, duration of stay in the MICU 3.1 days, in hospital 14.7 days, SAPS II was 30.3 points, death risk 0.17, death rate during ICU stay was 8.3%, during hospital stay 13.9% and the SMI 0.8% (0.74-0.88). Cardiac disease was the most common underlying condition (60%), while the small group of neurological conditions was remarkable for the high degree of severity and unfavourable prognosis. Both death rate and degree of disease severity increased with age. But the SMI was not significantly higher than 1.0 in both the elderly patients and the high-risk group of patients (on ventilator, renal replacement procedures, death risk > 0.5). CONCLUSIONS: Most patients in a MICU have underlying cardiac disease. Permanently available neurological consultation is essential. The high hospital death rate for elderly patients and those requiring respiratory support is a problem of disease severity, not of the quality of treatment. The risk of death is high on transfer to a general ward. Determination of the SMI is recommended for internal quality control in an ICU.

The ability of the Simplified Acute Physiology Score (SAPS II) to predict outcome in coronary care patients.

OBJECTIVE: To evaluate the applicability of the Simplified Acute Physiology Score (SAPS II) for coronary care patients. DESIGN: Prospective observational cohort study. SETTING: Medical ICU of a community teaching hospital. PATIENTS: 1587 consecutive patients admitted over a period of 18 months. MEASUREMENTS AND MAIN RESULTS: Patients were divided in two groups according to the primary admission diagnosis: general medical intensive care (ICU) patients and intensive coronary care (CCU) patients. Score prediction was tested using criteria suitable to evaluate the discrimination and calibration properties of SAPS II. Mean SAPS II score was 31.6 (+/- 20.1) in ICU and 28.3 (+/- 15.5) in CCU patients (p = 0.06), mean risk of death 0.206 and 0.134 (p = 0.001), and observed hospital mortality 17.8 vs 10.3%. The area under the receiver operating characteristic curve was 0.888 in ICU and 0.908 in CCU patients (p = 0.5). The correlation between predicted and observed hospital mortality was 0.62 (p = 0.001) in ICU and 0.66 (p = 0.001) in CCU patients. The calibration curves did not differ from each other. The probability of death in survivors and nonsurvivors was equally distributed in ICU and CCU patients (p = 0.5). CONCLUSION: We conclude that SAPS II is applicable to CCU patients in our unit.

[Analysis of outcome quality control in intensive care medicine using the Simplified Acute Physiology Score II]. / Analyse einer Ergebnisqualitätskontrolle in der Intensivmedizin mittels des Simplified Acute Physiology Score II.

AIM: the main aim of the study was to assess the applicability of the Simplified Acute Physiology Score II (SAPS II) to the evaluation of outcome quality within the framework of quality assurance in patients in a medical intensive care unit. The outcome parameter employed was hospital mortality, measured as mortality index (hospital mortality actually observed/predicted mortality), the predicted mortality being derived from the individual mortality risk calculated for each patient in accordance with SAPS II. METHOD: For the period of one year, the SAPS II score, the individual mortality risk, the mean scores, mortality risk, intensive care and hospital mortality, and the mortality index (99% confidence interval) were calculated with the aid of a specially developed program for all 1,114 patients kept under observation or treated for longer than 4 hours in the intensive care unit. The entries (data) were monitored by random checks for the correctness of the individual entries and overall completeness of patient inclusion. The applicability of the SAPS II for our own patient material was checked with the aid of Receiver Operating Characteristic curves. In compliance with the original SAPS II to include patients of a coronary care unit but not to evaluate them, only the 604 patients with the diseases of medical intensive care were taken into account for quality control. High-risk groups (patients older than 76, critically ill patients with a mortality risk of more than 0,5, patients receiving respiratory support) and individual diagnostic categories were considered separately as subgroups. RESULTS: In the entire group, the mean mortality risk was 21,1% the observed intensive care mortality 11,2%, the hospital mortality 18,0%, and the mortality index 0,86 (0,75 to 1,00). The mortality actually observed, therefore, corresponded to that predicted on the basis of the SAPS prognostic system. Also in the subgroups of elderly patients, and individual diagnostic categories (cerebral, bronchopulmonary cardiovascular, gastrointestinal diseases), the mortality index did not differ significantly from 1,0. A mortality index significantly less than 1,0 (observed mortality significantly lower than predicted mortality) was found in the sub-groups of the seriously ill, of patients receiving respiratory support, and in the diagnostic category of intoxications. The monthly analysis showed fluctuating mortality indices which, however, never differed significantly from 1,0. The surface under the ROC curve for the entire group was 0,89, and 0.81-0.99 for the various diagnostic categories. CONCLUSIONS: The prognostic system SAPS II can be employed to evaluate the quality of outcome measured by hospital mortality in patients of a medical intensive care unit, provided that the applicability of the score is demonstrated for the patient material involved, the outcome of the overall group and of the high-risk groups is referred to the accuracy and completeness of the entered data is checked, and the scoring systems accepted as quality standard.

[Predictive value of score parameters of the Simplified Acute Physiology Score (SAPS)-II for the duration of treatment of intensive care patients]. / Voraussagekraft von Scoreparametern des Simplified Acute Physiology Score (SAPS)-II für die Behandlungsdauer von Intensivpatienten.

Length of intensive care therapy and the total length of stay in hospital are important determinants of hospital costs. We therefore analysed the correlation between score parameters of SAPS-II with the time spent in the intensive care unit (ICU), and also in the hospital, for 604 general medical intensive care patients (ICU group) and 510 coronary care patients (CCU group). The mean stay in the ICU was 3.68 days for ICU patients and 2.67 days for CCU patients. The total stay in hospital was 13.5 days vs 16.1 days with a mortality of 18% (risk of death 0.21) in ICU and 10% (risk of death 0.13) in CCU patients. In patients who died, duration of therapy in the ICU was significantly longer than in surviving patients (5.88 vs 3.20 days in the ICU group and 3.65 vs 2.56 days in the CCU group). In contrast, total hospital stay was significantly shorter in patients who died (8.6 vs 14.5 days in the ICU group and 8.8 vs 16.9 days in the CCU group) (p = 0.001). The risk of death calculated from SAPS II was significantly correlated with the duration of intensive care. There was a significant indirect correlation between risk of death and the total hospital stay. In ICU patients duration of intensive treatment and hospital stay correlated with age, heart rate, maximum systolic blood pressure, body temperature, BUN, serum bilirubin, and sodium (all signs of systemic inflammatory reaction and organ dysfunction); in CCU patients length of intensive treatment and hospital stay correlated with body temperature, diuresis, BUN, bicarbonate, minimum systolic blood pressure (as signs of organ perfusion). A low Glasgow Coma Score was correlated with prolonged intensive care in all patients. In conclusion, score data, appear a suitable tool to predict the duration of intensive care treatment and length of hospitalization, in addition to outcome, and thus serve as gauge of efficiency.