Daily prognostic estimates for critically ill adults in intensive care units

A history of outcome prediction in the ICU

“…ICUs with superior or inferior SMRs, however, could not be distinguished by ICU clinicians and organizational researchers [20]. (4) Daily mortality estimates using ICU day 1-7 data were found to have a potential role in assessing individual prognosis [21]. (5) The use of APACHE II, APACHE III, MPM II, and SAPS II outside the United States resulted in conflicting conclusions about which model most accurately predicted mortality.…”

Outcome prediction in critical care: the Mortality Probability Models

“…One reason why MPM 24 has not been updated is the availability of robust models, such as APACHE [9] that already address this time period better than MPM [22]. Modeling outcomes beyond 24 h becomes progressively more difficult as the sample size shrinks dramatically (as patients leave the ICU alive or dead) and less datapoints are available for analysis as the initial flurry of laboratory and other testing subsides.…”

“…For example, in patients with perforated gastrointestinal viscus, subsequent development of multiple organ failure (not captured at admission) is a far better predictor of death than the APACHE-III score [23]. The use of sequential prognostic estimates has been explored by both APACHE [22] and MPM [14]. Timsit et al [24] proposed the use of SAPS-II plus an organ dysfunction score for outcome comparisons of patients remaining in the ICU for 72 h or longer.…”

Sind prognostische Scores nützlich?

“…Diesem Zweck dient der Leitartikel dieses Heftes, in dem Unertl und Kottler die derzeit am weitesten verbreiteten Prognose-Scores in übersichtlicher Weise vorstellen und kritisch werten [10]. Die [4,9], die Erfassung von Krankheitsverläufen [8], die Beurteilung des Behandlungserfolges und -aufwands einer Intensivstation [6], interne und externe Qualitätskontrollen [6], die Evaluierung neuer Therapieverfahren [3] sowie Kosten-Nutzen-Analysen. [8].…”

“…Die [4,9], die Erfassung von Krankheitsverläufen [8], die Beurteilung des Behandlungserfolges und -aufwands einer Intensivstation [6], interne und externe Qualitätskontrollen [6], die Evaluierung neuer Therapieverfahren [3] sowie Kosten-Nutzen-Analysen. [8]. Die individuelle Überlebenswahrscheinlichkeit läßt sich aber auch aus den Score-Verläufen nicht prognostizieren [7].…”