ICU Scoring Systems

22 02 2023

ICU scoring systems

Four decades of prognostic scores in intensive care

Intensive care units (ICUs) are people- and technology-intensive environments where timely and wise use of advanced monitoring and life support is crucial to revert or avoid life-threatening conditions. From their inception, this highly complex environment has been confronted with a need to demonstrate its effectiveness to healthcare stakeholders [1]. In the 1970s and 80s, increasing costs of intensive care, associated with poor outcomes of patients with multi-organ failure, urged intensivists and healthcare managers to look for metrics could concisely express ‘severity of illness’ and, thus, allow measurement of risk-adjusted outcomes [2].

In the early 1980s, the Acute Physiology and Chronic Health Evaluation (APACHE) system was a milestone in the history of ICU outcome prediction. This scoring system translated domains of pre-morbid conditions (age and co-morbidities), diagnoses and early physiologic derangements (organ failures, laboratory and physiological abnormalities) into a numeric expression of illness severity. In addition to the absolute value of the score, the APACHE system provided an estimate of the risk of death for each individual patient. APACHE was soon followed by the development of the Mortality Prediction Model (MPM) in the United States and the Simplified Acute Physiology Score (SAPS) in Europe. As the technologies improved, new treatments and protocols of care were applied, and the case-mix of the ICU changed (more elderly, co-morbidities and immunocompromised), scores needed to be updated to remain valid predictors of outcomes. The pioneering early versions of APACHE, SAPS and MPM were updated, with SAPS3, APACHE IV, and MPM0-III published, respectively, in 2005, 2006, and 2007 [3]. One of the important differences among these scores relates to the time when they are calculated. SAPS3 and MPM0-III use data from the first hour of ICU admission; whereas, APACHE IV uses the “worst” measurements from the first 24 h. Therefore SAPS3/MPM0-III potentially reflects the early severity of the non-resuscitated patient. APACHE IV provides more time for data collection and less missing data.

Finally, dynamic scores may be applied in the ICU. The most commonly used is the Sequential Organ Failure Assessment (SOFA) which was developed to define the degree of organ failure, to stratify risk particularly in patients with sepsis, and to monitor response to treatment. Scoring systems, although useful for individual risk assessment, are not as applicable for mortality prediction and ICU performance monitoring. They are often used to complement more general mortality scores in ICU.

Read the full article here:


  1. Power GS, Harrison DA (2014) Why try to predict ICU outcomes? Curr Opin Crit Care 20(5):544–549.

  2. Zimmerman JE, Kramer AA (2014) A history of outcome prediction in the ICU. Curr Opin Crit Care 20(5):550–556.

  3. Pellathy TP, Pinsky MR, Hravnak M (2021) Intensive Care Unit scoring systems. Crit Care Nurse 41(4):54–64.

  4. Harrison DA, Parry GJ, Carpenter JR, Short A, Rowan K (2007) A new risk prediction model for critical care: the Intensive Care National Audit & Research Centre (ICNARC) model. Crit Care Med 35(4):1091–1098.

  5. Paul E, Bailey M, Kasza J, Pilcher D (2016) The ANZROD model: better benchmarking of ICU outcomes and detection of outliers. Crit Care Resusc 18(1):25–36

  6. Yang S, Stansbury LG, Rock P, Scalea T, Hu PF (2019) Linking big data and prediction strategies: tools, pitfalls, and lessons learned. Crit Care Med 47(6):840–848.

  7. Kramer AA, Zimmerman JE, Knaus WA (2021) Severity of illness and predictive models in society of critical care medicine’s first 50 years: a tale of concord and conflict. Crit Care Med 49(5):728–740.

  8. Singer M, Deutschman CS, Seymour CW et al (2016) The third international consensus definitions for sepsis and septic shock (sepsis-3). JAMA 315(8):801–810.

  9. Straney LD, Udy AA, Burrell A, Bergmeir C, Huckson S, Cooper DJ, Pilcher DV (2017) Modelling risk-adjusted variation in length of stay among Australian and New Zealand ICUs. PLoS One 12(5):e0176570.

  10. Badawi O, Breslow MJ (2012) Readmissions and death after ICU discharge: development and validation of two predictive models. PLoS ONE 7(11):e48758.

  11. Ranzani OT, Simpson ES, Japiassú AM, Noritomi DT, Amil Critical Care Group (2016) The challenge of predicting pressure ulcers in critically ill patients. A Multicenter Cohort Study. Ann Am Thorac Soc 13(10):1775–1783.

  12. Kurtz P, Bastos LSL, Salluh JIF, Bozza FA, Soares M (2021) SAPS-3 performance for hospital mortality prediction in 30,571 patients with COVID-19 admitted to ICUs in Brazil. Intensive Care Med 47(9):1047–1049.

  13. Dongelmans DA, Pilcher D, Beane A, Soares M, Del Pilar Arias Lopez M, Fernandez A, Guidet B, Haniffa R, Salluh JIF (2020) Linking of global intensive care (LOGIC): an international benchmarking in critical care initiative. J Crit Care 60:305–310.

  14. Haniffa R, Isaam I, De Silva AP, Dondorp AM, De Keizer NF (2018) Performance of critical care prognostic scoring systems in low and middle-income countries: a systematic review. Crit Care 22(1):18

  15. Raffa JD, Johnson AEW, O’Brien Z, Pollard TJ, Mark RG, Celi LA, Pilcher D, Badawi O (2022) The global open source severity of illness score (GOSSIS). Crit Care Med 50(7):1040–1050.

The LOGIC Team