Rationale and Methodological Approach Underlying the Development of the Sequential Organ Failure Assessment (SOFA)–2 Score
Rui P. Moreno, Andrew Rhodes, Otávio T. Ranzani, Jorge I. Salluh, Joana Berger‐Estilita, Craig M. Coopersmith, Nicole P. Juffermans, John G. Laffey, Matti Reinikainen, Ary Serpa Neto, Miguel Tavares, Jean-François Timsit, M.D.P. Arias Lopez, Nish Arulkumaran, Elie Azoulay, Dipayan Chaudhuri, Dylan W. de Lange, Jan J. De Waele, Claúdia C. dos Santos, Bin Du, Sharon Einav, Ricard Ferrer, Tomoko Fujii, Hayley B. Gershengorn, Rashan Haniffa, Mohd Shahnaz Hasan, Steve Hollenberg, Mariachiara Ippolito, Christian Jung, М. Yu. Кirov, Inès Lakbar, Jeffrey Lipman, Vincent X. Liu, Suzana Margareth Lobo, Greg S. Martin, Philipp Metnitz, Sheila Nainan Myatra, Simon Oczkowski, José-Artur Paiva, Fathima Paruk, Pirkka T. Pekkarinen, David Pilcher, Lise Piquilloud, Anssi Pölkki, Halle C. Prescott, Annika Reintam Blaser, Ederlon Rezende, Chiara Robba, Bram Rochwerg, Edward J. Schenck, Cornelius Sendagire, Moses Siaw-Frimpong, Andrew J. Simpkin, Márcio Soares, Charlotte Summers, Wojciech Szczeklik, Jukka Takala, Jean‐Louis Vincent, J. Wendon, Fernando G. Zampieri, Manu Shankar‐Hari, Mervyn Singer
Abstract
Importance: The Sequential Organ Failure Assessment (SOFA) score was published in 1996 to describe organ dysfunction in critically ill adult patients in a readily quantifiable and sequential manner. Considerable changes have occurred over the last 3 decades in the use of organ support drugs and devices and in patient outcomes, necessitating revision of the score. Objectives: To develop definitions of organ dysfunction that reflect current understanding and to identify representative variables to generate a revised SOFA score (SOFA-2) of individual organ dysfunction. Evidence Review: A task force of experts in intensive care medicine and epidemiology generated definitions of organ dysfunction, identified relevant variables (physiological and laboratory data specific to the organ system, pharmacological and mechanical organ support), and proposed a 0 to 4-point grading of dysfunction severity through meetings, Delphi processes, and explicit rules, informed by data synthesis, including systematic reviews and meta-analysis. Variables were tested in 2 validation exercises using separate datasets totaling 3.34 million patients within 10 representative databases from diverse geographical and socioeconomic settings to assess distribution and predictive validity (mortality at intensive care unit discharge). Findings: A total of 60 experts participated, with 18 (30%) female participants. Overall, 65 countries were represented, with 33 (51%) from Europe and Central Asia, 13 (20%) from North America; and 8 (12%) from Latin America and the Caribbean. The physiological variables within the 6 organ systems used in the original SOFA score were retained, although some categories were renamed (ie, central nervous system was changed to brain, renal to kidney, coagulation to hemostasis, and hepatic to liver). Revisions of organ support drug and device variables were made to reflect current practice. Alternative variables were added for instances when laboratory data and/or organ support interventions would be inaccessible (eg, in some low-resource settings) or not indicated (eg, ceiling of treatment). Some point cutoff thresholds were modified based on evidence from systematic reviews and data analyses. Scores could not be developed for 2 additional organ systems (gastrointestinal and immune) due to insufficient data, complexity, or lack of content and predictive validity for the variables assessed. Explicit rules were developed to facilitate scoring consistency. Conclusions and Relevance: Through a methodologically robust development process, the SOFA-2 score offers updated definitions to describe organ dysfunction in adult patients requiring critical care and readily quantifiable criteria to grade the degree of dysfunction in individual organ systems. This score considers contemporaneous changes in patient management and outcomes.