Introduction

A specific clinical definition for acute respiratory distress syndrome (ARDS) is needed to inform clinicians about specific therapies or prognosis, and to facilitate investigators in developing reproducible research [1].

Prior definitions of ARDS

Like other syndromic illnesses such as depression or sepsis, a challenge in defining ARDS is the lack of a reference gold standard for its diagnosis. However, the definition can still be evaluated on the basis of feasibility, reliability, and other forms of validity [1]. In this way, the definition of ARDS has evolved over time with progressive refinements.

In 1967, the first widely accepted description of ARDS was published describing a case-series of 12 patients that developed a common pattern of respiratory failure stemming from a variety of insults. “Respiratory-distress syndrome” was clinically characterized by the acute onset of severe dyspnea, tachypnea, cyanosis refractory to oxygen, loss of compliance, and infiltration on the chest radiographs [2]. Informal definitions such as this were used for more than two decades until the Murray Lung Injury Score was published in 1988 [3], followed shortly thereafter in 1994 by the American-European Consensus Conference (AECC) definition of ARDS [4]. The AECC definition allowed knowledge on ARDS to remarkably increase, facilitating the completion of seminal epidemiological [5] and clinical [6, 7] studies. The AECC defined ARDS by the presence of the following four criteria: (1) acute onset; (2) hypoxemia, as indicated by the ratio of partial pressure of arterial oxygen to fraction of inspired oxygen [PaO2/FiO2] ≤200 mmHg; (3) bilateral infiltrates on frontal chest radiograph; (4) absence of left atrial hypertension. ARDS was considered the more severe form of an overarching entity called acute lung injury (ALI), defined by the same criteria, but with less severe hypoxemia (PaO2/FiO2 ≤300 mmHg) [4].

Issues with the AECC definition of ARDS

Despite its utility, it became apparent that the AECC definition also had limitations. These included imprecision in the exact time frame of ‘acute onset’; variation in the PaO2/FiO2 ratio according to FiO2 and mechanical ventilation settings; poor interobserver reliability of the chest radiographs interpretation; and the frequent misclassification of left atrial hypertension. In addition, the term ALI was often misused to indicate only less severe patients with PaO2/FiO2 200–300 mmHg, instead of being used to define the overall syndrome. Because of these concerns and others, the European Society of Intensive Care Medicine convened an international expert panel in 2011 in Berlin to develop a new definition of ARDS.

The Berlin definition of ARDS

According to the Berlin definition, ARDS is a form of acute diffuse lung injury occurring in patients with a predisposing risk factor, meeting the following criteria: (1) onset within 1 week of a known clinical insult or new/worsening respiratory symptoms; (2) presence of bilateral opacities on the chest radiographs, not fully explained by effusions, lobar/lung collapse, or nodules; (3) diagnosis of respiratory failure not fully explained by cardiac failure or fluid overload, with the need for objective assessment (e.g., echocardiography) to exclude hydrostatic edema if no risk factor is present; (4) presence of hypoxemia, defined by PaO2/FiO2 measured with a minimum requirement for PEEP of ≥5 cmH2O (or non-invasive continuous positive airway pressure ≥5 cmH2O for mild ARDS) and identifying three mutually exclusive categories of severity: mild with 200 mmHg < PaO2/FiO2 ≤ 300 mmHg, moderate with 100 mmHg < PaO2/FiO2 ≤ 200 mmHg, severe with PaO2/FiO2 ≤ 100 mmHg [8, 9].

Advances with the Berlin definition of ARDS

Several important issues were addressed in this new definition. The often misused term ALI was removed. A specific timing of onset was defined. The need for a predisposing risk factor was incorporated. The exclusion criterion based on the presence of hydrostatic edema was redefined. The radiological criteria were reformulated. The requirement of a minimum PEEP to establish the severity of hypoxemia according to PaO2/FiO2 was introduced. Moreover, the three mutually exclusive categories of mild, moderate, and severe ARDS were validated, as they were associated with increasingly severe disease using mortality, ventilator-free days, and the duration of mechanical ventilation as outcomes in survivors. Furthermore, the draft definition underwent testing and revision using predictive validity for mortality as a standard. This process resulted in a final definition that was simpler to use but had similar predictive validity and better distribution among categories [8, 10]. Finally, the reliability of the definition was addressed by the publication of a reference set of vignettes and chest radiographs to facilitate consistency in the interpretation [9].

Criticisms of the Berlin definition of ARDS

Criticisms have been raised about the Berlin definition. In particular, it has been shown that the measurement of PaO2/FiO2 using predefined ventilator settings after 24 h from the initial assessment would improve the stratification of the ARDS severity [11]. However, this approach could delay enrollment in clinical trials or delivery of beneficial treatments and also negates much observational research. Moreover, it has been suggested that adding compliance normalized to ideal body weight to the criteria would improve the risk stratification within each subcategory of severity [12]. These and other criteria including biomarkers, extravascular lung water, and CT scans were considered and rejected during definition development because of a lack of association data, impact on feasibility, or both. Only more observational data will allow further refinements of the definition of ARDS with the aim of retaining feasibility and improving its reliability and validity.

Uptake of the Berlin definition of ARDS

The uptake of the Berlin definition has been impressive. Initiated by the European Society of Intensive Care Medicine, the process was endorsed by the American Thoracic Society and the Society of Critical Care Medicine. Since its publication in 2012 it has been cited more the 1500 times and is the subject of more than 150 articles on PubMed. Furthermore, the Pediatric Acute Lung Injury Consensus Conference used the Berlin criteria for the development of the pediatric ARDS definition [13, 14]. Finally, the Berlin definition is used in a remarkable number of ongoing clinical trials in ARDS (Table 1).

Table 1 Title of the ongoing randomized clinical trials on ARDS, diagnosed according to the Berlin definition, the sponsor and the identification number
Full size table

Conclusion

The Berlin definition of ARDS shows qualitative improvements in a number of definition metrics, and better suits the needs of critical care clinicians and investigators as demonstrated by its immediate and wide uptake.