Rome Criteria for Exacerbation of Chronic Obstructive Pulmonary Disease: Not Built in a Day

To the Editor:

The Perspective article by Celli and colleagues is a much-needed effort to update the definition and severity classification of acute exacerbation of chronic obstructive pulmonary disease (AECOPD) (1). The current definition of AECOPD is limited by its reliance on patient reports, lack of specificity, and a severity classification defined by where patients are treated. The authors propose a number of criteria, which include both subjective and objective measures. Although a step in the right direction, several issues should be considered before this definition and classification schema are ready for prime time, either for use clinically or for use in research studies. The proposed definition justifiably argues for an upper limit for the duration over which symptoms develop. The lack of a lower threshold for the duration, however, makes it hard to differentiate day-to-day variation in symptoms from a true exacerbation at the time of onset. Unless symptoms are severe and of rapid onset, this definition may not facilitate early and timely detection of exacerbations. The existing definition, which requires a change in symptoms beyond usual, places the onus of detection of clinical deterioration on the patient. The main value, however, of this definition lies in that within-patient day-to-day variation is accounted for (2).

The proposed definition discounts cough as an important symptom, though many patients rank cough as more bothersome than dyspnea, and given that dyspnea can be modulated by the level of physical activity, cough may indeed be a more sensitive symptom than dyspnea is for early recognition of an exacerbation. Although a threshold of five for the visual analog scale for dyspnea is perhaps practical, it is important to make a distinction between chronicity and acuity of symptoms; it is not uncommon for individuals with severe COPD to have significant dyspnea at baseline. I presume the authors want to use a single value because this scale is not commonly used on a daily basis by patients. However, such a scale is, by its nature, especially when maximal severity is worded as “most dyspnea ever felt,” designed to measure change with the assumption that patients will have experienced the entire spectrum of the severity scale. Perhaps a visual analog scale recall over the preceding few days may help introduce a measure of change that is necessary to define an “exacerbation” or worsening. The suggested oxygen saturation criterion of less than 92% may be challenging to implement, given that the current recommendation is to maintain oxygen saturation at 88% and above.

The inclusion of self-reported physical exam findings may also be problematic. Not all patients can accurately and reliably count their heart rate and respiratory rate, especially if this is not measured over a certain minimum time period (3, 4). Do these physical exam findings need to be abnormal for a certain period of time? Although the authors suggest that technological advances will make these measurements easy, prior experience suggests that the signal-to-noise ratio is quite low for daily monitoring, and a snapshot provided by patients may result in frequent false alarms (5). The inclusion of laboratory values, although desirable for confirmation, will also imply that patients will need to get to a healthcare facility instead of diagnosis over the phone. The authors should be congratulated for this much-awaited move toward changing the current unsatisfactory definition of AECOPD, but Rome was not built in a day. The proposed definition needs refinement.