Published date: July 29, 2025
Professor Winfried Randerath is active at the Bethanien Hospital in Solingen, Germany and has for years been one of the leading experts within the European sleep world. He is frequently engaged as a speaker and is appreciated for his ability to structure, clarify and tie topics together. He has long held a leading role in the guideline work for the European Respiratory Society and the Baveno classification, the work discussed in this interview and one of his initiatives aimed at structuring and improving sleep care.
Prof. Randerath: Yes, in the first version we simply took the provocative approach of saying we don't need AHI to identify the need to treat a particular patient’s OSA, we only need it to identify the presence of the disease. AHI itself does not describe the severity of the condition. It does not describe the consequences, nor the possible prognostic effects. That's why we initially put it aside to a certain extent, leaving it out completely and concentrating only on what is potentially relevant for the patient.
With the first Baveno classification, we focused on symptoms and comorbidities as the primary factors to direct the need to treat. On one side, there are symptoms such as drowsiness or disturbed sleep. On the other, there are concomitant or secondary conditions, such as cardiovascular diseases. This led to a two-dimensional ABCD scheme, from no symptoms and no secondary diseases to pronounced symptoms and the presence of secondary diseases. However, we then realized AHI is a prognostic marker above a certain level and therefore should not be completely ignored. We know from many long-term studies that patients with a relatively high number of respiratory events are at increased risk, regardless of whether they have other symptoms. That led us to take AHI back into account in the case of a high degree of severity or a high number of respiratory events.
Prof. Randerath: The inclusion of AHI as one factor in the Baveno classification means that a much larger portion of patients fall into the risk group vs. an approach where only symptoms or external risk factors are considered. Therefore, bringing AHI into the equation casts a wider net, opening the door for a larger number of patients to require treatment. This results in more patients who are at risk and therefore more patients who need to be treated.
So, while if we focus only on the symptoms or only on existing risk factors, there are patients who do not fall into either category, i.e. who are asymptomatic or less symptomatic and do not yet have a significantly increased risk, but who do have a high number of respiratory events. In the long term, we know this has prognostic significance and will therefore influence treatment because then a larger portion of patients will require treatment.
Prof. Randerath: This is the consideration from various long-term studies that have set the AHI a little differently, but there was a relatively uniform picture that there is a prognostic factor for an AHI over 30. That's why we decided on this limit, which is of course arbitrary. The individual doctor will always make a different decision in one direction or another for a patient who has a higher or lower an AHI . He or she has this freedom, of course, but it was a decision based on the literature, which has often set a high risk above this limit.
Prof. Randerath: Compared to other criteria? Well, what do we have at the moment? Most current guidelines simply group patients into three AHI ranges—below 15, 15 to 30 and above 30. Some may go one step further and indicate treatment for patients with an AHI below 15 if symptoms are present, but that’s all.
In other words, the individual experience of the patient and their unique risk is not factored in. So, I think of the Baveno criteria is more like personalized medicine, not just medicine based on a single numerical value.
This article is one in a set from our conversation with Professor Randerath. To continue, read ”The impact of the modified Baveno classification on key patient groups”.
Privacy|Terms of use |©2025 Resmed|All rights reserved