How do acute worsening events influence daily life and healthcare-seeking behaviour in patients with COPD: an international multicountry qualitative study Authors: Lars Dijk, Merijn M G Driessen, Yoran H Gerritsma, Clare Bolton, Carla Da Silva, Janwillem W H Kocks Published: 2026 ABSTRACT AbstractObjectiveTo explore the impact of acute worsening events (AWEs), defined as clinically relevant deteriorations in peak expiratory flow, reliever use and/or symptoms, on patients’ daily lives and identify behaviours linked to their recognition and management.DesignA qualitative international substudy was conducted in the Netherlands, Spain, the USA, Canada and the UK (2023–2024).Participants19 patients with moderate-to-severe chronic obstructive pulmonary disease (COPD) from a randomised clinical trial participated. Interviews, triggered by AWEs and repeated 6 weeks later, were audio-recorded, transcribed and analysed.ResultsPatients varied in identifying bad days, reporting inability to perform physical activities, symptom worsening or the need for add-on treatment. Recognition of AWEs depended on their duration: shorter events (<2 days) often went unnoticed, while longer events were typically recognised. Most patients indicated they would seek healthcare if needed. Barriers included difficulty obtaining timely appointments (external) and a reactive attitude to worsenings (internal).ConclusionsAWEs significantly impact daily life but vary widely in recognition and thresholds for seeking care. Limited awareness of daily disease variations may contribute to both under-reporting of worsening events to healthcare providers and suboptimal self-management in COPD.Trial registration numberNCT05492877. STRENGTHS AND LIMITATIONS OF THIS STUDY The multicountry design increased methodological transferability by capturing perspectives from different healthcare systems and cultural contexts. The multicountry design increased methodological transferability by capturing perspectives from different healthcare systems and cultural contexts. Semi-structured qualitative interviews conducted shortly after acute worsening events enabled exploration of patient experiences while minimising recall bias. Semi-structured qualitative interviews conducted shortly after acute worsening events enabled exploration of patient experiences while minimising recall bias. The use of a predefined interview guide supported consistency across interviewers and study sites. The use of a predefined interview guide supported consistency across interviewers and study sites. Data collection relied on participants’ completion of symptom diaries to identify acute worsening events, which may have led to missed events. Data collection relied on participants’ completion of symptom diaries to identify acute worsening events, which may have led to missed events. Introduction Chronic obstructive pulmonary disease (COPD) is a common chronic lung disease that is characterised by persistent respiratory symptoms and airflow limitation.1 COPD is one of the leading causes of death worldwide2 and represents an important public health challenge that is both preventable and treatable.1 2 Exacerbations are an important feature of COPD and a major cause of morbidity, mortality and reduced health status.3 They impose a substantial burden on healthcare systems worldwide.3 Prevention of exacerbations is therefore often the goal for new medication being developed. A generally accepted definition of an acute COPD exacerbation is ‘a sustained worsening of the patient’s condition, from the stable state and beyond normal day-to-day variations that is acute in onset and may warrant additional treatment in a patient with underlying COPD’.4 Randomised controlled trials (RCTs) generally define acute exacerbations by (1) the use of antibiotics and/or oral corticosteroids (moderate exacerbations) as a treatment for acute worsening of symptoms, (2) hospital admissions or (3) mortality related to this (severe exacerbations).5 The use of this definition has limitations, including patient and healthcare provider subjectivity, as a substantial number of potentially impactful events go unreported and untreated. Patient diaries applied in RCTs have shown that periods of unreported events happen frequently and are therefore not recorded as exacerbations in healthcare systems.69 The impact of these unreported events on the daily life of patients with COPD is largely unknown, as are the reasons for not reaching out to healthcare professionals. Unreported events have been reported to be associated with worsening quality of life,10 a clinically important decline in health status11 and increased risk of subsequent emergency hospitalisation.12 Patient diaries have provided a potential outcome to register these unreported events. In a post hoc analysis of 20 RCTs, Vogelmeier et al investigated an algorithm for capturing ‘Acute Worsening Events’ (AWEs) based on deteriorations of peak expiratory flow (PEF), reliever use or symptoms that fulfil specific criteria in terms of reaching thresholds or slopes.13 A composite endpoint (COPDCompEx) was defined as the occurrence of either such an AWE, a moderate or a severe exacerbation, or a study dropout. These AWEs allow for capturing the missing unreported events, leading to a more complete endpoint for measuring treatment effects on the number of worsenings that patients experience. At 3 months of follow-up, the proportion of patients experiencing a COPDCompEx event increased over threefold versus exacerbations (ie, moderate or severe exacerbations) alone, showing a large amount of otherwise uncaptured worsening events.13 Incorporating COPDCompEx as an outcome measure in clinical trials may lead to more sensitive measurement of outcomes in patients with COPD. The higher event rate of COPDCompEx compared with exacerbations increases the statistical power of the trial, making it possible to reduce the length and required sample size. Before the widespread use, it is important to assess the impact on patients of AWE to establish the value as an outcome measure in future trials. While these AWEs are based on the patient’s diary data, the real (clinical) impact of these AWEs on patients daily living is not fully understood. Here, we performed a qualitative study using semi-structured interviews to provide more insight into the impact of AWEs, as defined by the COPDCompEx algorithm, on patients with COPD. Additionally, we investigated the behaviours associated with an AWE. The goals of this study were (1) to evaluate the impact of an AWE on the daily lives of patients with COPD and (2) to identify and explain the AWE-associated behaviours, including the underlying reasons for not consulting a healthcare professional (HCP). Methods Study design This qualitative study was conducted as an optional substudy for participants of the CRESCENDO trial (ClinicalTrials.gov reference: NCT05492877). The substudy ran in a selection of the countries participating in CRESCENDO, namely Canada (CAN), Spain (ESP), the Netherlands (NLD), UK and USA. The selection of countries for this substudy was guided by pragmatic considerations, including the willingness of local study sites to contribute to this substudy. Patients consented to the substudy between July 2023 and April 2024. The CRESCENDO trial is a phase IIa randomised, double blind, placebo controlled, parallel arm study to evaluate the efficacy and safety of mitiperstat (AZD4831) in adult participants with moderate-to-severe COPD, on dual or triple therapy, at risk of exacerbations between 40 and 80 years of age, without a current or past diagnosis of asthma (a full list of inclusion and exclusion criteria can be found on ClinicalTrials.gov reference: NCT05492877). In the CRESCENDO trial, participants filled out a daily eDiary in the morning and in the evening. eDiary variables were a measure of PEF (morning and evening), total doses for reliever use (sum total of morning and evening) and COPD symptoms in the evening, captured by the Breathlessness, Cough and Sputum Scale.14 Deterioration criteria were defined as either reaching a threshold for two consecutive days or deterioration based on daily rate of change for five consecutive days (ie, slope events). To record AWEs, concurrent worsening in at least two different diary variables was required with PEF always required to be one of these variables. This could either involve two variables meeting threshold criteria or at least one variable fulfilling a threshold criteria (eg, reliever use 1.75 times increase from baseline) and the slopes for all variables fulfilling the predefined criteria (eg, PEF 3% decrease rate per day). More detailed information on the algorithm can be found in the paper by Vogelmeier et al.13 Participants who signed up for this optional study were monitored for AWEs using COPDCompEx.13 Participants who had an AWE were interviewed two times. The first interview took place as early as possible after the AWE was triggered (mean±SD: 8.7±3.6 days). For one participant, their trigger AWE was still ongoing during their first interview. A second interview took place 6 weeks after the first interview (mean±SD: 42.7±4.71 days). AWEs that led to the first interview will further be referred to as trigger AWEs. Many of the interviewed participants had multiple AWEs (table 1), and around half of the participants interviewed had a subsequent event happen within the time window between interviews. These subsequent events provided extra insight into whether AWEs were recognised, and how they impacted the daily lives of the participants. As time is an important factor for remembering events, we considered AWEs that were still ongoing or ended <11 days before the second interview to keep a similar time window as between the trigger AWE and interview 1. Table 1Descriptive characteristics of participants at the time of enrolment in the CRESCENDO trialCharacteristicStudy population% missingn19Age (mean (SD))67.4 (6.0)0.0Sex (%)0.0Male14 (73.7)Female5 (26.3)Exacerbations (moderate or severe) in the past year (mean (SD))3.2 (2.5)21.1Smoking status (%)5.3Current6 (27.8)Former13 (72.2)Pack years (mean (SD))55.7 (27.3)5.3Medication regimen (%)0.0ICS/LABA2 (10.5)LAMA/LABAICS/LAMA/LABA1 (5.3)16 (84.2)COPD Assessment Test score at time of study enrolment (mean (SD))15.1 (7.3)5.3Country of residence (%)Canada5 (36.3)The Netherlands2 (10.5)Spain2 (10.5)UK8 (42.1)USA2 (10.5)AWE prevalence during the study (%)0.0One6 (31.6)Multiple13 (68.4)AWE, acute worsening event; COPD, chronic obstructive pulmonary disease; ICS, inhaled corticosteroid; LABA, long-acting beta₂-agonist; LAMA, long-acting muscarinic antagonist. Descriptive characteristics of participants at the time of enrolment in the CRESCENDO trial AWE, acute worsening event; COPD, chronic obstructive pulmonary disease; ICS, inhaled corticosteroid; LABA, long-acting beta₂-agonist; LAMA, long-acting muscarinic antagonist. Due to the research question looking at event recognition, the medical ethics committees agreed with not fully informing participants on the study goal or the actual reason for being invited to the interviews. Participants were informed that interviews were performed at random and that not all who consented to the sub-study would be interviewed. Theoretical perspective We used a combination of interpretivism and a grounded theory approach,15 where data were collected and analysed concurrently. This allowed us to explore patterns and develop conceptual insights regarding the impact of AWEs and associated behaviours based on the subjective experiences and interpretations of patients. While a formal, fully developed theory was not generated, this approach enabled a systematic identification of themes and relationships in the data. Participants All participants randomised into the CRESCENDO trial within selected study sites were eligible to sign up for this study. Consenting to this study was possible from the CRESCENDO screening visit until the fifth CRESCENDO study visit (after randomisation to control or intervention, which occurred at study visit 3). Participants were invited to take part in this substudy by study staff at their research site. In total, 52 participants expressed interest and signed up for participation. Participants were included if they experienced an AWE during the CRESCENDO trial and if the moderators were able to schedule an interview within the study time frame. No additional exclusion criteria were applied. Recruitment for the substudy was conducted by study staff at the research sites; therefore, no data were collected on the number of participants who declined participation. Participants were compensated for their time spent in interviews in accordance with local regulations. Semi-structured interviews Two semi-structured interviews of approximately 30 min each were conducted. The interview guide (online supplemental appendix A) specified which topics to discuss for each of the two interviews. In the first interview (directly after the trigger AWE), we discussed (1) current status of COPD, (2) perception of recent days, (3) perception of COPD symptoms and exacerbations and (4) reflection. In the second interview (6 weeks after the first interview), we discussed (1) perception of recent days, (2) health behaviour, (3) barriers to reaching out to healthcare providers. Interviews were performed remotely either via telephone or videoconferencing (Zoom) call. Interviews were performed by four local, native-speaking moderators, with the interviews in Canada and the USA being performed by the same moderator. Moderators from Spain (male, MD) and the Netherlands (male, MSc) were, respectively, a clinical doctor and a researcher. Moderators for Canada/USA (female, BCom) and UK (male, MSc) were qualitative researchers and experienced moderators, working as consultants. Moderators were trained on qualitative versus clinical interview techniques, study objectives and the interview guide. Over the duration of the study, moderators received feedback on interview technique and how to best capture further information that is important to the study objectives. After the revision of the interview guide, moderators were also provided with a summary of the first interview, to use as an extra guide for the second interview. For example, on specific wording used by the participant (eg, flare-up for an exacerbation), or relevant experiences that would require follow-up. Moderators did not establish a relationship with the participants prior to the study. At the start of the interview, moderators explained their role in the study to participants and that individual responses would not be shared with site staff. Interviews were audio-recorded and transcribed verbatim by independent companies. No field notes were taken during the interviews and transcripts were not returned to participants or moderators for commenting/correction. Canadian (French), Spanish and Dutch transcripts were translated to English. All interview transcripts were anonymised before analysis. Data analysis An initial basis for a coding tree was established based on the interview guide, after which the coding tree was continuously updated based on continuous analysis of the incoming interviews. Interviews were continuously analysed over the course of the study, but coding for final analysis was performed on both interviews of a participant together. This was done to better capture participants’ views and overarching patterns over the interviews. All data were coded by two researchers (authors MMGD (MSc) and LD (MSc)) with research experience in the respiratory field and training in qualitative research. Clinical oversight of the data analysis was provided by JWHK (author, MD, PhD). The first five interviews were coded by both researchers independently using Dedoose (V.9.2.6, SocioCultural Research Consultants, Los Angeles, California, USA). After each interview, the applied codes were discussed in a reconciliation meeting, and the coding tree was updated when necessary. After the fifth interview, consensus was reached and interviews were coded by a single researcher, with MMGD coding an additional 11 participants and LD, coding an additional 3 participants. Reconciliation meetings took place regularly, after every two to five interviews to further discuss new additions to the coding tree, or new views on existing codes. An overview of all the themes and codes that were identified can be found in figure 1, the codebook can be found as online supplemental appendix B,codebook. Overview of themes and codes that were identified. AWE, acute worsening event; COPD, chronic obstructive pulmonary disease. Presented participant quotations were selected in such a way to best represent the full participant sample, highlighting both views that were often encountered as well as views that might contradict. Participant quotations are identified using a coded patient number and the country. Descriptive statistics on the participant population were performed in R (V.4.4.1, using the RStudio IDE V.2024.04.2, R Foundation for Statistical Computing, Vienna, Austria). Participants did not provide feedback on the findings before reporting. The study was reported in accordance with Consolidated Criteria for Reporting Qualitative Research (COREQ) criteria16 (online supplemental appendix C, COREQ checklist) Patient and public involvement statement The initial interview guide was tested with two patients with COPD, one from the UK and one from the USA, contacted through the Patient Partnership Programme of AstraZeneca. These patients were not included in the main study. Results A total of 23 patients that provided consent for this substudy of the CRESCENDO trial experienced an AWE (44.2% of the 52 who consented for this substudy). Of these, 19 patients were interviewed (82.6%) for the study (CAN: n=5 (from two study sites), ESP: n=2 (two sites), NLD: n=2 (two sites), UK: n=8 (four sites), USA: n=2 (one site)). Thus, four patients with an AWE were not interviewed: one patient was hospitalised and not available for an interview. The other three patients had an AWE close to the end of their participation in the CRESCENDO study and could therefore not be interviewed within the duration of the study. A descriptive overview of the patient characteristics can be found in table 1. The focus of this paper relates to the impact and identification of an AWE on daily lives of patients with COPD (differences between good and bad days), and patients’ experiences during these events and their tendencies to reach out for help. The following overarching themes were identified during the interviews: (1) Heterogeneity within patients regarding the impact of COPD on daily living and the identification of worsenings (2) AWE recognition and (3) healthcare seeking behaviour during or after an AWE. Heterogeneity within patients regarding the impact of COPD on daily living Patients with COPD make for a very heterogeneous patient population. Not only in clinical characteristics but also in the way that they experience and cope with their COPD. From the interviews, it became clear that even in a population where the clinical characteristics are bound by strict selection criteria from clinical trials (ie, moderate-to-severe COPD and at high risk of exacerbations) there is a lot of variation in subjective experience and coping style. This subjective experience of patients is an overarching theme that runs through this study as a common thread and influences how AWEs are perceived and acted on. When asked how to describe bad days, or what constitutes a bad day, patients reported various answers. The most common ways it was defined was based on the patient’s ability to do normal daily activities, like going shopping, taking stairs or taking a shower or based on worsening symptoms, most often breathlessness (table 2 quotes 1–3). Table 2Participants terminology on ‘bad days’#Patient quotePatient identifier+country 1 Somedays I haven’t got the energy to do anything. [Uh]other days I-I can just potter about, and then when I am feeling sort of up, I can do the more physical type things like hooverP08, UK 2 A bad day is with my breathing—Where I can’t get a breathP12, UK 3 There are only two factors that make up a bad day. One is energy and the other is oxygen breathing. They’re the two factors. If my breathing is bad, my energy levels are low.P06, UK 4 It’s never terribly worse from one day to the next. Really.(…)I just don’t really have good and bad days. I don’t know what to tell you. I’ve been pretty(…silence…)unless I start trying to do something that is a lot of physical exertion and then it will get me.P19, USA 5 I call them “crises”, I had four of them. I had to take the emergency plan four times, to bring myself back to life and breathe.P01, CAN 6 I know that there’s exacerbation too… But exacerbation, for me, it’s when I need to go to the hospital and I am not able to, to recover from it.(…)I have periods when I have a low period. For me, that’s not a, an exacerbation. At that time, I know that I have to modify a bit my, my activities to recover from it.P05, CAN 7 Just not more than, I don’t know, 10 minutes or something.(…)Usually it’s just once in a while. It’s not consecutive days or anything.P19, USACAN, Canada. Participants terminology on ‘bad days’ CAN, Canada. Some patients were also unsure about what to say, indicating that they did not have bad days, or that all days were about the same (table 2 quote 4). The latter is interesting, given that this was asked at the time that they were experiencing a CompEx defined AWE. Patients mentioned both visible symptoms (such as breathlessness) when they described a bad day, others relate it more to the consequences of worsenings—rather than the ‘visible’ symptoms itself (ie, being admitted to the hospital (table 2 quote 8)). AWE recognition Terminology bad day First, we observed that the majority of patients did not use the term ‘exacerbation’ when describing bad days and/or periods of worsening. They used terms such as ‘crisis’ (table 2 quote 5) or were referring to bronchitis, others were referring to a lung attack or flare-up. The finding that patients had different perceptions of what constitutes a significant event is further demonstrated by the events they describe when asked about ‘a bad day’. Some patients immediately started describing periods of worsening requiring additional treatment (antibiotics and/or steroids or requiring hospital admissions). It showed a tendency of these patients to remember the most impactful events and describe a strong worsening of symptoms that required interventions, not recounting any other fluctuations or less severe, shorter, worsening of their symptoms. Other patients distinguished between different levels of bad days and also talked about smaller fluctuations in symptoms and energy (table 2, quote 6). When asked about the duration of bad days or a worsening, answers varied from 10 min (table 2, quote 7) to 15 days. Physical activity Many patients described bad days by the impact on their ability to do activities (table 3 quotes 1–2), and physical activity was often mentioned as a trigger for symptom worsening (table 3 quote 3). Table 3Physical activity#Patient quotePatient identifier+country 1 There’s days when, you know, I can’t even be bothered [uh]to go in the bathroom and sometimes I don’t even get a wash until maybe afternoon(…)even though I’ve slept and I’ve slept quite well, my energy levels are sometimes that low.P08, UK 2 It’s what helps me measure, measure how I, I feel. It’s the stairs.And after washing myself, whether it’s in the shower or the bath… I, with the humidity and the warm water… I, I feel like [um]that like, I know if I have more difficulty or not.P05, CAN 3 It’s only with physical exertion. If I do too much or want to do too much, I notice I’m running out of breath. Then, I take it easy for a while. But usually, I try to avoid [physical exertion].P14, NLD 4 It’s just running about. It’s taking dogs out. It’s going out with my wife, running around with granddaughters, things like that, shopping or taking dogs on walks. I tend to keep myself busy.P09, UK 5 I’ll go down and make my wife a little breakfast or something like that. Bring it up. We have a two-storey house, so I’ll go up. Maybe I get a little out of breath going up the stairs, but not too bad. I just do chores around, whatever. I watch TV, do chores around the house, go to the supermarket, go shopping with my wife, go see relatives, go see her daughter or grandkids.P18, USA 6 Always lazy morning. [Uh]I don’t motivate myself until around about this time. [Uh]and then I’ll-- I don’t even get dressed most days. If I’ve got nothing on, I won’t get dressed. I’ll go, that’s all. I have a shower, like come up in the shower when I got off phone from you. Put a clean pair of jammers on and that’s it as far as that’s concerned.P08, UK 7 it prevents me-- it prevents us--(…)we used to-- go to-(uh] shows to see plays, things like that, but now we do less.P02, CAN 8 Because I’ve seen people that do turn around and just back in that have got my symptoms and conditions and it just destroys you because you’re just stuck in a chair and that’s not good. That’s not life. Get a move. That’s my philosophy. Keep moving while you can.P09, UK 9 I just have to keep an eye out every day that I don’t overexert myself, that I get short of breath, that I start doing too much. But you just know that in the long run. I mean, you have to listen to your body.P14, NLD 10 I went out with my family on the Friday to the zoo with, you know, my family and the kids and everything.I couldn’t keep up with them. [uh] and I did feel tired, but so and I was at outdoors all day.(…)I pushed myself too much. Absolutely just… it was just like trying to be the bionic woman in not the right circumstances.P10, UK 11 If I had to go and walk the dog, which my husband walks my daughter’s dog, I would think, “Right. Come on. We’ve got to get on with it.” I would dash some inhaler before I go because of it being cold and I’d probably need some and take an inhaler with me and I'd probably need it when I got to the field or wherever I’m taking her.P07, UK 12 (on inhaler use) That could be about 3 or 4 times a day and, and different, different times. Again, it all depends where I am and what, what I’m doing.P07, UKCAN, Canada; NLD, the Netherlands. Physical activity CAN, Canada; NLD, the Netherlands. Across our patient population, a large variability in the extent to which individuals were still willing to engage in physical activity and push the limits of their abilities (table 3 quotes 4–6). Some patients still try to actively push their limits and do as much as possible, whereas others seem to have accepted that they are not able to do much physically (table 3 quotes 7–9). This difference in how patients cope with the limitations imposed by their COPD influences how likely patients are to push a bit too far and, for instance, also influences reliever use (table 3 quotes 11–12), potentially detected by the patient diaries in this study and triggering an AWE (in combination with deteriorations of PEF and/or symptoms). Among patients who indicated not having bad days, or only describing exacerbations that require treatment as bad days, physical activity was often a factor that did lead to increased symptoms. Acute worsening event recognition and associated behaviour Within our study, the duration of trigger AWEs varied from 2 to 50 days (min to max, median: 2, IQR 2–8.75), with 56% of all trigger AWEs being 2-day events. When later AWEs, before or during the second interview, are included, this percentage decreases to 36%. During the first interview, AWE recognition was assessed by asking patients how their COPD was over a past period, first inquiring in general and then asking about specific symptoms and their activities over this period. Later in the interview, patients were also asked when the last time was that they had bad days due to their COPD. About half of the patients mentioned a worsening matching with the dates of the trigger AWE, which showed that the recognition of AWEs on patients was mixed. The most important factor explaining AWE recognition was AWE duration, with strongly different responses on the above-mentioned questions between shorter 2-day events, which is the minimum duration of an AWE according to COPDCompEx criteria,13 and longer events (figure 2). Overview of AWE recognition. Duration of trigger AWEs >2 ranged between 5 and 50 days. AWEs, acute worsening events. Most patients appeared to clearly recognise the events that lasted longer than 2 days (range: 5–50 days), often reporting worsening symptoms and inability to perform activities over the same time window as the AWE (table 4 quotes 1–5). Others seemed to not have recognised the worsening, reporting their condition has been stable. These patients indicated not having bad days (table 4 quotes 7–8), or only described these (moderate) exacerbation events as bad days (table 2 quotes 5–6). Table 4Quotes on AWE recognition#Duration of AWEPatient quotePatient identifier+country 1 9 daysI’m having a hard time, but it’s because of my pneumonia.(…)tired, that’s for sure, very tiredP03, CAN(trigger AWE) 2 10 daysthis weekend I had two or three bad days where we [husband] were really strugglingP09, UK 3 6 daysLast week [uh]I felt really not well on a Sunday and the Sunday evening. Monday I just could not stop coughing at all, not at all.P10, UK 4 7 daysthe past few… I just did my little chores. I haven’t done physical activity more than that. I didn’t have enough energy for that.P03, UK(later AWE) 5 2 daysI couldn’t- I couldn’t even walk out to the parking lot. I couldn’t breathe well enoughP19, USA 6 2 daysI wouldn’t know when I last had a really bad day.P14, NLD 7 11 daysInterviewer: are there any occasions when you’ve had a worse, you know, like a bad day or a worse day that’s slow?Respondent: Not really, no.P13, UK 8 8 daysInterviewer: Do you sometimes have periods where it’s worse or some…Participant: Well so far [uh]no, not really.P04, CAN 9 2 daysit’s stable since-- For 2 weeks there, yes, it’s- it’s stable.P02, CAN (interview 1) 10 4 daysIt’s going less well. This week, it’s not going so well. I’m keeping an eye on it, every day, to see what’s going on. My wife is at -- encourages me to go [uh]to the emergency room, since my family doctor is two and a half hours away, we’ll forget about it. She’d like me to go to the ER, but I don’t like wandering around the ER. So—(…)It’s because I tell myself all the time, [you know], I don’t see myself showing up at the emergency room [uh], then that the symptom isn’t there. I don’t know if you understand.P02, CAN (interview 2) 11 5 daysOh, it’s going well, it’s going well. For sure through last week, I have had a bit more difficulty, but… that that’s like… I have days, sometimes when I have, I have a low period but that, I recover, I recover, I come back just fine.P05, CAN 12 50 daysIt was a miserable day, all told. I did nothing to go right. I dropped a knife. I was buttering a slice of bread and everything just couldn’t function at all. I couldn’t breathe. I seemed to be forever just thinking, now I need a minute. It was just a nightmare of a day.(…)and then the following day wasn't so bad. It wasn’t a good day, but it was a lot better than Friday.P06, UK 13 22 daysLast weekend, I had a really bad weekend. [Uh]I even phoned my son up and said, I wasn’t feeling very well [uh[, and I-I said, “I’m gonna go to bed.” And I said, “But phone me in the morning, just to make sure I’m okay.”P08, UKAWE, acute worsening event; CAN, Canada; NLD, the Netherlands. Quotes on AWE recognition AWE, acute worsening event; CAN, Canada; NLD, the Netherlands. Of all 2-day events, only one was recognised, where a patient indicated not feeling well and going to the hospital a few days later. One other event was not recognised at that time, but the patient indicated feeling bad in the week after the AWE event. None of the other 2-day AWEs were recounted by patients, with patients indicating that their COPD has been stable, and no bad days have been experienced recently (table 4 quote 6). As patients were also questioned about their recent COPD status at the second interview (6 weeks after the initial interview), we could look at the recognition of two separate AWEs for about half of the patients. Some patients did not recognise the initial trigger AWE but did recognise a later AWE during the second interview. Also here, the duration of the AWE seemed to be influential, as both patients had a minimum duration trigger event which was not recognised, followed by a longer later event which was clearly recognised from the interviews (table 4 quotes 9–10). The difference in impact was also clear from patients recounting over the past period, with patients who experienced a longer AWE describing impactful bad days, and many patients also feeling the need to initiate treatment with antibiotics and/or steroids or to go to the emergency room (ER)/hospital, escalating the event to an exacerbation. Healthcare seeking behaviour (barriers and facilitators) Patients reported multiple barriers and facilitators influencing the decision on whether (or not) to reach out to healthcare professionals. Both internal and external factors in healthcare-seeking behaviour were mentioned. External barriers In general, a couple of external barriers hindered patients from reaching out and getting treated during a worsening. The most indicated barrier was that patients found it difficult to get an appointment with their treating physician on short notice, whether this be their general practitioner, family doctor or pulmonologist (online supplemental table 1 quotes 1–4). I need to go to the emergency room to get treated, at least try to get a doctor’s appointment. Usually, I cannot get an appointment that quickly, so I end up going to the emergency room. (P18, USA) In exacerbations, I, I have rarely seen my doctor. […] Often, my family doctor isn’t available in those moments. (P04, CAN) This did not pose a hard barrier for most patients, as there was often a specialised COPD unit or ER available in case of emergency, or patients (specifically in the UK) had access to a rescue pack of antibiotics and/or steroids. However, in some cases not being able to get into contact with their familiar HCP did seem to increase the threshold to reach out, specifically in combination with bad experiences of long ER waiting times (online supplemental table 1 quotes 5–7). But… but that’s just it. I can’t go to the hospital and wait 2 or 3 hours if I’m struggling to breathe. I won’t go die in the waiting room. (P03, CAN) The other external barrier seemed to be that patients were unsure what to do when symptoms worsen, yearly check-ups were missing, or their current treatment was no longer in line with previous information provided by the HCPs—leaving them unsure of what to do (online supplemental table 1 quotes 8–10). Internal barriers The most frequently observed internal barriers that withheld patients from reaching out included reactive response to symptoms and self-management. Patients indicated that they tend to wait a while for it in the hope that it would recover without specific interventions, or that it would start improving the next day (online supplemental table 2, quotes 1–4). it was just one of those things that usually you figure you-you just feel bad, and you go to bed and wake up the next day, and you feel better, except I never did. I got worse. […] I waited too long. It’s my own fault. […] I-I let it go, but it was over the weekend and I didn’t feel like dealing with it. (P19, USA) This was often combined with limiting their activity and using over-the-counter medicine, in an effort to manage the worsening (online supplemental table 2 quotes 3–5). One patient also reported this reactive response and self-management due to not feeling adequately treated by their HCP (online supplemental table 2 quote 6). Further barriers were described by some patients not having any regular contact with an HCP, or not having a dedicated HCP at all, other patients just growing tired of constant healthcare visits (online supplemental table 2, quote 7–9), some not feeling adequately treated by their HCP or felt judged by their HCP (online supplemental table 2, quote 6, 11), and for a few patients by stigma, where they blamed themselves for their COPD: I guess you could say I’m a bit embarrassed to be stricken with this disease that I probably caused myself by smoking in my youth. (P18, USA) One patient also shared that their past experiences with healthcare formed a barrier to reaching out, ending up not treating their worsening: I just thought, see you go overnight because with it in the weekend as well. It would’ve been hard to speak to a doctor. You’d have to go through that 911. [Uh] And I’ve phoned them many at times when my husband was dying and you can't get through… So, that puts you off from ringing. (P08, UK) External facilitators Several facilitators were mentioned to make it possible for patients to reach out easily to an HCP. Most patients indicated that medical care in case of a worsening was easily available, whether this was their treating physician or a specialised COPD unit. There was often an option close by, and patients indicated that they mostly did not have to wait long for help (online supplemental table 3, quotes 1–4). they always make an appointment as soon as possible. […] that [um] always works pretty swiftly. That always works well with our GP. (P14, NLD) Patients from various countries indicated that they had a prescription for antibiotics and/or steroids available at home (ie, rescue-pack) or easily accessible, in case of an exacerbation (online supplemental table 3, quotes 5–7). These prescriptions make the barrier towards treating symptom worsening significantly smaller, since they can be used whenever the patient deems it necessary. Specialised COPD units also formed a strong facilitator when available, with many patients indicating that it has become substantially easier to get help as compared with before these units were created (online supplemental table 3, quotes 8). One anticipated external barrier was the potential costs induced when reaching out for healthcare. This was however not the case for any of our patients in the study, and many patients mentioned that all costs were covered by insurance (online supplemental table 3, quotes 9–10). Internal facilitators Patients generally indicated that they were happy with their HCP and that there was no hesitation to contact their HCP when they felt they needed to (online supplemental table 4, quotes 1–5). it’s great, I… I really trust her.[…]She is very well informed about my state so [uh] I find that she’s great at listening. (P05, CAN) Some patients discussed antibiotics and/or steroids treatment for exacerbations and had no issue starting or requesting treatment when they felt they needed to (online supplemental table 4, quote 6). A couple of patients indicated that they identified clear triggers or ‘crisis points’ for reaching out to healthcare, for example: That’s the key if my oxygen levels start to drop. […]If it goes down to 90, 91, I’ll start being concerned, and I would probably ring them. If it dropped below 90, then it would be immediately I would contact the virtual unit. (P06, UK) Discussion This international study aimed at (1) gaining more insight into the impact of an AWE on the daily lives of patients with COPD, and (2) identifying and explaining AWE-associated behaviour, including the underlying reasons for not consulting an HCP. Patients’ identification of bad days and worsenings showed large variability. Most of the 2 days AWEs were not noticed by patients, whereas longer (>2 days) AWE events were often reported. The majority of patients shared that they were contacting their HCP for healthcare when they felt that it was necessary. Difficulties in getting an appointment with a clinician at short notice was mentioned as an important external barrier, whereas a reactive stance to worsenings seemed to be a key internal barrier—both influencing the decision to consult an HCP. Overall, the impact of AWEs on daily living with COPD varied strongly, ranging from not being noticed to requiring emergency services and hospitalisation. We observed a distinct difference in AWE recognition based on event duration, with a strong contrast between minimum duration events (2 days) and longer events (>2 days). Minimum-duration-AWEs were not recognised by our patients, and did not substantially change behaviour of these patients. Longer AWEs were often recognised highlighted by clear instances of symptom worsening and an inability to perform normal day-to-day activities. In many cases, these longer AWEs were also acted on, with patients reaching out to their HCP, initiating treatment with antibiotics and/or steroids, or even contacting emergency services. The most influential factor for AWE associated behaviour was the recognition of the worsening. For patients that did not consult an HCP while presented with an AWE, barriers discussed by patients were mainly internal (eg, patients waiting for symptoms to recover without specific interventions or trying to manage them themselves). External barriers (eg, financial or relationship with HCP) rarely inhibited patients from reaching out when in need, and most patients would not hesitate to reach out when they felt the need to. The international nature of this study gave insight into similarities between patients’ perspective on AWEs across different countries. The recognition of worsenings and definition of ‘bad days’ did not differ across countries but could also be attributed to heterogeneity in patients with COPD. Although healthcare systems differ widely (eg, in terms of accessibility and/or distance between patient’s home and closest clinic, prescriptions of COPD rescue packs), in all countries, patients mentioned limited external barriers that inhibited them from reaching out to healthcare. Important key messages of this study were that the heterogeneity of patients with COPD led to different perspectives on bad days, and that the subjective experience and duration of symptoms is a driver for the recognition of AWEs. Not surprisingly, periods of worsening symptoms might be captured based on diaries in RCTs, but these events are often not reported to HCPs. The results from this study indicate that there is a need for personalised education on symptom burden in patients with COPD. A greater emphasis on patient’s understanding and recognition of early signs of flare ups should be given (by using tailored COPD management plans), as a timely and appropriate follow-up could prevent further decline in health status and reduce the risk of future events.11 12 Based on patient’s reported perceptions on living with COPD and their experiences with COPD-related healthcare—shortly after a trigger AWE and 6 weeks follow-up—the following aspects are important to understand the impact of AWEs on patients with COPD. Perception and one’s view of what constitutes a ‘bad day’ seemed to play a role in recognition of AWEs, where patients reported not having bad days, or recalling large events, requiring additional treatment and/or hospitalisations. This indicates that shorter fluctuations in symptoms are not considered as significant enough to report or are attributed to other factors like physical activity or bad weather. Physical activity was often mentioned as a cause of symptom worsening, also among these patients. Longer events were most impactful on patients with COPD, even if they did not escalate to the ‘typical’ moderate/severe exacerbations. Patients did not always feel the need to reach out, or there were some barriers in play that stopped them from reaching out, leading to impactful events that would go unnoticed when looking at exacerbations. The 2-day events seemed less impactful, as most of them were not recalled when asking about recent days and past bad days. The duration of the events might also explain this difference in impact, seen as it does not take long for the patients to start feeling better. In longer events, symptoms might get worse during the initial few days, whereas the peak within a 2-day event is quickly reached, after which symptoms start to improve again. This is in line with our most observed barrier, where patients tended to wait and see how symptoms progressed before taking action. When symptoms started to improve again the next day, that was often stated as enough reason to not reach out (table 4 quote 12). Most patients stated that they would not hesitate to reach out when they felt the need to. This observation is further influenced by the timing of our interviews, conducted on average 9 days after the trigger AWE. By this time, many shorter events would have resolved, reinforcing patients’ perception that immediate action was not required, further affecting their recognition of such AWEs. COPDCompEx was developed to allow for the design of clinical trials that have a shorter duration and/or require fewer participants, while reflecting the treatment effect on moderate or severe exacerbations in patients with COPD.13 Although the COPDCompEx endpoint, in large pooled datasets from various RCTs, highlighted clinically relevant metrics identifying COPD worsenings which mirror the treatment effect seen with exacerbations, it is important to assess the patient’s perspective of AWE to establish the value as an outcome measure in future trials. In our study, the duration of trigger AWEs varied from 2 to 50 days, with 56% of these events lasting only 2 days. Importantly, these 2 days AWEs were often not recognised by patients, whereas longer duration events (ranging between 5 and 50 days in our study) were often recognised by patients and seemed to have more impact on their daily living. Future research is needed to (1) investigate the nature (ie, driven by slope or threshold events) and frequency of AWEs in a broad population of patients with moderate-to-severe COPD, (2) validate the relevance of (frequent) 2 days AWEs on clinical outcomes and (3) to investigate whether 2-day AWEs accurately reflect treatment effect on moderate or severe COPD exacerbations. Additionally, future studies should investigate whether the nature and frequency of AWEs is linked to specific clinical phenotypes of COPD. A strength of this study is the international nature including five countries across the world, thereby enhancing the purposive sampling of a variety of patients. Furthermore, the variety of AWEs (short vs long) and number of AWEs observed during the study reflects a broad and comprehensive perception of the burden of AWEs on daily living with COPD. Patients were not informed that the trigger AWE was the reason for the first interview. This approach ensured that patients’ reports of their (daily) symptoms and healthcare-seeking behaviour were not influenced when speaking with the moderators. A limitation of this study is that, in some cases, the duration of events could have been influenced by non-adherence with the diary. Since the AWE calculations were dependent on patients completing the diary during the RCT, the duration of AWEs could have been cut short when patients did not complete their diary. When a severe worsening occurs, it is logical that patients’ first priority is not to fill out the diary two times a day. During the study treatment period, patients’ adherence to the daily eDiary was monitored, and if it dropped below 80%, study sites encouraged the participants to complete the eDiary. We acknowledge that it is possible that participants experienced a (short-duration) AWE that was not captured due to incomplete diary entries. The semi-structured interview format and grounded theory–inspired approach provided valuable insights and allowed the identification of key themes and conceptual patterns, but did not result in a complete theoretical framework explaining all relationships between AWEs and clinical characteristics. A clinical characteristic that might play a role in AWE recognition is COPD severity. Within our patients, there was a strong trend of patients who noticed AWEs but did not act on them having higher scores on the Clinical Assessment Test (CAT scores, range: 18–26). This could suggest a tendency among symptomatic patients to perceive ‘bad’ days as part of their everyday illness perception. In general, patients that did not notice any AWEs reported lower CAT scores (range: 7–13). It is worth emphasising that the AWEs are partly driven by symptoms (captured in the eDiaries). Considering the observed short duration of AWEs in our study, this finding could point to a relationship between the short duration of these worsenings and their lower impact on patient’s lives, as reflected by lower CAT scores. However, we should acknowledge that recall of symptom severity (regarding AWEs) may not accurately reflect the patient’s experience over time and there is considerable variation between patients. A cognitive bias might be introduced, where the highest and most recent intensity of a specific time period has the most influence on the recalled symptom level, as has been observed in patients presented with breathlessness. In a study by Jundberg et al,17 recalled breathlessness over a period of a 1-week period was most strongly related to reported peak breathlessness, rather than the last reported value. Recall for 1-day appeared to be closely linked to the mean breathlessness value for that day. In addition to symptom perception and recall, other factors may influence AWE recognition and response, including patients’ self-management strategies, health literacy and previous experiences with symptom worsening. These aspects were not explored in the current study but could affect how patients perceive, interpret and act on AWEs. The relationships between CAT score, illness perceptions at times of an AWE and variations in recall and self-management behaviour may therefore be interesting to explore in future research. A larger sample is needed to achieve sufficient power to examine these associations, which could, for example, be obtained by pooling data from RCTs collecting the variables needed for the AWE algorithm. In conclusion, the impact and patients’ perception of AWEs varies widely across individuals and strongly depends on the duration of these worsening events. AWEs that last for >2 days are generally recognised as bad days, and change patient behaviour (in terms of reduced physical activities and healthcare seeking behaviour). Overall, this study showed that patient’s recognition of early signs of worsenings varies widely in a population of patients with moderate-to-severe COPD. This is an important consideration in clinical practice while designing patient’ COPD action plans. Supplementary material