Summary
AERD symptom sequence varies and relates to onset age, diagnostic delay, and healthcare utilization burden.
Nasal polyps with NSAID sensitivity preceding asthma diagnosis may identify high‐risk patients requiring earlier specialist referral and biologic consideration.
Keywords: aspirin‐exacerbated respiratory disease (AERD), asthma, diagnostic delay, electronic medical records (EMR), nasal polyps, NSAID sensitivity, NSAID‐exacerbated respiratory disease, N‐ERD, Samters' triad
Abbreviations
- AERD
aspirin‐exacerbated respiratory disease
- ANP
sequence of the diagnoses: asthma, nsaid sensitivity, nasal polyps
- APN
sequence of the diagnoses: asthma, nasal polyps, NSAID sensitivity
- CRSwNP
chronic rhinosinusitis with nasal polyps
- EMR
electronic medical records
- FESS
functional endoscopic sinus surgeries
- IQR
interquartile range
- MHS
Maccabi Healthcare Services
- NAP
sequence of the diagnoses: NSAID sensitivity, asthma, nasal polyps
- NPA
sequence of the diagnoses: NSAID sensitivity, nasal polyps, asthma
- NSAID
non‐steroidal anti‐inflammatory drugs
- PAN
sequence of the diagnoses: nasal polyps, asthma, NSAID sensitivity
- PNA
sequence of the diagnoses: Nasal polyps, NSAID sensitivity, asthma
To the Editor,
Aspirin‐Exacerbated Respiratory Disease (AERD) is a chronic condition that impairs quality of life and may pose life‐threatening risks [1]. Patients typically suffer from severe asthma and chronic rhinosinusitis with nasal polyps (CRSwNP). Accurate and timely diagnosis is essential to ensure patient safety through education and safe medication use, and to enable the initiation of disease‐tailored therapies for uncontrolled asthma and recurrent nasal polyps. Yet, diagnosis delay remains a defining feature of AERD [1, 2]. The treatment of AERD is well‐established, safe, and effective. Consequently, the main challenge highlighted in the literature is the difficulty in diagnosing this disease [1, 2, 3].
Despite its clinical importance, the timing of symptom onset and the progression of AERD remain poorly characterized [4]. Clarifying these temporal dynamics is necessary, particularly given recent evidence that AERD follows a significantly more severe clinical course than aspirin‐tolerant asthma [5]. Understanding the progression and pace of disease development is crucial for facilitating earlier recognition and enabling timely, tailored treatment [1], especially since early biologic intervention, prior to irreversible epithelial damage, holds the potential to fundamentally alter the disease course [6].
Advances in digital health data offer new opportunities to improve the identification of AERD patients through the coexistence of three diagnostic codes: asthma, nasal polyps, and NSAID sensitivity [3, 4, 5, 7, 8]. Hence, the objective of this study is to investigate the diagnostic sequence and temporal intervals between AERD‐related conditions through analysis of EMR data.
Our cohort included 181 MHS patients diagnosed with asthma, nasal polyps, and NSAID sensitivity between 2002 and 2022, of whom 101 were females (56%) and 80 were males (44%). Diagnoses were identified based on ICD codes for asthma (493) and nasal polyps (471). Most patients had repeated spirometry testing (median 4 tests per patient), with similar patterns observed across all sequence groups (p = 0.889). This pattern suggests consistent diagnostic practices and ongoing clinical follow‐up across the cohort. NSAID hypersensitivity was based on clinician‐documented patient history, typically requiring at least two episodes. The study was approved by MHS Institutional Review Board.
Patients were classified into six groups based on the chronological order of symptom diagnosis. Diagnostic intervals, biomarkers, and healthcare utilization were analysed across groups (see Table 1). Consistent with previous studies, AERD was more prevalent in women [3, 4, 5, 7, 9], and symptom onset typically occurs after age 40. In half of the cohort, asthma was the initial symptom, while 26% presented first with nasal polyps and the remainder (24%) with NSAID sensitivity.
TABLE 1.
Demographic and clinical features.
| Characteristics | Statistical tests | All, N = 181 (100%) | Sequence of the diagnoses a (N = 181) | p | |||||
|---|---|---|---|---|---|---|---|---|---|
| 123, APN, N = 49 (27) | 132, ANP, N = 42 (23) | 213, PAN, N = 33 (18) | 231, PNA, N = 14 (8) | 312, NAP, N = 25 (14) | 321, NPA, N = 18 (10) | ||||
| Demographics | |||||||||
| Age of asthma onset, mean years (±SE) | One‐way ANOVA | 47 (±12.4) | 46 (±1.72) | 42 (±1.61) | 46 (±2.35) | 51 (±3.33) | 46 (±2.75) | 56 (±1.95) | 0.002 |
| Age of nasal polyp onset, mean years (±SE) | One‐way ANOVA | 49 (±12.2) | 49 (±1.65) | 52 (±1.75) | 43 (±2.18) | 43 (±3.51) | 51 (±2.59) | 52 (±2.40) | 0.008 |
| Age of NSAID sensitivity onset, mean years (±SE) | One‐way ANOVA | 48 (±12.4) | 53 (±1.66) | 46 (±1.61) | 51 (±2.34) | 46 (±3.39) | 43 (±2.59) | 46 (±2.78) | 0.006 |
| Diagnosis Interval (years), median (IQR), mean (±SE) | Kruskal–Wallis |
8 (4–12) 8 (±0.4) |
6 (4–9) 7 (±0.6) |
9 (5–15) 10 (±0.9) |
8 (4–10) 8 (±0.8) |
7 (4–12) 8 (±1.4) |
9 (5–11) 9 (±0.9) |
11 (8–13) 11 (±1.2) |
0.021 |
| Sex, female/male, N (%) | Chi‐square | 101 (56)/80 (44) | 31 (30.7)/18 (22.5) | 23 (22.8)/19 (23.8) | 12 (11.9)/21 (26.2) | 10 (9.9)/4 (5) | 14 (13.9)/11 (13.8) | 11 (10.9)/7 (8.8) | 0.169 |
| Asthma diagnosis by specialist, TRUE/FALSE, N (%) | Chi‐square | 124 (70)/55 (30) | 38 (31)/11 (20) | 27 (22)/13 (24) | 24 (19)/9 (16) | 7 (6)/7 (13) | 20 (16)/5 (9) | 8 (6)/10 (18) | 0.077 |
| Clinical features | |||||||||
| Number of physician visits, median (IQR), mean (±SE) | Kruskal‐Wallis |
390 (264–523) 416 (±16) |
413 (244–568) 436 (±33) |
320 (254–498) 395 (±31.2) |
354 (215–474) 372 (±32.7) |
365 (305–588) 483 (±80.3) |
433 (292–490) 422 (±34.8) |
417 (300–516) 436 (±55.2) |
0.695 |
| Total costs of physician visits, median (IQR), mean (±SE) | Kruskal–Wallis |
29,322 (20584–40,695) 32,438 (±1255) |
30,094 (19047–44,999) 33,369 (±2493) |
25,436 (19081–39,830) 30,858 (±2531) |
28,594 (17725–38,764) 29,047 (±2700) |
28,044 (23229–43,181) 37,629 (±6378) |
33,666 (23014–39,020) 33,093 (±2573) |
32,557 (23108–41,179) 34,864 (±4269) |
0.622 |
| Number of hospitalizations, median (IQR), mean (±SE) | Kruskal‐Wallis |
4 (2–7) 5.52 (±0.4) |
4 (3–8) 5.98 (±0.7) |
4 (2–6) 4.5 (±0.6) |
5 (2–9) 6.97 (±1.3) |
5 (2–5.25) 6.33 (±2.1) |
3 (2–6.25) 4.79 (±1.0) |
3 (2–6) 4 (±0.7) |
0.409 |
| Number of hospitalizations days, median (IQR), mean (±SE) | Kruskal‐Wallis |
6 (2–17) 17.83 (±2.9) |
8 (1.5–16) 17.0 (±4.1) |
5 (2–13) 16.1 (±7.4) |
6 (4–22.8) 17.0 (±4.0) |
10 (2–27.8) 41.5 (±24.7) |
3.5 (2–11.8) 15 (±5.8) |
11 (1–17) 11.2 (±3.0) |
0.661 |
| Number of ER visits, median (IQR), mean (±SE) | Kruskal–Wallis |
4 (2–8) 5.77 (±0.4) |
4 (3–8) 6.09 (±0.7) |
4 (3–7.5) 5.51 (±0.6) |
3 (1.75–7.5) 6.38 (±1.4) |
6 (1–8) 7.15 (±1.9) |
3 (2–6) 4.96 (±1.0) |
4 (2–6) 4.38 (±0.7) |
0.734 |
| Number MHS emergency department visits, median (IQR), mean (±SE) | Kruskal–Wallis |
4 (2–6) 4.54 (±0.4) |
3 (1.5–7) 4.93 (±0.9) |
4 (2.5–6) 4.78 (±0.6) |
3 (2–5) 4.18 (±1.1) |
6.5 (3.75–7) 5.62 (±0.7) |
3.5 (2–5) 4.36 (±0.9) |
2 (1–3) 2.29 (±0.6) |
0.150 |
| Number of spirometry procedures, median (IQR), mean (±SE) | Kruskal‐Wallis |
4 (1–10) 7.41 (±0.7) |
4 (3–11) 9.04 (±1.8) |
4.5 (0.25–10.5) 6.17 (±1.1) |
3 (0–10) 8.15 (±1.91) |
3 (1.25–9.75) 7.79 (±3.3) |
4 (2–9) 6.12 (±1.1) |
4 (1.25–8.25) 6.06 (±1.6) |
0.889 |
| Number of FESS procedures, median (IQR), mean (±SE) | Kruskal‐Wallis |
0 (0–1) 0.82 (±0.1) |
1 (0–2) 1 (±0.2) |
0 (0–1) 0.62 (±0.1) |
1 (0–2) 1.3 (±0.3) |
0 (0–0) 0.14 (±0.1) |
1 (0–1) 0.76 (±0.2) |
0.5 (0–1) 0.5 (±0.1) |
0.03 |
Note: Bold values indicate statistical significance (p < 0.05).
Sequence of the diagnoses: A, Asthma—1; P, Nasal polyps—2; N, NSAID sensitivity—3. APN = Asthma→Nasal polyps→NSAID sensitivity, etc.
The mean age at onset of each symptom was relatively similar. However, when examined separately by sequence group (see Table 1), significant differences in age at onset were observed for all three symptoms. For example, in the ANP group, mean asthma onset was 42 years, whereas in the NPA group it was 56 years. Similarly, the interval between first and last symptoms differed significantly: the overall mean was 8 years, ranging from 7 years (APN) to 11 years (NPA).
Characterizing these diagnostic intervals is of clinical importance, as it may help identify critical windows for earlier recognition and therapeutic intervention [5]. A shorter interval may reflect a more aggressive disease course, whereas a longer one may be associated with variations in clinical awareness, diagnostic practices, or access to specialized care [1, 9].
Differences in diagnostic practices may partly explain these temporal variations [2]. In most groups, asthma diagnosis was more frequently made by specialists. However, in the two groups in which asthma was the last symptom of the AERD triad to be diagnosed, fewer patients were referred to a specialist (NPA), or referral rates were similar (PNA). This observation suggests that earlier referral of patients with nasal polyp to a pulmonologist or allergy specialist may facilitate more timely recognition of asthma [7, 9].
Out of the six groups, two groups demonstrated a substantially delayed onset of NSAID sensitivity. In the APN group, mean age at NSAID sensitivity onset was 53 years, and in the PAN group, 51 years. Although the recorded timing of NSAID sensitivity in EMR data may not always reflect the true biological onset, this delay may reflect limited patient awareness of NSAID reactions [2]. Given the potential severity of such reactions, it may be advisable for health providers to explicitly inquire about NSAID use in patients with asthma and nasal polyposis, or to consider proactive sensitivity testing [2].
The six‐group classification enabled the identification of patterns that might have otherwise remained undetected, including both the age‐related differences in NSAID sensitivity and the distinct clinical trends observed in the PNA group. This group, the smallest (N = 14), showed a unique clinical profile. Although no statistically significant differences were observed in healthcare utilization, patients in the PNA group had the highest mean rates of physician visits, emergency department use, and hospitalizations, with large variability.
This trend is particularly interesting, as one might expect asthma, a condition that can be life‐threatening, to lead to greater healthcare utilization. In contrast, nasal polyps, which are not life‐threatening but substantially impair quality of life, appear to have a stronger impact on healthcare service use in this cohort. While the small sample size likely limited the ability to detect statistical power, these findings are nonetheless meaningful and may reflect a subgroup of patients with more severe or persistent disease courses.
The PNA group demonstrated the lowest mean number of FESS procedures, despite nasal polyps being the initial symptom in this group. In contrast, the PAN group, where nasal polyps also presented first, had the highest mean number of FESS procedures. This difference may reflect variations in disease management, referral patterns, or timing of intervention, and may partly explain the higher healthcare utilization observed in the PNA group. It is also possible that a delay in asthma diagnosis in the PNA group played a role, resulting in extended periods of uncontrolled inflammation and symptom progression [3]. Consequently, patients presenting with nasal polyps and NSAID sensitivity may represent a higher‐risk subgroup, in whom earlier evaluation for asthma and consideration of biologic therapies could be considered.
In conclusion, characterizing the diagnostic patterns and progression of AERD in real‐world settings may help reduce diagnostic delays. Patients presenting with nasal polyps and NSAID sensitivity prior to asthma may represent a distinct clinical subgroup. Recognition of such sequences may support earlier specialist referral and more timely consideration of targeted biologic therapy, with the potential to mitigate long‐term clinical and healthcare burden. While the six‐sequence classification provides a more detailed view of AERD progression, the relatively small subgroup sizes require cautious interpretation. Nevertheless, the sequence‐based analysis enabled identification of patterns that were not apparent when considering only the first manifestation of disease. Future studies may be conducted using broader, clinically‐oriented groupings based on the first manifestation of disease and on larger datasets. We hope this study may serve as a foundation for future EMR‐based investigations of diagnostic delay and subgroup patterns within heterogeneous AERD populations.
Author Contributions
Orli Weiser: investigation, conceptualization, study design, data collection and analysis, writing – original draft, editing and manuscript preparation. Guy Melamed: conceptualization, study design, data collection and analysis, manuscript preparation and review. Ronit Confino‐Cohen: conceptualization, manuscript preparation and review. Yossi Rosman: investigation, conceptualization, study design, data analysis, editing, manuscript preparation, review and supervision.
Funding
This work was supported by Maccabi Healthcare Services as the funding organization. KSM (Kahn‐Sagol‐Maccabi) Research and Innovative Center of Maccabi Healthcare Services, Retrospective Studies, 2023.
Ethics Statement
The research was approved by MHS Institutional Review Board (MHS‐0125‐23). Due to the retrospective design of the study and the use of de‐identified electronic medical records data, informed consent was not required.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgements
Declaration of Generative AI and AI‐Assisted Technologies in the Manuscript Preparation Process: During the preparation of this work the authors used ChatGPT in order to improve language clarity and readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy or ethical restrictions. The statistical analysis outputs generated in R and supporting the findings of this study have been deposited in a Zenodo repository and are available at: (10.5281/zenodo.20581336).
References
- 1. Haque R., White A. A., Jackson D. J., and Hopkins C., “Clinical Evaluation and Diagnosis of Aspirin‐Exacerbated Respiratory Disease,” Journal of Allergy and Clinical Immunology 148, no. 2 (2021): 283–291, 10.1016/j.jaci.2021.06.018. [DOI] [PubMed] [Google Scholar]
- 2. Buchheit K., Bensko J. C., Lewis E., Gakpo D., and Laidlaw T. M., “The Importance of Timely Diagnosis of Aspirin‐Exacerbated Respiratory Disease for Patient Health and Safety,” World Journal of Otorhinolaryngology—Head and Neck Surgery 6, no. 4 (2020): 203–206, 10.1016/j.wjorl.2020.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Kshirsagar R. S., Chou D. W., Wei J., and Liang J., “Aspirin‐Exacerbated Respiratory Disease: Longitudinal Assessment of a Large Cohort and Implications of Diagnostic Delay,” International Forum of Allergy & Rhinology 10, no. 4 (2020): 465–473, 10.1002/alr.22516. [DOI] [PubMed] [Google Scholar]
- 4. Dages K. N., Sofola‐James O., Sehanobish E., et al., “Sex, Ethnicity, Body Mass Index, and Environmental Exposures Associated With NSAID‐Exacerbated Respiratory Disease Symptom Sequence,” Journal of Allergy and Clinical Immunology. In Practice 11, no. 12 (2023): 3662–3669.e2, 10.1016/j.jaip.2023.07.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lee Y., Kim C., Lee E., et al., “Long‐Term Clinical Outcomes of Aspirin‐Exacerbated Respiratory Disease: Real‐World Data From an Adult Asthma Cohort,” Clinical and Experimental Allergy 53, no. 9 (2023): 941–950, 10.1111/cea.14362. [DOI] [PubMed] [Google Scholar]
- 6. Laidlaw T. M., “Aspirin‐Exacerbated Respiratory Disease in the Era of Biologics,” Annals of Allergy, Asthma & Immunology 135, no. 4 (2025): 366–373, 10.1016/j.anai.2025.06.001. [DOI] [PubMed] [Google Scholar]
- 7. Cahill K. N., Johns C. B., Cui J., et al., “Automated Identification of an Aspirin‐Exacerbated Respiratory Disease Cohort,” Journal of Allergy and Clinical Immunology 139, no. 3 (2017): 819–825.e6, 10.1016/j.jaci.2016.05.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Tao M., Roberts S., and Arnold M., “Algorithmic Identification of Patients With Aspirin‐Exacerbated Respiratory Disease Using an Electronic Health Record,” Otolaryngology—Neck Surgery 169, no. 2 (2023): 253–257, 10.1002/ohn.267. [DOI] [PubMed] [Google Scholar]
- 9. Peñalver M. J., Galleani C., Fernández C., and Crespo J. F., “Natural History of Nonsteroidal Anti‐Inflammatory Drug‐Exacerbated Respiratory Disease: A Retrospective, Single‐Centre Study,” Clinical and Experimental Allergy (2026), 10.1111/cea.70219. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy or ethical restrictions. The statistical analysis outputs generated in R and supporting the findings of this study have been deposited in a Zenodo repository and are available at: (10.5281/zenodo.20581336).