www.nature.com/npjpcrm ARTICLE OPEN Characteristics and treatment patterns of patients with asthma on multiple-inhaler triple therapy in Spain Miriam Barrecheguren1,2 ✉, Monica Monteagudo 3 , Marc Miravitlles 1,2 , Xavier Flor4,5, Alexa Núñez1,5, Jeisson Osorio 6 , Xavier Muñoz1,2 and Iñigo Ojanguren1,2,5 The aim of this observational, retrospective study was to describe characteristics, treatment patterns, and adherence among patients with asthma who initiated multiple-inhaler triple therapy (MITT) in Catalonia, Spain. This study used data of patients initiating MITT in 2016 from the SIDIAP (Information System for Research in Primary Care) database, which covers ~80% of the Catalonian population (5.8 million). Of 1,204 patients initiating MITT, 361 (30.0%) stepped down (discontinued ≥ 1 and continued ≥1 MITT component) and 89 (7.4%) stopped all three components of MITT for a period of 60 days during the following 12 months. In the follow-up period, 196 (16.3%) patients were considered adherent to MITT (>0.8 proportion of days covered [PDC]), with a mean (standard deviation) PDC of 0.52 (0.51) days. Given the low adherence and substantial rates of step down/ discontinuation among patients initiating MITT, there is an urgent need to implement strategies to improve treatment adherence/ persistence. npj Primary Care Respiratory Medicine (2022)32:11 ; https://doi.org/10.1038/s41533-022-00270-2 1234567890():,; INTRODUCTION to triple therapy (ICS/LABA/LAMA) may benefit patients on Asthma is a heterogeneous chronic inflammatory respiratory ICS/LABA with uncontrolled asthma. disease that is often both over- and underdiagnosed1–3. From The use of multiple inhalers as dual therapy in asthma is 1990 to 2015, the prevalence of asthma increased by 12.6% associated with lower adherence rates, lower medication persis- globally, with over 358 million people worldwide suffering from tence, increased asthma-related symptoms, and increased hospi- asthma in 20154. Prevalence rates vary significantly between talization compared with single-inhaler dual therapy18,19. countries or even between regions of the same country, with the Currently, there is limited real-world data regarding demo- prevalence rate in Spain estimated to be between 2.5 and 13.5% graphics, clinical characteristics, treatment patterns, discontinua- overall, and the prevalence in Barcelona reported to be 6.3–6.6%5–7. tion, and adherence of patients with asthma initiating MITT in Between 30 and 50% of patients with moderate/severe asthma Europe. Therefore, the aim of this study was to describe the have inadequately controlled disease despite attempts to characteristics of patients with asthma who initiate MITT, optimize adherence to inhaled corticosteroids in combination treatment patterns before and after MITT initiation, and treatment with a long-acting β2-agonist (ICS/LABA)8–12. In an observational, adherence to MITT in Catalonia, Spain. cross-sectional study conducted in Spain, the prevalence of uncontrolled severe persistent asthma among patients with asthma seen at pneumology and allergology hospital units based METHOD on a physician’s diagnosis was 3.9% (N = 36,649). Furthermore, the Study design level of asthma control was shown to be underestimated by This epidemiological study with retrospective analysis of long- physicians, which may reflect issues with adopting and imple- itudinal follow-up data used the SIDIAP database, which contains menting treatment guidelines in clinical practice13. The most anonymized electronic patient records for 5.8 million people common treatments recommended for both moderate/severe registered in 279 primary care centers of the Catalan Health asthma and uncontrolled severe persistent asthma are ICS/LABA Institute and covers approximately 80% of the population of alone or in combination with a short-acting β2-agonist (SABA) or a Catalonia20. The SIDIAP database contains information on patient leukotriene receptor antagonist13. The Global Initiative for Asthma demographics and medical records from 2006. Additionally, the (GINA) also recommends adding long-acting muscarinic antago- SIDIAP is linked to registers or databases that provide information nists (LAMA) as an additional controller for patients with regarding hospitalizations, medication prescribed in pharmacies uncontrolled asthma on at least medium-dose ICS/LABA therapy2. within the Catalan Health Institute hospitals (Pharmacy Register), The addition of a LAMA to ICS/LABA maintenance therapy as part and mortality (National Mortality Register). of a multiple-inhaler triple therapy (MITT) regimen14 or, more Patients initiating MITT were identified from January 1st to recently, as single-inhaler triple therapy (SITT) has been shown to December 31st, 2016. The index date was the date of initiation of improve lung function and reduce exacerbation rates in patients the third component of MITT, defined as overlapping prescriptions with uncontrolled asthma15–17. These data indicate that escalation for ICS, LABA, and LAMA (in two or three devices) for ≥60 days 1 Department of Pneumology, Hospital Universitari Vall d’Hebron, Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain. 2Ciber de Enfermedades Respiratorias (CIBERES), Barcelona, Spain. 3Institut Universitari d’Investigació en Atenció Primària (IDIAP) Jordi Gol, Barcelona, Spain. 4Centre d’Atenció Primària (CAP) Chafarinas, Institut Català de la Salut (ICS), Barcelona, Spain. 5Department of Medicine, Universidad Autónoma de Barcelona, Barcelona, Spain. 6Department of Pulmonary Medicine, Hospital Clínic-Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain. ✉email: mbarrecheguren@vhebron.net Published in partnership with Primary Care Respiratory Society UK M Barrecheguren et al. 2 Index date: day of first MITT episode experienced a 60-day gap in subsequent prescriptions for all three components (i.e., ICS, LABA, LAMA). The rationale of using a 60-day Pre-index Post-index gap to define discontinuation was based on requirements of the healthcare system in Catalonia where regular prescription of inhaled therapies by clinicians lasts for 2 months; after 2 months it 12 months 12 months is mandatory for the physician to renew or modify the prescription. The date of step down or discontinuation was 01 Jan 01 Jan 31 Dec 31 Dec defined as the last date of overlapping prescription of ICS, LABA, 2015 2016 2016 2017 and LAMA. Inclusion period Outcomes Demographics, comorbidities, and clinical characteristics including Study period lung function (FEV1 % predicted value and FEV1/FVC ratio), the percentage of eosinophils in peripheral blood, severe exacerba- Two devices tions, and treatments patterns were identified from the SIDIAP ICS/LABA LAMA database pre- and post-index. Severe exacerbations were defined Triple devices as the need for one hospitalization or emergency visit or the use Three devices of oral/systemic corticosteroids (or increase in the maintenance dose) for ≥3 days due to asthma, and were captured by the ICS LABA LAMA Catalan Health Institute hospitals. Additional outcomes assessed post-index included adherence to MITT, defined as the extent to which a patient acts in accordance with the prescribed interval Fig. 1 Study design. ICS inhaled corticosteroid, LABA long-acting and dose of a dosing regimen. In this study, treatment adherence β2-agonist, LAMA long-acting muscarinic antagonist, MITT multiple- was assessed in terms of PDC. 1234567890():,; inhaler triple therapy. Sample size and statistical analysis (Fig. 1). Data were collected for the 12 months prior to (pre-index) All patients in the SIDIAP registry (~5.8 million people) who met and 12 months following (post-index) the index date. the eligibility criteria were included in the study. Based on an estimated asthma prevalence of 3% in Spain, a sample size of Study population 1,117 patients was considered to be sufficient to address the main Patients included in this study were aged 18–75 years with at least study objective with a 95% CI. For patients who discontinued one protocol-defined ICD 10-CM (International Classification of MITT, assuming a 95% CI in a two-sided test, 195 patients in each Diseases, Tenth Revision, Clinical Modification21) asthma medical group (discontinued or not discontinued) were sufficient to find code, an overlapping prescription of at least 60 days for each of statistically significant differences with a minimum statistical the three components of MITT during the study inclusion period, power of 80%. and available medical records in the SIDIAP database pre- and Continuous variables are reported as a mean and SD and post-index. categorical variables are presented as number and percentage of Patients were excluded if they had a diagnosis of COPD patients. For spirometry data, paired analyses (when available) for (including emphysema and chronic bronchitis), cystic fibrosis, a small subgroup of patients are presented. heart failure, or lung cancer at any time in their medical records, or MITT discontinuation was calculated based on the number of had used MITT or LAMA monotherapy pre-index (to ensure patients who discontinued treatment in relation to the total exclusion of patients with COPD incorrectly classified as having number of patients in the 12 months pre-index date. asthma). Adherence to MITT was calculated according to the number of adherent patients in relation to the total number of patients in the 12 months post-index date. PDC was calculated during the Study endpoints 12-month post-index period by dividing the total number of days The primary objective of the study was to describe the that MITT was prescribed by the number of days post-index and demographics, comorbidities, clinical characteristics, and treat- ranged from 0 to 1. A patient was considered adherent when the ment patterns in the overall study population pre- and post-index. PDC value was >0.8. The secondary objectives were to describe discontinuation of A logistic regression model was developed to identify MITT post-index (including the number of patients discontinuing independent factors associated with adherence to and disconti- and the number of patients stepping down and subsequent nuation of MITT. The Hosmer–Lemeshow test22 was used and treatment patterns); describe patient characteristics according to independent variables were age, sex, BMI, smoking status, MITT persistence or discontinuation post-index; describe the comorbidities (pneumonia, sleep apnea syndrome, polyposis, characteristics and treatment patterns of patients who experi- allergic rhinitis, atopic dermatitis, conjunctivitis, gastroesophageal enced severe exacerbations pre-index; to describe treatment reflux, diabetes mellitus, hypertension, anxiety, depression), adherence in patients on MITT post-index, and to describe exacerbations, eosinophil count, chest X-ray, computerized characteristics and treatment patterns in patients who discon- tomography scan, allergy test, GP, nurse and pulmonologist visits, tinued MITT according to the timing of discontinuation (first pneumologist and allergy referrals, sick leave, sick leave due to a 90 days or Days 181–364). respiratory cause, influenza and pneumococcal vaccines, and Three different MITT treatment patterns were assessed post- previous treatment (ICS, ICS/LABA, ICS/LABA/anti-leukotrienes, index: persistence on MITT; step down from MITT; and disconti- ICS/LABA/OCS/anti-leukotrienes, ICS/LAMA, LAMA, LABA/LAMA, nuation of MITT. Persistence on MITT included patients who did LAMA/anti-leukotrienes, anti-leukotrienes, OCS, no treatment). not discontinue and did not step down; step down from MITT Adherence to MITT refers to the collection of over 80% of the included patients who discontinued at least one of the three MITT prescribed treatment from pharmacies, whereas discontinuation components but continued to be prescribed at least one refers to not having a prescription renewal by the doctor. Analyses component; and discontinuation of MITT included patients who were performed using the R (version 3.3.2) statistical package. For npj Primary Care Respiratory Medicine (2022) 11 Published in partnership with Primary Care Respiratory Society UK M Barrecheguren et al. 3 SIDIAP database N = 5,800,000 patients Exclusion: - Diagnosis of COPD - Cystic fibrosis, heart failure, or lung cancer at any time in their medical records - Use of MITT or LAMA monotherapy pre-index Patients initiating MITT in 2016 included in the study N = 1204 Patients persisting Patients stepping Patients with MITT down MITT discontinuing MITT N = 754 N = 361 N = 89 Patients with severe exacerbations initiating MITT in 2016 N = 667 Fig. 2 Flow chart of the patient selection from the SIDIAP database. COPD chronic obstructive pulmonary disease, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy, SIDIAP Information System for Research in Primary Care. statistical tests performed with the outcome variables, a patients who persisted (4.0% [3.2]) or stepped down (4.4% [3.0]) significance level of 0.05 was used, with no adjustments for from MITT (Supplementary Table 1). Prior to MITT initiation, the multiplicity. percentage of patients experiencing severe exacerbations was similar in patients who persisted, stepped down, or discontinued MITT (55.8% vs 55.1% and 52.8%, respectively) (Table 2). RESULTS Demographics, clinical characteristics, and treatment patterns Characteristics and treatment patterns pre- and post-index were generally similar across patients that stopped MITT in the Details of patient selection from the Information System for Research first 90 days (Supplementary Table 2) and those who stopped in Primary Care (SIDIAP) database are summarized in Fig. 2. Of 1,204 between Days 181–364 (Supplementary Table 3). patients included in this study, 851 (70.7%) were female and the mean (standard deviation [SD]) age at MITT initiation was 54.8 (13.8) Characteristics and treatment patterns of patients who years (Table 1). Body mass index (BMI), smoking status, and experienced severe exacerbations during the 12 months prior percentage of eosinophils measured in peripheral blood were to MITT initiation comparable pre- and post-index (Table 1). Mean (SD) forced In the pre-index period, 667 (55.4%) patients experienced at least one expiratory volume in 1 s (FEV1) % predicted was 70.8 (19.1) pre- severe exacerbation compared with 503 (41.8%) patients post-index index and 75.4 (19.6) post-index (Table 1). Among the 1,204 patients (Table 1). Among the patients who experienced severe exacerbations included, 881 (73.2%) had previously been receiving ICS/LABA and prior to initiation of MITT, 491 (73.6%) were female with a mean (SD) 162 (13.5%) were not receiving any treatment (Fig. 3). age at asthma diagnosis of 45.4 (15.8) years and a mean (SD) BMI of 30.4 (6.4) kg/m2 (Table 3). Additionally, post-index, 352 (52.8%) Patients stepping down and discontinuing MITT and patients in this subgroup experienced a severe exacerbation. There subsequent treatments was no change in the percentage of eosinophils measured in Among the patients that initiated MITT, 754 (62.6%) persisted with peripheral blood prior to and following MITT initiation (Table 3). Prior MITT, 361 (30.0%) stepped down (discontinued ≥ 1 and continued to MITT, 494 (74.1%) patients were previously treated with ICS/LABA ≥1 MITT component) and 89 (7.4%) discontinued MITT (stopped all and 80 (12.0%) patients received no treatment (Fig. 5). Post-index, three components of MITT for a period of 60 days) during the 169 (25.3%) patients received ICS/LABA and 29 (4.3%) patients did following 12 months. During the pre-index period, 589 (78.1%) not receive any treatment (Fig. 5). patients who persisted with MITT, 257 (71.2%) who stepped down and 35 (39.3%) who discontinued MITT were on ICS/LABA (Fig. 4a). Adherence to MITT: Independent factors associated with During the post-index period, among the patients that stepped adherence and discontinuation, and patient characteristics down, 305 (84.5%) were on ICS/LABA and, of the patients that Adherence to MITT was assessed using proportion of days covered discontinued MITT completely, 61 (68.5%) were not treated and 18 (PDC), which was calculated as follows: PDC = total days all drugs (20.2%) were receiving treatment with SABA or SAMA (Fig. 4b). available/days post-index. A patient was considered adherent when There were no clinically relevant differences between patients PDC was >0.8. that stepped down or discontinued MITT compared with those In the 12 months post-index, 196 (16.3%) patients were who persisted in terms of age, BMI, or smoking status pre-index adherent to MITT and the mean (SD) PDC was 0.52 (0.51) days. (Table 2) and post-index (Supplementary Table 1). Patients who The logistic regression model showed that greater adherence was persisted with MITT had similar FEV1 % predicted value and mean associated with having anxiety and depression, taking sick leave FEV1/forced vital capacity (FVC) ratio pre-index (Table 2) and post- due to a respiratory cause, and no previous ICS/LABA treatment index (Supplementary Table 1) compared with patients who did prior to initiation of MITT (Table 4). Although older age at not maintain MITT. Additionally, in patients who discontinued MITT initiation showed a significant association with greater treatment, the mean (SD) percentage of eosinophils measured in adherence, the odds ratio was only marginally above 1.0 (1.047; peripheral blood post-index was lower (3.2% [2.0]) compared with 95% confidence interval [CI]: 1.025–1.070) (Table 4). Published in partnership with Primary Care Respiratory Society UK npj Primary Care Respiratory Medicine (2022) 11 M Barrecheguren et al. 4 DISCUSSION Table 1.Demographics and clinical characteristics 12 months prior to and 12 months following MITT initiation. This retrospective observational study analyzed characteristics and treatment patterns of 1,204 patients with asthma initiating MITT in 12 months 12 months Catalonia, Spain. Overall, while persistence rates were relatively prior to MITT following MITT high, the rates of step down and discontinuation of MITT were (pre-index) (post-index) also high, and adherence was poor. Most of the patients who (N = 1204) (N = 1204) initiated MITT during the study period were on ICS/LABA prior to Demographics index date, while 13% were untreated. In the 12 months after initiating MITT, 63% of patients persisted on MITT, 30% stepped Female, n (%) 851 (70.7) down and 7% discontinued MITT. The proportion of adherent Age at asthma diagnosis, 45.2 (15.9) patients (>0.8 PDC) was only 16%; being older, having anxiety mean (SD) and/or depression, taking sick leave due to respiratory symptoms, Age at MITT initiation, 54.8 (13.8) and no previous treatment with ICS/LABA prior to initiation of mean (SD) MITT were all associated with greater adherence to treatment. Years from diagnosis to MITT, 9.6 (8.9) In our study, the unexpected finding that 14% of patients were mean (SD) not receiving any treatment prior to initiation of MITT could Urban location, n (%) 1,023 (85.0) perhaps be attributed to a late diagnosis of severe asthma or a Smoking status, n (%) failure of physicians to adhere to guideline recommendations Non-smokera 708 (58.8) 717 (59.6) leading to potential overtreatment. Additionally, the high rates of treatment discontinuation/stepping down might be associated Smoker 208 (17.3) 199 (16.5) with both physician behavior (e.g., justified discontinuation/step- Former smoker 247 (20.5) 272 (22.6) down due to lack of symptom control) and patient behavior (e.g., Unknown 41 (3.4) 16 (1.3) missed GP appointments, having a prescription but not filling it, BMIb n = 640 n = 536 and difficulties using multiple inhalers). BMI, mean (SD) kg/m2 30.3 (6.2) 30.9 (6.3) Suzuki et al published a similar observational cohort study in patients with asthma and asthma/chronic obstructive pulmonary Spirometryc, mean (SD) n = 387 n = 281 disease (COPD) overlap who initiated triple therapy in Japan23. FEV1 % predicted value 70.8 (19.1) 75.4 (19.6) Persistence rates among patients with asthma were lower than FEV1/FVC ratio 70.1 (13.1) 73.6 (13.3) those reported here (38.5% over 12 months)23. Furthermore, in our Common comorbidities, n (%) study, most patients (73%) initiating MITT were on ICS/LABA prior Anxiety/depression 459 (38.1) to the index date, compared with nearly 100% of patients in the Pneumonia 57 (4.7) observational study in Japan23. However, the sample size was much smaller. Studies on COPD using data from the SIDIAP Other respiratory infections 373 (31) database used in this study showed that 16.5% of patients Rhinitis 232 (19.3) initiated triple therapy 1 year after diagnosis24, and reported Hypertension 409 (34.0) similar persistence rates with 62.5% of patients continuing to Diabetes mellitus 139 (11.5) receive MITT 12 months following MITT initiation 25. Gastroesophageal reflux 112 (9.3) The benefits of initiating triple therapy on lung function, symptoms, and asthma control in patients with uncontrolled Conjunctivitis 12 (10.0) asthma have been demonstrated in multiple studies14–17. This Polyposis 76 (6.3) study was not designed to determine any change in clinical Severe exacerbationsd; n (%) 667 (55.4) 503 (41.8) outcomes before and after MITT initiation; however, a reduction in Blood eosinophils (%), n = 732 n = 717 the number of patients experiencing a severe exacerbation was mean (SD) 4.1 (3.4) 4.1 (3.1) seen in the 12 months following MITT initiation compared with the previous 12 months (55.4 and 41.8% of patients, respectively). BMI body mass index, FEV1 forced expiratory volume in 1 s, FVC forced vital capacity, SD standard deviation, MITT multiple-inhaler triple therapy. A reduction was observed in the aforementioned Japanese study a The number of non-smokers is higher following MITT initiation as the (64.0% vs 45.8%)23. Interestingly, in our study, a greater reduction smoking status for some patients only became available during that was observed in the subgroup of patients that had a severe period; exacerbation in the 12 months prior to MITT, with MITT reducing b reported for patients with a BMI record in their medical history; the number of patients experiencing exacerbations by half. c reported for patients with spirometry measurements in their medical Triple therapy with ICS/LABA plus tiotropium has been shown history; to improve lung function and reduce the risk of exacerbations in d a severe exacerbation was defined as the need for one hospitalization or patients with asthma26; however, the addition of tiotropium emergency visit or the use of oral/systemic corticosteroids (or increase in the maintenance dose) for at least 3 days due to asthma. requires the use of multiple inhalers with different dosing regimens14. More recently, the TRIMARAN, TRIGGER, CAPTAIN, and IRIDIUM studies15–17 evaluated the efficacy and safety of various triple ICS/LABA/LAMA combinations administered via The proportion of patients who were current smokers and those single inhalers. These studies demonstrated favorable effects on experiencing a severe exacerbation was higher in adherent asthma clinical and patient-reported outcomes with no new safety patients compared with non-adherent patients, whereas the concerns15–17. Observed differences between these studies (e.g., percentage of eosinophils in peripheral blood was slightly higher on asthma exacerbations) were likely due to differences in study in non-adherent versus adherent patients (Supplementary Table populations and outcome definitions used 27. 4). Being male, not receiving a previous allergy test, or no previous Adherence to therapy is complex and it is influenced by treatment prior to initiation of MITT were all associated with interdependent factors related to the patient, the healthcare discontinuation of MITT (Supplementary Table 5). Anxiety/depres- provider, and the healthcare system. In Spain, low adherence rates sion was marginally but non-significantly (p = 0.052) associated to triple therapy in COPD have been associated with high costs to with discontinuation of MITT (OR [95% CI]: 1.562 [0.996–2.448]). the Spanish National Health System28. In this study, adherence to npj Primary Care Respiratory Medicine (2022) 11 Published in partnership with Primary Care Respiratory Society UK M Barrecheguren et al. 5 100 80 73.2 Proportion of patients (%) 62.6 60 40 25.3 20 13.5 9.5 5.1 3.8 5.5 1.5 n = 754 n = 881 n = 305 n = 162 n = 61 n = 46 n = 18 n = 115 n = 66 0 Continued with MITT ICS/LABA* No treatment SABA or SAMA Other† 12 months prior to MITT (N = 1204) 12 months following MITT (N = 1204) Fig. 3 Treatment patterns 12 months prior to (pre-index) and 12 months following (post-index) MITT initiation in the overall population. Percentages may not total 100% due to rounding. *ICS/LABA category includes: ICS/LABA, ICS/LABA/anti-leukotrienes, ICS/LABA/OCS, ICS/ LABA/OCS/SAMA, ICS/LABA/anti-leukotrienes/OCS; †other category includes: ICS, ICS/LAMA, ICS/anti-leukotrienes, ICS/OCS, ICS/OCS/LAMA, LABA/LAMA, LABA/anti-leukotrienes, LAMA, LAMA/anti-leukotrienes, anti-leukotrienes, OCS, OCS/LAMA, OCS/LAMA/anti-leukotrienes. ICS inhaled corticosteroid, LABA long-acting β2-agonist, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy, OCS oral corticosteroid, SABA short-acting β2-agonist, SAMA short-acting muscarinic antagonist. a 12 months prior MITT 100 78.1 80 71.2 Proportion of patients (%) 60 43.8 39.3 40 20 13.3 11.7 9.9 11.1 8.4 3.6 3.9 5.6 n = 589 n = 257 n = 35 n = 75 n = 48 n = 39 n = 27 n = 14 n=5 n = 63 n = 42 n = 10 0 ICS/LABA* No treatment SABA or SAMA Other† Patients persisting (N = 754) Patients stepping down (N = 361) Patients discontinuing (N = 89) b 12 months following MITT 100 84.5 80 68.5 Proportion of patients (%) 60 40 20.2 20 15.5 11.2 n = 305 0 0 n = 61 0 n = 18 n = 56 n = 10 0 ICS/LABA* No treatment SABA or SAMA Other† Patients stepping down (N = 361) Patients discontinuing (N = 89) Fig. 4 Treatment patterns in patients that persist, step-down and discontinue MITT. a 12 months prior to (pre-index) and b 12 months following (post-index) MITT initiation. Percentages may not total 100% due to rounding. *ICS/LABA category includes: ICS/LABA, ICS/LABA/ anti-leukotrienes, ICS/LABA/OCS, ICS/LABA/OCS/SAMA, ICS/LABA/anti-leukotrienes/OCS; †other category includes: ICS, ICS/LAMA, ICS/anti- leukotrienes, ICS/OCS, ICS/OCS/LAMA, LABA/LAMA, LABA/anti-leukotrienes, LAMA, LAMA/anti-leukotrienes, anti-leukotrienes, OCS, OCS/ LAMA, OCS/LAMA/anti-leukotrienes. ICS inhaled corticosteroid, LABA long-acting β2-agonist, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy, OCS oral corticosteroid, SABA short-acting β2-agonist, SAMA short-acting muscarinic antagonist. Published in partnership with Primary Care Respiratory Society UK npj Primary Care Respiratory Medicine (2022) 11 M Barrecheguren et al. 6 Table 2. Demographics and clinical characteristics 12 months prior to Table 3. Characteristics of patients with severe exacerbations in the MITT initiation by MITT persistence. 12 months prior to MITT initiation and 12 months following MITT initiation. MITT MITT step MITT persistence down discontinuation 12 months prior 12 months (N = 754) (N = 361) (N = 89) to MITT (pre following MITT index) (post index) Demographics (N = 667) (N = 667) Female, n (%) 513 (68.0) 273 (75.6) 65 (73.0) Demographics Age at asthma diagnosis, 45.13 (15.6) 45.4 (16.3) 45.37 (16.2) mean (SD) Female, n (%) 491 (73.6) Age at MITT initiation, 55.30 (13.2) 54.0 (14.5) 54.25 (15.2) Age at asthma diagnosis, 45.4 (15.8) mean (SD) mean (SD) Years from diagnosis to 10.17 (9.2) 8.6 (8.2) 8.88 (8.5) Age at MITT initiation, 55.0 (13.8) MITT, mean (SD) mean (SD) Urban location, n (%) 642 (85.1) 309 (85.6) 72 (80.9) Years from diagnosis to MITT, 9.6 (8.5) mean (SD) Smoking status, n (%) Urban location, n (%) 573 (85.9) Non-smoker 439 (58.2) 218 (60.4) 51 (57.3) Smoking status, n (%) Smoker 134 (17.8) 52 (14.4) 22 (24.7) Non-smokera 412 (61.8) 414 (62.1) Former smoker 153 (20.3) 79 (21.9) 15 (16.9) Smoker 109 (16.3) 102 (15.3) Unknown 28 (3.7) 12 (3.3) 1 (1.1) Former smoker 131 (19.6) 146 (21.9) BMIa n = 388 n = 213 n = 39 Unknown 15 (2.2) 5 (0.7) BMI, mean (SD) kg/m2 30.3 (6.1) 30.2 (6.3) 30.0 (6.4) BMIb n = 376 n = 302 Spirometryb mean (SD) n = 244 n = 124 n = 19 BMI, mean (SD) kg/m2 30.4 (6.4) 30.8 (6.4) FEV1 % predicted value 69.4 (18.9) 72.9 (19.0) 74.2 (21.2) Spirometryc, mean (SD) n = 215 n = 135 FEV1/FVC ratio 69.1 (13.1) 71.5 (12.6) 74.5 (14.7) FEV1 % predicted value 70.6 (19.1) 74.7 (20.1) Common comorbidities, n (%) FEV1/FVC ratio 69.6 (13.6) 73.8 (13.5) Anxiety/depression 281 (37.3) 150 (41.6) 28 (31.5) Common comorbidities, n (%) Pneumonia 42 (5.6) 11 (3.0) 4 (4.5) Anxiety/depression 272 (40.8) Other respiratory 225 (29.8) 118 (32.7) 30 (33.7) infections Pneumonia 32 (4.8) Rhinitis 142 (18.8) 74 (20.5) 16 (18.0) Other respiratory infections 233 (34.9) Hypertension 254 (33.7) 125 (34.6) 30 (33.7) Rhinitis 131 (19.6) Diabetes mellitus 90 (11.9) 41 (11.4) 8 (9.0) Hypertension 243 (36.4) Gastroesophageal reflux 71 (9.4) 36 (10.0) 5 (5.6) Diabetes mellitus 79 (11.8) Conjunctivitis 67 (8.9) 44 (12.2) 10 (11.2) Gastroesophageal reflux 72 (10.8) Polyposis 53 (7.0) 21 (5.8) 2 (2.2) Conjunctivitis 77 (11.5) Severe exacerbationsc, 421 (55.8) 199 (55.1) 47 (52.8) Polyposis 51 (7.6) n (%) Severe exacerbationsd, n (%) 667 (100.0) 352 (52.8) Blood eosinophils (%), n = 453 n = 232 n = 47 3.0 Blood eosinophils (%), n = 425 n = 417 mean (SD) 4.0 (3.5) 4.4 (3.4) (1.9) mean (SD) 4.1 (3.6) 4.1 (3.3) BMI body mass index, FEV1 forced expiratory volume in 1 s, FVC forced vital BMI body mass index, FEV1 forced expiratory volume in 1 s, FVC forced vital capacity, SD standard deviation, MITT multiple-inhaler triple therapy. capacity, SD standard deviation, MITT multiple-inhaler triple therapy. a a Reported for patients with a BMI record in their medical history; The number of non-smokers is higher following MITT initiation as for b reported for patients with spirometry measurements in their medical some patients, the smoking status only became available during that history; period; c b a severe exacerbation was defined as the need for one hospitalization or reported for patients with a BMI record in their medical history; emergency visit or the use of oral/systemic corticosteroids (or increase in c reported for patients with spirometry measurements in their medical the maintenance dose) for ≥3 days due to asthma. history; d a severe exacerbation was defined as the need for one hospitalization or emergency visit or the use of oral/systemic corticosteroids (or increase in the maintenance dose) for ≥3 days due to asthma. triple therapy in the 12 months following MITT initiation was 16.3% and the mean (SD) PDC was 0.52 (0.51) days. A similar low adherence rate was reported in the Japanese study discussed previously with a PDC of 0.51 in the asthma-only group as well as patient understanding/awareness, limited supervision, or follow- in the asthma-COPD overlap group23. It has been reported that up with a healthcare professional, lack of awareness of disease persistence and adherence rates are lower when using multiple severity, and complexities for patients and for caregivers in using inhalers versus single-inhaler dual therapy in both asthma and multiple inhalers with different dosing regimens (e.g., once vs COPD18,28,29. While the reasons for poor adherence in this study twice daily)30–33. Additionally, an observational study conducted in are unknown, possible explanations include the high cost of Australia showed that over 80% of patients stepped-up to GINA asthma medications and requirements for copayments (even in Step 5 therapies while showing poor adherence to ICS/LABA34. publicly funded healthcare systems such as in Spain), lack of Based on this, further analyses to determine the PDC before npj Primary Care Respiratory Medicine (2022) 11 Published in partnership with Primary Care Respiratory Society UK M Barrecheguren et al. 7 100 80 74.1 Proportion of patients (%) 63.1 60 40 25.3 20 12.0 9.6 4.3 4.0 1.6 5.2 n = 421 n = 494 n = 169 n = 80 n = 29 n = 27 n = 11 n = 66 n = 37 0 Continued with MITT ICS/LABA* No treatment SABA or SAMA Other† 12 months prior to MITT (N = 667) 12 months following MITT (N = 667) Fig. 5 Treatment patterns in the 12 months prior to (pre-index) and 12 months following (post-index) MITT initiation among patients with severe exacerbations. Percentages may not total 100% due to rounding. *ICS/LABA category includes: ICS/LABA, ICS/LABA/anti- leukotrienes, ICS/LABA/OCS, ICS/LABA/OCS/SAMA, ICS/LABA/anti-leukotrienes/OCS; †other category includes: ICS, ICS/LAMA, ICS/anti- leukotrienes, ICS/OCS, ICS/OCS/LAMA, LABA/LAMA, LABA/anti-leukotrienes, LAMA, LAMA/anti-leukotrienes, anti-leukotrienes, OCS, OCS/ LAMA, OCS/LAMA/anti-leukotrienes. ICS inhaled corticosteroid, LABA long-acting β2-agonist, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy, OCS oral corticosteroid, SABA short-acting β2-agonist, SAMA short acting muscarinic antagonist. Table 4. Independent factors associated with adherence to MITT determined by the logistic regression model. Variable and reference arm Estimated adjusted OR of adherence to MITTa 95% CI p-value in the model Age at MITT initiation: older vs younger 1.047 1.025–1.070 <0.001 Having anxiety and depression: yes vs no 1.817 1.045–3.159 0.034 Previous sick leave of respiratory cause: yes vs no 2.616 1.235–5.541 0.012 Previous ICS/LABA use: yes vs no 0.528 0.308–0.904 0.020 Adherence to MITT defined as PDC > 0.8. BMI body mass index, CI confidence interval, GP general practitioner, ICS inhaled corticosteroids, LABA long-acting β2-agonist, LAMA long-acting muscarinic antagonist, MITT multiple-inhaler triple therapy, PDC proportion of days covered, OCS oral corticosteroids, OR odds ratio. a Higher OR is associated with a greater likelihood of being adherent to MITT. Independent variables were age, sex, BMI, smoking status, comorbidities (pneumonia, sleep apnea syndrome, polyposis, allergic rhinitis, atopic dermatitis, conjunctivitis, gastroesophageal reflux, diabetes mellitus, hypertension, anxiety, depression), exacerbations, eosinophil count, chest X-ray, computerized tomography scan, allergy test, GP, nurse and pulmonologist visits, pneumologist and allergy referrals, sick leave, sick leave due to a respiratory cause, influenza and pneumococcal vaccines, and previous treatment (ICS, ICS/ LABA, ICS/LABA/anti-leukotrienes, ICS/LABA/OCS/anti-leukotrienes, ICS/LAMA, LAMA, LABA/LAMA, LAMA/anti-leukotrienes, anti-leukotrienes, OCS, no treatment). initiation of MITT might give important insights into the low disease. Finally, spirometry and blood eosinophil data were limited adherence rates reported here. Notably, although adherence to due to the small number of patients with available data. MITT was low in this study, most patients (62.6%) persisted on Overall, results from this study in patients with asthma initiating MITT, which may indicate that patients experienced benefits on MITT showed that a significant proportion of patients step down, triple therapy. Nonetheless, the poor adherence rates we observed discontinue MITT, and are poorly adherent during the first highlight a need for improvement. Offering SITT could benefit 12 months of treatment. The poor adherence and substantial these patients and improve adherence. Interestingly, in this study rates of step down/discontinuation we observed in this study, greater adherence was associated with having anxiety and/or highlights an urgent need to implement strategies to improve depression, taking sick leave due to a respiratory cause, and no treatment adherence. Addressing factors such as accuracy of previous ICS/LABA treatment. Studies have shown that treatment diagnosis, inhaler technique, complexity of dosing regimens, and adherence tends to be higher among patients with uncontrolled regularly checking treatment adherence in patients currently on and severe asthma35,36. The fact that taking sick leave was ICS/LABA prior to initiating MITT may all be beneficial. associated with higher adherence may indicate that patients in this study were perhaps more symptomatic or perceived their asthma as severe. Reporting summary Strengths of our study include the large sample size, good Further information on research design is available in the Nature representation of the population, and lack of a recall bias as this is Research Reporting Summary linked to this article. a database study. However, as this was an observational retrospective study, it has inherent limitations including under- reporting of exacerbations, miscoding biases, and invalid/ DATA AVAILABILITY unknown data. To mitigate the impact of invalid/unknown data, The data that support the findings of this study belong to the Catalan Healthcare. we have used a range of diagnostic codes that delimited the study System and are not publicly available. Anonymised individual participant data and Published in partnership with Primary Care Respiratory Society UK npj Primary Care Respiratory Medicine (2022) 11 M Barrecheguren et al. 8 study documents can be requested for further research from www. 25. Monteagudo, M. et al. Treatment pathways before and after triple therapy in clinicalstudydatarequest.com. COPD: A population-based study in primary care in Spain. Arch. Bronconeumol. 57, 205–213 (2021). 26. Singh, D. et al. Determinants of response to inhaled extrafine triple Received: 30 June 2021; Accepted: 22 December 2021; therapy in asthma: Analyses of TRIMARAN and TRIGGER. Respir. Res. 21, 285 (2020). 27. Ojanguren, I. & Pilia, M. F. Efficacy and safety of once-daily single-inhaler triple therapy in patients with inadequately controlled asthma: The CAPTAIN trial. Breathe 17, 200279 (2021). REFERENCES 28. Miravitlles, M. et al. Estimation of the clinical and economic impact of an 1. Kavanagh, J., Jackson, D. J. & Kent, B. D. Over- and under-diagnosis in asthma. improvement in adherence based on the use of once-daily single-inhaler triple Breathe 15, e20–e27 (2019). therapy in patients with COPD. Int. J. Chron. Obstruct. Pulmon. Dis. 15, 1643–1654 2. GINA. Global Strategy for Asthma Management and Prevention, www.ginasthma. (2020). org (2021). 29. Stoloff, S. W., Stempel, D. A., Meyer, J., Stanford, R. H. & Carranza Rosenzweig, J. R. 3. Aaron, S. D. et al. Reevaluation of diagnosis in adults with physician-diagnosed Improved refill persistence with fluticasone propionate and salmeterol in a single asthma. JAMA 317, 269–279 (2017). inhaler compared with other controller therapies. J. Allergy Clin. Immunol. 113, 4. GBD 2015 Chronic Respiratory Disease Collaborators. Global, regional, and 245–251 (2004). national deaths, prevalence, disability-adjusted life years, and years lived with 30. Rodriguez-Garcia, C., Lourido-Cebreiro, T. & Gonzalez-Barcala, F. J. The ATAUD disability for chronic obstructive pulmonary disease and asthma, 1990–2015: A Study: The need to improve adherence. Arch. Bronconeumol. 55, 509–510 (2019). systematic analysis for the Global Burden of Disease Study 2015. Lancet Respir. 31. Laba, T. L. et al. Cost-related underuse of medicines for asthma-opportunities for Med. 5, 691–706 (2017). improving adherence. J. Allergy Clin. Immunol. Pr. 7, 2298–2306 e2212 (2019). 5. 5.0., G. Guía Española para el Manejo del Asma (2020), https://www.gemasma.com 32. Melero Moreno, C., Quirce, S., Huerta, A., Uria, E. & Cuesta, M. Economic impact of (2020). severe asthma in Spain: multicentre observational longitudinal study. J. Asthma 6. Morell, F. et al. La población de asmáticos ambulatorios y su control tras adaptar 56, 861–871 (2019). el tratamiento a las recomendaciones internacionales (ASMACAP I). Arch. Bron- 33. Martinez-Moragon, E. et al. Economic cost of treating the patient with asthma in coneumol. 43, 29–35 (2007). Spain: The AsmaCost study. Arch. Bronconeumol. 45, 481–486 (2009). 7. To, T. et al. Global asthma prevalence in adults: Findings from the cross-sectional 34. van Boven, J. F. M. et al. Trajectory analyses of adherence patterns in a real-life world health survey. BMC Public Health 12, 204 (2012). moderate to severe asthma population. J. Allergy Clin. Immunol. Pr. 8, 1961–1969 8. Bateman, E. D. et al. Global Initiative for Asthma 2016-derived asthma control e1966 (2020). with fluticasone propionate and salmeterol: A Gaining Optimal Asthma Control 35. Gon, Y., Ohyanagi, N. & Kobayashi, A. The association between control level and (GOAL) study reanalysis. Ann. Allergy Asthma Immunol. 123, 57–63 e52 (2019). self-reported treatment adherence across different treatment types in Japanese 9. Bernstein, D. I. et al. Fluticasone furoate (FF)/vilanterol (100/25 mcg or 200/25 asthma patients. Respir. Investig. https://doi.org/10.1016/j.resinv.2021.02.003 mcg) or FF (100 mcg) in persistent asthma. J. Asthma 52, 1073–1083 (2015). (2021). 10. Lee, L. K., Obi, E., Paknis, B., Kavati, A. & Chipps, B. Asthma control and disease 36. Souza-Machado, A., Santos, P. M. & Cruz, A. A. Adherence to treatment in severe burden in patients with asthma and allergic comorbidities. J. Asthma 55, 208–219 asthma: Predicting factors in a program for asthma control in Brazil. World Allergy (2018). Organ J. 3, 48–52 (2010). 11. Sulaiman, I. et al. A randomised clinical trial of feedback on inhaler adherence and technique in patients with severe uncontrolled asthma. Eur. Respir. J. https:// doi.org/10.1183/13993003.01126-2017 (2018). ACKNOWLEDGEMENTS 12. David, J., Trudo, F., Sidall, J. & Small, M. Burden of asthma among patients This study was funded by GlaxoSmithKline (GSK; HO-18-18550). Editorial support (in adherent to ICS/LABA: A real-world study. J. Asthma 56, 332–340 (2019). the form of writing assistance, including preparation of the draft manuscript under 13. Quirce, S., Plaza, V., Picado, C., Vennera, M. & Casafont, J. Prevalence of uncon- the direction and guidance of the authors, collating and incorporating authors’ trolled severe persistent asthma in pneumology and allergy hospital units in comments for each draft, assembling tables and figures, grammatical editing, and Spain. J. Investig. Allergol. Clin. Immunol. 21, 466–471 (2011). referencing) was provided by Lucia Correia and Chloe Stevenson, of Fishawack Indicia 14. Kerstjens, H. A. et al. Tiotropium in asthma poorly controlled with standard Ltd, UK, part of Fishawack Health, and was funded by GSK. combination therapy. N. Engl. J. Med. 367, 1198–1207 (2012). 15. Virchow, J. C. et al. Single inhaler extrafine triple therapy in uncontrolled asthma (TRIMARAN and TRIGGER): Two double-blind, parallel-group, randomised, con- AUTHOR CONTRIBUTIONS trolled phase 3 trials. Lancet 394, 1737–1749 (2019). 16. Kerstjens, H. A. M. et al. Once-daily, single-inhaler mometasone-indacaterol- M. Monteagudo contributed to the conception/design of the study, acquisition of glycopyrronium versus mometasone-indacaterol or twice-daily fluticasone-salme- data, and data analysis/ interpretation. I.O. and M.B. contributed to conception/ terol in patients with inadequately controlled asthma (IRIDIUM): A randomised, design of the study, and data analysis/interpretation. M. Miravitlles, X.F., J.O., A.N., and double-blind, controlled phase 3 study. Lancet Respir. Med. 8, 1000–1012 (2020). X.M. contributed to data analysis/interpretation. All authors were involved in the 17. Lee, L. A. et al. Efficacy and safety of once-daily single-inhaler triple therapy (FF/ preparation and review of the manuscript and approved the final version to be UMEC/VI) versus FF/VI in patients with inadequately controlled asthma (CAP- submitted. All authors take complete responsibility for the integrity of the data and TAIN): A double-blind, randomised, phase 3A trial. Lancet Respir. Med. 9, 69–84 accuracy of the data analysis. (2020). 18. Stempel, D. A. et al. Adherence to asthma controller medication regimens. Respir. Med. 99, 1263–1267 (2005). COMPETING INTERESTS 19. O’Connor, R. D. et al. Asthma-related exacerbations, therapy switching, and I.O. has received travel grants, consulting fees, speaking fees or research grants therapy discontinuation: a comparison of 3 commonly used controller regimens. from AstraZeneca, Bial, Boehringer Ingelheim, Chiesi, GSK, Mundipharma, Novartis, Ann. Allergy Asthma Immunol. 95, 535–540 (2005). Teva, Sanofi, MSD and Puretech; I.O. is a researcher supported by the “Pla Estratègic 20. Bolibar, B. et al. SIDIAP database: electronic clinical records in primary care as a de Recerca i Innovació en Salut PERIS)” 2016–2020 (SLT008/18/00108;G60594009). source of information for epidemiologic research. Med. Clin. 138, 617–621 (2012). M.B. has received speaker fees from CSL Behring, Grifols, Menarini, GSK and 21. CMS. 2020 ICD-10-CM Coding Guidelines, https://www.cms.gov/Medicare/Coding/ Boehringer Ingelheim, consulting fees from GSK, Novartis and Boehringer ICD10/2020-ICD-10-CM (2020). Ingelheim. X.M. has received fees as a speaker, scientific advisor, or participant of 22. Fagerland, M. W., Hosmer, D. W. & Bofin, A. M. Multinomial goodness-of-fit tests clinical studies from AstraZeneca, Boehringer Ingelheim, Chiesi, Faes, GSK, Menarini, for logistic regression models. Stat. Med. 27, 4238–4253 (2008). Mundipharma, Novartis, and Teva. M. Miravitlles has received speaker fees from 23. Suzuki, T., Fairburn-Beech, J., Sato, K. & Kaise, T. Clinical characteristics, treatment AstraZeneca, Boehringer Ingelheim, Chiesi, Cipla, Menarini, Rovi, Bial, Zambon, CSL patterns, disease burden, and persistence/adherence in patients with asthma Behring, Grifols and Novartis, consulting fees from AstraZeneca, Boehringer initiating inhaled triple therapy: Real-world evidence from Japan. Curr. Med. Res. Ingelheim, Chiesi, GSK, Bial, Gebro Pharma, CSL Behring, Laboratorios Esteve, Opin. 36, 1049–1057 (2020). Ferrer, Mereo Biopharma, Verona Pharma, Teva, pH Pharma, Novartis and Grifols 24. Monteagudo, M. et al. Clinical characteristics and factors associated with triple and research grants from Grifols. X.F. has received speaking fees from AstraZeneca, therapy use in newly diagnosed patients with COPD. NPJ Prim. Care Respir. Med. Chiesi, GSK, Mundipharma and Teva. M. Monteagudo, A.N., and J.O. declare no 31, 16 (2021). competing interests. npj Primary Care Respiratory Medicine (2022) 11 Published in partnership with Primary Care Respiratory Society UK M Barrecheguren et al. 9 ETHICAL APPROVAL AND INFORMED CONSENT Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims The study was conducted in compliance with local legislation for observational trials in published maps and institutional affiliations. and it was approved by the IDIAP Jordi Gol Clinical Research Ethics Committee. As this is an observational study based on a database that uses anonymised patient records, patient informed consent is not applicable. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, ADDITIONAL INFORMATION adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Supplementary information The online version contains supplementary material Commons license, and indicate if changes were made. The images or other third party available at https://doi.org/10.1038/s41533-022-00270-2. material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Correspondence and requests for materials should be addressed to Miriam Creative Commons license and your intended use is not permitted by statutory regulation or Barrecheguren. exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permission information is available at http://www.nature.com/ reprints © The Author(s) 2022 Published in partnership with Primary Care Respiratory Society UK npj Primary Care Respiratory Medicine (2022) 11