Demographics, Treatment Patterns, and Morbidity in Patients with Exercise-Induced Bronchoconstriction: An Administrative Claims Data Analysis

Introduction

Exercise-induced bronchoconstriction (EIB) is the transient narrowing of the lower airways during or after exercise and commonly occurs in patients with asthma but can also occur in the absence of asthma.^1–4

The decline in lung function that occurs during or after exercise was initially termed exercise-induced asthma but has more recently been subdivided and reclassified as EIB without asthma (EIB_alone) or EIB with asthma (EIB_asthma).^2,5 EIB occurs in up to 90% of patients with a confirmed asthma diagnosis,⁶ and is more frequent among patients with poor asthma control or more severe asthma.³ EIB_alone is particularly prevalent in children and in patients with atopy or rhinitis,⁵ and is associated with atopy in athletes.⁷ In children aged ≤16 years, the prevalence estimates of EIB with or without asthma vary between 3–35%,³ with one estimate of the global mean prevalence of EIB in children/adolescents with asthma as high as 46%.⁸ The wide range of these estimates is due to variability in, and lack of consensus on, the diagnostic criteria and methods of clinical confirmation, as well as variability in testing conditions (environmental conditions such as relative humidity, temperature, pollutant levels¹ or exercise conditions such as type and intensity of exercise⁵).

There is broad agreement that the aims of pharmacologic treatment of EIB_alone and EIB_asthma should be prevention and/or amelioration of exercise-related bronchoconstriction and symptoms.^1,4,9 For patients with EIB_alone, use of inhaled short-acting β₂-agonists (SABA) taken 5–20 minutes before exercise and/or if symptoms develop post-exercise is generally recommended.^1,10 In contrast to the usual management of EIB_alone, which targets symptom prevention and/or relief with SABA only,^1,4 patients diagnosed with EIB_asthma should be treated with daily maintenance therapy for optimal asthma control,^1,9,10 according to the severity of their asthma, and following clinical practice guidelines. Patients with EIB_alone may also be prescribed daily inhaled corticosteroid (ICS) or leukotriene modifier (LM) treatment if they use SABA on a frequent or even daily basis.^1,4 Chronic/daily use of SABA or long-acting β₂-agonist (LABA) alone for patients with EIB_alone should be avoided, because of the risk of developing tachyphylaxis to the bronchoprotective and bronchodilatory effects of β₂-agonists and subsequent risk of serious adverse effects including exacerbations requiring hospitalization.^1,4,11–13

Other treatment options include low-dose ICS-formoterol taken as needed and before exercise in patients with EIB_asthma,^9,14 inhaled short-acting anticholinergics,^4,15 and mast cell stabilizers (inhaled cromolyn sodium or nedocromil sodium) given before exercise. Although mast cell stabilizers are recommended for patients with EIB_alone or EIB_asthma who continue to have EIB despite SABA use or who require an inhaled SABA daily or more frequently,⁴ they attenuate EIB only by ~50%, have a limited duration of effect and no bronchodilator activity,¹ and are only available in the USA in a formulation requiring nebulization.

Real-world treatment patterns and clinical outcomes of EIB in patients with or without asthma in the USA have not been studied in detail to date. The aim of this study, therefore, was to compare demographic characteristics, treatment patterns, morbidities, healthcare resource use, and costs between patients with EIB_alone versus those with EIB_asthma in a real-world setting.

Materials and Methods

Data Source and Patients

This was a retrospective analysis of administrative claims data from the IBM^® MarketScan^® Research Databases in the USA, which includes the Commercial, Medicare Supplemental, and Medicaid databases. These databases contain healthcare data for patients insured commercially or via the Medicare or Medicaid programs. They include detailed cost, use and outcomes data for healthcare services, including prescription drug claims, provided in inpatient or outpatient settings. The online analytic platform Treatment Pathways 4.0 was used to access the MarketScan databases.

Unique identifiers are used to link person-level enrollment data to medical and outpatient prescription drug claims. All database studies using US data that are conducted by AstraZeneca comply with the United States Health Insurance Portability and Accountability Act of 1996 Privacy Rule, which allows for use of health information that neither identifies nor provides a reasonable basis to identify an individual. This study used fully de-identified data and as such was not classified as research involving human participants as defined by 45 CFR 46.104(d)(4). For the use of fully de-identified US data, AstraZeneca complies with the procedures set forth in Sections 164.514 (a)–(b)1ii of the Health Insurance Portability and Accountability Act of 1996 Privacy Rule, which allows for use of health information that neither identifies nor provides a reasonable basis to identify an individual. Therefore, approval from an institutional review board was not sought.

Patients aged ≥4 years with a SABA fill between January 1, 2011 and December 31, 2016 were assessed for inclusion. A random SABA claim within the analysis period was chosen as the index date. Randomization was conducted through the random number generator (RANUNIT function) using SAS 9.4. Eligible patients were required to have continuous enrolment in the MarketScan databases for 1 year prior to (pre-index period) and 1 year following this index date (post-index period), 1 inpatient or 2 outpatient non-diagnostic claims (ie visits) and a separate diagnostic code for EIB (International Classification of Diseases (ICD)-9-CM: 493.81; ICD-10-CM: J45990), asthma (ICD-9-CM: 493.xx, except 493.81; ICD-10-CM: J45xx, except J45990), or both, at any time in the pre-index through 60 days in the post-index period, and ≥1 additional SABA fills, or for patients with a single SABA fill (index fill only), ≥1 maintenance medication fills in the post-index period. Thus, patients in the EIB_alone group had a diagnostic code for EIB, but not asthma, and patients in the EIB_asthma group had diagnostic codes for both EIB and asthma. Patients with diagnostic codes on an inpatient claim or non-diagnostic outpatient claim any time during the pre- or post-index periods associated with chronic lower respiratory diseases other than asthma, primary eosinophilic disorders, cancer, or autoimmune conditions were excluded from the analysis.

Data Collected

The following patient data were collected: demographics (age, sex), diagnoses of interest, and medication claims in the 1-year post-index period (SABA, maintenance medication, and systemic corticosteroids [SCS; oral or injectable] claims per National Drug Code).

Data on healthcare resource use and costs in the 1-year post-index period were gathered: both all-cause and disease (EIB and/or asthma)-specific healthcare service utilization by place of service, including medical services (inpatient admission, outpatient medical services) and outpatient prescriptions. Disease-specific medical costs included EIB claims in the EIB_alone group and both EIB and asthma claims in the EIB_asthma group.

Healthcare costs were calculated based on amounts paid for adjudicated claims; this included insurer and health plan payments, as well as patient cost-sharing in the form of copayments, deductibles or coinsurance. All costs were inflation-adjusted using the medical care component of the Consumer Price Index (CPI), and were standardized to annualized 2017 US dollars (US$).

Statistical Analysis

Descriptive statistics were used to summarize patient demographics, annual medication claims, and all-cause and disease-specific healthcare costs. Comparisons between patients with EIB_alone and those with EIB_asthma for treatment patterns and healthcare utilization were made using the healthcare Student’s t-test and chi-squared (χ²) tests, and for costs using unpaired Student’s t-tests, at a pre-determined alpha significance level of 0.05.

Results

A total of 13,480 patients with EIB_alone and 14,862 with EIB_asthma were identified and included in this analysis (Figure 1). Of note, the population of patients having diagnostic codes for both asthma and EIB made up 1.3% of the population with a diagnostic code for asthma (N=1,134,143). Patients in the EIB_alone group were older (mean [standard deviation (SD)] 20.4 [13.6] vs 17.8 [13.6] years) and had a greater proportion of females (60.7% vs 54.7%) than the EIB_asthma group (both χ² p<0.001; Table 1). EIB_alone was more common than EIB_asthma in all the age groups except in the 4–11 years’ group, where there were fewer patients with EIB_alone than with EIB_asthma (20.9% vs 34.2%; χ² p<0.001) (Table 1).

Table 1 Demographic Characteristics of Patients Aged ≥4 Years with a Diagnosis of EIBalone or EIBasthma

Figure 1 Study population selection from the IBM® MarketScan® Research Databases in the US, which includes the Commercial, Medicare Supplemental, and Medicaid databases. aPresence of chronic lower respiratory illness other than asthma, primary eosinophilic disorders, cancer, or an autoimmune disorder resulted in exclusion. Abbreviations: EIBalone, exercise-induced bronchoconstriction (EIB) without asthma; EIBasthma, EIB with asthma; SABA, short-acting β2-agonist.

Treatment Patterns

The annual mean [SD] SABA claims were lower in the EIB_alone than the EIB_asthma group (1.9 [1.4] vs 2.5 [2.2]; p<0.001), but the median number of SABA claims was 2.0 in both groups.

Significant between-group differences (p<0.0001) were observed according to annual quantity of SABA claims: in the EIB_alone group, more patients had filled one or two SABA claims compared with the EIB_asthma group (44.4% vs 34.8% and 31.3% vs 26.9%, respectively; Figure 2). In contrast, significantly fewer patients in the EIB_alone group than EIB_asthma group had 3 or 4 or more SABA fills in the post-index period (12.8% vs 15.9% and 11.6% vs 22.4%, respectively).

Figure 2 SABA treatment patterns compared between patients aged ≥4 years with a diagnosis of EIBalone (n=13,480) or EIBasthma (n=14,862). *P<0.0001 for comparisons between EIBalone and EIBasthma for each level of SABA fills (χ2 tests). Abbreviations: EIBalone, exercise-induced bronchoconstriction (EIB) without asthma; EIBasthma, EIB with asthma; SABA, short-acting β2-agonist.

Maintenance medication fills were less common in the EIB_alone than the EIB_asthma group (Figure 3). The most common maintenance therapy regimen in both groups was a LM alone or in combination with other maintenance treatments, recorded for 40.2% of patients in the EIB_alone group and 45.0% of patients in the EIB_asthma group (between-group difference p<0.0001). Between-group differences were significant (p<0.0001) for all categories of maintenance medication prescription (Figure 3). Maintenance medication prescriptions were higher in the EIB_asthma than EIB_alone group for ICS in combination with other maintenance medication, ICS only or ICS plus LM only (montelukast, zafirlukast, zileuton), ICS-LABA only, and ICS-LABA in combination with any other maintenance medication, while prescriptions for LM only were higher with EIB_alone versus EIB_asthma (all p<0.0001) (Figure 3).

Figure 3 Asthma maintenance treatment prescriptions for patients aged ≥ 4 years with EIBalone (n=13,480) or EIBasthma (n=14,862). Rates of specific maintenance medication use include patients who use other maintenance medications throughout the 12-month post-index period (indicated by “+ other maintenance medication”); “only” implies no other use of maintenance medication other than the maintenance medication(s) specified. *P<0.0001 for comparisons between EIBalone and EIBasthma for each maintenance therapy category (χ2 tests). Abbreviations: EIBalone, exercise-induced bronchoconstriction (EIB) without asthma; EIBasthma, EIB with asthma; ICS, inhaled corticosteroid; LABA, long-acting β2-agonist; LM, leukotriene modifier.

The proportion of patients with ≥1 annual claim for SCS was lower in the EIB_alone group than the EIB_asthma group (29.4% vs 32.0%; p<0.0001) (Table 2). However, when analyzed by type of SCS (oral versus injectable), with the between-group difference being significantly higher only for oral corticosteroids (22.5% [EIB_alone] vs 26.6% [EIB_asthma]; p<0.0001). Approximately 3 weeks’ worth of oral corticosteroid prescriptions were filled in the post-index year in both groups: a mean of 20.7 days (SD 30.8) in the EIB_alone group and 19.8 days (28.8) in the EIB_asthma group (p=0.188). Chronic SCS use was rare and was observed in a similar proportion of patients in both groups (both 0.2%; p=0.448).

Table 2 Systemic Corticosteroid Prescriptions for Patients Aged ≥4 Years with a Diagnosis of EIBalone or EIBasthma

Healthcare Resource Utilization

More than two-thirds of patients in both groups had disease-specific outpatient healthcare professional (HCP) office visits in the post-index period, fewer in the EIB_alone than the EIB_asthma group (65.1% vs 72.3%; p<0.0001) (Table 3). While uncommon overall, fewer patients with EIB_alone than patients with EIB_asthma required urgent care visits in the post-index period year (0.3% vs 2.0%; p<0.0001). Emergency department (ED) visits were much less common among patients with EIB_alone than patients with EIB_asthma (1.0% vs 13.0%; p<0.0001). Fewer patients with EIB_alone had a disease-related hospitalization (0.4%) than patients in the EIB_asthma group (2.0%; p<0.0001).

Table 3 Disease-Specific Healthcare Resource Use for Patients Aged ≥4 Years with a Diagnosis of EIBalone or EIBasthma

All-Cause and Disease-Related Costs

All-cause total healthcare costs were higher in the EIB_asthma than the EIB_alone group (US$6511 vs US$5432; P<0.0001). This was likely attributable to significantly higher outpatient costs, such as HCP office visits and pharmacy prescriptions costs in the EIB_asthma versus EIB_alone group (Table 4). However, all-cause inpatient costs were higher in the EIB_alone than the EIB_asthma group (mean costs per patient per year: US$25,542 vs US$21,013; P<0.0001).

Table 4 All-Cause and Disease-Specific Healthcare Costs (US$) for Patients Aged ≥4 Years with a Diagnosis of EIBalone or EIBasthma

Disease-specific healthcare costs for EIB_alone represented 11.6% of all-cause healthcare costs in this group, while for the EIB_asthma group it was 26.2%. Although disease-specific inpatient and ED costs were higher in the EIB_alone than EIB_asthma group (mean costs per patient per year: US$18,110 vs US$15,463; P<0.0001, and US$822 vs US$764; P=0.002, respectively), the overall total healthcare costs were higher in the EIB_asthma group (US$1708 vs US$633 in the EIB_alone group; P<0.0001). This was driven by significantly higher medication costs (SABA and maintenance medication) and outpatient visit costs in the EIB_asthma versus EIB_alone group (Table 4).

Discussion

Prescription claims and healthcare resource utilization data from 28,342 US patients with EIB, of whom 14,862 also had a diagnosis of asthma, were analyzed in this study. To the best of our knowledge, this is the first investigation of its kind in which real world evidence data is analyzed to better understand the burden of disease in patients with EIB. Systemic corticosteroid claims, healthcare resource utilization, and cost of care were high in both cohorts, suggesting considerable unmet need in patients with EIB.

Almost one-third of patients with EIB_alone had ≥1 SCS claim (29.4% of patients), with 3-weeks’ worth of oral corticosteroids being filled in the post-index year. Healthcare resource use data confirmed the extent of unrecognized morbidity in EIB_alone; 65.1% of patients in this group had outpatient HCP office visits associated with EIB. In the EIB_asthma group, 32.0% of patients had ≥1 SCS claim in the post-index period, and similar to the EIB_alone population, this group also filled close to 3 weeks’ worth of oral corticosteroids. Additionally, the majority of patients with EIB_asthma (72.3%) had ≥1 outpatient office visit and 13.0% attended an ED for an asthma exacerbation.

Analysis of treatment patterns may help explain the observed morbidity associated with EIB in both sets of patients. The EIB_asthma group relied heavily on SABA use, with 22.4% filling 4 or more SABA claims in the year. This would equate to at least 400 doses annually (each canister contains 200 puffs, ie 100 doses), constituting an average use of SABA at least once daily. Regular use of SABA may not only lead to an increase in bronchial hyperresponsiveness, but also a tolerance to the bronchodilator effect and an increase in bronchoconstriction in response to the stimulus of exercise.^4,11,13 Patients with EIB_asthma should be treated for optimum asthma control, ie with maintenance therapy, to reduce the frequency and severity of EIB and the use of SABA.^1,4,9,10 Although over 90% of patients with EIB_asthma were receiving maintenance therapy, the reliance on SABA suggests their asthma was uncontrolled. Indeed, the level of SCS use in this group adds further support to this explanation.

Although there was no difference in the median number of SABA fills per year between the EIB_asthma and EIB_alone populations (2.0), the majority of patients in the EIB_alone group had one or two refills (~76%). This is what would be expected if patients were using their SABA intermittently prior to exercise, as recommended,^1,4 since one canister provides 100 doses and two canisters over 52 weeks would equate to usage of 3.8 times a week, which seems reasonable for prophylactic use during regular exercise. However, 80% of patients with EIB_alone in our analysis were also prescribed daily maintenance medication. Taken together with the number of SCS claims in the EIB_alone group, our analysis raises several important issues that should be brought to the attention of US healthcare providers: patients classified as EIB_alone may be misdiagnosed, have unrecognized asthma, or be placed on inadequate maintenance medication. EIB needs to be identified objectively as per current guidelines,^2,4 and such rigor in making this diagnosis most likely occurred in a very small proportion of the groups identified in this study. For many assumed to have EIB, symptomatic assessment alone can be inadequate for assessing both the presence and severity of EIB. The effective treatment of EIB may require daily ICS with or without additional ICS prior to exercise, where mast cell degranulation due to airway drying results in inflammatory mediator release.¹ In fact, it has been proposed that many EIB_asthma patients have active EIB due to poor adherence to regular ICS;⁴ thus, as-needed ICS in EIB may be a problem in those whose asthma is of a severity requiring daily maintenance ICS to achieve asthma control. Furthermore, the increased disease severity observed in this study for patients with EIB_alone may point to this condition having a more severe phenotype than is currently recognized in US clinical practice; these patients may be more similar to EIB_asthma patients or those transitioning into active asthma.

We did observe differences in healthcare resource utilization between patients with EIB_alone and those with EIB_asthma. Patients with EIB_alone had fewer disease-related hospitalizations and outpatient, urgent care, and ED visits. The higher outpatient (medical and pharmacy) costs in the EIB_asthma group could be attributable to greater vigilance in clinical assessment and determining treatment plans for patients with an established asthma diagnosis, thus helping to minimize the cost of any disease-specific inpatient admissions. Since patients with EIB_alone did not have an asthma diagnosis, inpatient costs may have been higher due to a more thorough clinical evaluation needed for differential diagnosis, especially as between-group differences in disease-specific inpatient costs were not attributable to differences in mean length of stay (data not shown).

Equally important for both patient populations, more attention may need to be paid to EIB symptoms so that insufficient and/or ineffective treatments can be modified to decrease morbidity. Only a very small proportion of all patients in this real world medical and pharmacy claims database identified as having an asthma diagnosis and a diagnosis of EIB (1.3% of the 1,134,143 patients with asthma), even though it is well established that EIB and asthma frequently co-occur in up to 90% of patients with asthma.¹⁶ This finding suggests EIB may be ignored or downplayed if asthma is already diagnosed. Much the same as providers under-coding for allergic asthma in patients with a pre-existing asthma diagnosis,⁸ failure to enter a diagnostic code for EIB when treating a patient with asthma for exercise-related symptoms may occur. Moreover, this low prevalence of EIB in the asthma population may have also resulted from patients not exercising, and therefore not experiencing EIB, or from fear of exercising and inducing EIB.

Overall, our morbidity results indicate that the currently recommended EIB treatment paradigm of SABA prior to exercise or in response to exercise-related symptoms with or without daily maintenance therapy is not optimal.^1,4 For patients with EIB_asthma, the level of SABA use related to exercise is excluded from National Asthma Education and Prevention Program (NAEPP) recommendations for determining asthma control¹⁷ and is not specifically mentioned in terms of determining asthma control in the Global Initiative for Asthma (GINA) 2021 guidelines.⁹ This suggests that the consequent potential risk of morbidity associated with SABA as prevention and/or treatment of EIB may be overlooked by clinicians and patients, as illustrated by our results. Under-explored alternative approaches include the use of as-needed ICS-formoterol or an ICS whenever SABA is taken, for patients with any severity of asthma. In a 6-week randomized study in 59 patients with mild EIB_asthma, as-needed budesonide-formoterol was superior to as-needed SABA alone in reducing EIB, and non-inferior to daily budesonide maintenance therapy plus as-needed SABA.¹⁴ While as-needed budesonide-formoterol is not recommended for the management of EIB, and the NAEPP 2020 focused update does not include any revisions on their recommendations for the management of EIB_asthma,¹⁷ the 2021 GINA strategy document acknowledges that when as-needed budesonide-formoterol is used to manage mild asthma, EIB can be managed by using the same inhaler as-needed prior to exercise,⁹ thus avoiding the need for a separate inhaler prior to exercise, and mitigating the risks associated with SABA overuse.

Limitations of this study include the reliance on administrative claims databases, which could introduce the potential for coding and data entry errors, misclassification or misdiagnosis of asthma status and healthcare resource utilization, and incomplete assessment of cost of care. Furthermore, EIB should be diagnosed on the basis of changes in lung function provoked by exercise, not on the basis of symptoms.⁴ The patient selection criterion of requiring those with only the index SABA fill to also have had ≥1 maintenance medication claim could have biased our observations, although this was to ensure that patients with a one-off prescription of SABA were not included. Additionally, our findings may not be generalizable to patients that have healthcare insurance types not in the database or patients that have no insurance coverage. It is therefore possible that the two groups we assessed did not represent the full spectrum of patients with EIB_alone or EIB_asthma. Because the data available for analysis were inherently limited by the information available in the database, we were not able to adjust for age, comorbidities, socioeconomic status, or other variables likely to influence clinical morbidity and asthma- and all-cause healthcare costs. The cost analysis of inpatient admissions was based on a small sample size without the ability to adjust for outliers, and should be interpreted cautiously. Additionally, it was not possible to determine the role of maintenance treatment adherence on the lack of asthma control observed in these patient populations. Studies with a more detailed evaluation of demographic, clinical and treatment data would be helpful in understanding the full extent and causes of morbidity in patients with EIB. Most of these limitations may be expected from a claims-based study, but because this analysis was performed on the MarketScan database, it represents actual prescribing practices and healthcare resource use in a real-world setting.

Conclusions

To our knowledge, this is the first real-world study demonstrating the considerable burden of disease associated with EIB with or without a concomitant asthma diagnosis. Our findings underscore that there may be a lack of recognition, as well as inadequate management, of EIB. Prospective research in clinical practice settings that analyze methods of diagnosis and the relationships between EIB and asthma control are needed to determine if new therapeutic paradigms for patients with EIB can impact the associated morbidity.