ABSTRACT
ABSTRACT
Objective: To assess asthma control and associations with health-related quality of life (HRQoL) and economic outcomes among patients with asthma and allergic comorbidities treated with inhaled corticosteroids (ICS) and long-acting beta-agonists (LABA) combination therapy. Methods: Data from the 2011–2013 US National Health and Wellness Survey were used to identify patients with asthma currently treated with ICS and LABA combination therapy (N = 1923). Patients were included if they self-reported a physician diagnosis of asthma and at least one allergic/asthma-related comorbid condition (e.g., nasal allergies, atopic dermatitis). Asthma Control Test scores categorized patients as very poorly (scores ≤ 15; 29.3%), not well (16–19; 25.1%), or well controlled (20–25; 45.7%). Outcomes included HRQoL (SF-36v2; SF-12v2), work productivity and activity impairment, healthcare utilization (HRU), and annual indirect and direct costs. Generalized linear models, controlling for covariates, examined whether outcomes differed by asthma control. Results: Over half of the patients had very poorly or not well-controlled asthma (54.4%). Patients with very poorly controlled versus well-controlled asthma reported significantly greater decreases in HRQoL, greater overall work impairment, and higher HRU (all, p < 0.05). Very poorly controlled patients incurred over double the indirect costs and nearly one and a half times the direct and total costs of well-controlled patients. Conclusions: Increasing level of asthma control was related to improved HRQoL and lower costs. The considerably high prevalence of uncontrolled asthma among patients on ICS and LABA suggests poor treatment adherence or unmet needs in current treatment and may require step-up therapy in appropriate patients according to clinical guidelines.
Introduction
Asthma is characterized by chronic wheezing and airflow obstruction, which occurs because of swelling and inflammation in the lining of the bronchial tubes [1]. Asthma can be triggered by contact with an allergen; common triggers include dust mites, animal dander, and cockroach debris [2]. Between 2001 and 2009, the prevalence of asthma in the US general population increased from 7.3% to 8.2%, with the prevalence of asthma at 7.7% for US adults [3]. According to the National Health, Lung, and Blood Institute (NHLBI), for adults, poorly controlled asthma entails frequent asthma symptoms (i.e., symptoms experienced >2 days per week requiring a short-acting beta-agonist to control) and may be associated with a progressive decline in pulmonary function [4].
Previous research indicated that adults with poorly controlled asthma had significantly lower HRQoL, with higher work productivity loss and activity impairment than those with well-controlled asthma [5, 6]. Asthma has also been associated with poorer health status. The rate of depression among patients with asthma was significantly higher than that for matched controls [7]. Additionally, after controlling for comorbidities (e.g., gastroesophageal reflux disease and sleep apnea), asthma patients reported poor sleep quality, which was even worse with more severe symptoms [8]. Regardless of asthma severity, poor sleep quality was found to be linked to inadequate asthma control and lower HRQoL. Overall, these research findings suggest that achieving asthma control is integral to improving HRQoL and health status. However, this may be challenging, given almost 40% of US adults with asthma had poor asthma control [5].
Asthma is also associated with a considerable economic burden. Data from the 2003 and 2005 Medical Expenditures Panel Survey (MEPS) showed that costs associated with asthma among adults in the US totaled $18 billion annually in 2008 US dollars [9]. Among US adults, those with poorly controlled asthma had significantly higher healthcare resource use (HRU; i.e., emergency room [ER] visits, hospitalizations, and visits to healthcare providers [HCPs]) than adults with well-controlled asthma [5, 6]. Nevertheless, only a third of US adults with asthma were found to be actively managing their symptoms with long-term control medications [11]; indicating that asthma may be sub-optimally treated in the US. Insufficient treatment may serve to increase the burden of this disease.
Long-term maintenance medications are used daily with the goal being to achieve symptom control for persistent asthma [4]. In contrast, short-term treatments (e.g., short-acting beta-agonists) are available to quickly relieve acute incidents of airflow obstruction. Patients with persistent asthma need a combination of these medications. NHLBI guidelines recommend that adult asthma patients step up 1 step for not well-controlled asthma [10]. Patients not well controlled on step 3 preferred treatment (low-dose ICS +LABA OR medium-dose ICS) need to step up to step 4 therapy (medium-dose ICS +LABA or alternatively low-dose ICS +either Leukotriene Receptor Antagonist [LTRA],Theophylline, or Zileuton). Patients not well controlled on step 4 preferred treatment (medium-dose ICS +LABA) need to step up to step 5 therapy (high-dose ICS + LABA AND consider omalizumab for patients who have allergies).
Despite available studies on the burden of asthma in the US, noticeable gaps still exist. Most studies underestimate asthma burden by exclusively focusing on just symptoms or costs, rather than a comprehensive assessment of burden. Furthermore, all forms of asthma and asthma medication regimens are frequently aggregated together when estimating prevalence and burden. This approach is problematic, as asthma is a complex disease with multiple phenotypes that can differ by medication requirements and age of onset, among other characteristics [11]. ICS and LABA combination therapy is the mainstay of controller treatment for moderate-to-severe asthmatics [4], yet the burden of these patients, particularly among those with asthma-related and allergic comorbidities, is unexplored. Clarifying the extent of the burden attributed to asthma and related comorbidities, by level of asthma control, in patients on ICS and LABA combination therapy will provide clinicians, patients, and payers with important information on additional unmet need in patients optimized on these therapies. Such information can help to guide better and more cost-effective treatment decisions in asthma patients.
Objectives
The current study had three key objectives. First, the study examined level of asthma control among patients with asthma and asthma-related or allergic comorbidities on ICS and LABA combination therapy. Second, this study examined the association between level of asthma control and health burden, controlling for patient characteristics, among patients with asthma and asthma-related or allergic comorbidities on ICS and LABA combination therapy. Third, this study examined the association between level of asthma control and economic burden, controlling for patient characteristics, among patients on ICS and LABA combination therapy.
Methods
Data source
This retrospective observational study included data from the 2011–2013 US National Health and Wellness Survey (2011 NHWS N = 75,000; 2012 NHWS; N = 71,157 and 2013 NHWS; N = 75,000). The NHWS is a representative, large-scale survey of the adult population (≥18 years) assessing health status and outcomes across a wide array of diseases. The NHWS uses random stratified sampling framework to ensure that the demographic composition is similar to the US general population [5]. To maximize sample size of asthma patients, 3 years of NHWS data were pooled. Since it was possible for a single respondent to have participated in consecutive survey years, only the most recent year of survey data of the respondent were kept in such instances. The current study received approval from the Essex Institutional Review Board (Lebanon, New Jersey). Informed consent was obtained from all participants.
Sample
Patients with asthma currently on ICS and LABA combination therapy were included in the analyses (N = 1923). Asthma was defined by a self-reported physician diagnosis of asthma. Asthma with asthma-related or allergic comorbidities (here after referred to as “allergic asthma”) was defined by a self-reported physician diagnosis of asthma and at least one of the following comorbid conditions: chronic hives, dermatitis, nasal allergies, hay fever, atopic dermatitis, eczema, or skin allergies. ICS and LABA combination therapy included patients who were currently on a combination inhaler (ICS and LABA in one inhaler/device) or those on two separate inhalers (one ICS only inhaler plus one LABA only inhaler); patients may also have been taking other asthma prescription medications.
Measures
Asthma control
Patients were grouped by asthma control, according to their Asthma Control Test scores (scores ≤15: very poorly controlled; 16–19: not well-controlled; 20–25: well-controlled asthma) [12].
Demographics and patient characteristics
Participants provided data on age, gender, race/ethnicity (non-Hispanic black, Hispanic, other, vs. non-Hispanic white), marital status (married/living with partner vs. not married), insurance (do not have vs. have), education (university degree vs. less than university degree), household income (<$25K, $25–$50K, $50–$75K, or decline to answer vs. $75K+), and employment status (employed vs. unemployed).
Participants self-reported smoking status (currently smoke, former smoker vs. never smoker), alcohol use (drink vs. do not drink alcohol), and exercise behavior (exercise vs. do not exercise for 20 minutes or more, one or more times in the past month). Body mass index (BMI; underweight, overweight, obese, unknown vs. normal weight) and comorbidities (Charlson Comorbidity Index [CCI] scores) [13], were calculated. Higher total CCI index score indicate more comorbidity burden on the patient. Additionally, participants were asked how many years since they were diagnosed with asthma, time(s) of the year asthma was experienced (spring, summer, autumn, and/or winter), frequency of problems in the seasons when experiencing asthma, and main causes of asthma attacks (e.g., allergies, exercise, etc.).
Health burden
Health status. Eight health domain scores (bodily pain, general health, mental health, physical functioning, role emotional, role physical, social functioning, and vitality), physical component summary score (PCS), mental component summary score (MCS), and Short Form-6 Dimension (SF-6D) health utilities were derived from the Medical Outcomes Study Short Form-12 Item Health Survey version 2 (SF-12v2; survey year 2011) or the Short Form-36 Item Health Survey version 2 (SF-36v2; survey years 2012–2013) [14–18]. The norm-based MCS, PCS, and health domains have a population mean of 50 and a standard deviation of 10 [16]. Previous research reported a population mean of 0.79–0.81 for health utilities [19]. Additionally, 3 points for MCS and PCS, 5 points on health domains, and 0.041 points for health utilities represent minimally important differences (MIDs) [16, 20].
Psychological symptoms. Participants self-reported presence or absence of depression, anxiety, and sleep problems (insomnia and sleep difficulties) were experienced in the past 12 months.
Economic burden
Work productivity and activity impairment. The Work Productivity and Activity Impairment-General Health Scale (WPAI-GH v2.0) [21] includes the following metrics: absenteeism (% of time missed from work due to health in the past 7 days), presenteeism (% of impairment while at work due to health in the past 7 days) overall work impairment (aggregate of absenteeism and presenteeism), and activity impairment (% of impairment in daily activities due to health in the past 7 days).
HRU. The frequency of various forms of all-cause (i.e., for any reason) HRU in the past 6 months was self-reported, specifically, ER visits, hospitalizations, and traditional HCP visits (includes general/family practitioners, internists, and specialists).
Estimated annual direct costs. All-cause HRU costs were calculated from MEPS data by applying 2012 age-specific unit cost to each HRU event for each NHWS respondent [22].
Estimated annual indirect costs. Similar calculations were performed for estimates of indirect costs. In this case, 2012 age- and gender-specific wages from the Bureau of Labor Statistics were applied to work hours lost with presenteeism and absenteeism from the WPAI-GH [23].
Estimated annual total costs. For employed participants, estimated total costs included the total of direct and indirect costs. Total costs for unemployed respondents only included direct costs.
Statistical analysis
Overall one-way analysis of variance (ANOVA; for continuous variables) and chi-square tests (for categorical variables) were used to compare demographics, health characteristics, and health and economic outcomes by asthma control groups. All multiple pair-wise comparisons were conducted with t-tests (continuous variables) or binomial proportion tests (z-tests; categorical variables) with Bonferroni correction.
Generalized linear models (GLMs) were then used to predict differences in outcomes between asthma control groups, controlling for age, gender, ethnicity, education, income, partner status, insurance, BMI category, smoking, alcohol use, exercise, CCI scores, and years diagnosed with asthma. GLMs specifying identity link distributions were used to predict health domains, physical and MCSs, and health utilities. GLMs specifying a binominal distribution with logit link functions were used to predict the presence/absence of depression, anxiety, or sleep problems. GLMs specifying negative binominal distributions with log-link functions were used to predict WPAI-GH scores, HRU visits, and costs (due to pronounced skew). Adjusted means were reported, along with 95% confidence intervals (CIs) for all multivariable models. Any two-sided p-value, less than 0.05, was considered statistically significant.
Results
Demographic and patient characteristics
Among all asthma patients, the mean age was 49.8 years, 66.3% were female, and 75.9% were white (Table 1). There were 563 (29.3%) very poorly controlled, 482 (25.1%) not well-controlled, and 878 (45.7%) well-controlled asthma patients. Compared with well-controlled asthma patients, a higher proportion of very poorly controlled asthma patients were single, had less than a 4-year college degree, and had lower income. Compared to well-controlled asthma patients, a higher proportion of very poorly controlled asthma patients were obese, current smokers, and did not exercise. In terms of comorbidity burden, a higher proportion of very poorly controlled asthma patients had a score of 3+ on the CCI than not well or well-controlled asthma patients. The most common allergic conditions across the groups were nasal allergies, hay fever, skin allergies, and eczema. Compared with well-controlled asthma patients, very poorly controlled asthma patients had a longer history of asthma and experienced asthma problems more on a daily basis. Although asthma patients in this analysis were all treated with ICS and LABA, many were also taking other asthma medications. Among the other asthma medications taken by allergic asthma patients, 75.2% were on a short-acting beta-agonist (SABA), 25.1% on leukotriene antagonists, 8.7% on an anticholinergic, 1.1% on a theophylline, 1.1% on anti-IgE, and less than 1% on an oral corticosteroids (See Table A1 in Appendix).
Table 1. Demographic and patient characteristics by asthma control among diagnosed asthma patients with asthma-related/allergic comorbidities on ICS and LABA combination therapy.
Health and economic burden of allergic asthma
Among asthma patients in the study sample, mean MCS was 45.9, PCS was 42.5, and health utilities was 0.66; all were below established population norms of 50 for MCS and PCS and 0.79–0.81 for health utilities (Table 2). All of the mean scores for the eight health domains were also below population norms of 50. More than a third of the allergic asthma patients experienced any anxiety disorder or depression, and half experienced sleep problems.
Table 2. Health and economic burden by asthma control among diagnosed asthma patients with asthma-related/allergic comorbidities on ICS and LABA combination therapy.
Among asthma patients, the mean percentage of overall work impairment was 25.6%, which was largely driven by presenteeism (22.6%), rather than absenteeism (6.3%) (Table 2). The mean percentage of activity impairment was 40.0%. Additionally, on average, patients visited a traditional HCP approximately eight times in the past 6 months and incurred annual total costs of $33 228.
The eight health domain scores, MCS, PCS, and health utilities were lowest among patients with very poorly controlled, followed by patients with not well controlled, and highest in patients with well-controlled asthma (Table 2). Patients with very poorly controlled and not well-controlled asthma had greater incidences of experiencing any anxiety disorder, depression, and sleep problems than patients with well-controlled asthma.
WPAI-GH scores (presenteeism, overall work impairment, and activity impairment), HRU (ER visits and hospitalizations), and annual estimated costs (indirect, direct, and total) were highest among patients with very poorly controlled, followed by patients with not well controlled, and lowest in patients with well-controlled asthma (Table 2).
Adjusted associations between asthma control and HRQoL
After adjusting for covariates, very poorly controlled, compared with not well-controlled and well-controlled, asthma patients had lower MCS, PCS (Figure 1), and healthTable 3 utilities (adjusted means = 0.61 vs. 0.66 and 0.70, p < 0.001). Not well-controlled patients also had significantly lower MCS, PCS, and health utilities (adjusted means = 0.66 vs. 0.70, p < 0.001) than well-controlled patients. Findings for the eight health domain scores followed a similar pattern (Figure 1). Notably, patients with very poorly controlled, compared with patients with well-controlled asthma, scored more than eight points lower on role-physical and nearly eight points lower on social functioning. Compared with those who were well-controlled, very poorly controlled patients experienced 1.8 times as much of any anxiety disorder, 1.9 times as much depression, and 1.7 times as much sleep problems (Table 3).
Published online:
06 June 2017Figure 1. HRQoL by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.001, except for very poorly controlled vs. not well-controlled on vitality, which differed at p = 0.001.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ijas20/2018/ijas20.v055.i02/02770903.2017.1316394/20180202/images/medium/ijas_a_1316394_f0001_b.gif"})
Figure 1. HRQoL by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.001, except for very poorly controlled vs. not well-controlled on vitality, which differed at p = 0.001.
Table 3. Experience of any anxiety disorder, depression, or sleep difficulties in the past 12 months by asthma control among diagnosed asthma patients with asthma-related/allergic comorbidities on ICS and LABA combination therapy, adjusting for covariates.
Adjusted associations between asthma control and economic outcomes
After adjusting for covariates, very poorly controlled asthma patients, compared with those not well-controlled and well-controlled, had greater overall work impairment and activity impairment (Figure 2). Very poorly controlled asthma patients, compared with those not well-controlled and well-controlled, had more ER visits and hospitalizations (Figure 3). Very poorly controlled asthma patients, compared with those well-controlled, had higher numbers of HCP visits.
Published online:
06 June 2017Figure 2. Work productivity and activity impairment by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. Presenteeism and overall work impairment were calculated for employed respondents only. All pair-wise comparisons between asthma control groups significantly differed at p < 0.01.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ijas20/2018/ijas20.v055.i02/02770903.2017.1316394/20180202/images/medium/ijas_a_1316394_f0002_b.gif"})
Figure 2. Work productivity and activity impairment by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. Presenteeism and overall work impairment were calculated for employed respondents only. All pair-wise comparisons between asthma control groups significantly differed at p < 0.01.
Published online:
06 June 2017Figure 3. Six-month HRU by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.05, except HCP (not well vs. very poorly controlled; p = not significant) and ER (not well vs. well-controlled; p = not significant) visits.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ijas20/2018/ijas20.v055.i02/02770903.2017.1316394/20180202/images/medium/ijas_a_1316394_f0003_b.gif"})
Figure 3. Six-month HRU by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.05, except HCP (not well vs. very poorly controlled; p = not significant) and ER (not well vs. well-controlled; p = not significant) visits.
Annual direct costs incurred by very poorly controlled asthma patients were 1.5-fold higher than annual direct costs for well-controlled patients (Figure 4). Annual indirect costs incurred by very poorly controlled asthma patients were 2.2-fold higher than annual indirect costs for well-controlled patients. Annual total costs incurred by very poorly controlled asthma patients were 1.5-fold higher than annual total costs for well-controlled patients. Annual indirect and total costs were similarly higher for very poorly controlled asthma patients, compared with not well-controlled patients.
Published online:
06 June 2017Figure 4. Annualized estimated costs by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.05, except for direct costs for not well vs. very poorly controlled (p = 0.06).
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ijas20/2018/ijas20.v055.i02/02770903.2017.1316394/20180202/images/medium/ijas_a_1316394_f0004_b.gif"})
Figure 4. Annualized estimated costs by asthma control, adjusted for covariates. Models were adjusted for age, gender, ethnicity, education, income, partner status, insurance, BMI, smoking, alcohol, exercise, and years diagnosed with asthma. Error bars depict 95% CIs. All pair-wise comparisons between asthma control groups significantly differed at p < 0.05, except for direct costs for not well vs. very poorly controlled (p = 0.06).
Discussion
The current study aimed to examine associations between level of asthma control with HRQoL and economic outcomes, among asthma patients with asthma-related or allergic comorbidities on ICS and LABA combination therapy. Unlike prior research, this study provided a more comprehensive account of the health and economic burden specific to adults with asthma and related comorbidities on ICS and LABA combination therapy. Findings of less than optimal levels of asthma control and the positive impact of better asthma control on HRQoL, WPAI, HRU, and costs highlight considerable unmet need in this population.
Prior studies reported that poorly controlled symptoms were found among a sizeable percentage (40%) of US adults with asthma [5]. The current study found that among allergic asthma patients on an ICS plus LABA, over half (54.4%) reported very poorly controlled or not well-controlled asthma. It may be that patients with allergic asthma who are not well-controlled have comorbidities that were not well treated. However, in multivariable regression analyses, the CCI, a measure of comorbidity burden, was included as a covariate to control for the effect of CCI comorbidities. Thus, comorbidities or untreated comorbidities cannot fully account for high uncontrolled asthma finding. Another possibility is that patients on ICS plus LABA regimen were not adherent to their treatment and thus, there was a high prevalence of uncontrolled asthma found in the current study. It may be possible that there were differences in the formulations of fixed-dose combination medications used in the controlled vs. uncontrolled groups leading to possible differences in the effectiveness of fixed-dose combination medications in the controlled vs. uncontrolled groups. The higher rate of SABA and ipratropium inhalers in the uncontrolled group could be a result of patients confusing controller inhalers with rescue inhalers, though, we cannot be certain of this occurrence. Additionally, physicians may not be consistently abiding by guidelines to step-up therapy for uncontrolled patients with persistent asthma. Particularly, NHLBI guidelines (for the US) indicate that, as symptom severity and frequency increase, the dose of ICS (plus LABA) should likewise increase [4]. Furthermore, additional treatment options for appropriate allergic asthma patients should be considered such as biologics and oral corticosteroids if needed (Steps 5 and 6) [4].
The current study's findings on allergic asthma agreed with prior research showing asthma to be associated with lower HRQoL [5, 6, 8]. Specifically, very poorly controlled compared with not well-controlled and well-controlled allergic asthma patients had lower MCS, PCS, and health utilities that exceeded MIDs. Not well-controlled patients also had significantly lower MCS, PCS, and health utilities than well-controlled patients; however, only health utilities reached MIDs thresholds. Moreover, differences in general health, bodily pain, physical functioning, role emotional, role physical, and social functioning health domain scores exceeded MIDs between very poorly controlled and well-controlled patients [16, 20]. Also, in line with previous research, uncontrolled asthma patients experienced poorer sleep quality [8]. Thus, the large health-related burden of asthma was replicated in this study for allergic asthma.
Consistent with previous research, patients with uncontrolled asthma (i.e., very poorly controlled or not well-controlled asthma) had greater work productivity loss, activity impairment, and HRU [5, 6]. As a result of work productivity loss and HRU, very poorly controlled patients were found to have total costs over one and half times higher than patients with well-controlled symptoms. Specifically, the indirect costs of very poorly controlled patients were over double than those incurred by well-controlled patients. The direct costs of very poorly controlled patients were almost one and a half times higher than those of well-controlled patients. Thus, results from the present study contributed to the evidence demonstrating the high costs ascribed to asthma, especially for those with uncontrolled symptoms, and particularly for the allergic asthma subpopulation [5, 6, 9]. Overall findings underscored the considerable economic burden associated with poor asthma control among allergic asthma patients.
Implications of study findings can be contextualized using Link and Phelan's fundamental social causes of health theoretical framework, which posits that social factors (e.g., economic hardship, low educational attainment, racial discrimination, sexism, etc.) can fundamentally cause inequalities in health [24, 25]. According to this framework, high socioeconomic status and its associated resources of money, knowledge, power, and prestige are beneficial to maintaining a health advantage. Indeed, the current study found that there was a higher rate of uncontrolled asthma in the lower income and education brackets. Although not directly assessed in the current survey, it is possible the cost of medication may impact adherence which would impact asthma control. Additionally, the considerably high costs associated with uncontrolled asthma may place a particularly higher burden on low-income relative to high-income patients and their families. Thus, an appreciation of the high economic burden of uncontrolled asthma within the context of socioeconomic disparities further illustrates the impact of poor asthma disease management, and may help clinicians better understand potential individual and societal-level gains from long-term asthma control improvement.
Strengths and limitations
This study examined the health and economic burden of asthma from the patient's perspective. Patient self-report can be more readily used to assess subjective outcomes, such as HRQoL, that cannot be evaluated using alternative means, like claims databases or are expensive to collect in clinical trials. Moreover, the study contributes to the limited real-world data on asthma patients with asthma-related or allergic comorbidities treated with ICS and LABA. By conceptualizing burden broadly to encompass health and economic outcomes, this study provided a more complete account of the unique burden experienced by patients with asthma and related comorbidities and demonstrated that asthma control has meaningful effects on both health and costs. This study contributed to the better understanding of the burden specifically associated with asthma in combination with asthma-related or allergic comorbidities and highlighted the need for improvement in disease management, including the assessment of treatment adherence and consideration of other treatment options.
There were some limitations to consider when making inferences about the results. Particularly, causal conclusions cannot be established from the findings, given this was a cross-sectional survey. Further, it was unknown if patients were being adherent to their prescribed treatment regimen. Thus, future research should examine treatment adherence in order to better understanding the factors contributing to the high prevalence of uncontrolled asthma among allergic asthma patients on ICS and LABA found in the current study. Recall bias may have also been introduced, due to the self-reported response format. Similarly, it was not possible to confirm the patient-reported responses regarding diagnosis and treatment, among other study variables. Certain psychiatric diagnoses such as depression maybe particularly susceptible to self-reporting bias and may be under-reported. To better control for these shortcomings, future research could supplement self-report measures with more objective sources of data (e.g., medical records) to validate participants’ responses.
The survey may selectively under-represent the asthma population, due to age-related limitations (e.g., very frail elderly patients are less likely to complete the survey) and limited access to the Internet (e.g., very low-income individuals and elderly asthma patients may not have computer access). While the NHWS was designed to mimic the demographic characteristics of the general adult population, whether it adequately represents the population of adults with asthma cannot be verified.
Conclusions
Among asthma patients on ICS and LABA, over half had very poorly or not well-controlled asthma. The substantial number of uncontrolled asthma patients demonstrates the importance of appropriate disease monitoring by the treating clinician, including assessment of treatment adherence and consideration for stepping up therapy. Findings revealed lower levels of asthma control (i.e., very poorly controlled and not well-controlled) were associated with worse HRQoL, greater work productivity loss, activity impairment, HRU, and indirect, direct, and total costs. Even a modest increment in asthma control was associated with less work productivity loss, HRU, and costs. Overall, the current study highlights a considerable unmet need in asthma patients on ICS and LABA and should help inform both healthcare practitioner and payor understanding and decision-making with regard to alleviating disease burden.
| Asthma Control | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Very poorly controlled asthma N = 563 | Not well controlled asthma N = 482 | Well controlled asthma N = 878 | |||||||
| Total N = 1923 | (a) | (b) | (c) | ||||||
| Mean / % | SD / N | Mean / % | SD / N | Mean / % | SD / N | Mean / % | SD / N | ap-value | |
| Demographics | |||||||||
| Age | 49.8 | 15.4 | 51.4b | 14.2 | 48.3a | 16.4 | 49.6 | 15.5 | 0.004 |
| Gender | |||||||||
| Female | 66.3% | 1274 | 65.9% | 371 | 66.0% | 318 | 66.6% | 585 | 0.949 |
| Male | 33.7% | 649 | 34.1% | 192 | 34.0% | 164 | 33.4% | 293 | |
| Race/Ethnicity | |||||||||
| Non-Hispanic white | 75.9% | 1459 | 71.0%c | 400 | 75.5% | 364 | 79.2%a | 695 | 0.016 |
| Non-Hispanic black | 10.1% | 195 | 11.2% | 63 | 9.8% | 47 | 9.7% | 85 | |
| Hispanic | 7.1% | 136 | 9.4%c | 53 | 6.8% | 33 | 5.7%a | 50 | |
| Other ethnicity | 6.9% | 133 | 8.3% | 47 | 7.9% | 38 | 5.5% | 48 | |
| Education | |||||||||
| <4-yr college degree | 60.3% | 1159 | 74.2%b.c | 418 | 62.0%a.c | 299 | 50.3%a,b | 442 | <0.001 |
| 4-yr college degree or higher | 39.7% | 764 | 25.8%b,c | 145 | 38.0%a,c | 183 | 49.7%a,b | 436 | |
| Annual Household Income | |||||||||
| < $25k | 23.9% | 460 | 32.5%b,c | 183 | 26.8%a | 129 | 16.9%a | 148 | <0.001 |
| $25k to $50k | 26.3% | 505 | 30.0%c | 169 | 26.1% | 126 | 23.9%a | 210 | |
| $50k to $75k | 19.6% | 377 | 15.8%c | 89 | 18.3% | 88 | 22.8%a | 200 | |
| $75k or more | 24.6% | 474 | 18.1%b,c | 102 | 23.9%a | 115 | 29.3%a | 257 | |
| Decline to Answer | 5.6% | 107 | 3.6%c | 20 | 5.0% | 24 | 7.2%a | 63 | |
| Marital Status | |||||||||
| Single | 42.9% | 825 | 47.2%c | 266 | 44.0% | 212 | 39.5%a | 347 | 0.013 |
| Married/living with partner | 57.1% | 1098 | 52.8%c | 297 | 56.0% | 270 | 60.5%a | 531 | |
| Has Health Insurance | 89.5% | 1722 | 85.8%c | 483 | 87.6%c | 422 | 93.1%a,b | 817 | <0.001 |
| Health Characteristics | |||||||||
| Charlson Comorbidity Index (CCI) | 0.98 | 1.5 | 1.5b,c | 1.8 | 0.88a,c | 1.3 | 0.67a,b | 1.2 | <0.001 |
| CCI categories: 0 | 51.5% | 991 | 34.3%b,c | 193 | 51.9%ac | 250 | 62.4%a,b | 548 | <0.001 |
| CCI categories: 1 | 24.3% | 468 | 27.4%c | 154 | 26.80% | 129 | 21.1%a | 185 | |
| CCI categories: 2 | 12.2% | 234 | 17.1%b,c | 96 | 11.6%a,c | 56 | 9.3%a,b | 82 | |
| CCI categories: 3+ | 12.0% | 230 | 21.3%b,c | 120 | 9.8%a,c | 47 | 7.2%a,b | 63 | |
| Allergic Comorbidities | |||||||||
| Chronic hives | 2.2% | 43 | 2.70% | 15 | 1.20% | 6 | 2.50% | 22 | 0.231 |
| Nasal allergies | 76.1% | 1464 | 77.10% | 434 | 75.70% | 365 | 75.70% | 665 | 0.819 |
| Atopic dermatitis | 2.9% | 55 | 3.00% | 17 | 2.70% | 13 | 2.80% | 25 | 0.952 |
| Dermatitis | 3.3% | 63 | 3.00% | 17 | 2.70% | 13 | 3.80% | 33 | 0.529 |
| Eczema | 14.9% | 286 | 14.20% | 80 | 15.40% | 74 | 15.00% | 132 | 0.86 |
| Skin allergies | 20.3% | 391 | 23.1%c | 130 | 22.20% | 107 | 17.5%c | 154 | 0.019 |
| Hayfever | 43.9% | 845 | 46.00% | 259 | 41.70% | 201 | 43.80% | 385 | 0.376 |
| Body Mass Index (BMI) Category | |||||||||
| BMI: underweight | 1.1% | 22 | 1.8% | 10 | 1.2% | 6 | 0.7% | 6 | <0.001 |
| BMI: normal weight | 21.5% | 414 | 16.9%c | 95 | 17.8%c | 86 | 26.5%a,b | 233 | |
| BMI: overweight | 25.2% | 484 | 24.0% | 135 | 27.0% | 130 | 24.9% | 219 | |
| BMI: obese | 49.0% | 943 | 53.5%c | 301 | 51.5% | 248 | 44.9%a | 394 | |
| BMI: unknown | 3.1% | 60 | 3.9% | 22 | 2.5% | 12 | 3.0% | 26 | |
| Smoking Status | |||||||||
| Current Smoker | 17.1% | 329 | 25.2%b,c | 142 | 18.7%a,c | 90 | 11.0%a,b | 97 | <0.001 |
| Former Smoker | 35.7% | 687 | 37.8% | 213 | 36.5% | 176 | 33.9% | 298 | |
| Never Smoker | 47.2% | 907 | 36.9%b,c | 208 | 44.8%a,c | 216 | 55.0%a,b | 483 | |
| Drinks Alcohol | 64.4% | 1239 | 55.8%b,c | 314 | 63.7%a,c | 307 | 70.4%a,b | 618 | <0.001 |
| Exercise 20+ minutes 1+ times in past month | |||||||||
| Exercise: 0 times | 40.6% | 781 | 50.1%b,c | 282 | 38.6%a | 186 | 35.6%a | 313 | <0.001 |
| Exercise: 1+ times | 59.4% | 1142 | 49.9%b,c | 281 | 61.4%a | 296 | 64.4%a | 565 | |
| Years of asthma diagnosis | 23.9 | 17.0 | 26.0b,c | 17.7 | 22.1a | 15.9 | 23.6a | 17.0 | 0.001 |
| Time of the year suffer from asthma | |||||||||
| In the spring | 88.8% | 1707 | 88.3% | 497 | 88.4% | 426 | 89.3% | 784 | 0.798 |
| In the summer | 82.5% | 1586 | 88.3%b,c | 497 | 82.0%a | 395 | 79.0%a | 694 | <0.001 |
| In the autumn | 85.1% | 1637 | 87.2% | 491 | 83.2% | 401 | 84.9% | 745 | 0.182 |
| In the winter | 80.3% | 1544 | 80.3% | 452 | 81.3% | 392 | 79.7% | 700 | 0.777 |
| Frequency of problems in the seasons when experience asthma | |||||||||
| Daily | 28.0% | 538 | 48.5%b,c | 273 | 24.5%a,c | 118 | 16.7%a,b | 147 | <0.001 |
| 4-6 times a week | 14.2% | 274 | 21.8%c | 123 | 18.0%c | 87 | 7.3%a,b | 64 | |
| 2-3 times a week | 20.0% | 384 | 19.7%b,c | 111 | 29.9%a,c | 144 | 14.7%a,b | 129 | |
| Once a week | 8.7% | 168 | 4.3%b,c | 24 | 9.3%a | 45 | 11.3%a | 99 | |
| 2-3 times a month | 14.1% | 271 | 4.4%b,c | 25 | 12.2%a,c | 59 | 21.3%a,b | 187 | |
| Once a month or less often | 15.0% | 288 | 1.2%b,c | 7 | 6.0%a,c | 29 | 28.7%a,b | 252 | |
| Main causes of asthma attacks | |||||||||
| Allergies/Allergic reaction | 77.6% | 1493 | 80.5% | 453 | 75.7% | 365 | 76.9% | 675 | 0.143 |
| Exercise/Physical activity | 49.6% | 954 | 57.2%c | 322 | 53.9%c | 260 | 42.4%a,b | 372 | <0.001 |
| Illnesses/Conditions (e.g. colds, flu) | 45.6% | 876 | 51.7%c | 291 | 44.6% | 215 | 42.1%a | 370 | 0.002 |
| Pollutants/Airborne irritants | 60.1% | 1155 | 69.8%b,c | 393 | 60.0%a | 289 | 53.9%a | 473 | <0.001 |
| Second hand smoke | 36.0% | 692 | 38.4% | 216 | 34.2% | 165 | 35.4% | 311 | 0.341 |
| Stress/Anxiety/Emotional states | 36.9% | 709 | 51.3%b,c | 289 | 38.6%a,c | 186 | 26.7%a,b | 234 | <0.001 |
| Weather/Temperature | 62.0% | 1192 | 74.1%b,c | 417 | 61.8%a,c | 298 | 54.3%a,b | 477 | <0.001 |
| None of these | 2.2% | 42 | 1.4% | 8 | 2.5% | 12 | 2.5% | 22 | 0.338 |
| Asthma control | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Very poorly controlled asthma N = 563 | Not well-controlled asthma N = 482 | Well-controlled asthma N = 878 | |||||||
| Total N = 1923 | (a) | (b) | (c) | ||||||
| Mean/% | SD/N | Mean/% | SD/N | Mean/% | SD/N | Mean/% | SD/N | ap-value | |
| Health-related quality of life | |||||||||
| SF-36/SF-12 summary scores | |||||||||
| SF-36 mental component summary score (MCS) | 45.9 | 11.7 | 41.8b,c | 12.2 | 45.2a,c | 11.1 | 48.8a,b | 10.9 | <0.001 |
| SF-36 physical component summary score (PCS) | 42.5 | 11.6 | 35.7b,c | 11.4 | 43.0a,c | 10.5 | 46.7a,b | 10.2 | <0.001 |
| SF-6D health utilities | 0.66 | 0.14 | 0.59b,c | 0.12 | 0.65a,c | 0.11 | 0.71a,b | 0.13 | <0.001 |
| SF-36/SF-12 norm-based domain scores | |||||||||
| Bodily pain | 43.1 | 11.3 | 37.5b,c | 10.9 | 43.3a,c | 10.5 | 46.7a,b | 10.4 | <0.001 |
| General health | 41.8 | 11.4 | 35.9b,c | 10.9 | 41.7a,c | 10.1 | 45.6a,b | 10.7 | <0.001 |
| Mental health | 45.5 | 11.4 | 41.8b,c | 12.0 | 44.9a,c | 10.3 | 48.3a,b | 10.9 | <0.001 |
| Physical functioning | 43.4 | 11.6 | 36.4b,c | 11.5 | 44.0a,c | 10.5 | 47.5a,b | 10.0 | <0.001 |
| Role emotional | 44.9 | 12.6 | 38.7b,c | 14.1 | 44.9a,c | 11.5 | 48.9a,b | 10.2 | <0.001 |
| Role physical | 43.6 | 11.8 | 36.2b,c | 11.5 | 43.7a,c | 10.7 | 48.2a,b | 10.0 | <0.001 |
| Social functioning | 44.3 | 12.0 | 37.8b,c | 12.0 | 44.2a,c | 11.2 | 48.5a,b | 10.4 | <0.001 |
| Vitality | 44.3 | 10.7 | 40.2b,c | 10.3 | 43.6a,c | 9.6 | 47.4a,b | 10.5 | <0.001 |
| Experience of disorders in past 12 months | |||||||||
| Any anxiety disorder | 37.7% | 725 | 44.4%c | 250 | 44.2%c | 213 | 29.8%a,b | 262 | <0.001 |
| Depression | 36.5% | 702 | 48.1%b,c | 271 | 40.0%a,c | 193 | 27.1%a,b | 238 | <0.001 |
| Sleep problems (insomnia, sleep difficulties) | 50.5% | 971 | 61.5%b,c | 346 | 53.9%a,c | 260 | 41.6%a,b | 365 | <0.001 |
| Work productivity and activity impairment | |||||||||
| Absenteeism %b | 6.3 | 16.9 | 12.3b,c | 22.4 | 6.1a | 17.2 | 3.7a | 12.9 | <0.001 |
| Presenteeism %c | 22.6 | 26.6 | 37.5b,c | 30.8 | 23.4a,c | 25.1 | 15.7a,b | 22.4 | <0.001 |
| Overall work impairment %b | 25.6 | 29.4 | 42.2b,c | 33.1 | 26.3a,c | 28.2 | 17.9a,b | 24.9 | <0.001 |
| Activity impairment % | 40.0 | 32.1 | 59.1b,c | 28.8 | 39.6a,c | 29.7 | 28.0a,b | 29.3 | <0.001 |
| Healthcare resource use in past 6 months | |||||||||
| Number of visits to any traditional healthcare provider | 8.1 | 10.2 | 9.8b,c | 12.6 | 8.1a | 10.2 | 7.0a | 8.2 | <0.001 |
| Number of ER visits | 0.48 | 1.3 | 0.79b,c | 1.7 | 0.41a | 1.3 | 0.32a | 1.0 | <0.001 |
| Number of hospitalizations | 0.21 | 0.72 | 0.39b,c | 1.1 | 0.17a,c | 0.52 | 0.10a,b | 0.41 | <0.001 |
| Annualized estimated costs ($) | |||||||||
| Total indirect costsb | 8716 | 12 533 | 15 374b,c | 16 926 | 8394a,c | 10 811 | 5974a,b | 9725 | <0.001 |
| Total direct costs | 29 122 | 47 301 | 41 631b,c | 67 375 | 27 222a,c | 40 265 | 22 143a,b | 30 886 | <0.001 |
| Total costd | 33 228 | 48 311 | 47 202b,c | 68 124 | 31 297a,c | 41 234 | 25 328a,b | 31 917 | <0.001 |
| 95% CI for OR | |||||
|---|---|---|---|---|---|
| OR | Lower | Upper | bp-value | Adjusted means | |
| Experience of any anxiety disorder | |||||
| Very poorly controlled | 1.8 | 1.4 | 2.3 | <0.001 | 41.9% |
| Not well-controlled | 1.7 | 1.3 | 2.2 | <0.001 | 41.4% |
| Well-controlleda | 29.1% | ||||
| Experience of depression | |||||
| Very poorly controlled | 1.9 | 1.5 | 2.5 | <0.001 | 43.1% |
| Not well-controlled | 1.5 | 1.2 | 1.9 | 0.001 | 37.3% |
| Well-controlleda | 28.2% | ||||
| Experience of any sleep problems (insomnia, sleep difficulties) | |||||
| Very poorly controlled | 1.7 | 1.3 | 2.2 | <0.001 | 58.0% |
| Not well-controlled | 1.5 | 1.2 | 1.9 | 0.001 | 54.2% |
| Well-controlleda | 44.6% | ||||
| Asthma control | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Very poorly controlled asthma N = 563 | Not well-controlled asthma N = 482 | Well-controlled asthma N = 878 | |||||||
| Total N = 1923 | (a) | (b) | (c) | ||||||
| % | N | % | N | % | N | % | N | ap-value | |
| Any ICS/LABA fixed-dose combination | 97.2 | 1870 | 97.5 | 549 | 96.7 | 466 | 97.4 | 855 | 0.676 |
| Fluticasone/salmeterol | 59.4 | 1142 | 57.0c | 321 | 53.1c | 256 | 64.4a,b | 565 | <0.001 |
| Budesonide/formoterol fumarate | 17.7 | 340 | 20.6c | 116 | 19.9c | 96 | 14.6a,b | 128 | 0.005 |
| Fluticasone propionate/salmeterol (HFA) | 15.1 | 290 | 14.9 | 84 | 16.6 | 80 | 14.4 | 126 | 0.537 |
| Mometasone furoate/formoterol fumarate dihydrate | 5.1 | 98 | 5.0 | 28 | 7.1 | 34 | 4.1 | 36 | 0.060 |
| Any LABA | 2.8 | 53 | 2.5 | 14 | 3.3 | 16 | 2.6 | 23 | 0.676 |
| Salmeterol | 1.8 | 35 | 1.6 | 9 | 2.3 | 11 | 1.7 | 15 | 0.673 |
| Formoterol fumarate | 0.9 | 18 | 0.9 | 5 | 1.0 | 5 | 0.9 | 8 | 0.964 |
| Any ICS | 2.8 | 53 | 2.5 | 14 | 3.3 | 16 | 2.6 | 23 | 0.676 |
| Mometasone | 0.8 | 15 | 0.9 | 5 | 0.8 | 4 | 0.7 | 6 | 0.902 |
| Fluticasone propionate | 0.8 | 15 | 0.5 | 3 | 0.8 | 4 | 0.9 | 8 | 0.721 |
| Beclometasone dipropionate | 0.4 | 7 | 0.4 | 2 | 0.4 | 2 | 0.3 | 3 | 0.976 |
| Flunisolide | 0.1 | 1 | 0.2 | 1 | 0.0 | 0 | 0.0 | 0 | 0.299 |
| Triamcinolone | 0.1 | 1 | 0.2 | 1 | 0.0 | 0 | 0.0 | 0 | 0.299 |
| Budesonide | 0.1 | 2 | 0.2 | 1 | 0.0 | 0 | 0.1 | 1 | 0.669 |
| Fluticasone propionate | 0.1 | 2 | 0.2 | 1 | 0.2 | 1 | 0.0 | 0 | 0.426 |
| Ciclesonide | 0.1 | 1 | 0.0 | 0 | 0.2 | 1 | 0.0 | 0 | 0.224 |
| Fluticasone propionate (HFA) | 0.5 | 9 | 0.0 | 0 | 0.8 | 4 | 0.6 | 5 | 0.123 |
| Any SABA | 75.2 | 1446 | 85.1b,c | 479 | 80.3a | 387 | 66.1a | 580 | <0.001 |
| Albuterol (generic) | 38.1 | 732 | 47.4b,c | 267 | 38.8a,c | 187 | 31.7a,b | 278 | <0.001 |
| Albuterol sulfate (branded; HFA) | 20.4 | 393 | 25.0c | 141 | 23.0c | 111 | 16.1a,b | 141 | <0.001 |
| Albuterol (branded) | 8.5 | 164 | 10.5 | 59 | 8.1 | 39 | 7.5 | 66 | 0.134 |
| Ipratropium bromide | 5.1 | 99 | 9.1c | 51 | 5.6c | 27 | 2.4a,b | 21 | <0.001 |
| Albuterol sulfate | 6.6 | 127 | 9.1c | 51 | 7.5 | 36 | 4.6a | 40 | 0.002 |
| Albuterol sulfate (branded; HFA) | 5.9 | 114 | 8.3c | 47 | 6.6 | 32 | 4.0a | 35 | 0.002 |
| Albuterol (HFA) | 4.5 | 87 | 4.4 | 25 | 3.7 | 18 | 5.0 | 44 | 0.552 |
| Levalbuterol | 3.7 | 71 | 6.6b,c | 37 | 2.9a | 14 | 2.3a | 20 | <0.001 |
| Pirbuterol | 1.2 | 23 | 1.1 | 6 | 1.5 | 7 | 1.1 | 10 | 0.830 |
| Any leukotriene antagonists | 25.1 | 483 | 29.3c | 165 | 23.9 | 115 | 23.1a | 203 | 0.023 |
| Montelukast | 24.0 | 461 | 27.7 | 156 | 22.6 | 109 | 22.3 | 196 | 0.047 |
| Zafirlukast | 1.1 | 22 | 1.6 | 9 | 1.2 | 6 | 0.8 | 7 | 0.367 |
| Any anticholinergic | 8.7 | 168 | 16.9b,c | 95 | 8.5a,c | 41 | 3.6a,b | 32 | <0.001 |
| Tiotropium | 6.8 | 131 | 13.1b,c | 74 | 6.6a,c | 32 | 2.8a,b | 25 | <0.001 |
| Ipratropium bromide | 2.2 | 43 | 4.6c | 26 | 2.1 | 10 | 0.8a | 7 | <0.001 |
| Theophylline | 1.4 | 27 | 2.0 | 11 | 1.7 | 8 | 0.9 | 8 | 0.223 |
| Anti-Ige | |||||||||
| Omalizumab | 1.1 | 21 | 1.6 | 9 | 0.8 | 4 | 0.9 | 8 | 0.385 |
| Oral corticosteroids | |||||||||
| Prednisone | 0.4 | 7 | 0.9 | 5 | 0.0 | 0 | 0.2 | 2 | 0.039 |
| Prednisolone | 0.0 | 0 | 0.0 | 0 | 0.0 | 0 | 0.0 | 0 | N/A |
Acknowledgements
The authors acknowledge the literature review and editing assistance of Martine C. Maculaitis, PhD, on behalf of Kantar Health. Poster presentations of sections of this work were given at the Academy of Managed Care Pharmacy (AMCP) Annual Meeting on April 19–22, 2016 and American Thoracic Society (ATS) Conference on May 13–18, 2016. A poster presentation of a third section of this work was also given at the American College of Chest Physicians (CHEST) Annual Meeting on October 22–26, 2016.
Declaration of interests
The study reported on this manuscript was funded by Novartis Pharmaceuticals Corporation. Lulu Lee is an employee of Kantar Health, who received funding from Novartis to conduct and report on this study. Engels Obi, Brandee Paknis, and Abhishek Kavati are employees of Novartis, who funded this study. Bradley Chipps serves as a consultant for Novartis, AstraZeneca, Boehringer Ingelheim, Circassia, Genetech, and Teva.
Appendix
Table A1. Asthma treatment by asthma control among diagnosed asthma patients with asthma-related/allergic comorbidities on ICS and LABA combination therapy.