FormalPara Key Summary Points

Why carry out this study?

Although some patients with chronic obstructive pulmonary disease (COPD) may benefit from regimens that include inhaled corticosteroids (ICS), ICS-containing treatments are often overprescribed and can increase the risk of pneumonia.

To investigate whether combination therapy with the long-acting muscarinic antagonist (LAMA)/long-acting β2-agonist (LABA) tiotropium/olodaterol could provide better clinical outcomes than LABA/ICS, this large non-interventional database study assessed the risk of COPD exacerbations, pneumonia, and escalation to triple therapy in patients with COPD who initiated maintenance therapy with either combination.

What was learned from the study?

Overall, tiotropium/olodaterol was associated with a lower risk of COPD exacerbations, pneumonia, and escalation to triple therapy versus LABA/ICS, both individually and as a combined risk of any one of these events occurring; the combined risk was reduced versus LABA/ICS irrespective of baseline eosinophil count or exacerbation history.

These results support and expand on those from previous randomized controlled trials that report a lower risk of exacerbations in subsets of patients with COPD treated with LAMA/LABA versus LABA/ICS.

Our findings highlight the important role of LAMA/LABA in the management of COPD and implicate it as a strong alternative to LABA/ICS to avoid ICS overuse and reduce exacerbations in patients with COPD.

Plain Language Summary

There are several different inhaled medicines for people with chronic obstructive pulmonary disease (COPD), including long-acting muscarinic antagonists (LAMAs), long-acting β2-agonists (LABAs), and inhaled corticosteroids (ICS). These medicines can be prescribed on their own or together. The most suitable choice of medicine depends on the person. For example, some people may benefit from treatments that include ICS, such as those who also have asthma or high levels of eosinophils—a type of white blood cell that can indicate inflammation in the body. However, ICS can also increase the risk of pneumonia. In this study, we compared treatment with a LAMA/LABA combination (called tiotropium/olodaterol) against LABA/ICS. We looked at which was more effective, also taking into account the risk of side effects. Using a large US database, we looked at information from over 40,000 patients with COPD who started treatment with either tiotropium/olodaterol or LABA/ICS. Overall, tiotropium/olodaterol reduced the risk of experiencing a worsening of symptoms (known as an exacerbation) compared with LABA/ICS. Tiotropium/olodaterol was also associated with a lower risk of pneumonia and the need to increase (or “escalate”) treatment to triple therapy with LAMA/LABA/ICS versus LABA/ICS. Looking at exacerbations, pneumonia, and escalation to triple therapy together, the risk of any one of these events was also reduced. This was true for patients with low or high eosinophil levels and for those who had a low or high number of exacerbations before starting treatment. Overall, these results suggest that a LAMA/LABA combination like tiotropium/olodaterol is a strong alternative to LABA/ICS.

Digital Features

This article is published with digital features, including a graphical abstract, summary slide, and plain language summary, to facilitate understanding of the article. To view digital features for this article go to https://doi.org/10.6084/m9.figshare.13677457.

Introduction

Multiple therapies can be used to treat chronic obstructive pulmonary disease (COPD) as long-term therapy, including long-acting bronchodilators (long-acting muscarinic antagonists [LAMAs] and/or long-acting β2-agonists [LABAs]) and inhaled corticosteroids (ICS). The Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2020 strategy report recommends a stepwise approach to pharmacologic treatment, starting with a LAMA or LABA for most patients with COPD, and escalating to dual bronchodilation as the next step [1]. However, many patients remain symptomatic with monotherapy [2,3,4]. Recent guidelines from the American Thoracic Society go further, strongly recommending dual LAMA/LABA therapy over LAMA or LABA monotherapy for patients with COPD with dyspnea or exercise intolerance [5].

The optimal choice of treatment for COPD may depend on several factors. For instance, some patients, such as those with high eosinophil levels, an increased exacerbation risk, or a history of asthma, may benefit from regimens that include ICS [1, 5, 6]. The GOLD 2020 strategy report recommends that initial therapy with LABA/ICS is limited to patients with eosinophil levels of at least 300 cells/µL who are at high risk of exacerbations and are symptomatic [1].

Despite these recommendations, ICS are sometimes overprescribed and may be appropriate only in a subset of users [7,8,9,10,11,12,13,14,15]. In particular, ICS overuse is common in the USA, as shown by a cross-sectional study in the US Veterans Affairs system in which a quarter of patients with COPD without an identifiable indication for ICS had filled two or more prescriptions for ICS in the past year [16]. Notably, ICS use is widely associated with an increased risk of pneumonia and related use of healthcare resources and cost [8, 17,18,19,20], making it particularly important that ICS are prescribed appropriately. For example, triple therapy with LAMA/LABA/ICS has been associated with an increased risk of pneumonia and higher costs compared with dual therapy with LAMA/LABA combinations [5, 21]. Hence, escalation to triple therapy should be reserved for a specific population of patients with COPD, namely symptomatic patients with a high risk of exacerbations, in whom the treatment benefits may outweigh the risks [5, 22, 23].

Given the increasing evidence around the need for more tailored use of ICS in people with COPD [1, 5, 6, 24], it is important to investigate if non-ICS-containing regimens can provide the same or better clinical outcomes without the added risk of side effects associated with ICS. Treatment with LAMA/LABA combinations, including tiotropium/olodaterol, has been shown to significantly improve lung function versus LABA/ICS combinations [3, 25,26,27]. This non-interventional database study therefore aimed to assess the risk of COPD exacerbations, pneumonia, and escalation to triple therapy, both individually and as a combined risk of any one of these events occurring, in patients with COPD who initiated maintenance therapy with tiotropium/olodaterol versus any LABA/ICS combination.

Methods

Data Sources

This study (NCT04138758) was conducted using data from the HealthCore Integrated Research Database (HIRD; January 2013–March 2019), a large administrative healthcare claims database maintained by HealthCore that has previously been used for the study of numerous diseases, including COPD [28,29,30,31,32]. The study was conducted using administrative claims in the form of a limited data set pursuant to data agreements between HealthCore and participating health plans in compliance with the US Health Insurance Portability and Accountability Act. The study did not require ethics committee approval, nor were subjects required to provide informed consent.

HIRD includes longitudinal medical and pharmacy claims data for health plan members residing in the USA. Member enrollment, medical care (professional and facility claims), outpatient prescription drug use, outpatient laboratory test result data, and healthcare utilization may be tracked for health plan members in the database dating back to January 2006, and with diagnoses recorded in International Classification of Diseases, version 10 (ICD-10) since October 2015. Laboratory data are available for those tests that have been performed using two large, national reference laboratories (Quest and LabCorp).

Study Population and Design

Patients in each cohort were required to have at least one prescription for a fixed-dose combination (FDC) inhaler of tiotropium/olodaterol or LABA/ICS between 1 January 2013 and 31 March 2019, with the first prescription defined as the index date. Patients were also required to have at least one diagnosis of COPD at any time prior to the index date and at least 1 year of continuous health plan eligibility prior to the index date. Patients were excluded for the following reasons: age less than 40 years at the index date; diagnosis of asthma in the year prior to the index date, lung cancer, interstitial lung disease, or lung transplant; or pre-index use of tiotropium/olodaterol, LABA/ICS, or LABA/LAMA/ICS in free or fixed form.

Study Subpopulations

All analyses were conducted for the total population, as well as in subgroups of patients at high or low risk of exacerbation based on (1) previous history of exacerbations in the year preceding cohort entry (low exacerbation history, 0 hospitalizations and 0–1 outpatient exacerbations; high exacerbation history, at least 1 hospitalization or at least 2 outpatient exacerbations); and (2) blood eosinophil count (baseline eosinophils less than 300 cells/µL; baseline eosinophils at least 300 cells/µL) as identified on the basis of the laboratory result value that was closest but prior to the index date (within 6 months).

Exposure Measures

Exposure measures were based on pharmacy dispensation of tiotropium/olodaterol and LABA/ICS over up to 1 year of follow-up. Exposure was based on the current use of tiotropium/olodaterol or LABA/ICS as defined by the dispensation date plus the day’s supply recorded at the time of dispensation.

A gap of 15 days was allowed between dispensations to allow for delays in obtaining medication refills and for continued use beyond the days supplied where medication has been missed because of imperfect adherence. The 15-day gap was increased in the sensitivity analyses to 30 and 60 days.

Patients were followed from the index date until the earliest of the following: first occurrence of a COPD exacerbation, community-acquired pneumonia, escalation to triple therapy, switch in treatment, discontinuation of COPD treatment, the end of the study period, the end of continuous health plan eligibility, or (for the main analyses) 1 year after cohort entry. The treatment segment ended at the earliest of the following events: (1) 15 days after the end of the observed day’s supply for the medication received on the index date without a subsequent dispensing of COPD medication (i.e., discontinuation); (2) initiation of triple therapy; or (3) any other change in use of study medication, including to a different combination therapy, from a fixed-form to a free-form combination therapy, or from combination therapy to monotherapy.

Study Outcomes

The primary outcome was to compare the risk of first COPD exacerbation after initiating treatment with tiotropium/olodaterol versus LABA/ICS. COPD exacerbation was defined as a COPD-related hospitalization (severe exacerbation) or an emergency department visit for COPD and/or prescription of an antibiotic for a respiratory condition the same day as an oral corticosteroid (moderate exacerbation). To increase the specificity of our definition of exacerbation, the prescription of either an antibiotic or oral corticosteroid alone was not included. Severe and moderate exacerbations were considered as a composite for the main analyses.

Secondary outcomes included the risk of hospitalization for community-acquired pneumonia, risk of pharmacy dispensation indicating escalation to triple therapy, and risk of any of the following events occurring: COPD exacerbation, hospitalization for pneumonia, or escalation to triple therapy. Pneumonia was defined as hospitalization with one of the following diagnostic codes: 481.x–486.x, 487.0, 507.x, 507.0, 507.1, 507.8, 510.0, 510.9, 511.0, 513.0, 514.x, 517.1, 519.8, 530.84 (as per International Classification of Diseases, version 9, Clinical Modification [ICD-9-CM]; used up to October 2015) and J10.0, J11.0, J12–J18, J22, J69, J85.0, J85.1, and J86 (ICD-10; used from October 2015 onwards) (pneumonia without hospitalization was not captured). These definitions from hospitalization data have been used in several previous studies in patients with COPD [33] and are detailed in Table S1 in the supplementary material. As a result of the nature of the ICD diagnostic codes (e.g., J10.0: “influenza due to other identified influenza virus with pneumonia” [ICD-10]), it was not possible to verify pneumonia as the primary driver of hospitalization. Escalation to triple therapy was defined as any combination of pharmacy dispensations resulting in the overlapping use of a LAMA, LABA, and ICS through any fixed or free combination for at least 1 day.

Statistical Analysis

Outcome rates are described using incidence rates (IRs) and associated 95% confidence intervals (CIs). Differences in the risk of a first COPD exacerbation, pneumonia, escalation to triple therapy, or any one of these events (COPD exacerbation, or pneumonia, or escalation to triple therapy) for tiotropium/olodaterol use relative to LABA/ICS use were assessed using hazard ratios (HRs) and associated 95% CIs. Cox proportional hazard models were performed as an as-treated analysis to derive HRs; a first treatment carry-on analysis was conducted as a sensitivity analysis. Confounding was controlled via fine stratification and reweighting of an exposure propensity score and by including potential confounding factors in the models for variables that remained imbalanced after reweighting. Data trimming was applied on the basis of the overlap of propensity score distributions [34].

Data were analyzed separately and for subgroups based on blood eosinophil levels (for those with available results) and exacerbation history using the same approach. Calculation of propensity scores and fine stratification and reweighting were repeated within the patient subgroups to create weighted populations.

Propensity scores including both pre-specified and high-dimensional, data-derived variables were calculated; scores were calculated on the basis of the following covariates: sex, age, calendar year of cohort entry, season of index date, US census region of residence, insurance type, medication history, exacerbation history and pre-specified chronic comorbidities, in addition to the most frequently occurring diagnoses, medications, and procedures observed. Adjustment for the propensity score used the fine stratification method to create a weighted pseudo-population [34]. Ten strata were defined on the basis of the distribution of the propensity score within tiotropium/olodaterol-exposed individuals for the region of overlap with the LABA/ICS group, and stratum-specific weights were applied to create balance between the exposure groups. The balance of patient characteristics between the treatment groups was described before and after propensity score application. The Cox proportional hazard model was further adjusted for patient characteristics found to be imbalanced after application of the propensity score, where imbalance was defined as standardized differences greater than 10%. Several sensitivity analyses and bias analyses were performed, as described in the supplementary methods.

Results

Study Cohort

Data were available from 237,328 patients with a diagnosis of COPD who had received at least one dose of study medication from 1 January 2013 to 31 March 2019 (Fig. 1). After removing patients who did not meet the inclusion/exclusion criteria, there were 2684 tiotropium/olodaterol and 59,301 LABA/ICS users remaining. Of those, 361 tiotropium/olodaterol users (13.5%) and 6586 LABA/ICS users (11.1%) had at least one eosinophil test result recorded within 6 months prior to the index date. Based on pre- and post-index health plan eligibility data, the median time of enrollment was 61.1 months for tiotropium/olodaterol and 60.3 months for LABA/ICS. In the LABA/ICS group, the mean (± standard deviation) dose of ICS was 254.51 ± 242.31 µg and the majority of users (n = 56,558; 95.4%) received LABA/ICS via a single inhaler.

Fig. 1
figure 1

Formation of the study cohort. Percentage values show the proportion of individuals lost from the study cohort at each step compared with those with at least one COPD diagnosis (n = 237,328 [100%]). COPD chronic obstructive pulmonary disease, ICS inhaled corticosteroids, LABA long-acting β2-agonist, LAMA long-acting muscarinic antagonist, Olo olodaterol, Tio tiotropium

Full size image

Baseline Characteristics

Baseline characteristics before and after reweighting are shown in Table 1. Prior to reweighting, the mean age of patients was 64.8 and 65.0 years for tiotropium/olodaterol and LABA/ICS, respectively; 45.6% of tiotropium/olodaterol users and 49.7% of LABA/ICS users were female. The majority of patients had not received previous maintenance treatments for COPD (tiotropium/olodaterol group, 69.2%; LABA/ICS group, 82.0%). The most common previous COPD treatment was LAMA monotherapy (tiotropium/olodaterol, 22.4%; LABA/ICS, 12.9%). More patients in the tiotropium/olodaterol group reported previous LAMA/LABA combination therapy versus those in the LABA/ICS group (6.4% versus 1.1%).

Table 1 Baseline patient demographics in patients with COPD receiving tiotropium/olodaterol versus LABA/ICS
Full size table

After reweighting for stratified propensity scores, the total pseudo-population consisted of 42,953 patients (tiotropium/olodaterol, 2600; LABA/ICS, 40,353). Overall, after reweighting, baseline characteristics were balanced between users of tiotropium/olodaterol and LABA/ICS, as demonstrated by low values of standardized difference (Table 1). There was still some imbalance in prior exacerbation history between tiotropium/olodaterol and LABA/ICS users, particularly for severe exacerbations (tiotropium/olodaterol versus LABA/ICS: one severe exacerbation in year prior to cohort entry, 13.2% vs. 18.3%; at least two exacerbations, 2.7% vs. 4.7%). To account for this imbalance, severe exacerbations were adjusted for in the model.

Risk of COPD Exacerbations

Tiotropium/olodaterol users had a lower adjusted IR (aIR [95% CIs]) per 100 person-years of exacerbations (59.53 [53.68, 66.03]) than LABA/ICS users (88.67 [86.45, 90.94]) (see Table 2 for adjusted and unadjusted data). The risk of exacerbations was also lower with tiotropium/olodaterol compared with LABA/ICS (adjusted HR [aHR] 0.76 [95% CI 0.68, 0.85]) (Fig. 2).

Table 2 Incidence rate of exacerbation, pneumonia, escalation to triple therapy, or a combination of these events (exacerbation, or pneumonia, or escalation to triple therapy) in patients receiving tiotropium/olodaterol versus LABA/ICS
Full size table
Fig. 2
figure 2

Risk of exacerbation, pneumonia, escalation to triple therapy, or a combination of these events (exacerbation, or pneumonia, or escalation to triple therapy) in patients receiving tiotropium/olodaterol versus LABA/ICS. Hazard ratios were derived using Cox proportional hazard models. The Cox proportional hazard model was further adjusted for patient characteristics found to be imbalanced after application of the propensity score, where imbalance was defined as standardized differences greater than 10%. aHR adjusted hazard ratio, CI confidence interval, COPD chronic obstructive pulmonary disease, ICS inhaled corticosteroids, LABA long-acting β2-agonist, Olo olodaterol, Tio tiotropium

Full size image

Similar results were observed for patients stratified by baseline exacerbation frequency and for patients with baseline eosinophil levels less than 300 cells/µL (Table 2 and Fig. 3). Notably, the observation that tiotropium/olodaterol users had a lower risk of exacerbations than LABA/ICS users tended to be more pronounced in patients with baseline eosinophils less than 300 cells/µL (aHR 0.54 [95% CI 0.36, 0.82]) versus patients with at least 300 cells/µL (aHR 0.83 [0.50, 1.38]).

Fig. 3
figure 3

Subgroup analyses for the risk of exacerbations in patients receiving tiotropium/olodaterol versus LABA/ICS. Hazard ratios were derived using Cox proportional hazard models. The Cox proportional hazard model was further adjusted for patient characteristics found to be imbalanced after application of the propensity score, where imbalance was defined as standardized differences greater than 10%. aHR adjusted hazard ratio, CI confidence interval, ICS inhaled corticosteroids, LABA long-acting β2-agonist, Olo olodaterol, Tio tiotropium

Full size image

Across sensitivity analyses, the aHR estimates varied from 0.69 to 0.91 (see Table S2 in the supplementary material). This was generally consistent across the subgroup analyses. For the bias analysis, only minimal changes in the aHR were observed when correcting for the parameters shown in Table S3 in the supplementary material.

Risk of Pneumonia

Tiotropium/olodaterol users had a lower aIR (95%CI) per 100 person-years of hospitalization for pneumonia (8.57 [6.61, 11.10]) than LABA/ICS users (12.54 [11.76, 13.36]) (see Table 2 for adjusted and unadjusted data). The risk of pneumonia was also lower with tiotropium/olodaterol compared with LABA/ICS (aHR 0.74 [95% CI 0.57, 0.97]) (Fig. 2).

Risk of Escalation to Triple Therapy

Tiotropium/olodaterol users had a lower aIR (95% CI) per 100 person-years of escalation to triple therapy (20.81 [17.53, 24.47]) than LABA/ICS users (114.79 [112.50, 117.25]) (see Table 2 for adjusted and unadjusted data). The risk of being escalated to triple therapy was also lower with tiotropium/olodaterol compared with LABA/ICS (aHR 0.22 [95% CI 0.19, 0.26]) (Fig. 2).

Combined Risk of COPD Exacerbation, or Pneumonia, or Escalation to Triple Therapy

When looking at the combined risk of any one of a COPD exacerbation, pneumonia, or escalation to triple therapy, tiotropium/olodaterol users had a lower aIR (95% CI) per 100 person-years (79.04 [72.32, 86.56]) than LABA/ICS users (207.21 [203.81, 210.75]) (see Table 2 for adjusted and unadjusted data). The risk of a COPD exacerbation, or pneumonia, or escalation to triple therapy was also lower with tiotropium/olodaterol compared with LABA/ICS (aHR 0.45 [95% CI 0.41, 0.49]; Fig. 2).

Similar results were observed for patients stratified by baseline exacerbation frequency and eosinophil levels (Table 2 and Fig. 4). For eosinophils, there was no notable difference between tiotropium/olodaterol and LABA/ICS in patients with fewer than 300 cells/µL (aHR 0.39 [95% CI 0.29, 0.53]) versus patients with at least 300 cells/µL (aHR 0.54 [0.34, 0.86]).

Fig. 4
figure 4

Subgroup analyses for the risk of exacerbation or pneumonia or escalation to triple therapy in patients receiving tiotropium/olodaterol versus LABA/ICS. Hazard ratios were derived using Cox proportional hazard models. The Cox proportional hazard model was further adjusted for patient characteristics found to be imbalanced after application of the propensity score, where imbalance was defined as standardized differences greater than 10%. aHR adjusted hazard ratio, CI confidence interval, ICS inhaled corticosteroids, LABA long-acting β2-agonist, Olo olodaterol, Tio tiotropium

Full size image

Discussion

This non-interventional, real-world study showed that, in patients who initiated maintenance therapy with tiotropium/olodaterol versus LABA/ICS, there was a reduction in the risk of COPD exacerbations, community-acquired pneumonia and escalation to triple therapy, as well as a 54% reduction in the combined risk of any one of these events. The combined measure provides a useful and clinically relevant comparison of the two treatment options, given that the prescription of a treatment may be based on avoiding a number of possible events.

These findings are consistent with a growing body of evidence reporting that LAMA/LABA combination therapy is associated with a comparable or lower risk of exacerbations compared with LABA/ICS, while ICS use is linked with an increased risk of pneumonia [35,36,37,38]. For example, the FLAME trial reported that patients with COPD and a history of exacerbation receiving LAMA/LABA (indacaterol/glycopyrronium) had fewer exacerbations and a lower incidence of pneumonia than those receiving LABA/ICS (salmeterol/fluticasone) over 1 year of follow-up [35]. Similarly, in LANTERN, patients with moderate-to-severe COPD and a history of at least one exacerbation in the previous year who received indacaterol/glycopyrronium had a significantly lower rate of moderate-to-severe exacerbations and a threefold lower incidence of pneumonia compared with those receiving salmeterol/fluticasone [37]. In AFFIRM COPD, there was a comparable reduction in exacerbation risk, but a lower incidence of pneumonia, with LAMA/LABA (aclidinium/formoterol) versus salmeterol/fluticasone in patients with stable, moderate-to-severe COPD [38]. Additionally, several large observational studies have shown no benefit of LABA/ICS over LAMA or LAMA/LABA in terms of reducing exacerbation risk [39,40,41]. For example, in a large retrospective study in the USA evaluating 5384 patients on LAMA/LABA and 473,388 on LABA/ICS, both treatments were similarly effective in terms of exacerbation rates [39]. Similarly, in a recent real-world study comparing 1977 patients with COPD initiating LAMA/LABA with 1977 initiating LABA/ICS, LAMA/LABA was as effective as LABA/ICS in preventing exacerbations, but with a lower incidence of severe pneumonia [40]. Also, in a population-based observational study of 12,366 patients with COPD initiating LAMA and 12,366 initiating LABA/ICS, patients with low blood eosinophil levels (at most 300 cells/µL) experienced no difference in exacerbation risk with either treatment [41]. In contrast, in both the IMPACT and ETHOS trials, which compare triple LAMA/LABA/ICS therapy with LABA/ICS and LAMA/LABA, a lower rate of moderate or severe COPD exacerbations was observed with LABA/ICS versus LAMA/LABA [42, 43]. However, it is worth noting that these patients were more severe in terms of exacerbation history and lung function than those in the current study [42, 43]. Our results confirm that LAMA/LABA is the preferred option as initial therapy in patients with infrequent exacerbations.

There is also evidence to support the safe withdrawal of ICS where appropriate [24]. Recent guidelines from the European Respiratory Society (ERS) conditionally recommend withdrawing ICS in patients with COPD without a history of frequent exacerbations and strongly recommend treatment with one or two long-acting bronchodilators if ICS are withdrawn [24]. In support of this, in patients from the WISDOM trial with COPD and a history of exacerbations who received LAMA/LABA/ICS triple therapy (tiotropium/salmeterol/fluticasone propionate), the risk of moderate or severe exacerbations was similar among those discontinuing versus continuing ICS [36]. Notably, the ERS guidelines strongly recommend not withdrawing ICS in patients with blood eosinophil counts of 300 cells/µL or higher [24].

Given that LABA/ICS are currently recommended for symptomatic patients with higher exacerbation risk [1], lower exacerbation rates may be anticipated in LAMA/LABA versus LABA/ICS users, which has the potential to confound any comparison of these treatment regimens. However, the reduction in exacerbation risk with tiotropium/olodaterol versus LABA/ICS in our study occurred irrespective of baseline exacerbation history, indicating reductions in patients with both low and high exacerbation history. Similarly, prior therapy did not influence our results, as shown by a sensitivity analysis that included only those patients who had not received prior monotherapy; in this analysis, only marginal variations in the risk estimates for COPD exacerbations were observed overall and in the subgroup analyses by exacerbation history, suggesting that prior monotherapy did not modify the observations. Lastly, eosinophil count has the potential to influence exacerbation outcomes; however, our study showed that patients benefit from initiating therapy with tiotropium/olodaterol compared with LABA/ICS, irrespective of baseline eosinophil count. The reduction in COPD exacerbation risk with tiotropium/olodaterol versus LABA/ICS was more pronounced in patients with low baseline eosinophil levels (less than 300 cells/µL) compared with those with high baseline eosinophils (at least 300 cells/µL). Of note, a previous UK observational study reported no difference in exacerbation risk with LABA/ICS versus LAMA initiation in patients with COPD and low baseline eosinophil levels; however, a reduction in risk with LABA/ICS versus LAMA was observed in patients with high baseline eosinophils [41].

The increased risk of pneumonia with LABA/ICS versus tiotropium/olodaterol in our study is consistent with findings from a previous retrospective claims analysis, which showed that pneumonia risk was higher for ICS users versus non-users in Medicare patients with COPD in the USA [44]. An increased risk of pneumonia with LABA/ICS versus LAMA/LABA or LAMA was also observed in two previous real-world, clinical practice, observational studies [40, 41]. It is worth noting that the incidence rate of pneumonia was slightly higher in our study compared with previous studies [41, 45]; however, it was similar to that reported in a recent observational study by Suissa et al. [40]. Large variations in pneumonia rates in patients with COPD have been reported, for example between countries [46], which therefore makes it difficult to compare incidence rates between studies.

For escalation to triple therapy, our observation of a reduction in risk with tiotropium/olodaterol versus LABA/ICS is in line with results from a previous retrospective US health insurer database study showing that patients with COPD initiating LAMA/LABA (umeclidinium/vilanterol) had a longer time before escalation to multiple-inhaler triple therapy (and higher adherence) than those initiating fluticasone propionate/salmeterol [47].

A key strength of this study is the large number of patients (n > 40,000) in the main cohort and that only new initiators of tiotropium/olodaterol and LABA/ICS combinations were identified, allowing the study to reflect treatment outcomes relating to initial usage. Despite the substantial difference in the numbers of tiotropium/olodaterol and LABA/ICS users, cohorts were balanced after fine stratification, reweighting, and trimming by exposure propensity score, and large numbers of patients could be retained in the analysis. After reweighting, an imbalance in prior history of severe exacerbations remained between tiotropium/olodaterol and LABA/ICS users; however, this was accounted for by adjusting the model for severe exacerbations. To control for any confounding by treatment group, we used fine stratification and reweighting by exposure propensity score. However, this was only possible for measured covariates, and therefore an impact of residual confounding by unmeasured confounders cannot be ruled out. For instance, claims data are unlikely to capture any lifestyle factors that are less critical to insurance billing, as assessed in the bias analysis. In addition, eosinophil test results were only available in a subset of users, who may not be representative of the total study population. A further limitation is the lack of lung function data, which could have introduced unmeasured confounding by not allowing accurate reweighting by COPD disease severity. The model was, however, adjusted for severe exacerbations.

In addition, the definition used for moderate exacerbations did not include the prescription of either an antibiotic or an oral corticosteroid alone, and therefore under-reporting of COPD exacerbations was possible. However, it is more likely that only true exacerbation events were captured. A further limitation was that it was not possible to verify pneumonia as the primary driver of hospitalization because of the nature of the ICD diagnostic codes (e.g., J10.0: “influenza due to other identified influenza virus with pneumonia” [ICD-10]). The inclusion criteria present another potential limitation, since patients were included if they had at least one prescription of an FDC inhaler; those with only one prescription have an increased possibility of non-use or discontinuation when compared with multiple prescriptions. The prescriptions dispensed by a pharmacy but not taken by patients could lead to misclassification of exposure. Pharmacy dispensation data should more closely reflect patient use than physician prescribing given that the patient has taken the effort to obtain the medication. Quantitative bias analyses were used to formally describe the extent to which some of these issues are present. The analyses suggest that the impact of residual confounding due to unmeasured obesity or smoking status did not have a meaningful impact on study results. It is worth noting, however, that bias analysis parameters are informed by literature and clinical expert opinion; hence, any analysis is limited by such assumptions.

There is an inherent limitation in that the study population is limited to those with health insurance and hence the results may not be generalizable to uninsured patients with COPD. However, the sensitivity analyses that were conducted suggest that there was a relatively small variation in the risk estimates for COPD exacerbations overall and across the subgroup analyses. A further limitation is the possibility of information bias due to misclassification of outcomes, exposure, or missing data. In addition, for the subgroup analyses of COPD exacerbations and pneumonia, the relatively small sample sizes did not allow for strong conclusions to be drawn.

Conclusion

This analysis shows that treatment with tiotropium/olodaterol is associated with a lower risk of COPD exacerbations, pneumonia, and escalation to triple therapy, both individually and as a combined risk of any one of these events occurring, versus LABA/ICS in patients with COPD. Our findings support and expand on those from previous randomized controlled trials reporting a lower risk of exacerbations in subsets of patients with COPD treated with LAMA/LABA versus LABA/ICS.

These results highlight the important role of LAMA/LABA in the management of COPD, and implicate it as a strong alternative to LABA/ICS to avoid ICS overuse and reduce exacerbations in patients with COPD.