Clinical characteristics, use and switch of drugs for obstructive airway diseases among patients with COPD experiencing an exacerbation: a retrospective analysis of Italian administrative healthcare data

Abstract

Background

Chronic obstructive pulmonary disease (COPD) represents an important health challenge, despite being preventable and manageable thanks to up-to-date recommendations. In Italy, the pharmaceutical care of COPD patients is still ill-timed and inaccurate. This study aimed to describe the treatment of COPD patients in Italy and possible switches following an exacerbation.

Methods

This observational retrospective analysis of Italian administrative healthcare data from the Fondazione Ricerca e Salute (ReS) database identified patients aged ≥ 45 years with COPD in 2019 and 2020. At least 6 years of look-back period and absence of concomitant asthma were required. COPD patients were categorized by treatment (SI–single/MI–multiple inhalers, TT–triple therapy, DT–dual therapy, other respiratory treatments, untreated) at index date (first dispensation during accrual period). Occurrence of moderate/severe exacerbation during one-year preceding index date and treatments during one-year preceding the exacerbation (possible switch) were evaluated.

Results

From ~ 4.7 million beneficiaries of the Italian National Health Service in 2019 and 2020, respectively, 105,828 and 103,729 (43 and 41 × 1,000 inhabitants aged ≥ 45 years) were identified as having COPD. Of 2019/2020 patients: 3.4%/5.2% received SI-TT, 20.7%/17.5% MI-TT, 35.9%/38.1% DT, 33.0%/33.1% other treatments, and 7.0%/6.0% were untreated. Males were prevalent and median age was > 73 years for all groups. Of 2019/2020 cohorts, heart failure and coronary artery disease affected 24/20%, 18/17%, and 11%/16% patients with SI-TT, MI-TT, DT, and other treatments, respectively. A previous moderate/severe exacerbation (2019/2020 patients) occurred to 60.5%/56.6%, 39.9%/37.4%, 30.8%/29.2% and 31.9%/29.7% patients treated with SI-TT, MI-TT, DT, and other treatments, respectively. Of 2019/2020 patients experiencing moderate/severe exacerbation: 6.0%/7.0% receiving DT, 5.1%/7.0% receiving other treatments and 4.5%/10.0% untreated, switched to SI-TT; 23.7%/16.9% receiving DT, 21.4%/17.7% receiving other treatments and 15.4%/12.0% untreated, switched to MI-TT.

Conclusions

COPD patients receiving TT were older and had more comorbidities, especially cardiovascular diseases, than patients receiving DT or other treatments. The limited number of patients switching after exacerbation suggests that many COPD patients may be inappropriately treated. Ensuring early and adequate treatment, combination of in-hospital and outpatient management, and integration of specialist and primary care is pivotal for the appropriate clinical management of COPD patients.

Background

Chronic obstructive pulmonary disease (COPD) is a chronic heterogeneous lung disease and a leading cause of morbidity and mortality worldwide [1, 2]. The economic and social burden of COPD is substantial and increasing due to the exposure to COPD risk factors and aging of the population [1]. Diagnosis of COPD is still not optimal and often late, and its pharmacological treatment has been frequently found inappropriate [1]. The COPD care pathway is currently built on international guidelines (Global Initiative for Chronic Obstructive Lung Disease—GOLD) [1, 3], which suggest a pharmacological strategy based on the severity of symptoms and the frequency of acute exacerbations [4]. Exacerbation of COPD is a worsening of respiratory symptoms characterized by increased dyspnea, cough and sputum production [5]. The most recent guidelines [1] recommend a dual therapy (DT) including a long-acting beta agonist (LABA) and a long-acting muscarinic antagonist (LAMA), as both initial (group B or E) and follow-up treatment (patient on bronchodilator monotherapy with persistent dyspnea or with persistent exacerbations). A triple therapy (TT) (i.e., LABA/LAMA/inhaled corticosteroids – ICS) should be considered: as initial treatment in patients experiencing at least 2 moderate or 1 severe exacerbation in the previous year (group E) if the blood eosinophil count in cells/μL (EOS) is higher than 300, regardless of symptoms; as follow-up treatment in case of persistent exacerbation on bronchodilator monotherapy, EOS higher than 300, or in case of persistent exacerbations on DT and EOS higher than 100 [1]. When possible, DT and TT should be prescribed as single inhaler (SI), which is more convenient and effective than multiple inhaler (MI) [6]. However, de-escalation of ICS should be considered if pneumonia or other considerable side effects occur [1].

In Italy, TT was available only as a MI until the introduction of the SI strategy in 2018. Specific Italian guidelines on COPD do not exist. However, in 2021 the Italian Medicines Agency (AIFA) released a document, called Nota AIFA 99 [7], which synthetizes international recommendations and presents as national address line for general practitioners (GPs) and specialist pneumologists. The Nota AIFA 99 is currently not updated with the latest version of international guidelines and restricts to pneumologists the prescription of SI-TT, which could lead to a delayed treatment initiation and suboptimal integration between specialist and primary care [8].

The rationale of this study is due to the fact that, despite the widespread recommendations and health policies, real-world studies have shown that patients with COPD are still misdiagnosed, late diagnosed, and treated with a pharmacological strategy inconsistent with the stage at diagnosis or with follow-up conditions, worldwide and specifically in Italy [2, 9,10,11]. Compared to the 2019/2020 GOLD versions [12, 13], the 2023 GOLD version has undergone small but important changes, especially regarding the insertion of the recommendation of TT as starting therapy. This descriptive retrospective observational cohort study of Italian administrative healthcare data aimed to identify patients with COPD and describe them in terms of treatment approach in 2019 and 2020, and how treatment strategies changed following an acute exacerbation, in the light of the GOLD guidelines of the time and soon after the marketing and reimbursement in Italy of the SI-TT.

Methods

Data source

The database of Fondazione Ricerca e Salute (ReS) is an administrative healthcare database collecting and integrating the administrative healthcare data that Italian Local and Regional Healthcare Authorities annually convey to the Italian Ministry of Health, upon specific agreements for reimbursement purposes respondent to the universal coverage characteristic of the Italian National Health Service (Servizio Sanitario Nazionale – SSN). The ReS database includes about 5 million inhabitants per year (from 2014 to 2021), which represents around 8% of the Italian population, and has been demonstrated as consistent in the distribution by age group with that obtained by the Italian Institute of Statistics (figure S1) [14]. More information about Fondazione ReS and the database ReS are available in the online supplementary material (box S1). The datasets analysed during the current study are not publicly available, because they are owned by the Italian Regional/Local Health Authorities; only Fondazione ReS has access to the database population. Data were anonymized at the source and analysed in aggregated form according to European Regulations 2014/536 [15] and 2016/679 [16], to which the Regional/Local Authorities owners of the data have agreed. Informed consent was waived according to the specific Italian Privacy Authority’s provision [17]. Ethical approval was waived according to the European Regulation 2014/536 [15], and it is not expected by the most recent national legislation on observational studies [18], which requires the Ethic Committee’s positive opinion only for prospective observational studies, while it does not refer to retrospective ones.

Study design and cohorts’ selection

Patients aged ≥ 45 years, variously distributed throughout northern, central and southern Italy, and with at least 6 years of look-back period (i.e., analysable until 2013) were identified from 1st January 2019 to 31st December 2019 and from 1st January 2020 to 31st December 2020 (accrual periods; figure S2) according to criteria listed in table S1. The look-back period was necessary to reliably describe the comorbidities of interest. After having excluded patients meeting asthma identification criteria (table S1), the study population was categorized according to the first treatment strategy for COPD dispensed in the accrual periods into the following cohorts: patients treated with TT (SI or MI), DT (SI or MI), other treatment strategy with drugs for obstructive airway diseases (table S2). When the period between the end of the treatment coverage of the first dispensation of LABA or LAMA or ICS and the following dispensation of LABA or LAMA or ICS useful to deliver a DT or a TT fell within 30 days, the DT or TT were considered as a MI strategy. The index date was the first dispensation of one of the treatment strategies that defined each cohort, i.e., a fixed dose combination, one of the drugs included in an open combination, one single drug (figure S2). Since administrative data have been considered as a reliable data source to identify and study patients with COPD [19], Fondazione ReS, supported by a team of high-level independent clinical experts, has validated the algorithms used to select the study population and the acute exacerbations of COPD.

Baseline characteristics

Cohorts were described at baseline (i.e., index date and look-back periods) in terms of sex, age and comorbidities of interest (i.e., arterial hypertension, dyslipidaemia, diabetes, heart failure, coronary artery disease, cardiac arrythmias, depression, cerebrovascular diseases, neoplasia, chronic kidney disease and chronic liver disease - identification criteria in table S4).

Occurrence of COPD exacerbations and switch of treatment strategy

Within each cohort, patients experiencing a COPD exacerbation in the 12 months preceding index date were identified. A moderate COPD exacerbation was defined by at least one dispensation of antibiotic within 30 days before or after the dispensation of systemic corticosteroid (table S3). A severe exacerbation was defined by at least one overnight hospitalization due to COPD or due to COPD-related causes and with COPD as secondary diagnosis (table S3). In case of more than one COPD exacerbation per patient, the most recent one was considered for the therapeutic strategy analysis.

To verify whether a switch of treatment strategy occurred after the exacerbation, the last treatment strategy that was dispensed to each patient within the year preceding the COPD exacerbation was assessed.

Statistical analyses

Continuous variables are described as means ± standard deviation (SD), and median (first quartile; third quartile). Categorical variables are provided using frequencies and percentages within each category. Data extraction was performed using Oracle SQL Developer, Italian version 18.1.0.095 (California, United States) and subsequent. Descriptive statistics were performed through Microsoft Excel Office 365 (Micro-Soft Company, Washington, United States).

Results

From the ReS database (around 4.7 million inhabitants in 2019 and in 2020), after having excluded patients with asthma, people with COPD in Italy were 105,282 in 2019 and 103,729 in 2020, namely a prevalence of 43 and 41 × 1,000 inhabitants aged ≥ 45 years, respectively (Figs. 1 and S3). Among them, at index date in 2019 and 2020, respectively, 3.4% and 5.2% received a SI-TT, 20.7% and 21.6% a MI-TT, 35.9% and 36.6% a DT, 33.0% and 32.1% other treatment strategies, while 7.0% and 6.0% were untreated (Figs. 1 and S3).

Fig. 1

Selection of patients with chronic obstructive pulmonary disease and treated with respiratory drugs in 2019. COPD: chronic obstructive pulmonary disease; ICS: inhaled corticosteroid; LABA: long-acting beta agonist; LAMA: long-acting muscarinic antagonist. Figure 1 shows the flowchart of the identification of patients with chronic obstructive pulmonary disease (COPD) and treated with drugs for obstructive airway diseases (ATC code R03) in 2019, by treatment strategy

In 2019 and in 2020, patients with COPD were mostly males, especially those treated with SI-TT (77.7% and 77.0%) – Tables 1 and S5. Median ages are highest in the cohorts of patients treated with MI-TT (76 (69; 82) and 77 (70; 84) years). Overall, the prevalence of treatment increased with age (Tables 1 and S5).

Table 1 Description of patients with COPD and treated with respiratory drugs in 2019, by treatment strategy. Shows the characterization of patients with COPD and treated with drugs for obstructive airway diseases, and rate of exacerbations during the year preceding index date in 2019, by treatment strategy

The highest proportions of patients affected by at least one comorbidity of interest were found in the cohorts treated with SI-TT, except for depression and chronic kidney diseases that were prevalent in patients with other treatments (Tables 1 and S5). Arterial hypertension was the most common comorbidity among all treatment groups, followed by dyslipidemias and diabetes.

During the year preceding index dates in 2019 and in 2020, respectively, a moderate/severe exacerbation of COPD occurred in 60.5% and 56.6% of patients receiving SI-TT, 39.9% and 37.4% of patients receiving MI-TT, 30.8% and 29.2% of patients receiving DT, and 31.9% and 29.7% of patients dispensed with another treatment strategies, respectively (Tables 1 and S5). A previous severe COPD exacerbation occurred mainly in patients treated with SI-TT (36.5% in 2019 and 36.3% in 2020) (Tables 1 and S5).

Of the 2019 and 2020 cohorts, respectively, patients treated with TT before a moderate/severe exacerbation increased from 8.4% to 32.4% and from 10.6% to 31.3% after a moderate/severe exacerbation (Figs. 2a, b). Out of patients treated with TT at index date in 2019 and 2020, respectively, 29.7% and 23.9% came from DT before the exacerbation.

Fig. 2

a Distribution (%) of patients with COPD and treated/untreated with respiratory drugs in 2019, following an exacerbation. Figure 2a shows the percentage distribution of patients with COPD and treated or untreated with drugs for obstructive airway diseases in 2019, by treatment strategy preceding and following a moderate/severe exacerbation. Figure 2b Distribution (%) of patients with COPD and treated/untreated with respiratory drugs in 2020, following an exacerbation. Figure 2b shows the percentage distribution of patients with COPD and treated or untreated with drugs for obstructive airway diseases in 2020, by treatment strategy preceding and following a moderate/severe exacerbation

Tables 2 and 3 show that among 2019 and 2020 populations, regardless of the severity of exacerbation, most of patients supplied with TT before the COPD exacerbation continued to be treated with TT following exacerbation, mainly as MI (71.8% and 59.8%, respectively), but with increased dispensation rates in 2020. Most of patients supplied with DT before the exacerbation continued to be treated with DT (63.8% and 70.0% in 2019 and 2020, respectively), as well the majority of patients supplied with other treatment strategies before the exacerbation continued to be treated with other treatment strategies (50.6% and 52.8%). Whereas, more than half of patients untreated preceding an exacerbation were supplied with other treatment strategies following the exacerbation (51.3% and 47.8%). Also, Tables 2 and 3 show that the switch to a SI-TT was higher following a severe than a moderate exacerbation.

Table 2 Switch of treatment strategy following the exacerbation in 2019, by treatment pre-exacerbation. Shows the switch of treatment strategy following the exacerbation which occurred during the year preceding index date in 2019, by treatment strategy dispensed pre-exacerbation

Table 3 Switch of treatment strategy following the exacerbation in 2020, by treatment pre-exacerbation. Shows the switch of treatment strategy following the exacerbation which occurred during the year preceding index date in 2020, by treatment strategy dispensed pre-exacerbation

Discussion

This analysis of a large Italian administrative healthcare data in 2019 and 2020 showed a slice of the Italian real-life clinical practice. The study population in 2020 was slightly lower than in 2019, probably because of the less healthcare resource utilization during the early phase of the Sars-Cov2 pandemic. The demographic and clinical characteristics of patients with COPD in this study were similar to those identified in other Italian and international studies, in terms of age, sex and prevalence of comorbidities of interest, especially cardiovascular and metabolic diseases and depression [2, 4, 10, 11, 20,21,22]. This supports the robustness of findings from this study. Patients treated with TT had higher median ages and comorbidity rates, especially cardiovascular ones, compared to patients receiving DT, highlighting the high disease burden of patients with COPD treated with TT. Comorbidities negatively impact the clinical course and prognosis of COPD [23], with subsequent poor quality of life, high healthcare resource consumption, especially increased risk of hospitalizations, and mortality [24].

Overall, findings suggest an underprescription of TT in both cohorts. Nevertheless, the 2020 cohort of COPD patients showed a slightly higher rate of SI-TT dispensation (5.2% vs 3.4%) compared to the 2019 cohort, which was significant (p value < 0.05) for patients switching from both DT and other treatment strategies. This may be due to the increased availability of SI-TTs in Italy over the years since its first marketing in 2018. Compared to MI-TT, SI-TT was dispensed to patients slightly older (median age 76 vs 74 years) and with more comorbidities of interest (92.5% vs 90.9% patients with at least one comorbidity). This finding should be relevant in the light of the need for a simpler therapy for patients often exposed to an unbearable polypharmacy [1, 13]. The other findings overlap between the 2019 and 2020 cohorts, although during the SARS-CoV2 pandemic, the healthcare utilization for non-COVID19 conditions, especially chronic diseases, substantially decreased worldwide [25].

TT as starting therapy was not recommended by GOLD guidelines effective during the study period [12, 13]. Whereas the current version of the guideline recommends TT as initial treatment in patients with at least 2 moderate exacerbations or 1 severe exacerbation during the previous year and EOS ≥ 300 [1]. At index date, in 2019 and 2020, the proportions of patients treated with SI-TT were very low (3.4% and 5.2%, respectively), while MI-TT was dispensed more frequently (20.7% and 17.5%) after both moderate and severe exacerbation. Given the still current version of the Nota AIFA 99 that restricts the prescription of SI-TT to pneumologists, and the high probability that a patient with COPD primarily require the GP’s intervention, it is possible to assume that MI-TT were mostly prescribed in the primary care setting. Despite only more recent guidelines explicitly recommend the SI strategy, when possible [1, 6, 26], the clinical benefits, also in terms of improved therapy adherence which reduces all-cause mortality and incidence of exacerbations, of the SI-TT over a MI-TT were already known [13]. As per this study, the significant increased dispensation of SI-TT from 2019 to 2020 could be due to the even greater acceptance by pneumologists and reimbursement by the SSN of this strategy, which was marketed in Italy only since 2018. However, dispensation rates of SI-TT in Italy remain lower than that of MI-TT, because the recently released Nota AIFA 99 still restricts the SI-TT prescription to pneumologists [7], A further study of the database of Fondazione ReS on more recent data will be useful to assess whether the therapeutic approach in Italy has changed following the updated guidelines recommendations and the released Nota AIFA 99, and possible savings following an increased use of SI-TT.

The highest dispensation rates of TT are seen among patients older and more affected by the analyzed comorbidities, especially SI-TT, suggesting that TT is considered by Italian clinicians the last therapeutic chance in case of more fragile patients (i.e., older and with many comorbidities), and that the escalation should occur, though infrequent, after whatever exacerbation, namely followed by a hospitalization or not. Most patients in TT are treated with MI-TT, partly because of the only recent availability of the SI-TT in Italy and partly because the MI-TT is the only TT strategy that can be prescribed also by GPs. As also shown by this study, these patients can be considered as fragile patients, who largely need treatments improving the therapeutic adherence, because they are frequently treated with more than 5 drugs (i.e., polypharmacy). Moreover, these patients are mainly and in the first place cared and followed-up in the primary setting. Findings from this study are of utmost importance to highlight that GPs should be ever more involved in both the training of patients with COPD about appropriate treatment strategies [20] and the prescription of the SI-TT. These findings must be interpreted also in the light of the scarcity of diagnostic tools (e.g., spirometry) in primary care ambulatories, on which the appropriate prescription of the TT must be based. Finally, in order to optimize the patient journey and the collaboration between specialists and GPs, the Nota AIFA 99 should be constantly updated with the most recent guidelines and modified to allow GPs prescribing the SI-TT as starting and/or follow-up therapy, as it is in other major European countries. Findings from clinical studies cannot be generalized to real-world patients with COPD, because the two populations are not overlapping. This study of the ReS database provides important insights about the real-world clinical practice in Italy, which are useful for the multidisciplinary and multiprofessional teams that are designated to review the therapeutic strategy of patients with COPD, if needed, to the end of an ever more personalized medicine [1, 27]. The proportion of 7.0% and 6.0% study population that remained untreated in 2019 and 2020, respectively, is worth of attention, despite the lack of the out-of-pocket purchase of drugs reimbursed by the SSN within administrative data, which, however, accounted for the 5% of the overall pharmaceutical expenditure in 2019 [28]. Also, this study has provided a low rate of escalation especially after an acute severe exacerbation. The 2019 and 2020 GOLD guidelines, which should have been followed during the study period [12, 13], already recommended as follow-up strategy an early escalation from monotherapy to the LABA/LAMA strategy in case of poorly controlled symptoms or persistent exacerbations, and the addition of an ICS (i.e., LABA/ICS) in case of EOS ≥ 300 or EOS ≥ 100 and at least 2 moderate or 1 severe exacerbations per year. Nevertheless, even though the study population could not be stratified by COPD severity because of the limitations of administrative data, our findings suggest a possible misinterpretation of international recommendations by Italian clinicians; indeed, it is likely that the high number of patients treated with TT included less severe patients who did not require the addition of an ICS. Escalation should always be reviewed, and de-escalation of ICS or switch should be considered if there is a lack of clinical benefit and/or side effects occur. The DT LABA/ICS is no longer recommended by current guidelines [1], but the Nota AIFA 99 still recommends it. For the best Italian clinical practice, we highlight the need for a routinely update of the Nota AIFA 99 with the latest version of the internationally approved guidelines and the most recent evidence.

Strengths and limitations

The main strength of this study is that it is based on the analysis of the ReS database, which includes real-world data of about 5 million SSN beneficiaries who represent about the 8% of the Italian population and are likely representative of patients cared in the Italian clinical practice. Administrative data have been considered reliable for the identification of people with COPD [19], especially when potentially confounding patients with asthma or occasionally treated are excluded, like in this study. Therefore, this analysis has identified a fair number of patients with COPD and described their treatment in the Italian real-life clinical practice during the study periods, achieving the set aims.

However, certain limitations should be considered when interpreting the results. The use of in-hospital diagnoses among the selection criteria could have underestimated patients with COPD, and the use of dispensations of drugs for obstructive airway diseases could have misclassified this cohort. Given the lack of clinical parameters (e.g., EOS values and spirometry tests coming from electronic medical records or the primary care database), neither the correction of misclassified COPD cases nor the stratification by COPD severity could be done.

This was a descriptive retrospective cohort study that was aimed to describe patients treated with drugs for obstructive airway diseases through different strategies, especially by assessing the dispensation of TT and the potential switch from DT to TT. A further study could be designed to describe the direct escalation also from monotherapy to TT, which is a frequent practice worldwide, often inconsistent with clinical guidelines [29]. Moreover, the diagnosis based on which the prescribed drugs are reimbursed by the SSN is unknown; therefore, it is likely that a small number of drugs for obstructive airway diseases dispensed to the study population may not have been intended to treat COPD. Likewise, the use of only drug supplies (i.e., corticosteroids and antibiotics for systemic use) as a proxy of moderate exacerbation could have overestimated its rate. On the contrary, the use of in-hospital criteria as a proxy for severe exacerbation was likely more reliable to identify these cases. As regards moderate exacerbation, we have chosen to define the moderate exacerbation through the association of corticosteroids and antibiotics because the absence of clinical parameters in the administrative data has led us to be as much restrictive as possible when the criteria (i.e., dispensation of drugs in this case) are not specific for the studied disease, and this association is a common Italian clinical practice.

The impact of the treatment strategies described and of the switch after an exacerbation on outcomes, such as exacerbation rate and survival, between users and non-users of TT was not provided. Indeed, the absence of clinical parameters (e.g., EOS values and results of spirometry tests) in the administrative database may lead to a challenging comparison and interpretation of these outcomes.

Conclusions

Findings from this study on Italian real-life data from primary and specialist clinical practice settings suggest the urgent need of a targeting training of GPs and pneumologists and their collaboration on the COPD treatment approach, which has been demonstrated still not in line with current international guidelines and recommendations. The TT was prescribed to patients older and affected by many comorbidities, suggesting the absence of a controlled escalation and a multidisciplinary assessment of these fragile patients. A more precise and an earlier diagnosis and severity definition of COPD must be stressed to set up the earliest and most appropriate treatment strategy, both at the beginning and during follow-up. In Italy, the national health and reimbursement policy, through the Nota AIFA 99, must be constantly updated based on the latest international and evidence-based guidelines to ensure an immediate and direct application of the best care pathway of patients with COPD since the primary care setting, by appropriately training also GPs and allowing them to prescribe the SI-TT throughout the care pathway. This would improve the collaboration between specialist and primary care to the end of better patient’s outcomes, adherence and healthcare resource consumptions.