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.
link