Patterns of treatment with antiplatelet therapy after an acute coronary syndrome: Data from a large database in a community setting

Abstract

Introduction

Despite rates of death due to cardiovascular diseases having declined over the past decades, acute coronary syndrome (ACS) remains a leading cause of mortality worldwide.^1,2 Clinical presentation of ACS encompasses ST-segment elevation myocardial infarction (STEMI), non-ST-segment myocardial infarction (NSTEMI) and unstable angina based on electrocardiographic criteria and troponin elevation values.³

Since the activation of platelets and their subsequent aggregation have a pivotal role in the propagation of arterial thrombosis, antiplatelet therapy is a key target in the treatment of ACS.^4,5 Therefore, the European Society of Cardiology guidelines recommend dual antiplatelet therapy (DAPT) with aspirin plus P2Y₁₂ receptor inhibitors to reduce the risk of both acute ischaemic complications and recurrent atherothrombotic events in patients with ACS.^6–8

In the last decades, a large number of randomized clinical trials (RCTs) provided robust evidence on the substantial reduction in mortality and morbidity achieved through the improvement of intervention techniques and shortening of symptom-to-intervention times, along with the introduction of new antithrombotic drugs and prolonged long-term management of ACS patients.^9,10

However, RCTs usually enrol highly selected populations and whether findings from these studies can be transferred to real-world patients has been largely debated, due to major differences in baseline clinical characteristics, patients' medical complexity and reproducibility of outcomes of treatment.¹¹

Different aspects of current management in daily practice can be better assessed by means of registries, surveys and observational epidemiological studies.¹¹ In particular, large-scale registries provide a wealth of information to fill important gaps in the current available evidence, and, in contrast to RCTs, they document the real utilization of procedures and drugs, allowing a comparison between clinical practice and guideline statements. The degree of physicians' adherence to ACS management guidelines and recommendations in Coronary Care Units (CCUs), or more generally in cardiology wards, has been reported as high.^11,12 Further, high levels of adherence have been clearly associated with favourable outcomes.¹³ However, some patients with ACS do not enter clinical trials nor cardiology registries, but these subjects are often the most seriously ill and complex.

A prior analysis of the large Italian real-world database Ricerca e Salute (ReS; Research and Health Foundation, in partnership with CINECA, previously defined as ARCO database) showed that more than 20% of patients discharged alive after an ACS during 2014 were not prescribed an antiplatelet agent and that about 25% of patients treated at discharge did not show prescription continuity over the one-year follow-up.¹² However, data concerning the characteristics of patients not treated with antiplatelet drugs after an ACS hospitalization are scarce. It has been hypothesized that advanced age of the population, gender, comorbidities, economic status and care organization may play a role in physicians' decision to treat or not.^12–16 Yet, few studies addressed the questions of the causes of under-treatment with antiplatelet after an ACS.

Therefore, this analysis aimed to provide a more detailed description of real-world patients with and without antiplatelet treatment after an ACS, their outcomes at one-year follow-up and the related integrated cost, assessing a large healthcare administrative database.

Methods

Study design and data

This observational retrospective study analysed the administrative healthcare claims database ReS, formerly ARCO database, a population-based patient-centric database including about 20 million Italian inhabitants. The ReS database includes information on demographic variables, in-hospital diagnosis and procedures, pharmaceutical prescriptions and out-of-hospital diagnostic/therapeutic procedures.

In-hospital diagnoses and procedures were collected from hospital discharge administrative data and assessed according to the Italian version of the ninth International Classification of Disease (ICD-IX-CM).^17,18 The list of codes is reported in the Supplementary Material Table 1 online.

Pharmaceutical data consisted of dose, number of packages, dispensing date and cost of medicinal products and generics reimbursed by the Italian National Health System (INHS),¹⁸ according to the Anatomical Therapeutic Chemical classification (Supplementary Table 1).¹⁹

Demographic information about patients was anonymous, in accordance with the Italian law regarding the protection of privacy,^20,21 therefore ethical approval for conducting this observational analysis was unnecessary.

Cohort selection and follow-up

Out of a population of more than 12 million inhabitants, a cohort of 26,834 patients discharged after an ACS was identified from 1 January 2014 to 31 December 2014 (accrual period). The hospital discharging unit, whether CCU, Cardiology Unit or other (i.e. Internal Medicine Unit, Geriatric Unit, etc.), cannot be deduced from the available data. ACS patients were defined as patients discharged alive at least once during the accrual period with primary diagnosis (ICD-IX-CM code) of ACS.

After this first hospital discharge (index date), patients were followed up throughout one year. Some patients (6.3%) were lost to follow-up due to migration to other local health units, probably to find healthcare assistance closer to their needs.

Clinical characteristics and pharmaceutical indicators

ACS patients were characterized according to whether or not they received a coronary revascularization by means of ICD-IX-CM codes describing coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI).

In addition, possible comorbidities, such as diabetes, hypertension, depression, neoplasia and chronic obstructive pulmonary disease, were evaluated through an analysis of the 365 days before the index date.

Concerning antiplatelets use, patients were categorized into the following subgroups: aspirin alone, clopidogrel alone, prasugrel alone, ticagrelor alone, the free combination of clopidogrel and aspirin, the fixed combination of clopidogrel and aspirin, the free combination of aspirin with prasugrel or ticagrelor, no treatment with antiplatelet agents.

Non-antithrombotic cardiovascular therapy, including beta-blockers, calcium antagonists, angiotensin converting enzyme inhibitors or angiotensin receptor blockers, and lipid lowering agents were also analysed.

Re-hospitalizations (ordinary and day-hospital regimens) were assessed over the follow-up period, according to the occurrence of hospital discharges due to cardiovascular and non-cardiovascular causes.

Healthcare costs

Free filled cardiovascular and non-cardiovascular drug prescriptions (reimbursed by the INHS), outpatient specialist services (Italian national tariffs) and hospitalizations (Italian national tariffs for the hospital care supply, DRGs-TUC 2008 version) were used as indicators to analyse healthcare costs over the one-year follow-up. All estimated costs of drugs, hospitalizations and outpatient procedures and visits were also integrated, in order to give the total mean per capita expenditure, by follow-up time, group of treated and untreated with antiplatelet patients and performance of coronary revascularization at the index hospitalization. Mean cost was calculated per capita up to the end of the one-year follow-up or the death of the patient. Therefore, only direct healthcare resources charges to the INHS were considered.

Statistical analysis

When administrative data are analysed, the number of patients/events is so large that even minimal differences can determine the obtainment of the conventional level of statistical significance, often without any correspondent convincing clinical significance. For this reason, we preferred, as in other studies using similar databases,^22,23 to not provide detailed p values but just to describe the nominal differences.

Results

Out of a population of 12,185,229 inhabitants with available demographic and administrative data for 2014 and 2015, 26,834 individuals (22.0 per 10,000) were hospitalized for ACS. In-hospital all-cause mortality was 6.4%. Therefore, 25,129 (93.6%) ACS patients were discharged alive and represent the study cohort. Among them, 11,796 (46.9%) were treated invasively with PCI or CABG during the index admission (Figure 1).

Figure 1.

Patient disposition.

ACS: acute coronary syndrome

Open in new tabDownload slide

At least one antiplatelet drug was prescribed during the first month after the index event to 76.9% (n = 19,333) of the entire cohort. The remaining 23.1% (n = 5796) were not prescribed an antiplatelet medication during the first month; 7.8% (n = 454) of them were prescribed with an oral anticoagulant. During the one-year follow-up, 3,846/5,796 subjects (15.3%) received an antiplatelet drug; 7.7% of patients (n = 1950) did not receive any antiplatelet prescription over the whole year following the index event (Table 1).

Mean (±SD) and median age were 70 ± 12 and 71 years, and 75 ± 12 and 77 years, respectively, for treated and not treated patients with antiplatelet agents during one month after the index event. The mean (±SD) and median age of patients untreated during the whole year after the ACS event were 77 ± 13 and 80 years (Table 2). The prevalence of female gender was 30.8%, 41.6% and 51.3%, respectively of treated, untreated during the 1st month, and not treated over 1 year of follow-up.

Detailed characteristics of the study population by age, gender and comorbidities are reported in Table 2. According to comorbidities analysis, patients treated with antiplatelets over the one-month follow-up and who did not undergo revascularization were slightly more complicated than those who underwent CABG/PCI.

In the group of one-month follow-up treated and re-vascularized patients (n = 10,644) the most used antiplatelet therapeutic regimens were aspirin in free combination with clopidogrel (31.9% of group), ticagrelor (26.2%) and prasugrel (13.9%), followed by clopidogrel (7.5%) and aspirin alone (6.4%). Whereas, among not re-vascularized subjects (n = 8689), aspirin alone (26.3%) followed by freely combined with clopidogrel (25.4%) and ticagrelor (12.3%), and clopidogrel alone (13.5%) were most prescribed. DAPT was prescribed to 49.5% (n = 12,445) of the entire cohort of patients discharged alive. Particularly, it was prescribed to 71.5% (n = 8436) of subjects treated invasively during their index ACS episode (n = 11,796), and to 30.1% (n = 4009) of medically managed ACS patients during their event index. Concomitant non-antithrombotic cardiovascular therapies were also investigated during the first month after the index date, in patients treated and not treated with antiplatelet drugs, and over the year of follow-up. In Table 1, the five most frequently prescribed antiplatelet drugs are presented.

All-cause one-year re-hospitalizations occurred in about half of all groups of treated and untreated subjects; particularly, 47.9% of the patients in the cohort treated with antiplatelets, versus 67.9% of untreated after one month and 62.4% of untreated at one year. Among cardiovascular causes of re-hospitalizations, the five most frequent were ‘Other forms of chronic ischemic heart disease’, ‘Acute myocardial infarction’, ‘Angina pectoris’, ‘Other acute and subacute forms of ischemic heart disease’ and ‘Heart failure’. They were differently distributed in the groups of treated and untreated patients during the one-month follow-up (Table 3).

The analysis of one-year healthcare costs showed that hospitalizations were the main cost driver for all groups, accounting for about 83% to 91% of the overall expenditure (Table 4). These were followed by pharmaceutical costs (from about 4% to 11% of overall expenses) and outpatient services costs (from about 4% to 6% of the total expenditure).

Discussion

The most relevant findings of this analysis of the large community database ReS can be summarized as follows:

1. About 23% of patients discharged alive after an ACS event did not receive an antiplatelet agent during the first month after hospital discharge.

2. About 66% of patients discharged after an ACS event without antiplatelet prescription during the first month of follow-up received antiplatelet therapy subsequently.

3. Of patients discharged alive after an ACS event 7.7% did not receive any antiplatelet prescription for the whole year of follow-up.

4. DAPT in patients treated invasively was prescribed in 71.5% of cases. In medically managed patients, DAPT was prescribed on 30.1% of cases.

5. The female gender constituted almost half of untreated groups in the first month and in the year of follow-up.

6. Patients aged 70 to 89 were largely represented in the untreated group, both at one-month and at one-year follow-up.

7. The re-hospitalization rate was markedly higher in subjects not treated with antiplatelets during the first month of follow-up (65.4% of re-vascularized and 68.6% of not re-vascularized, vs. 41.1% of treated patients re-vascularized and 56.2% of not re-vascularized).

8. The costs for the INHS were higher for untreated patients than for subjects prescribed an antiplatelet agent at discharge.

It must be emphasized that, though keeping in mind the weakness of the comparison between cardiology registries and big data, in-hospital mortality after ACS is low (6.4%), but it is higher than that reported from registries (2–3%).^24–26 Also, coronary revascularization rates are different, and lower than those of Italian cardiology registries (46.9% vs. 67.7%).

Results of this analysis show an antiplatelet under-treatment after hospital discharge for ACS, markedly different from and higher than that observed from Italian cardiology registries.^24,27 These observations, interestingly, cover roughly the same period of time as ours and refer to the antiplatelet prescription at discharge. The rate of patients who do not take even a single antiplatelet agent in the first month and then during a whole year of follow-up is higher than that reported in the cardiology registries (respectively 23.1% and 7.7% vs. 3.1%). Moreover, the use of DAPT administered both to invasively and to medically managed ACS patients is different compared with that described in the cardiology registries. Indeed, the present report shows a DAPT prescription of one-month follow-up of 64%, considerably lower than what is reported from Italian cardiology registries.^24,27  Table 5 reports a comparison between data of the ReS database and those of cardiology registries conducted in Italy,²⁵ the UK, Sweden²⁸ and the Danish National Registry.²⁹ Rates of coronary revascularizations during ACS and of prescriptions of DAPT appear to be higher when data are collected in the cardiology ward by cardiologists.

Even though our epidemiological study does not allow to address definitively the cause of this under-treatment, in part this gap may be explained by the fact that a proportion of patients (nearly 8%) were prescribed with an oral anticoagulant. Elderly patients with a documented atrial fibrillation admitted for chest pain who develop a troponin leak are usually left with their chronic anticoagulant therapy without adding an antiplatelet agent. Even if difficult to be precisely quantified, in Italy some hospitals provide discharged patients with a small amount of drug prescribed at discharge, possibly resulting in a missed prescription in the first month after discharge.

Given the lack of crucial information such as the hospital ward, contraindications to treatment and a more detailed clinical picture of each patient, it is not possible to specifically evaluate the reason for under-treatment. It is, however, possible to comment on some of the demographic and clinical circumstances associated with it.

The percentage of female patients increases in the group of under-treated patients. The rate of women discharged with antiplatelet agents and treated with revascularization compared with those non-revascularized is consistently lower and increases slightly when compared with patients treated with antiplatelet agents within the first month and those treated after the first month. Although in our population the rates of women treated at discharge appear to be high, ranging between 29% and 39%, they do not differ substantially from those reported by cardiology registries (about 31%), whereas, when considering women untreated during the entire follow-up year, the rate reaches 51.3% of this subgroup. It is interesting to note that with increasing age, the rate of under-treatment, both for coronary re-vascularized and not, increases. This is evident in particular for females. The issue of under-treatment of older women proposes again a gap in their management that has been described for years.^30,31

Even considering the possible biases in the precise definition of the rate of under-treatment with antiplatelet agents, and specifically DAPT, the analysis of the ReS database surely suggests that these Class 1A guideline recommended treatments are under-prescribed in clinical practice.

The rate of one-year untreated patients is of particular interest. As this analysis suggests, this is a population predominantly of female gender, elderly, and with many comorbid conditions. These patients are not only untreated with antiplatelet agents but, in general, poorly treated with all cardiovascular therapies. Thus, it may be hypothesized that elderly patients with comorbidities (e.g. recent bleeding, high risk of bleeding, difficulties to warrant pharmacological compliance, etc.), and probably frequently admissions in internal medicine instead of cardiology wards, represent a population at high risk of under-treatment.

A group of patients deserving attention is that composed of the 5796 subjects not treated with antiplatelet drugs during the first month after discharge. Of these, 66.4% were treated with these drugs during the following months, and only 19.9% underwent coronary revascularization at the index event. Their co-morbidities do not differ substantially from those of ACS patients treated at discharge even if each single considered comorbidity was slightly more frequent in the group of patients treated after the first month. ACS patients untreated during the first month, in essence, were poorly treated also during the whole follow-up, and more severely ill probably due to the increased rate of hospital readmissions. For them the most used antiplatelet regimens after the first month of follow-up are aspirin (often as monotherapy) and clopidogrel (alone or in combination with aspirin).

Re-hospitalizations and healthcare costs

The costs for hospital admissions were definitely the major cause of total expenditure for the management of this clinical condition. Patients not prescribed an antiplatelet agent at discharge were more costly than those for whom this prescription was assured. Actually, hospital readmissions of untreated patients were markedly more frequent than those of subjects treated at discharge. However, between patients untreated, those untreated during the first month of follow-up were the more expensive, compared with those never treated over the year. The lack of antiplatelet treatment at discharge is probably a proxy of the clinical complexity of the group of subjects not treated at discharge. The same is not true for those never treated, whose costs were among the lowest, together with those of the patients treated early and re-vascularized.

Limitations

Besides the limitations discussed above, several others must be mentioned.

The ReS database collects administrative information of about 30% of the Italian population. However, as already shown through previous observations,¹² they reliably represent the whole Italian population in terms of demographic characteristics (Supplementary Figure 1).

As already mentioned, the administrative data do not provide a complete set of clinical variables, therefore not allowing an appropriate assessment of the severity of the study population. The lack of clinical information (contraindications, therapy discontinuation, etc.) and of the knowledge of the hospital ward where the index event took place (i.e. Cardiology, Internal Medicine, Geriatric or others) does not allow a complete evaluation of the rate of and the reasons for under-treatment.

Moreover, these results do not permit the distinction between STEMI and NSTEMI, which usually have a substantially different revascularization rate. Any imbalance between the two types of ACS could artificially influence the global rate of coronary revascularization treatments that were observed in this analysis, but the large sample size should act as a corrective for this risk.

Conclusions

Administrative healthcare databases, such as the ReS database, provide, in essence, information very different from that obtained by RCTs or cardiology registries. Therefore, it is not surprising that the rate of patients undergoing coronary revascularization during a hospitalization for ACS appears to be lower than that observed in cardiology registries. Similar considerations can be applied to the results on patients not treated with antiplatelet agents (more than double than that observed in randomized clinical trials and cardiology registries). Yet some interesting issues not captured in observational registries are worthy of note. The most impressive finding from this analysis is the high percentage of ACS patients untreated with antiplatelet agents during the first month after their index event. A considerable proportion of these patients is treated with coronary revascularization and receives antiplatelet therapy after the first month. Yet, these subjects present a very high clinical complexity, as shown by the high rate of re-hospitalization they underwent in the 12-month follow-up and by the high healthcare costs. In view of the prognostic improvement and the economic saving that could derive, it is desirable that these untreated patients with a severe clinical picture be better characterized in terms of contraindications to treatment, other comorbidities, concomitant therapies and clinical setting of assistance. It is also desirable in order to implement initiatives to improve adherence to the recommendations of current guidelines, including structured rehabilitation programmes, and useful also to reinforce patients' education on drug adherence and, as recently suggested, to exert a non-pharmacological platelet inhibition activity.³²

Author contributions

All authors provided substantial contribution to interpretation of data, drafting the article and revising it critically for important intellectual content, and final approval of the version to be submitted.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: CC: received consultation fees and honoraria from Sanofi, Bayer, Pfizer, BMS, Daiichi Sankyo, Boehringer Ingelheim. HPF: has received occasional fees for collaborations as medical writer, lectures, congresses, from: Bayer, Daiichi-Sankyo, Boehringer Ingelheim, Pfizer, Sanofi, Maccann Medical Complete Srl, Health and Life, IMS Health, Clinical Forum Srl, Medi K Srl. APM received grants for the participation in committees of studies supported by Bayer, Cardiorentis, Sanofi, Fresenius and Novartis, outside the contents of this manuscript. LD, AP, GR, SC, CP and NM have no conflicts to declare.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this work was partially supported by an unrestricted grant from Sanofi Italy.

Supplemental Material

Supplemental material for this article is available online.

References