Self-reported prevalence rates of risk factors for CHD in Portugal, including overweight and obesity, hypercholesterolemia, hypertension, diabetes and current smoking, were abstracted from the AMALIA study on the prevalence and distribution of cardiovascular risk factors in Portugal.33 The continuous datasets of the National Health and Nutrition Examination Survey (NHANES) were used to calculate comparable rates for the US, applying the same age inclusion criteria, time period and definitions of risk factors as in AMALIA whenever possible.33,34 Rates for the US were calculated for adults aged 40 years or older surveyed in the 2005-2006 and 2007-2008 NHANES cycles. Calculations of rates were based on data of self-perceived hypertension, hypercholesterolemia, and diabetes, defined as the participant's awareness of or current medication use for that condition, and smoking status was defined as one of the following: current smoker, ex-smoker or non-smoker. However, overweight/obesity prevalence rates for the US were based on actual measurements of weight and height, whereas AMALIA used self-reported measures. Age-adjusted prevalence rates were computed for comparison between the countries, standardized to the 2010 US population (Table A1 in Appendix A. Supplementary Material). US rates were computed using SAS 9.4 with code available on the NHANES tutorial website.35

CHD prevalence rates are not available for Portugal. Instead estimates of myocardial infarction (MI) prevalence were retrieved from the 2005-06 National Health Survey cycle and standardized to the 2010 US population controlling for gender and age.36 Both CHD and MI estimates for the US were abstracted from statistical reports available from the American Heart Association (AHA) for 2010.2 Crude per capita hospitalizations rates were produced based on the numbers reported by the AHA for the US, and by the Directorate-General of Health for Portugal divided by estimates of the total population.37–39

Crude rates of all-cause mortality and mortality from cardiovascular disease (CVD), CHD and MI were calculated from vital statistics data and estimates of the total populations.37,38,40,41 Age- and gender-adjusted mortality rates were obtained by direct standardization to the 2010 US population (Table A1 in Appendix A. Supplementary Material). Rate ratios were computed for analysis of within- and between-country variations in death rates.

A list of procedures and medical device categories and another of active substances and therapeutic uses were drawn up by interventional cardiologists who gave their subjective assessment as to which were technological breakthroughs in the field of interventional cardiology between January 1980 and February 2015. Several cardiovascular devices reviewed were outside the scope of CHD treatment, but were included for a better understanding of the dynamics of the two healthcare systems. After this list was drawn up, the first device model approved for each procedure and device category in each of the two countries was then determined. Similarly, the first drug brand approved for marketing for each active substance and therapeutic use was identified.

In the US, approval dates of most medical devices and drugs are available from FDA databases.30–32 The only approval date that had to be retrieved from a medical journal was of the first percutaneous transluminal coronary angioplasty (PTCA) balloon. In Portugal, approval and marketing dates of drugs are well documented but medical device data are not. We therefore relied on a combination of different sources of information to obtain dates that devices obtained the CE (Conformité Européenne) mark. The CE mark is the certification attributed to certain products, including medical devices, when approved for marketing in the EU. These sources included medical device industry representatives, online press releases and news agencies, medical device industry associations in the EU and Portugal (Eucomed and the Portuguese Association of Medical Device Companies [APORMED]), and scientific journals. However, approval dates of some devices were not available for Portugal and the dates of first use were considered as a time reference of their approval. These dates refer to the first use in the interventional cardiology ward of Santa Cruz Hospital, which is a reference hospital and pioneer in many interventional cardiology procedures in Portugal.42 Both approval and marketing dates of drugs were abstracted from INFARMED, but only marketing dates were used for comparison with the US in the results.

First device models and drug brands were determined according to the earliest approval date among their device or active substance class in each country. For comparison between countries, the lag in access to healthcare technology between the US and Portugal was defined as the difference between approval/first use/marketing dates, calculated as the US date minus the date for Portugal. February 13, 2015, the most recent date for which data are available, was used to compute the difference if the technology was still unavailable in one of the countries. In the event of missing information on the specific day or month of dates, the last day of the month or the month of June, respectively, were used to calculate the difference.

In order to characterize the available human and infrastructure resources, information was collected on the number of cardiothoracic surgeons, cardiologists, hospitals with cardiac catheterization laboratories and hospitals with cardiac surgery facilities in 2000 and in 2010 for the US and Portugal (personal contact with the hospitals).43–45 Per capita rates were then computed based on estimates of the total populations of the two countries.37,38

Figures 1 and 2 depict the lag in access to selected devices and drugs between the US and Portugal. Brand names and their approval/first use/marketing dates have been omitted from this article but are available as supplementary material (Tables A4 and A5 of Appendix A. Supplementary Material).

PTCA: percutaneous transluminal coronary angioplasty; TAVR: transcatheter aortic valve repair. The lag in access to devices was determined based on approval dates for the US available from the Food and Drug Administration (in one case from a medical journal), and, for Portugal, dates of CE marking retrieved from personal communication with medical device industry representatives, online press releases, EUCOMED, and scientific journals, and first use dates in Santa Cruz Hospital (raw data available in Table A4 in Appendix A. Supplementary Material).'> PTCA: percutaneous transluminal coronary angioplasty; TAVR: transcatheter aortic valve repair. The lag in access to devices was determined based on approval dates for the US available from the Food and Drug Administration (in one case from a medical journal), and, for Portugal, dates of CE marking retrieved from personal communication with medical device industry representatives, online press releases, EUCOMED, and scientific journals, and first use dates in Santa Cruz Hospital (raw data available in Table A4 in Appendix A. Supplementary Material).' title='Lag in access to selected interventional cardiology devices between the US and Portugal. Dashed line represents median time lag (US-Portugal). N/A: not available; PT: Portugal; PTCA: percutaneous transluminal coronary angioplasty; TAVR: transcatheter aortic valve repair. The lag in access to devices was determined based on approval dates for the US available from the Food and Drug Administration (in one case from a medical journal), and, for Portugal, dates of CE marking retrieved from personal communication with medical device industry representatives, online press releases, EUCOMED, and scientific journals, and first use dates in Santa Cruz Hospital (raw data available in Table A4 in Appendix A. Supplementary Material).'/>

Figure 1.

Lag in access to selected interventional cardiology devices between the US and Portugal. Dashed line represents median time lag (US-Portugal). N/A: not available; PT: Portugal; PTCA: percutaneous transluminal coronary angioplasty; TAVR: transcatheter aortic valve repair. The lag in access to devices was determined based on approval dates for the US available from the Food and Drug Administration (in one case from a medical journal), and, for Portugal, dates of CE marking retrieved from personal communication with medical device industry representatives, online press releases, EUCOMED, and scientific journals, and first use dates in Santa Cruz Hospital (raw data available in Table A4 in Appendix A. Supplementary Material).

(0.2MB).

INFARMED (raw data available in Table A5 in Appendix A. Supplementary Material).'> INFARMED (raw data available in Table A5 in Appendix A. Supplementary Material).' title='Lag in access to selected interventional cardiology drugs between the US and Portugal. Dashed line represents median time lag (US-Portugal). GP: glycoprotein; N/A: not available; N/C: not marketed; PT: Portugal. Lag in access to active substances was determined based on approval dates available for the US from the Food and Drug Administration and marketing dates for Portugal from INFARMED (raw data available in Table A5 in Appendix A. Supplementary Material).'/>

Figure 2.

Lag in access to selected interventional cardiology drugs between the US and Portugal. Dashed line represents median time lag (US-Portugal). GP: glycoprotein; N/A: not available; N/C: not marketed; PT: Portugal. Lag in access to active substances was determined based on approval dates available for the US from the Food and Drug Administration and marketing dates for Portugal from INFARMED (raw data available in Table A5 in Appendix A. Supplementary Material).

(0.17MB).

Access to medical devices in the US is slower than in Portugal, with only four of the 16 devices reviewed being available first in the US. Moreover, on average, it took 22 months longer to approve the first device model in the US than in Portugal (median time difference 27 months). The drug-eluting balloon for percutaneous coronary intervention (PCI) was the most recent device approved in Portugal, in 2009, and this is still unavailable in the US.

New drugs are more readily available in the US than in Portugal. Only four out of the 14 drugs included in this study were available first in Portugal. On average, drugs were available 51 months sooner in the US (median time difference 61 months). However, up to February 13, 2015, nicorandil and ivabradine had not been approved in the US but have been marketed in Portugal since 2007, and ranolazine had not been marketed in Portugal, despite approval in Portugal in 2008 and in 2006 in the US.

Table 4 provides information regarding diagnostic and treatment facilities as well as medical specialties available in each country. Relative to Portugal, in 2010 the US had nearly double the number of hospitals per capita with cardiac catheterization facilities and four times the number of hospitals per capita with cardiothoracic surgery services. However, while the number of cardiothoracic surgeons per capita was higher in the US, there were more cardiologists per capita in Portugal.

Of the medical devices studied, the majority were approved earlier in the EU/Portugal than in the US, showing that the EU approval system is more time efficient than that of the US (the FDA), as reported by other researchers.10,26,28 This study adds to the current evidence that this time advantage remains when considering the first model of a new technological family.

Earlier approval does not always translate into earlier use of devices. The first medical device approved for intracoronary brachytherapy (ICB) received the CE mark in 1999 but only diffused into Portugal in January 2001. This was motivated by the FDA's approval of the Novoste Beta-Cath system (Novoste Corp.) in that year, establishing ICB as an adjuvant therapy of PCI for in-stent restenosis (ISR).46 Nevertheless, it ceased to be used within six months of its adoption. In the US, ICB saw moderate and decreasing use due to a combination of factors, including the approval of drug-eluting stents (DES) two years later, which significantly reduced the occurrence of ISR, the logistic complexity associated with ICB, and the increasing evidence favoring DES over ICB to treat ISR.19,47–49 Dangas et al. reported that most hospitals no longer possess the necessary facilities to provide ICB in the US.50

DES were rapidly adopted in both countries after their approval. Routine implantation of DES began immediately after approval, in 2002 in Portugal and in 2003 in the US.51,52 By 2003, Portugal had the highest rate of DES implantation in Europe, in 55% of PCI procedures, and by the third quarter of 2005 the use of DES per PCI peaked at nearly 90% in the US.51,53 However, it should be noted that Portugal has the lowest primary PCI rates per capita in Western Europe and, until 2007, thrombolysis was more commonly used than angioplasty in the treatment of ST-elevation myocardial infarction.54,55 The number of PCIs per in-patient case in Portugal is less than half of that in the US.5 This is also consistent with the established capacity to treat patients assessed by the number of specialized physicians and existing infrastructure in each country (Table 4). Considerable differences were observed between the two countries regarding the per capita number of cardiothoracic surgeons and specialized facilities. Only the number of cardiologists per capita was comparable. Moreover, the balance between the number of cardiac catheterization laboratories, cardiac surgery centers, interventional cardiologists and cardiothoracic surgeons, and the potential number of patients to treat is an important but complex topic of discussion with impact on access and quality of care.51,56,57 Although outside the scope of this paper, this is an important issue, given the public vs. private nature of the two healthcare systems, their economic capacity, and observed differences in CHD incidence (Tables 2 and A2).

To our knowledge there are few publications reporting on the real-world implantation of left ventricular support (LVS) devices, specifically the Impella 2.5 (Abiomed Inc.) and TandemHeart (CardiacAssist Inc.). Results from the Europella, EUROSHOCK and USpella registries indicate that the US was faster in adopting the Impella, perhaps in response to early European experience with these devices. Three years after the approval of the Impella (data collected between July 2004 and December 2007), 144 elective high-risk PCI patients with prophylactic LVS using the Impella were registered in the Europella registry over 10 European centers, including one from Portugal.58 Hence, on average, five patients per center per year were implanted with this device. Maini et al. reported the first results in the USpella registry of 276 high-risk PCI patients across 34 US centers.59 The authors do not clearly specify the reporting period of the data but given the initial collection date (March 2010) and the submission date of the article (June 2011), it can be assumed that these numbers represent a one-year span. Thus, on average, eight patients per center per year were implanted with the Impella 2.5. It is worth noting that the intra-aortic balloon remains the most frequently used device for LVS.60–62

Transcatheter aortic valve replacement (TAVR) devices also saw early adoption and rapid increase in use following market approval in both countries.63,64 According to the TAVR registry in Portugal, TAVR use plateaued in 2010 with more than 20 implants per center (6.1 implants per million population).64 In the US, Brennan et al. observed a significant increase in the volume of TAVR use based on TAVR records collected between 2008 and 2013.63 In 2012, one year after the first TAVR device was approved by the FDA, approximately 20 TAVR procedures per center were performed (15.8 implants per million population).

There was a five-year gap between approval of the MitraClip device (Abbott Vascular) in Portugal and the US. However, according to numbers presented at a scientific conference in April 2014, the first case treated using the MitraClip in Portugal was in late 2012, four years after its market approval.65 Since then, a total of nine cases have been treated with this device.65 In the US, adoption appears to have been much faster, given that more than five hundred cases had been treated in the US by the end of August 2014, around a year after its approval, according to figures from the TVT registry presented by Sorajja et al. at the American College of Cardiology 2015 Scientific Sessions.66 Furthermore, by August 2014, there were 76 centers in the US providing MitraClip therapy (around 0.25 centers per million population) as opposed to two in Portugal (0.19 centers per million population).67,68

In contrast to devices, information on approval dates of drugs in Portugal is readily available. In addition to approval dates, INFARMED provides information on effective marketing dates of drugs. On average, marketing of drugs in Portugal was delayed by 81 months from approval (median 65 months lag). Although almost 60% of the selected drugs were approved earlier in Portugal, 61% were marketed earlier in the US (Figure 2 and Table A5 in Appendix A. Supplementary Material). In particular, outpatient prescription of clopidogrel in Portugal started in 2003, the same year this drug became eligible for reimbursement and the same year that effective marketing is reported by INFARMED.12 By contrast, in the US, sales of clopidogrel, marketed as Plavix, have increased since its approval in 1997.69 Longer launch times in Portugal may be driven by both implementation of maximum prices and reimbursement regulations that aim to control the governmental budget, but also depend on the launch strategies of pharmaceutical companies, which ultimately decide when to launch new drugs onto the market. Given the pricing rules of each country, there will be a specific order to be followed by each proposed drug, so as to optimize pricing. External (cross-country) price referencing means that low prices tend to spill over across countries, often delaying launches of new drugs by companies in countries with low expected prices and small volume sales such as Portugal.13

Our results show that CHD deaths were significantly lower in Portugal compared to the US, but no statistical difference was observed for MI deaths between countries. These observations, combined with the fact that MI is the most acute manifestation of CHD, for which the use of health technologies by trained professionals and adequately equipped health facilities may prolong both duration and quality of life, suggest that differences between Portugal and the US in available resources and access to health technologies (faster access to drugs, earlier adoption and/or faster diffusion of devices, and more specialized health facilities and professionals in the US compared to Portugal) may play an important role in quality of care, mitigating differences in epidemiologic risk profiles. Nevertheless, only Portugal exhibited a statistically significant decrease in MI mortality in the decade under study, suggesting that healthcare for these patients has significantly improved in Portugal. Therefore, it is possible that the trends of approval lags observed in Figures 1 and 2, increasingly favorable to Portugal in more recent years, may have contributed to this improvement.

Our study has some limitations. First, rates based on self-perceived measurements likely underestimate the prevalence of risk factors. However, additional evidence confirms that most risk factors are less prevalent in Portugal than the US. Only hypertension appears to be more prevalent in Portugal (further details available as supplementary material). Furthermore, risk profiles were determined considering the overall populations of the two countries. The actual profiles of the CHD populations were not studied, which could explain the differences in usage patterns and mortality rates. Second, routine in-hospital use of clopidogrel, ticlopidine, and abciximab in Portugal began prior to the marketing date provided by INFARMED.53,70–72 Particularly in the case of the drugs indicated for exclusive hospital use (e.g. glycoprotein IIb/IIIa inhibitors), these reports substantially contradict INFARMED's marketing dates. Therefore, information on marketing should be regarded with caution with regard to drugs that are to be used in-hospital only. Third, the brief discussion on adoption and diffusion of healthcare technology is based on key scientific papers for a few technologies. A deeper understanding of the adoption mechanisms is thus required. Furthermore, death certificates may inaccurately state the cause of death and fail to describe the effectiveness of technologies. There is thus a need for better knowledge of the diffusion patterns and effectiveness of health technologies, ideally through the development and analysis of equivalent datasets from the two healthcare systems.

The authors declare that no experiments were performed on humans or animals for this study.

The authors declare that no patient data appear in this article.

The authors declare that no patient data appear in this article.