| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 1999;16:2-8
© 1999 Elsevier Science NL
Department of Cardiac Surgery, Hammersmith Hospital, London, W12 0NN, UK
Received 7 January 1999; received in revised form 15 March 1999; accepted 23 March 1999.
Corresponding author. Tel.: +44-181-3833490; fax: +44-181-3831915
e-mail: rkh.wyse{at}rpms.ac.uk
Abstract
Objectives: Cardiac surgery is the first of any therapeutic discipline to attempt to establish a major international registry on a patient-by-patient basis. Being the first necessarily imposes a challenge since no rules or guidelines exist. Possibilities and boundaries must be identified. Methods: The EACTS database committee established ECSUR (The European Cardiac SUrgical Registry) with the aim of collecting patient-by-patient information from centres across Europe. It was decided to collect patient data in three areas; adult cardiac surgery, paediatric cardiac surgery and thoracic surgery. Currently, these three areas are in different stages of development, with multinational data collection now having already commenced in adult cardiac surgery. Results: After just over a year of data collection ECSUR has now received data from 98 centres in 30 countries, encompassing approximately 100 000 patients. Three national cardiac surgical registries currently send their data to ECSUR, and agreement has been reached with three other countries for prospective data sharing during 1999. Conclusion: This paper describes how ECSUR was established, the on-going status of the project, and how the project has now developed a global perspective through a joint European and US initiative to undertake international dataset unification for surgeons worldwide, by the definition of core adult cardiac, paediatric cardiac and thoracic surgical datasets.
Key Words: Adult cardiac surgery • Paediatric cardiac surgery • Thoracic surgery • Database
1. Introduction
In 1994 the European Association for Cardiothoracic Surgery (EACTS) established a database committee with the view of exploring how a variety of patient-related information might be shared by its members. The central notion was to provide a knowledge base for the better management of patients by EACTS members. From these discussions a strategy emerged to develop a major EACTS cardiac surgical registry. The aim was to centralise cardiac surgical data from a large number of cardiac surgical centres across Europe, and to provide appropriate analysis for participants. The project encompassed three specialties: adult cardiac, paediatric cardiac and thoracic, surgery. Funding was successfully sought for this initiative from the DG XII division of the European Commission. Based at the Hammersmith Hospital in London, the project commenced in December 1996. It was called ECSUR (The European Cardiac Surgical Registry).
Up to this point there were already two on-going pan-European initiatives both designed solely to collect overall annual numbers of cases in adult cardiac surgery (Professor Unger in Austria), and in paediatric cardiac surgery (Professor Munro in the UK). What ECSUR sought to achieve was patient-by-patient data collection allowing, for the first time, an approach to data collection and analysis that was directly aimed at improving patient selection and management.
In fact, at the commencement of the ECSUR project, several national patient-by-patient registries already existed in Europe in various stages of development, of sophistication, and in the range of the individual patient data they collected. At this stage, adult cardiac surgery data were being collected on a national basis by Germany, the Czech Republic, Norway, Sweden and in the UK. In addition, plans were underway to establish adult cardiac surgery registries in Belgium, Holland, Denmark and Switzerland. While no adult cardiac surgical registry outside of the US (STS database) was fully operational at this time (1996), it was clear that many felt the registry approach represented an important initiative to undertake on a national level. Now, in late 1998, most of the registries described above are either fully operational, or will become so during 1999. In addition, initiatives for establishing national adult cardiac surgery registries are currently underway in Ireland, France, India, Australia, Japan, and Hungary, and with several other countries actively considering the possibility.
It was envisaged that ECSUR would collect patient data both from these international registries, and also from individual centres. It was acknowledged that validation issues would differ between these two types of sources since most national registries would have already gone through quality control procedures of varying rigour. However, it was unlikely that individual centres would, without prompting, be so diligent in data validation in their day-to-day routines, so procedures would have to be put into place to rectify this for inclusion of their data into the registry.
It was recognised that for the most to be achieved from the ECSUR project some level of data unification was required across Europe; ideally with all centres collecting the same dataset, and using the same, or comparable, surgical diagnostic and procedure codes. The issue becomes most evident when the construction of the individual national registries is considered. Belgium currently collects just 10 data variables from each of its 30 centres, with Holland (
20), Germany (80), and the Czech Republic (500) being examples of countries independently choosing to adopt vastly different levels of complexity in their nationwide collection of adult cardiac surgical data. While ECSUR was designed from the outset for maximal participation this current variability in practice and sophistication across Europe meant that a modular approach to datasets within ECSUR would be necessary, from the simple to the more complex. The modular approach has the particular advantage that centres and national registries would find it straightforward to start to participate in the ECSUR project using a basic ‘minimum’ dataset (the entry level), and later they could easily choose to migrate to a more complex dataset (see below) as future needs changed.
Consequently the EACTS database committee advocated offering the approach of staged levels of ECSUR participation for centres. The collection of the basic ‘minimum’ dataset containing just eight variables, combines basic patient demographics with surgical procedures performed. The second level of complexity includes risk factor stratification. The third level includes outcomes. The fourth level is for research, using the existing infrastructure of patient data collection, centralised within ECSUR, as a framework for ad-hoc research projects across groups of centres and/or countries, and according to specifically-designed multi-centre clinical trial protocols.
2. Methods
Establishment of the ECSUR registry proceeded in five concurrent stages:
2.1 Development of a contacts database
In order to approach centres to send their patient data it became necessary to develop a sophisticated contacts database. What was needed was a centre-orientated database spanning the whole of Europe, and this had never been compiled before. The EACTS already maintained some information, but only on their membership, not the centres themselves. These records not only excluded non-members, but did not address which surgical units were actually currently operational in each country. The ECSUR team used several methods to collate contact information. These included writing to several EACTS members in each of 50+ nations for them to send their current intelligence of the contact information for all surgical centres in their own country. National associations and registries, where available, were contacted for their lists of members. Delegate lists from a number of previous cardiac surgical conferences were also scanned for additional surgical units. The result of these efforts has meant that the comprehensive contacts database has now become the central information base and administrative core for the ECSUR team.
Once armed with a substantial list of European cardiac surgical centres (>700) the ECSUR team sent a data template to each, requesting the following information:
Direct response rate from centres was high (see Section 3), and this data template was also later made available on the ECSUR website, with many responses subsequently received by this route.
The ECSUR contact database, held in Microsoft Access 97, was formed into two linked sections: centres and individuals. As well as providing simple and cohesive means to interact with participants, the contacts database is now regularly used to mail merge, for mass mailings of ECSUR information, or for internally managing our in-house developed data entry software used externally to assist units in their own patient data collection. Each centre was given a unique identifier within the contacts database, and for the identification of incoming patient data.
2.2 Establishment of the hardware and software infrastructure to manage the data flow
At the commencement of the project, the ECSUR team estimated that between 500 000 and 750 000 cardiac surgical procedures were conducted throughout continental Europe in 1996, some 3500 each weekday. It was decided that monthly, quarterly, or even yearly batches of data would be acceptable in the compilation of this patient data for subsequent analysis. However, the two powerful dual processor servers with RAID architecture installed at ECSUR are capable of being converted to daily automatic telephone on-line use if required. The ECSUR computers run Microsoft Windows NT 4.0 as an operating system, and have client machines for use by the staff. Received patient data are stored in Access 97, used as a front end. Each centre is given its own subdirectory for the storage of its data. Access 97 affords an easy medium to check the patient data for omissions or errors. The strategy is then to pass the quality-controlled data onto a more permanent and robust database, Microsoft SQL, which is a more appropriate repository than Access 97 for very large datasets.
Several methods for receipt of patient data were also developed. These included:
The possibility of developing Internet forms for data transfer was eventually disregarded on the basis that it would be inherently suspect in achieving 100% of all cases conducted in any centre; it was considered far more robust to retrieve data as a regular by-product of the normal patient management and reporting structure of the surgical unit.
Whenever possible, the ECSUR team has encouraged centres to transfer their data on electronic medium (Floppy discs, CD-ROMs, or by e-mail). In fact, ECSUR went further and developed free software, written in Access 97, to assist centres in collecting adult cardiac surgical data, for producing local unit reports, and for transferring it regularly to ECSUR (see Section 2.4).
2.3 Publicity
Awareness of the ECSUR initiative is vital for centres to start to send in their data. Widespread publicity has been approached in several ways:
2.4 The commencement of data inflow
After extensive pilot testing, the ECSUR project started receiving adult cardiac surgical patient data on October 1st, 1997. Data arrived from centres in a variety of formats, and contained a variety of surgical coding systems. There was a wide range of complexity: from simple datasets, including the ECSUR adult cardiac minimum dataset (see below), to highly complex home-grown systems.
Since a large number of centres across Europe had little or no appropriate patient management software, data collection software for the ECSUR adult cardiac surgical minimum dataset was developed by the ECSUR team. It was launched at the 1998 annual EACTS conference in Brussels. The software, running on Access 97, was configured to help centres to collect the minimum dataset to send to ECSUR. In addition, report writing capabilities were deliberately incorporated into the software so each centre could also use it to help run their department by producing local reports on their activities. The view was that this approach would ensure 100% entry of all patients operated upon by each unit. Floppy diskettes containing the software were given free to several hundred centres. Each diskette was unique. Each contained details of the individual centre, and its unique centre-identifier; so ECSUR would always know from which centre data was being received.
Negotiations to receive data from various national registries were commenced. In some cases this requires the protracted satisfaction of specific national legal and data protection issues. In some cases it has required changes to national law.
2.5 The unification of datasets
It quickly became clear that cardiac surgical centres in Europe, without any central and consistent model, had come to individual, and often idiosyncratic, decisions about what data they chose to collect on each patient, and what diagnostic and surgical procedural coding systems (if any) they adopted. However, to be meaningful, patient registries need a cohesive data structure.
Early efforts were made within ECSUR to make various coding systems correspond one to another; we found that more than 20 separate coding systems were being used across Europe. There was clearly a need to move towards a unified ECSUR standard. Even more important, it was recognised that ECSUR needed to identify and define a unified dataset for each specialty to give centres a clear statement about what data to collect. For those centres already collecting comprehensive patient information, using sophisticated patient management systems, it would simply be a matter of getting their local software to send to ECSUR only those variables that had been identified as core elements.
Work proceeded in defining standard core datasets for adult cardiac, paediatric cardiac, and thoracic surgery. This work is still underway and will be described later. However, the first standard produced by ECSUR was the adult cardiac surgery minimum dataset. This contains the following elements:
This minimum dataset aims to record basic information about the patient demographics and the procedure performed. As such it represents the lowest common denominator that is hoped all centres, no matter what their level of computer sophistication, will be able to provide for ECSUR. The Belgian national adult cardiac surgery registry is structured much along the same minimalist lines as this.
The content of the ECSUR adult cardiac surgery minimum dataset links directly to a much larger dataset containing risk factors, and pre-, peri-, and post-operative details (see later). The aim is to allow centres to choose at which level they wish to share data with ECSUR. It was hoped and indeed anticipated that many centres will wish to upgrade to the larger dataset as their computer systems become more sophisticated.
3.1 Development of a contacts database
ECSUR is currently in direct contact with centres in 68 countries, and has now received patient data from 30 of these countries (see Section 3.2).
As of February 1999, and now with a more worldwide perspective, the number of the centres with details listed in the ECSUR contacts database was 1352, and the number of individual cardiac surgeons in the linked ‘individual’ cardiac surgeons ECSUR database was 3951. These figures exclude any entries from USA and Canada since our colleagues in the STS already collate their local information.
The number of completed data templates received to date now exceeds 500, although ECSUR has already received data on almost all of the remainder of the centres from various other sources, particularly the existing national registries and societies in each country. Of the information ECSUR has received it appears that approximately 80% of these centres currently conduct cardiac surgery (with or without thoracic surgery also), the remainder are thoracic only.
A by-product of developing a comprehensive contacts database was that it allowed ECSUR to conduct some demographic analysis. For example, analysis of the population demographics of each country shows that none of the 15 EU member states has a ratio of national population to number of adult cardiac surgery centres of higher than 1.1 million per centre. The lowest ratio was in Belgium (333 267 with 30 centres covering a population of almost 10 million inhabitants). No other EU country had a ratio below 600 000. Only two countries, Netherlands (15 centres) and Portugal (nine centres) had ratios exceeding 1 million. These figures for EU member states compare favourably with some other countries, e.g. (Hungary: 1 469 857, seven centres; Turkey: 1 900 000, 31 centres; Bulgaria: 2 116 750, four centres; Ukraine: 4 014 000, 13 centres; Romania: 4 553 400, five centres; Kasachstan: 8 517 500, two centres; India: 9 476 000, 105 centres; Uzbekistan: 20 708 000, one centre). Twelve non-EU European countries had ratios exceeding 2 million; nine had achieved infrastructures with ratios under 1 million (Czech Republic, Cyprus, Estonia, Iceland, Israel, Latvia, Monaco, Slovenia and Switzerland). Such demographic information has important implications for deposition of centres within countries, and for the wider organisation of cardiac and thoracic surgeons under training. As such it is useful to national societies, surgical centres, EACTS and governments alike.
3.2 The commencement of data inflow
After just over a year of data collection, ECSUR has now received adult cardiac surgical data on approximately 100 000 patients from 98 centres in 30 countries (see Appendix B).
Three national registries, Belgium, Norway and UK now share their adult cardiac surgical data with ECSUR. In fact, annual adult cardiac surgical data was received for the following years from the national registries in Belgium (1997, 30 centres), Norway (1996 and 1997, seven centres) and the UK (1997, 14 centres). Negotiations to receive adult cardiac surgical data from other national registries have so far led to successful agreements with a further three to share their data with ECSUR during 1999. Several other national registries have indicated the likelihood that they will soon agree for their data to be shared with ECSUR. Furthermore, ECSUR is also actively helping several other countries inaugurate their own new national cardiac surgical registries; these will be established both for their own local uses, and to share their patient data with ECSUR.
Over this time ECSUR has received adult cardiac surgical data according to a wide variety of formats, particularly in Excel and Access. The development and distribution of data collection software for the ECSUR adult cardiac surgical minimum dataset in September 1998 gave a focus for many centres to start collecting data according to a standardised format. The large number of centres now using this software will be approached every 3 months for their accumulated patient data.
Analysis of ECSUR data will be sent to participating centres and national registries on a regular basis. While the current, descriptive report did not aim to present any actual data analysis, a number of manuscripts of this type are now in preparation based on the adult cardiac surgical data so far received by ECSUR. Furthermore, interactive data analysis will be placed on the ECSUR website if participants wish it.
Paediatric cardiac surgical data from centres across Europe has been jointly collected for several years by two members of the EACTS database committee (Bohdan Maruszewski in Warsaw and Martin Elliott in London). Initiatives described elsewhere in this paper should lead to the retrospective and prospective unification of this data into the new international paediatric cardiac surgical dataset, and to have it correspond directly to three separate and currently developing STS (US Society for Thoracic Surgery) paediatric datasets.
The collection of patient-by-patient thoracic data remains at the planning stage, both in ECSUR and in the STS. Some progress on the international thoracic surgical dataset has already been accomplished. This will correspond directly to the STS thoracic dataset but may be somewhat smaller in size for international use (see below).
3.3 The unification of datasets
This falls into three categories: adult cardiac, paediatric cardiac, and thoracic surgery. The STS has coordinated adult cardiac and paediatric cardiac databases in the USA and Canada for several years, according to strictly defined data templates. Being mindful of these STS initiatives the ECSUR team felt there was an opportunity, never before available, to move towards worldwide standards for adult cardiac, paediatric cardiac, and thoracic surgical datasets, by comparing, amalgamating and distilling the crucial elements from the US and European approaches.
With several EACTS members, including the ECSUR team, sitting on the STS database committee, it was mutually decided to establish joint STS-ECSUR subcommittees for each of the three specialties to set about defining worldwide standard datasets. The aim is to recommend two international datasets to use for standardised reporting of both adult cardiac, and thoracic, surgery on a worldwide basis: (1) minimum, and (2) more comprehensive (see next section). However, there will probably be only one international standardised (minimum) dataset for paediatric cardiac surgery (although the STS themselves are also developing larger datasets for their own internal use). Furthermore, the format of the datasets of these three specialties will be unified, so that centres conducting more than one type of surgery can simply report all their cases according to a standardised structure, particularly the patient demographics.
3.3.1 Adult cardiac surgery
ECSUR has already developed the adult cardiac surgical minimum dataset (described earlier), and many centres now regularly send ECSUR their patient data in this format. To add to this, the joint STS-ECSUR database subcommittee for adult cardiac surgery is now tasked to produce, by mid 1999, a unified larger dataset for worldwide use. Containing information on patient demographics, risk factors, pre-operative assessment, diagnosis, surgical procedures, operative details, post-operative course and outcomes, the aim is to produce a dataset that will:
The notion of the staged, modular approach for data collection mentioned earlier will remain in the sense that centres worldwide be still be able to choose at what level of complexity to collect their patient data, and share with ECSUR. The ECSUR minimum dataset will remain the entry level and will also form the core of the larger international dataset, thus ensuring total consistency between them. Similarly the STS and European database initiatives will unify by mapping directly to the international dataset.
3.3.2 Paediatric cardiac surgery
The European Congenital Heart Defects Database (ECHDD), operating out of London and Warsaw has been collecting data from paediatric centres across Europe for several years. Recent developments in the field include the construction of two alternative datasets. One of these (STS) is still under development but will probably consist of three levels of complexity. As with adult surgery, the joint STS-ECSUR paediatric surgical database sub-committee is charged with the task of producing a single international dataset from these various separate initiatives. For practical reasons, worldwide, it is likely that it will not only correspond closely to the smallest STS paediatric dataset (which contains basic patient demographic, diagnostic, procedural, operative and outcomes information), but also link directly with the two current European alternative paediatric datasets. This new international paediatric cardiac surgical dataset will be made available by the joint STS/ECSUR database committee in September 1999.
3.3.3 Thoracic surgery
As with the other two specialties, a joint STS-ECSUR thoracic surgical database sub-committee is tasked to create worldwide standard datasets. Again, the aim is to strike a balance between the practicalities of data collection by units worldwide, and the type of information that needs to be analysed to give better insight into patient selection and management. The joint STS/ECSUR database subcommittee will make available the international minimum thoracic surgical dataset by mid 1999, with the larger dataset (including risk factors) being launched during Autumn 1999. As with the other two specialties, the unifying aim of these core datasets is to gain rapid adoption by centres worldwide, whether or not they choose to collect additional patient information.
4. Discussion
This report, only part way through the initial start-up phase of the ECSUR project, is necessarily a snapshot of current and past activities. What is clear is that the establishment of an international registry, the first in any therapeutic area, is a massive undertaking. It has also required the ECSUR team to become involved in areas not envisaged at the planning stage. These include major efforts to define several new datasets, assisting national societies to establish national registries, and expansion of a European initiative to a world-wide perspective. These will be discussed in turn.
The requirement to produce clearly defined adult, paediatric and thoracic datasets might have been predicted, or at least suspected before ECSUR launch. However, it was only after dialogue with hundreds of centres that the lack of any cohesiveness in their approach became apparent; essentially no-one knew. We also identified the remarkable fact that there was concurrent use of more than 20 separate adult cardiac surgical procedural coding systems across Europe. Also, the lack of cohesive approach of the national registries was similarly striking, and reflected whether the prime instigators in each country were interested in basic assessment of activities, or of full risk factor analysis. Because of the understandably parochial viewpoints of these many centres and national bodies, it was clear that ECSUR needed to take a pan-European perspective, and that it would fall solely on us to work on developing cohesive European standards.
Part of the work on producing these standards was to assist a number of national societies to develop their own national registries, where they never existed before. When they each start to collect data, the intention is that each of these new national registries will also share their data with ECSUR. Several national registries are now in various stages of development, and ECSUR has advised them with both the technical aspects and the actual content of the dataset to be collected. We advise these new national registries to establish regular quality control and validity checks for completeness and accuracy on the data they receive from the centres in their country. In turn, each national society, their centres, and ECSUR all benefit from the efforts to generate high quality, validated data since it permits reliable, trustworthy and meaningful data analyses. Patient confidentiality laws must be adhered to, but vary in different countries. As in ECSUR, we advise national registries to maintain appropriate protection of the privacy issues of patients, surgeons and surgical centres. In the analysis phase, normally raw data is not shared, but relevant aggregates of all results are made available to participating surgeons and centres and, if requested, with a specific private report for each participant to help them benchmark themselves against the anonymized results of the other participants. To advise centres and national registries, on all these issues, within ECSUR there are skills of commercial and therapeutic area patient registries, computer technologies, cardiological and cardiac surgical coding systems, resource utilisation, cost-effectiveness, and a substantial understanding of national issues and healthcare systems in many countries. When they each start to collect data, the intention is that each of these new national registries will also share their data with ECSUR.
The expansion of the European initiative to a world-wide perspective was unanticipated, but is logical. Trying to counter the lack of cohesiveness of datasets across Europe coincided with our discussions with the STS database committee. From these discussions has emerged a shared determination to create datasets for adult cardiac surgery, paediatric cardiac surgery, and thoracic surgery that, for the first time, offer a clear unified global approach to everyone collecting data in cardiac and thoracic surgery. The aim is to provide an opportunity for each centre and national registry to collect a rational, pre-defined dataset, whether or not they choose for their own local use also to collect more. Any diversity remaining across these datasets is unimportant so long as a core dataset is shared by all, in each specialty.
As a result of our intention to produce world-wide datasets, ECSUR has expanded the remit of its contact database to collect information from centres in all five continents, and indeed already does so. We estimate there are approximately 2500 cardiac surgical centres world-wide, and expect our remit to encompass the 1500 outside of USA and Canada (with the STS subserving the same function there). ECSUR has already started to collect patient data from several centres in Asia and Australasia, and was instrumental in helping establish a national adult cardiac surgical registry in India. The expanded global role for ECSUR is consistent with the fact that the membership of EACTS is drawn from some 56 countries worldwide.
It is only when these core international datasets are published and adopted by many centres and national registries that the real value of this cohesive international approach for data collection will become apparent. It is our view that when ECSUR starts to receive data according to these unified larger datasets then local, national, and international idiosyncrasies of risk factor stratification will emerge. Centres will be able to conduct any number of important analyses such as to benchmark themselves (anonymously) for any procedure according to patient demographics and risk factors.
In the meantime ECSUR will continue to collect data from the many collaborating centres and national registries that already send it to us. The ECSUR team will also start to write analyses for participating centres, for participating national registries, and for publication in the medical literature. To receive data from 98 centres (covering approximately 100 000 adult cardiac surgery patients) in just over one year exceeded our expectations, and is starting to approach the critical mass of data needed to commence meaningful analysis. If requested, ECSUR would be able to place such analyses on its website, and these could (if centres wished) be accessed interactively using unique participant passwords. In theory, a centre could conduct its entire unit analysis from the ECSUR website (even if it had only sent us its data as paper records).
ECSUR is an on-going initiative. As it grows the main benefits will not only come to the national registries, but to the participating centres and the individual cardiac surgeons within them. We urge all readers who do not yet share their data with ECSUR to do so.
Acknowledgments
This work was supported by a grant from the DG XII division of The European Commission.
Appendix A
| ||||||||||||||||||||||
Appendix B
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
|
|
 |
|
  A. T. Pezzella Global statistics/outcomes. J. Thorac. Cardiovasc. Surg., September 1, 2006; 132(3): 726 - 726. [Full Text] [PDF]
|
|
|
|
 |
|
 W. B. Dobkowski Myocardial Protection During Minimally Invasive Cardiac Surgery Seminars in Cardiothoracic and Vascular Anesthesia, December 1, 2002; 6(4): 319 - 324. [Abstract] [PDF]
|
|
|
|
 |
|
 A. T. Pezzella Progress in international cardiac surgery: emerging strategies Ann. Thorac. Surg., February 1, 2001; 71(2): 407 - 408. [Full Text] [PDF]
|
|
|
|
 |
|
 R. K. Wyse and K. M Taylor Developing a Cardiothoracic Surgical Registry in Asia Asian Cardiovasc Thorac Ann, December 1, 1999; 7(4): 255 - 258. [Full Text] [PDF]
|
|
|
|
|
|
A T. Pezzella The Emerging Global Cardiothoracic Surgical Community: Need for Increasing Communication and Cooperation Asian Cardiovasc Thorac Ann, December 1, 1999; 7(4): 360 - 360. [Full Text] [PDF]
|
|
|
|
 |
|
 D.J. Wheatley Cardiothoracic surgery in Europe: politics, pressures and practice Eur. J. Cardiothorac. Surg., December 1, 1999; 16(6): 593 - 601. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
