Bibliometric data in clinical cardiology revisited

Authors: T Opthof A A M Wilde

Publish Date: 2011/04/28

Volume: 19, Issue: 5, Pages: 246-255

PDF Link

Abstract

In this paper we assess the bibliometric parameters of 37 Dutch professors in clinical cardiology Those are the Hirsch index hindex based on all papers the hindex based on first authored papers the number of papers the number of citations and the citations per paper A top 10 for each of the five parameters was compiled In theory the same 10 professors might appear in each of these top 10s Alternatively each of the 37 professors under assessment could appear one or more times In practice we found 22 out of these 37 professors in the 5 top 10s Thus there is no golden parameter In addition there is too much inhomogeneity in citation characteristics even within a relatively homogeneous group of clinical cardiologists Therefore citation analysis should be applied with great care in science policy This is even more important when different fields of medicine are compared in university medical centres It may be possible to develop better parameters in the future but the present ones are simply not good enough Also we observed a quite remarkable explosion of publications per author which can paradoxical as it may sound probably not be interpreted as an increase in productivity of scientists but as the effect of an increase in the number of coauthors and the strategic effect of networksIn April 2009 we published a bibliometric analysis of the work of 28 Dutch professors in clinical cardiology 1 We scored the number of papers excluding abstracts and letters and analysed their citation as well In addition we determined the Hirsch index 2 hindex Hirsch proposed the hindex as an alternative parameter for measuring the scientific output of an individual 2 Hirsch stated ‘I propose the index h defined as the number of papers with citation number higher or equal to h as a useful index to characterise the scientific output of a researcher’ The hindex is a very simple and useful albeit imperfect parameter When an author has an hindex of 50 it means that ‘he’ has published 50 papers that have each been cited 50 times or more The remainder of ‘his’ papers are cited less than 50 times Therefore the effect of a very large number of citations to a single paper eg to a clinical trial in which the scientist was just ‘one of the 100 authors’ is mitigated whereas authors who publish many papers in low impact journals eg for educational reasons are not punished compared with a system that is based on citation per paperAt that time with 1 June 2008 as the date of assessment we observed a huge variability in hindex The highest hindex was 95 for Dr PW Serruys The lowest hindex was 17 Obviously and very different from financial indices the hindex can only increase with time Thereby it may favour the ‘scientific establishment’ to which professors in clinical cardiology may be considered to belong It has been suggested by Hirsch himself 2 that this may be solved by dividing the hindex by ‘scientific age’ which could be defined as the time since publication of a first paper Such a correction may be a disadvantage for anyone taking an alternate path in his/her career It may also be a disadvantage for young medical students who are successful early in their career and then proceed with intensive internships with little scientific output for a substantial number of years Finally it is a drawback for those who have invested a considerable time over many years to raise a family as was pointed out by Spaan 3 When we nevertheless applied this correction we found Dr Serruys again at poll position with 297 hindex 95 divided by 32 his first paper being published in 1977 Now another professor received the red lantern with a score of 074In 2009 we made the choice to publish these data with the names of these professors in alphabetical order thereby downtuning the impression of quality ranking We have been criticised for not publishing our data anonymously Although we have sympathy for that point of view we feel that the readability of our paper at that time 1 and again now would suffer from anonymous data Moreover the data are available for those with entrance to the Web of Science of Thomson Reuters which is normally the case in academic institutions One of the reasons is that we have emphasised that it is far from certain that even a relatively homogeneous group of scientists as professors in clinical cardiology can be assessed with similar methods To assess this it is unavoidable that the subfields are uncovered and this requires linkage to individuals As long as the methods have not been validated with adequate and published research we feel that citation analysis should be discouraged in complex organisations as university medical centres and universities This position has been at the heart of a clearcut disagreement with the previous management of our own university medical centre Academic Medical Center Amsterdam including its research council 4 5

Keywords: