Bibliometrics of global malaria vaccine research

Authors


K. C. Garg, National Institute of Science, Technology, and Development Studies (NISTADS), Dr. K. S. Krishnan Marg (Pusa Gate), New Delhi 110012, India. E-mail: gargkc@nistads.res.in

Abstract

Objectives:  This study evaluates malaria vaccine research carried out in different parts of the world during 1972–2004 using different bibliometric indicators.

Method:  Data have been downloaded from PubMed for the period 1972–2004 using the keywords (malaria* or plasmodium or falciparum) and (vaccine*) in the title and abstract fields. The study examined the pattern of growth of the output, its geographical distribution, profile of different countries in different subfields and pattern of citations using GOOGLE Scholar.

Results:  Malaria vaccine research output is gradually increasing. The USA, followed by the UK and Australia contributed the highest number of papers. Publication activity has decreased in Switzerland and Sweden, but has increased in Brazil and China. The majority of the countries have focused on the development of asexual blood stage malaria. Citations per paper and incidence of high-quality papers for the USA, the UK, Papua New Guinea and Denmark are more than the average. The majority of the prolific institutions are located in the USA, the UK, France and Australia.

Conclusion:  The last two decades have witnessed considerable growth in research output in this field, while a successful malaria vaccine still remains elusive. Interestingly, the countries like the USA, the UK and Australia that lead in the quantity, quality and citation of this output are often not those directly affected by malaria.

Introduction

The human race has stepped into the 21st century and, despite tremendous progress in malaria research during the past two decades, this disease still represents a major threat to populations in many parts of the world. As a result of poor epidemiology, the exact extent of the malaria problem, as with many other communicable diseases, is not known.1 However, according to experts, it is estimated that nearly 300–500 million people suffer from malaria each year and almost 2 million die—mostly children younger than 5 years old, with sub-Saharan Africa accounting for 90% of them.2 After the launch of the malaria eradication programme in different countries, the incidence of malaria and the death rate as a result of it dropped considerably during the 1970s.3 However, this disease emerged with a renewed vengeance in the decades since then. In addition to increased international air traffic, administrative lapses and poor financial support, two main reasons responsible for the resurgence of malaria are insecticide-resistant mosquitoes and drug-resistant parasites.4 With the worsening of the malaria situation worldwide, new international programmes have been initiated; these include the Multilateral Initiative on Malaria (MIM), Roll Back Malaria (RBM), the Global Fund for HIV, Tuberculosis and Malaria (Global Fund), The Medicines for Malaria Venture and the Malaria Vaccine Initiative.5

Few bibliometric studies dealing with malaria research have been reported in the literature in the past. Maclean et al.6 and Lewison et al.7 estimated the financial resources going into malaria research. Garg et al.8 estimated the quantum of malaria research output during 1990 and 2000 using PubMed (the online edition) and the Commonwealth Agricultural Bureaux International (CABI) CD-ROM incorporating the Tropical Disease Bulletin (TDB). Lewison and Srivastava5 mapped the malaria research output during the years 1980–2004 using the Science Citation Index (SCI). However, none of these studies deals with the status of malaria vaccine research, which constitutes approximately 9% of the total malaria research output.5 The present study is a step in the direction of filling that gap.

Objectives

The purpose of the present study is to analyse the scientific research output on malaria vaccine research published during 1972 to 2004. It aims to map the quantum of the world output in the field of malaria vaccine research. Specific objectives of the study are as follows:

  • • the pattern of the publication output during 1972–2004 in blocks of 3 years each and their language of publication;
  • • the geographical distribution of the research output and study of the transformative activity profile of different countries during 1983–1993 and 1994–2004;
  • • the activity profile of prolific countries in different subfields of malaria vaccine research;
  • • the pattern of citation of the research output and identification of highly cited papers;
  • • the citation impact of different countries using different bibliometric indicators;
  • • identification of the most prolific institutions involved in malaria vaccine research and study of their citation impact.

Methods

The data source for the present study was PubMed (the online edition), published by the National Library of Medicine (Bethesda, MD, USA). This is the world's leading English language abstracting and indexing service, providing information on all aspects of medicine, nursing, dentistry, veterinary medicine, the health care system and the preclinical sciences from more than 5000 biomedical journals published in the USA and 80 other countries.

The keywords used for searching were (malaria* or plasmodium or falciparum) and (vaccine*) in the title and the abstract fields for the period 1 January 1972 to 31 December 2005, yielding 2662 records. To ensure that no records for the year 2004 were left out of the analysis, we downloaded records up to 2005. A ‘screen dump’ of the actual search strategy with PubMed is given in Fig. 1.

Figure 1.

A ‘screen dump’ of the actual search strategy with PubMed

The search strategy yielded 2662 records that dealt with malaria vaccine until 2005. The bibliographic details for each record included: document type, title of the article, authors(s), affiliation and name of the journal where it was published. From these downloaded records, we deleted 241 records that were published during 2005 and 2006. From the remaining 2421 records, we also removed 101 records that were published as bibliographies, editorials, interviews, historical articles, lectures, letters and news items. The remaining set of 2320 publications was then manually reviewed in order to eliminate records that did not concern malaria vaccine research. In all, 313 such records were identified that did not deal with malaria vaccine. These 313 records dealt with different aspects of malaria research, such as the history of malaria, prevention strategies and treatment of malaria, seasonal profile of malaria infection, use of insecticide-treated nets against malaria, effect of changing environment on malaria control and drug development for malaria, etc., diseases such as acquired immunodeficiency syndrome/human immunodeficiency virus (AIDS/HIV), cancer, leptospirosis, schistosomiasis and tuberculosis, and different aspects of health hazards during international travel. After excluding these 313 records, the remaining 2007 records were assigned the subfields of the malaria vaccine research. The data set of 2007 records was converted into a database in FOX-PRO 25 (Microsoft, USA). The database created was further enriched with information on country of publication of the journals, subfield of vaccine research and the citations received by each paper. (The citations were examined using GOOGLE Scholar, as Science Citation Index (SCI) was not available at our work place at the time of undertaking the study.)

Results

Pattern of publication output

A total of 2007 publications scattered over 352 journal titles originating from 40 different countries were published on the topic of malaria vaccine. About one-third of these journals were publications from the USA. Appendix 1 lists the most commonly used journals where research on malaria vaccine was published. These journals published 1149 (~57%) papers and most of them originated from the scientifically advanced countries, unlike malaria research,8 where the journals originating from China and India are also listed among the most commonly used journals. The average number of publications per year was ~61. The pattern of output during the period 1972–2004 in 11 blocks each of 3 years is depicted in Fig. 2. As expected, the number of publications was less in the first five blocks, i.e. 1972–1986. This number started increasing during the later blocks and reached its peak in the year 1990–1992. After a slight decline during the periods 1993–1995 and 1996–1998, it started increasing again. The highest number of publications was in the last block, i.e. 2002–2004. English is the most preferred language of communication followed by Chinese and French. The number of articles published in English, Chinese and French were 1916 (95.5%), 32 (1.6%) and 26 (1.3%), respectively. The remaining 33 (1.6%) papers were published in 12 other different languages.

Figure 2.

Pattern of publications output

Geographical distribution of research output

During the period under study, 59 countries were involved in malaria vaccine research. Of these, 11 countries contributed to approximately 83.5% and the remaining 48 countries contributed approximately 16.5% of the total research papers. The USA, UK and Australia together contributed more than half (~57%) of the total publication output.

Distribution of the research output for different countries in three blocks of 11 years each (Table 1) indicates that only 12 countries produced 56 papers in the first block (1972–1982). The main contributors among these countries were the USA and UK. However, in the second block (1983–1993), many more countries joined the research effort, raising the number of countries to 37 and the publication output to 663, almost 12 times that of what it was in the first block. In the last block (1994–2004), the number of countries increased further to 50 and the output to 1288, almost twice as much as the output in the second block.

Table 1.  Output in malaria vaccine research in different blocks
Country1972–19821983–1993 (TAI)1994–2004 (TAI)Total 1983–2004Change in TAIGrand total (%)
  1. TAI, Transformative Activity Index.9

USA23267 (109)456 (96)723(–) 13746 (37.2)
UK1576 (103)142 (99)218(–) 4233 (11.6)
Australia259 (111)97 (94)156(–) 17158 (7.9)
Switzerland254 (167)41 (65)95(–) 10297 (4.8)
France128 (87)67 (107)95(+) 2096 (4.8)
Colombia20 (83)52 (109)72(+) 2672 (3.6)
India216 (68)53 (116)69(+) 4871 (3.5)
Sweden26 (132)32 (84)58(–) 4858 (2.9)
Brazil3 (17)50 (143)53(+) 12653 (2.6)
China4 (25)44 (139)48(+) 11448 (2.4)
Japan7 (48)36 (127)43(+) 7943 (2.1)
Others11103 (94)218 (103)321 332 (16.5)
Total5666312881951 2007 (100)

The Transformative Activity Index (TAI) suggested by Guan and Ma9 was developed to compare the relative change in the output during the last two blocks, i.e. 1983–1993 and 1994–2004. The TAI has been calculated only for the last two blocks, as the publication output in the first block of five countries is zero. The methodology to calculate TAI is similar to that used for calculating the activity index and was used earlier by Karki and Garg10 and Kumar and Garg11 in their studies on alkaloid chemistry research in India and computer science research in India and China, respectively.

Mathematically,

TAI = [(Ci/Co)/(Wi/Wo)] × 100

where Ci is the number of publications of the specific country in the ith block, Co is the total number of publications of the specific country during the period of study, Wi is the number of publications of all countries in the ith block and Wo is the number of publications of all the countries during the period of study.

The values of TAI for different countries (Fig. 3) suggest that the publication activity has decreased considerably for Switzerland and Sweden. In the case of Switzerland, the decrease is the highest and is relatively high compared with the USA, UK and Australia. For the other countries shown in Fig. 3, there is an increase in the publication activity as reflected by the values of the TAI. The increase is highest for Brazil and China.

Figure 3.

Change in the values of the Transformative Activity Index (TAI) for different countries

Activity profile of different countries in different subfields of malaria vaccine

Based on the life cycle of malaria parasite, three main types of malaria vaccines currently under development are (i) pre-erythrocytic, including liver stage, (ii) asexual blood stage and (iii) transmission-blocking vaccines. In addition, research is also being undertaken on the development of a multi-stage malaria vaccine. Based on the above, the total published output has been divided into five classes, which include three types of malaria vaccine and their trials, and others that included popular and general information articles on malaria vaccine.

To compare the relative research effort between different countries during the period of study, the Activity Index (AI) was used as the absolute publication output is affected by the size of the country as well as the size of the subfield. The AI was first suggested by Frame12 and later elaborated upon by Schubert and Braun.13 The AI characterizes the relative research effort a country devotes to a given subfield. It is the ratio of the country's share of the world's publication output in the given field to the country's share of the world's publication output in all science fields, expressed as percentage. An AI score = 100 indicates that a country's research effort in the given discipline corresponds precisely to the world average. An AI > 100 reflects higher than average effort and an AI < 100 indicates a lower than average effort by the country.

Here,

AI = [(Nij/Nio)/(Noj/Noo)] × 100

where Nij is the total number of publications of a country in a subfield j, Nio is the total number of publications of the country i in all the subfields, Noj is the total number of publications for all the countries in the subfield j and Noo is the total publication output for all subfields for all countries.

The values of the AI for different subfields of malaria vaccine for countries producing 2% or more papers are given in Table 2. The values of the AI (Table 2) suggest that certain countries concentrate their research effort only in one subfield of malaria vaccine, while some other countries distribute their research effort in more than one subfield. Figure 4 presents the activity profile of different countries in malaria vaccine research. The USA, China and Japan concentrate their research effort in two subfields. Australia, Sweden, Brazil, China and Japan focused their research effort on the development of an asexual blood stage malaria vaccine. However, other subfields on which the publications of China and Japan concentrated were trials of a malaria vaccine and a transmission blocking vaccine, respectively. The USA emphasis was on pre-erythrocytic and transmission blocking vaccines and the UK, Switzerland and Colombia placed emphases on a transmission blocking vaccine, a pre-erythrocytic vaccine and trials of a malaria vaccine, respectively.

Table 2.  Publication output (activity index) of different countries in malaria vaccine research
CountryPre-erythrocytic, including liver stage vaccineAsexual blood stage vaccineTransmission blocking vaccineTrial of malaria vaccineOthersTotal
USA204 (137)180 (77)53 (137)147 (100)162 (92)746
UK32 (69)73 (99)21 (174)33 (72)74 (134)233
Australia15 (48)78 (157)1 (12)23 (74)41 (110)158
Switzerland39 (202)20 (65)0 (0)19 (99)19 (83)97
France20 (105)33 (109)1 (20)15 (79)27 (119)96
Colombia13 (91)25 (110)0 (0)26 (183)8 (47)72
India9 (64)25 (112)3 (82)10 (71)24 (143)71
Sweden2 (17)38 (208)1 (33)10 (87)7 (51)58
Brazil5 (47)23 (138)1 (36)8 (76)16 (128)53
China5 (52)19 (126)2 (80)13 (137)9 (79)48
Japan5 (58)21 (155)6 (269)3 (35)8 (79)43
Others   51   97   15   90   79332
Total4006321043974742007
Figure 4.

Activity profile of different countries in malaria vaccine research

Pattern of citation of research output

An examination of the citations of the research output indicates that 2007 papers received 35 299 citations during the period 1972–2004. The average rate of citation was ~18. Analysis of the citation data indicates that, of the 2007 published papers, 264 (~13%) papers did not have any citation and the remaining 87% had one or more citations. Of the cited papers, ~26% papers were cited between one and five times and the remaining 74% were cited more than five times. Thirty-six papers (~2%) received more than 100 citations. The number of papers being cited more than the average is 546. Of the 264 non-cited papers, 191 were published prior to 2000 and the remaining 73 during 2000–2004. Of course, it is very possible that papers published during the later period will get citations in the future. The results of the citation analysis are given in Table 3. Appendix 2 lists highly cited papers that received 125 or more citations. Of the 25 highly cited papers, 14 are from the USA, five from the UK, two from Kenya and one each from Denmark, Sweden, Spain and Tanzania. The majority of the highly cited US papers are from the National Institutes of Health.

Table 3.  Distribution of citations
Number of citationsNumber of papersTotal citationsNumber of citationsNumber of papersTotal citations
0264011–203845724
112112121–301804523
210120231–40903200
312236641–50542416
48232851–75865162
58643076–100373201
6–103642902> 100366724
Total200735 299
Average 18

Prolific countries and their citation impact

Countries that have published 1% or more of the papers have been considered prolific. The following indicators of impact suggested by Nagpaul14 and used by Garg and Padhi15 have been used for inter-comparison of quality.

Citations per paper (CPP).  Based on the publication output and the number of citations received by these papers, citation per paper for different countries and different institutions has been calculated. Citation per paper has been calculated by using the following formula:

(total number of citations for a country or an institution/total number of papers for that country or institution).

Number of high-quality papers (NHQ).  Citation per paper for different countries and institutions was calculated as the pattern of citation varied from one country to another country. Papers that received more than twice the average citations have been considered as high-quality papers.

Relative Quality Index (RQI).  This indicator is the ratio of the proportion of high-quality papers (NHQ%) to the proportion of the publications (TNP%).

NHQ% = [(number of high-quality papers for a country or an institution/total number of high-quality papers)] × 100
TNP% = [(total publications output of a country or an institution/total publication output for all countries or institutions)] × 100

The measure relates to the incidence of high-quality papers for a country or an institution. A value of RQI > 1 indicates higher than average value, whereas a value of RQI < 1 indicates lower than average quality.

Table 4 lists 17 countries with their total number of publications, total citations, CPP, NHQ and RQI for different countries. The average value of CPP is 18. The value of CPP is the highest for the USA. Other countries that have a higher than average value for CPP are the UK, Spain, Papua New Guinea and Denmark. In the case of Australia, the value of CPP is close to the average. For the other countries listed in Table 4, the value of CPP is less than average.

Table 4.  Prolific countries and impact of their output
CountryTNPTNCCPPNHQRQI
  1. CPP, citations per paper; NHQ, number of high-quality papers; RQI, Relative Quality Index; TNC, total number of citations; TNP, total number of papers.

USA74617 926241381.46
UK2334 663 20391.32
Australia1582 67617220.92
Switzerland971 0161140.32
France961 0481140.33
Colombia721 0251460.66
India71497 720.22
Sweden587461350.68
Brazil53402 8
China4849 1
Japan434811120.36
Netherlands375161440.85
Germany333341020.48
Thailand323141020.49
Papua New Guinea224661951.65
Denmark214051931.13
Spain204022020.80
Others1652 33314140.67
Total200735 29918254

The standing of different countries judged from the values of the RQI indicates that the USA, the UK, Papua New Guinea and Denmark have more than average incidence of high-quality papers, as the value of the RQI is more than 1 and, for the other countries, the incidence of high-quality papers is less than average. For Papua New Guinea, the value of the RQI is significantly higher compared with other countries, which implies that it has outperformed other countries listed in Table 4.

Prolific institutions and their citation impact

Total publication output during 1972–2004 came from 420 institutions spread among 59 countries. However, most of these institutions were concentrated in the USA, the UK, France and Australia. The number of institutions from these four countries was 117, 44, 24 and 17, respectively. The remaining 218 institutions were spread throughout 55 other countries. Institutions that have contributed 1% or more of the total output have been considered as prolific. Based on this parameter, 21 institutes fall into this category (Table 5) and are scattered in nine countries. Of the total output, these 21 institutions contributed 908 publications (45%). Of these, eight institutions from the USA contributed 24% of papers, four institutes from the UK contributed 6% of papers and two institutes each from Australia and Columbia contributed 4% and 3% of publications, respectively. The remaining five institutions, one each from Sweden, France, India, Switzerland and Papua New Guinea, contributed 8% of publications.

Table 5.  Most prolific institutions and the impact of their output
InstituteTNPTNCCPPNHQRQI
  1. CPP, citations per paper; NHQ, number of high-quality papers; RQI, Relative Quality Index; TNC, total number of citations; TNP, total number of papers.

 1  National Institutes of Health (USA)1123 96835312.2
 2  Naval Medical Research Institute (USA)772 67035222.3
 3  Walter Reed Army Institute of Research (USA)691 62424111.3
 4  Queensland Institute of Medical Research (Australia)659191441.7
 5  Centers for Disease Control and Prevention (USA)618521481.0
 6  Naval Medical Research Center (USA)541 48127131.9
 7  New York University (USA)531 35526172.5
 8  University of Stockholm (Sweden)497041450.8
 9  John Radcliffe Hospital (UK)341 30138102.3
10  University of Edinburgh (UK)341 02330112.6
11  Institut Pasteur Paris (France)324761541.0
12  International Center for Genetic Engineering and Biotechnology (India)323821220.5
13  London School of Hygiene and Tropical Medicine (UK)313481120.5
14  Universidad Nacional De Colombia (Colombia)316652251.3
15  F. Hoffmann-La Roche & Co (Switzerland)302871020.5
16  University of Maryland (USA)293741320.5
17  National Institute for Medical Research (UK)246182662.0
18  Walter and Eliza Hall Institute of Medical Research (Australia)244912162.0
19  Papua New Guinea Institute of Medical Research (Papua New Guinea)234632051.7
20  Fundacion Instituto De Inmunologia De Colombia (Colombia)22125 600.0
21  University of Hawaii (USA)222351110.4
Total90820 361221671.4
Other 399 institutions109914 93814870.6
Grand total200735 299182541.0

Table 5 lists 21 institutions with their total number of publications, total citations, CPP, NHQ and RQI. These impact indicators show that the value of CPP for 11 institutions is more than average and for the remaining 10 institutions is less than average. Of the 11 institutions for which the value of CPP is more than average, five are from the USA, three from the UK and one each from Colombia, Australia and Papua New Guinea. The highest value of CPP is for the John Radcliffe Hospital (UK), followed by the National Institutes of Health (USA) and the Naval Medical Research Institute (USA). All institutions except the University of Stockholm (Sweden), International Center for Genetic Engineering and Biotechnology (India), London School of Hygiene and Tropical Medicine (UK), F. Hoffmann-La Roche & Co (Switzerland), the University of Maryland (USA), Fundacion Instituto De Inmunologia De Colombia (Colombia) and the University of Hawaii (USA) stand out because of a higher than average incidence of high-quality papers as the value of the RQI is > 1.

Discussion

Malaria vaccine research constitutes about 9% of total malaria research output. Bibliometric analysis of global malaria vaccine research is important to obtain crucial policy insights into this aspect of malaria prevention research. Despite the fact that the research and development of vaccines for malaria has become the focus of considerable research endeavour during the last two decades, a successful malaria vaccine still remains elusive. However, according to recent studies, the phase I and phase II trials of a malaria vaccine (RTS, S/AS02) are showing some promising results and it is hoped that a malaria vaccine will be available in the near future.16,17 The meagre output of publications in malaria vaccine research in the first five blocks (1972–1986) indicates that the advances in malaria vaccine research are fairly recent developments. There have been attempts globally in the last two decades to develop vaccines, not only against malaria but also other diseases as well. Large investments were made in the 1990s, especially to promote malaria vaccine research, as is evident from global consortia such as the Global Fund for HIV, Tuberculosis and Malaria (Global Fund) and the Malaria Vaccine Initiative, etc. This is also evident from the steady increase of publication output after 1986, with the highest number of publications in the last block (2002–2004), with an increased number of countries involved in malaria vaccine research. As in other fields of science and technology, the USA is the leading country in publication output, possibly as a result of federal funding from different agencies such as the National Institutes of Health and the Ministry of Defence for Malaria Research. Although in absolute terms the outputs for the USA, UK and Australia have increased in the last block (1994–2004), their proportion in the world output as reflected by the values of the TAI has decreased during the same block. The proportional outputs for Switzerland and Sweden have also decreased in the last block, but a steep rise in proportional output has been observed for Brazil and China. One of the possible reasons behind the rise in China's output may be that China also started publishing in the English language, which was not the practice earlier. Different countries concentrate their research efforts in different sub-disciplines of malaria vaccine research. The development of an asexual blood stage malaria vaccine has received more attention compared with both pre-erythrocytic and transmission blocking malaria vaccines. This finding is in accordance with that which has been stated by Breman et al.18 Three-quarters of the papers are cited more than five times, while one-quarter received an above-average number of citations. The pattern of citation varies among different countries and most of the highly cited papers originated from the USA. Citation per paper is also highest for the USA. However, most of the developed countries have a less than average value of citation per paper, which implies that their work is not often quoted. The value of RQI is highest for Papua New Guinea followed by the USA. Most of the prolific institutions are situated in the USA. Among the prolific institutions, the John Radcliffe Hospital (UK) had the highest value of citation per paper, while the University of Edinburgh (UK) had the highest value for RQI.

Key Messages

Implications for Policy

  • • The recent growth in published literature on malaria vaccine research indicates further scope for research and development until a successful vaccine is developed.
  • • The growth in published literature also indicates the important role of funding in encouraging further research towards vaccine development.
  • • Unless countries that need the malaria vaccines the most invest substantially in indigenous development of a malaria vaccine, they will be reduced to mere markets for importing vaccines.
  • • Countries such as India, with significant research capacity and investments in the field, are ideally placed to take proactive measures in the development of effective and affordable vaccines for domestic and other price-sensitive markets.

Implications for Practice

  • • Researchers from malaria-affected countries may wish to join the growing research activity to benefit from the increasing global interest and funding for vaccine development against malaria.
  • • Countries that lack a research and development infrastructure to develop the vaccine locally may use bibliometric trends and classifications to analyse the inherent diversity in approaches to vaccine development.

Acknowledgements

The authors would like to thank the two anonymous referees for their valuable comments and suggestions.

Ancillary