A Bibliometric Analysis of Health Services Research Publications: Trends and Characteristics

Transcription

1 Facultad de Ciencias Médicas Universidad Nacional de La Plata Centro Interdisciplinario Universitario para la Salud A Bibliometric Analysis of Health Services Research Publications: Trends and Characteristics Paganini, José María. Universidad Nacional de La Plata. Centro Interdisciplinario Universitario para la Salud. La Plata. Argentina. Raiher, Sandra. Centro de Estudios de Estado y Sociedad. Ciudad Autónoma de Buenos Aires. Argentina. September 2006 Summary The purpose to achieve equity, quality and efficiency in health services requires the contribution of scientific knowledge to facilitate appropriate policy decisions. Although this is a valid requirement to all countries, it has even a greater priority in countries with scarce resources that need to accelerate its achievements towards equity and the millennium development goals. This bibliometric analysis of Health Services Research publications of the last 25 years indicates that there is a limited scientific production in Latin American and Caribbean countries in comparison with that in some developed countries. Key words: <HEALTH SERVICE RESEARCH> <BIBLIOMETRICS> <BIBLIOMETRIC INDICATORS> <SCIENTIFIC PRODUCTION INDICATORS> <MILLENIUM DEVELOPMENT GOALS> <HEALTH EQUITY> <HEALTHCARE QUALITY> <EFFICIENCY> <HEALTH SERVICES>

2 Introduction and Research Objectives Health Services Research (HSR) is an important component of Health Research. Its main purpose is to produce scientific knowledge for the development of evidence-based policies and programs towards equity, quality and efficiency in health.(1) The latest studies in this field have been indicating that, although there is an urgent need to improve the equity, quality and efficiency of health services, HSR is the least developed area of health research.(3)(4) (5) The main objective of this paper is to apply the bibliometric analysis to the HSR papers published in scientific journals with the aim of identifying tendencies, gaps and characteristics in comparing scientific production of some developed countries with Latin American and Caribbean Countries. These results are expected to be a contribution towards the support of new developments in this area. 1 Material and Methods Four data source are utilized. For a historical overview covering the period, selected papers retrieved to produce the PAHO/WHO publication Health Services Research: an anthology are analyzed here.(1) Medline, Pub Med and LILACS databases are further considered for a fourteen-year period (1991 to 2004) with some adjustment as necessary, depending on data availability. The historical overview involves the analysis of 275 papers. Although the articles published in the PAHO Anthology were limited to 100 articles, a "short list" of 275 HSSR papers was available to the authors including studies published from 1914 up to Although those publications are not a random representative selection of articles, according to the editorial committee they are believed to be representative of much of the best in the field when they were published, and as such they should be instructive. (1) 1 This analysis has been partially produced with the support of the AHPSR WHO as a contribution to Publication of the Alliance for Health Policy and Systems Research, Strengthening health systems: the role and promise of health policy and systems research. Geneva, Alliance HPSR,

3 Medline 2 database (National Libraries of Medicine, USA) contains approximately 13 million health-related references to journal articles. It includes information from around 4500 Journals and 70 countries. PubMed 3 database provides free access to MEDLINE, and includes access to additional selected life sciences journals not in MEDLINE and Links to the full-text articles at participating publishers' web sites. Thematic access to both bases was done through the Major Medical Subject Heading (MeSH) Health Services Research (HSR). All publications (clinical trials, editorials, letters, meta-analysis, practice guidelines, randomized control trials and reviews) are included with papers published in English, French, German, Italian, Japanese, Russian, and Spanish languages. LILACS 4 database (Latin American and Caribbean Health Scientific database) is the PAHO/WHO BIREME Regional Scientific Information network based in Sao Pablo, Brazil. It collects information from around 670 periodic journals, monographs, thesis and no conventional publications, with more than 350,000 bibliographic citations, since The period from 1991 to 2004 includes a total of 264,994 articles, including periodic journals (76.7 %), monographs (17 %), thesis (5.5 %) and non-conventional publications (0.8 %). The number of published articles adjusted by population indicates that Costa Rica, Uruguay and Ecuador are the three countries with a higher percentage (from 33.5 to 19.3 per 100,000 population) followed by Chile, Cuba, Argentina and Brazil (from 10.8 to 5.3). The countries with relative fewer publications are Peru and Mexico (2.4 to 1.7). This study is a descriptive comparative analysis of HSR publications by year, by country groups and by language. The following indicators are utilized: ratio of HSR original articles/total database, by year of publication, by groups of countries, by author s country of residence, by country of study and by language of publication. RESULTS Descriptive analysis of HSR articles by year of publication Although the PAHO/WHO Anthology on HSR is not a representative study, it is interesting to note that out of the total articles selected for the period 1914 /1990 (275), more than half of them (140 articles), that is, 51% were produced in the 90 s decade. (The distribution of the 275 selected articles includes one article published in 1914, two others published in the 30 s decade; 22 published in the 50 s decade; 35 in the 60 s decade; 73 in the 70 s decade, and 140 in the 80 s decade (two articles have no year identification). 2 Medline Database. Available in 3 PubMed Database. Available in 4 LILACS Database. Available in 3

4 The analysis of the Medline database for the period 1991/2004 indicates that, from the total of 6,474,082 articles included in the database, only 15,914 (2.4 per thousand) correspond to the Major subject heading (MeSH) Health Services Research (Table 1). The distribution of HSR articles per year and the number of articles per 1000 is as follows. Table 1. Medline Database.HSR articles per 1000 over total database Year HSR Articles Total database Articles per thousand Total The trend analysis from 1991 to 2004 indicates a clear increase of the number of HSR articles published: from 609 HSR articles in 1991 to 1471 in 2004 (241% increase). Nevertheless, when comparing this trend with the total articles in the database it is clear a slight relative increase, that is, from 1.56 in 1991 to 2.51 in

5 From these data we can infer a first conclusion indicating that, in spite of the increase in the total HSR articles per year over the last 14 years, this increase is still very small in comparison with the total Health Care publications. The LILACS database is also analyzed by identifying the number of publications under the subheading Health Services Research. (See Table 2). In the fourteen-year period, there is a total of 304 HSR articles, that is, 1.10 per thousand over the 276,403 total articles in the database. It is also interesting to point out that, unlike the Medline database, the trend analysis of the number of publications shows no increase and the ratio per thousand indicates a tendency to decrease. Table 2. LILACS database. HSR articles per 1000 over total database Year HSR Articles Total Database Articles per Thousand Total

6 Comparing those indicators with the Medline database, the difference between both databases is clear. In the Medline database, dominated by publications by authors from developed countries, the mean number of HSR publications per 1000 for the 14 years is 2.4 with a tendency to increase from 1.56 in 1991 to 2.51 in 2004 per On the contrary, in the LILACS database, dominated by authors from Latin American and Caribbean countries, not only is the number of HSR publications per 1000 twice lower than the Medline database in the last 14 years but also has a tendency to decrease over the years. (See Graph 1) Graph 1. Tendency of HSR articles over total database in developed (Medline) and developing conuntries (LILACS) ,5 3 3,06 2,98 3,05 2,5 2 1,5 1 0,5 0 2,61 2,64 2,4 2,41 2,22 2,07 2,12 2,14 2,21 2,1 1,67 1,56 1,52 1,27 1,31 1,36 1,31 1,22 1,25 1,01 0,89 0,62 2,51 0,25 0,34 Developing countries Developed countries This is another strong indication of the low priority given to HSR activities in developing countries in comparison with central developed countries. HSR publications by country of residence of authors The analysis of the number of HSR articles by author s country of residence in both PAHO/WHO publication and Medline database indicates a strong participation of articles that come from developed countries/central economies (95 percent). The concentration of publications in few developed countries is even more evident in the Medline database where only two countries (USA and England) produce 75% of the total articles. 6

7 This is a clear hint, in our opinion, not only of the different cultural and economic situations that influence the production of scientific knowledge but also of the different interests, incentives and opportunities that authors encounter when producing and publishing HSR papers. Publications by country location of the study Considering that the previous analysis of number of publications by country of residence of authors could limit information of the actual country where the research study was carried out, the PubMed database was analyzed by selecting 1811 articles in a systematic random sample of the 27,040 for a 10 year period ( ) The analysis of all selected articles revealed that 1715 (94.7%) publications reporting HSR projects were executed in developed countries (USA 993; UK 320; Canada 279; Australia 56; Spain 29; Denmark 10; Switzerland 17; France 11). Studies located in USA and England accounted for 77% of the total from developed countries. If Canada and Australia are included, they amount to 96% of this total. The studies from developing countries were a total of 96 (5.3%) (India 49, South Africa 13, Senegal 9, Chile 4, Mexico 5, Argentina 2, others 14). This result confirms the previous findings in the Medline database indicating a very low participation of developing countries in the production and publication of HSR articles. HSR articles by language of publication From the total articles in the Medline database, the most representative languages are English with 17,901 articles (91.6%), French with 401 (2.05%), German with 384 (1.9%) and Spanish with 236 (1.2%). While in LILACS database, the most representative language is Spanish with 113 (47%), following by Portuguese with 84 (35%) and English with 45 (18%). Collaboration index among authors The analysis of this index could provide information about some characteristics of HSR production that in theory ought to be a teamwork approach with an interdisciplinary content and with participation of several institutions. In the PAHO/WHO historical database the mean number of authors is 2.7 per article. Out of the total, 171 articles (62%) were produced by more than one author, and 105 (38%) were produced by only one author. The mean numbers of authors by year from Medline and LILACS database are shown in Table 3. 7

8 Table 3. Mean numbers of authors of the HSR articles in the Medline and LILACS database Year Medline database LILACS database It is interesting to note that the Medline database indicates a higher mean number of authors per article over the LILACS database. At the same time, the Medline database shows a slight increase in the mean number of authors per article. No clear tendency is observed in the LILACS database. 7. Conclusions and Limitations One of the main conclusions of the empirical analysis of HSR publications is the important gap identified between developed and developing countries in the volume of scientific papers in HSR. There is a permanent and continuous 5/95 difference (only 5% of the HSR publications belong to poor and developing countries), even more dramatic than the 10/90 gap (Global Forum for Health Research). 8

9 The two databases analyzed (Medline and PubMed) identify that a few central or developed countries using English as the main language of publication produce 95% of articles. At the same time, it is clear that in the Medline database there is a tendency to increase the percentage of HSR articles over the total. This tendency does not appear in the analysis of the LILACS database that includes only developing countries. This conclusion suggests that HSR is failing to attract talent and resources in developing countries vis-à-vis other disciplines, even though the potential of HSR to improve health could be greater in relation to other disciplines. There is certainly the possibility that some of the research activities and production of knowledge in most developing countries appear in grey publications or as evaluative research projects within technical programs. Nevertheless, the lack of communication and of quality control in most of the scientific publications and current databases makes these efforts practically useless and some times counterproductive by contributing to wrong or biased policy decision-making and a waste of scarce resources. Finally, among other possible limitations of this study, the main limitation could be the language bias with a strong English influence. It is recognized that not all scientific papers produced in developing countries have the same possibility to be published and to integrate currently available databases. 9

10 References 1.- Pan-American Health Organization. Health Services Research: an anthology. Washington, D.C., PAHO, (PAHO. Scientific Publication, 534) 2.- National Library of Medicine. NICHSR History of Health Services Research. HSR Class Manual Available online at: Global Forum for Health research. The 10/90 Report on Health Research Geneva: GFHR, Global Forum for Health research. The 10/90 Report on Health Research Geneva: GFHR, Global Forum for Health research. The 10/90 Report on Health Research Geneva: GFHR, Hanney SR, Gonzalez-Block MA, Buxton MJ, and Kogan M. The utilisation of health research in policymaking: concepts, examples and methods of assessment. Health Research Policy and Systems, 2003; 1:2. Available at: Task Force on Health Systems Research. Informed choices for attaining the Millennium Development Goals: towards an international cooperative agenda for health systems research. Lancet 2004; 364: Gonzalez Block M, and Mills A. Assessing Capacity for Health Policy and Systems Research in Low and Middle Income Countries. Health Research Policy and Systems 2003; 1 (1): Commission on Health Research for Development. Health Research: Essential Link to Equity in Development. New York: Oxford University Press, Nuyens Y. Recent developments in HSR Worldwide. Advisory Gropup on Health Systems research and Development. WHO: Geneva, 1993 (MAL/GAG/WP/1) 11.-Advisory Committee on Health Research. Report of ACHR peer review group on the Report of the Ad Hoc Committee on Health Research Relating to Future Intervention Options. Geneva: WHO, (Document ACHR 33/95.7) 12.- Ad Hoc Committee on Health Research Related to Future Intervention Options. Investing in Health research and development. Geneva: WHO, (Document TDR /Gen/96.1) Nuyens Y. Workshop on Priority setting for Essential National research (PSENHR) review of Processes, Mechanisms and Outcomes of PSENHR. Geneva: Council on Health research for development, Advisory Committee on health Research. A research policy agenda for science and technology to support global health development. Geneva: WHO, (Document WHO/RPS/ACHR/98.1) 10

11 15- Council on Health Research and Development (COHRED). Essential National Health Research and Priority Setting: Lessons Learned. Geneva: COHRED, (Document 97.3) Okello D, Chongtrakul P. A manual for research priority setting using the ENHR strategy. Geneva: COHRED, (Document ) Working Group on Priority Setting, Council on Health Research for Development. Priority setting for health research: lessons from developing countries. Health Policy Plan 2000; 15(2): The International Conference for Health Research for development, Bangkok 10/13 October Conference report, Geneva Global Forum for health Research. Annual Meetings. Available online at: Ghaffar A, De Francisco A, and Matlin S, editores. The Combined Approach Matrix: A priority-setting tool for health research. Geneva: Global Forum for Health Research, Available online at: The Collaborative Training Programme (CTP) Resource Modules on Health Research for Policy, Action and Practice, Version 2, Available online at: Lomas J, Fulop N, Gagnon D, and Allen, P On being a Good Listener: Setting Priorities for Applied Health Services research. The Milbank Quarterly, 2003; 81(3). Available online at: Klein R, Williams A. Setting priorities: what is holding us back inadequate information or inadequate institutions? In: Coulter A, Ham C. The global challenge of health care rationing. Buckingham: Open University Press, 2000; Bowling A, Jacobsen B, Southgate L. Health services priorities. Exploration in consultation of the public and health professionals on priority setting in an inner London health district. Soc Sci Med 1993; 37: Coburn A. The Role of Health Services Research in developing State Health Policy. Health Aff (Millwood). 1998;17(1): Ginzberg E, editores. Health Services Research. Key to Health Policy. Cambridge: Harvard University Press, Plsek P. Appendix B. Redesigning health Care with Insights from the Science of Complex Adaptative Systems. P In: Crossing the Quality Chasm: A New Health System for the 21 st Century. Washington D.C.: Institute of Medicine, ,- Daniels N, Sabin JE. Limits to health care: Fair procedures, democratic deliberation and the legitimacy problem for insurers. Philos Public Aff. 1997; 26(4):

12 29.- Varkevisser C, Pathmanathan I, Brownlee A. Proposal Development and Fieldwork. In Designing and Conducting Health Systems Research Projects. IDRC; WHO (Health Systems Research Training Series 2,1). Available online at: Vol2/part1-2%202003ed/Designing%20and%20conducting%20health%20system%20research%20projects,%20volume %201,%20Proposal(1).pdf 30.- Whitehead M. The concepts and principles of equity and health. Int Journal of Health Serv, 1992; 22: ISEqH. International Society for Equity in Health. Definitions. Available online at: Paganini JM. Salud y Equidad. La Plata: Centro INUS, p. 14. (Publ. Cientifica # 1) Report of the Advisory Committee on Health Research. Rev Panam Salud Publica. 1997; 2(6): Available online at: Paganini JM. Guía para el análisis del sector salud / Guidelines for the Analysis of Health Sector. Paper p. 38. Maestria Salud Publica. Centro INUS. Facultad de Ciencias Médicas. UNLP. Argentina. Health sector analysis. Paper. INUS WHO Choice project. Chosing Interventions that are Cost-Effective. Geneva: WHO, González-Block MA. Health policy and systems research agendas in developing countries. Health Research Policy and Systems 2004; 2:6. Available online at: Araújo Ruiz JA. Informetría, bibliometría y cienciometría: aspectos teórico-prácticos. ACIMED 2002, 04. Available online at: Cibermetrics. Bibliographic and patent databases for use in scientometrics, bibliometrics... Databases. Available online at: Martínez de Sousa J. Diccionario de bibliología y ciencias afines. Madrid: Fundación Germán Sánchez Ruipérez, Velho L. Indicadores científicos: aspectos teóricos y metodológicos. In: E. Martínez, ed. Ciencia, tecnología y desarrollo: interrelaciones teóricas y metodológicas. Caracas: Nueva Sociedad, UNESCO; Spinak E. Indicadores cienciométricos. Ci. Inf., 1998; 27(2):

13 42.-Sengupta IN. Bibliometrics, Informetrics, Scientometrics and Librametrics: an overview. Libri, 1992; 42(2): Vinkler P. An attempt of surveying and classifying bibliometric indicators for scientometric indicators for scientometric purposes. Scientometrics, 1998; 13(5-6): Spinak E. Quantitative analyses of scientific literature and their validity for judging Latin American Production. Bulletin of PAHO, 1995; 29(4): Moravcsik MJ. The classification of science and the science of classification. Scientometrics, 1986; 10(3-4): Farkas J. The science of science as a new research field and its function in prediction. In: Whitley R, ed. Social processes of scientific development. Boston: Routledge & Kegan, Garfield E. Análisis cuantitativo de la literatura científica y sus repercusiones en la formulación de políticas científicas en América Latina y el Caribe. Bol. Oficina Sanit Panam, 1995; 118(5): Harter SP. "The Impact of Electronic Journals on Scholarly Communication: A Citation Analysis." The Public-Access Computer Systems Review, 1996; 7(5) 49.- Loet Leydesdorff. Indicators of Innovation in a Knowledge-based Economy. Cybermetrics, 2001; 5(1) Available online at: Niessen L., Grijseels E, Rutten F. The evidence based approach In health policy and health care delivery. Inst. Of Medical Technology Assessment, Rasmus University, Rotterdam. p