COMMENTARY
Further reflections on open access
Kimberly Parker*
Department of Knowledge Management and Sharing, World Health Organization, 20 avenue Appia, 1211 Geneva 27, Switzerland
The round table in last month's Bulletin stated that "it is only through removing the barriers to access to global research that health improvements can be accelerated".1 Similar thinking inspired the founding publishers and WHO to establish the Health Access to Research programme (HINARI) in 2001. I agree with the round table's authors that the topic is a critical reminder for policy-makers as the world continues to struggle with those barriers.
HINARI is a public-private partnership and our main concern is to make research available to developing countries. The partnership focuses on that goal and does not attempt to evaluate the relative advantages of any particular publishing model.
I would like to correct a few small errors in the way HINARI is described in this round table. HINARI is portrayed as a means for developing country institutions to have access to commercial publisher journals. However, in addition to commercial publishers, HINARI has a wide range of open access and local publishing partners such as MedKnow, the Public Library of Science (PLoS) and Bioline International, with which two of the authors of this round table are affiliated.
The authors state the importance of exposing research "from both research communities in developing countries as well as from 'international' research".1 In fact, HINARI and its related programmes - Access to Global Online Research in Agriculture (AGORA) and Online Access to Research in the Environment (OARE) - currently have 127 publishers that usually charge subscription fees and 89 publishers that distribute their content without subscription fees. All three programmes do more than just providing access to published research. The programmes also supply training on how to search the scientific literature and use its resources to inform decisions. Another important function of HINARI, AGORA and OARE is to collaborate with other international agencies and partners to build the technological infrastructure that institutions need to access information.
I am puzzled by the concern expressed by the authors that selection of content for HINARI is not driven by science (i.e. contents are determined by whatever publishers wish to donate). The content initially offered by each publisher is further expanded and refined through ongoing feedback from the programme and users with a view to including all that is relevant - for HINARI this is biomedical and related social sciences literature. Such an inclusive policy aligns with the open access goal of providing access to what is published and letting the reader decide what is of most interest.
The authors state that access is available only from registered libraries and on provision of a password controlled by libraries. This is not quite accurate. HINARI registers many tiny research institutes, very few of which actually have a library and some of which are research institutes of only one or two people. Passwords are provided to directors of institutions and librarians, if there is a library. If there is no library, the director may designate another person to assist in the password distribution. In this way, even the smallest AIDS research station can have the benefit of the information access and HINARI becomes their library.
The authors quote a letter to PLoS Medicine noting an analysis done at Universidad Peruana Cayetano Heredia on access to HINARI in April 2007.2 This letter mentions that one possibility for a drop in HINARI users was an increase in users of commercial content aggregations like ProQuest and Ebsco. Further studies are needed: both to capture the complex picture of information access in developing countries as people gain more access to the world's research output and to design the spectrum of responses that are required to address their needs. HINARI, AGORA and OARE review their activities on a regular basis and use the results to ensure the achievement of objectives and improve the quality of service.
HINARI is only one way to lower the barriers to information access and there are many others. It is unlikely that any one solution will meet everyone's needs. The key is to develop and perfect many overlapping solutions and methods, so that all researchers can access the scientific information published globally by their predecessors and contemporaries, however that access is provided. 
Competing interests: I am the programme manager of HINARI.
References
1. Chan L, Arunachalam S, Kirsop B. Open access: a giant leap towards bridging health inequities. Bull World Health Organ 2009;87:631-5. doi:10.2471/BLT.09.064659
2. Villafuerte-Gálvez J, Curioso WH, Gayoso O. Biomedical journals and global poverty: is HINARI a step backwards? PLoS Med 2007;4:e220. PMID:17594171 doi:10.1371/journal.pmed.0040220
* Correspondence to Kimberly Parker (e-mail:parkerk@who.int).