The Biosafety Clearing House - Safe, Clear and Friendly?1
by Beth Burrows, Edmonds Institute and Lim Li Lin, Third World Network
As a central database of biosafety information, the Biosafety Clearing House (BCH) being constructed for the Parties to the Cartagena Biosafety Protocol and the interested public should be expected to meet the highest standards of equity, usability, reliability, transparency, and security. This analysis raises concerns about the BCH as it currently exists in "pilot" form. Also included are suggestions for ameliorating the problems identified.
Countries that cannot - for whatever reason - use the electronic BCH fully will be disadvantaged, in some cases permanently.
1.1 It is NOT correct to say that whatever appears on the BCH can be accessed in another form, particularly in the case of data derived from BCH interfaces with other electronic databases. The owners of those interfaced databases, for example, have no legal duty to provide CPB Parties who may be non-electronic users of the BCH with complete copies of whatever is in their databases.
Further, copies of material from other databases would not be "comparable" to what might be found by using the electronic BCH because, at the very simplest level, the paper user would not be able to find the relevant data as quickly as the electronic user, and might effectively have less time to make decisions and request further information of exporters.
Finally, paper and CD versions of selections from the BCH would not be sufficient replacement for the BCH itself (as was once suggested). Who is competent to judge for all which information to select and distribute on paper and who will be responsible to ensure accuracy and comprehensiveness of information, or if key information is omitted leading to faulty biosafety decision-making or unnecessary trade disputes?
1.2 For developing and least developed countries, the costs of acquiring compatible equipment and software, trained personnel, and internet, telephone, and electricity connection is more expensive or proportionately/comparatively more expensive than in many developed countries. This from the outset will disadvantage the poorer countries.
Further, in many countries, while equipment and software and reasonable access to internet facilities may be available, it may not be sufficient for the task of accessing and reading all parts of the BCH, including interfaces with other databases. For some countries, lack of the latest or reasonably updated/upgraded equipment and software may mean that some documents and data are inaccessible on or through the BCH, and/or where poor internet and telephone connections are facts of life, that the cost of being on-line is expensive or prohibitive, thereby leading to longer periods of time required for decision-making, higher costs overall, and a decision-making context in which good science may be judged too expensive and time-consuming to pursue.
1.3 The BCH must be made available in all the languages of the UN; and this must include material found on any of the links that are created from the BCH to other sites or other databases. There should be no exceptions to this, lest some CPB Parties inadvertently be given a comparative advantage in their biosafety decision-making.
(a) Capacity building - Create a fund and the training capacity to ensure that all countries can fully use the electronic BCH. This, of course, would mean sufficient funds be made available for equipment purchase and maintenance, electricity generation (or provision of equipment that can use solar energy), adequate and fast telephone lines, and hiring and/or training of personnel.
(b) The MOP must consider that any country that is a Party to the CPB and does not have capacity - for whatever reason - to fully use the electronic BCH will be allowed to disallow the importation of any GMOs (or products thereof) until such time as the Party is able to avail itself of the electronic BCH.
2.1 Unfortunately, the BCH is difficult to read and to follow. While the toolkit and any CDs or trainings that are being held for users will doubtlessly be of help to some future users, nothing will avail if the BCH itself continues to be difficult to read. At the moment, its font sizes are much too small to be readable, even for readers whose native tongue is English. The BCH should be designed for maximum readability.
With little interactivity at the site and a layout and font size that promote squinting but not understanding, especially by anyone new to the site or to electronic media, the site will seem unusable to many.
2.2 The language of the site (e.g., "portal" etc.) is not communicative to everyone.
Solution: Redesign the pages to have more white space and much larger fonts. Make the entire site more interactive. Make the language simpler, less jargon-ridden. Consider that in many places the users may not have sufficient knowledge about computer use. It should not be assumed that users - even governmental users - will all have attained the highest standards of professionalism in both biosafety and information technology.
2.3 The site should be designed with the most naive user in mind. It must be easy to navigate. One never has a sense of what to do next; it always seems a guessing game. When clicking on an option to find some information, one is immediately confronted with a search page. This is immensely disconcerting as it puts off the casual browser (particularly bearing in mind that the BCH is also meant for public use). It assumes that one has to know what he/she is looking for.
Potential solution: After logging on, the site user might like to be asked at the outset: "Have you been here before?" If he/she answers "yes", the next question might be "How do you want to proceed?" and the choice of answers might be "with the aid of menus and questions (like this page)" or "with no further aids"?
Whatever choice the user made about aids, the user would be told what kind of equipment and software is necessary for using the site (This will prevent users with equipment/software that cannot handle the size of the site from proceeding.) There should be some instructions for those whose find their computer and software too outmoded for the site.
2.4 Although it is certainly useful that trainings and CDs are being provided for new users, current CDs and trainings may not help future users. Those being trained now tend to be management and diplomatic personnel and not those who will be entering and retrieving data from the website once the Protocol comes into force. It would seem imperative that the site itself contains an interactive program for new users to go through the site, ask questions, and test their knowledge.
Solution: There must be training provided for the actual personnel that will be responsible for the BCH in every country.
2.5 Another problem hindering usability is the fact that the toolkit is also difficult to read or follow. It seems designed for the technical person, not for the secretarial personnel most likely to be interfacing with the BCH. The toolkit cannot replace training modules at the site. The language of the toolkit is too difficult for someone who just wants to learn how to use the site.
Solution: Simpler language, greater interactivity, availability in all the languages of the UN -- all these would help the toolkit.
3. Reliability and transparency
The information in the BCH comes from many sources. Much of the information will only be in the form of "summaries" and much of it may come from other databases. While the breadth of sources may lead to greater information, it also creates profound problems.
3.1 It is imperative that the site user knows at every point where the information he/she is viewing comes from, what are the standards for its accuracy, and how he/she may obtain fuller information or pose questions. Incomplete or misleading information can be as dangerous as incorrect information.
In large measure, entry of data into the BCH is an act of compliance with the CPB. Where errors occur in the data found on the BCH, there must be a method of verifying who made the error (and possibly, who neglected to catch the error).
There may be liability implications to some errors; information has to be traceable back to its source. Simply stating that liability must be borne by the person or entity that has submitted the information is insufficient-it may place liability on the responsible party but this does not prevent misinformation from being used or relied upon. The mechanism of having the Secretariat check back with the Party entering data before that data appears on the site is also not a sufficient mechanism for having accurate data.
The problem of liability is most acute in the case of data coming from (the interface with) databases that may have different standards for completeness, different ideas of accuracy, and different legal consequences for inaccuracy than those connected with the BCH as stipulated by the MOP.
Potential interim solutions:
(a) Ensure at every point that the source of each bit of the data is apparent to the site user, that there is an easily-accessible name connected with the data, and an email address and a phone number so that users with questions have a process to seek answers.
(b) Request that the MOP (i) set a time line for responding to requests about BCH data, and (ii) create a mechanism for correcting data, and for resolving data disputes.
(c) Further, the MOP should create a mechanism whereby users - including non-Parties and members of the public - may challenge the accuracy of data and have their challenges duly considered and resolved. Without such mechanisms, biosafety errors may persist in the data of the BCH for years and influence many (costly) decisions about transboundary movement.
(a) Decisions related to ensuring the quality of data must reside with the MOP, not with the Secretariat or with the managers of other sites with which the BCH interfaces, or with the BCH site designer (or later, with the site manager).
(c) The MOP will have to decide what constitutes sufficient data and what are the mechanisms for ensuring data is complete and accurate, whose is the job of ensuring that accuracy and completeness, and what are the penalties for entering incomplete or inaccurate data or refusing to provide complete and accurate data. This kind of discussion, of course, leads us into unresolved issues of liability related to data entry.
(d) The MOP may wish to consider that entering data on the site is a tacit agreement to provide complete and accurate information to site users. The MOP will also have to make the relevant decisions about the locus of liability for incomplete, misleading or incorrect information.
3.2 Another problem presented by the interface with other databases is this: by creating interfaces only with those databases which comply with guidelines for "interoperability", the designers of the BCH site have effectively made a political decision for the MOP.
(a) The decisions as to which databases may be included in the BCH must be decisions on the part of the MOP, not on the part of site designers.
(b) The decisions about what is the standard for inclusion on the site and which is the desirable form of data must be made by the MOP. The technical specifications for the BCH must not be the engine that drives the decisions of the MOP -- it must be the other way around.
The site must be made consistent with (i) MOP needs for and decisions about BCH readability, (ii) MOP-created mechanism for ensuring that the decisions of the Parties are in no way altered by the writing of technical specifications for the BCH, and (iii) MOP decisions about sharing the interoperability guidelines with whomever the MOP designates, including non-Parties and NGOs. It may be that MOP decisions vis-a-vis the BCH will have more to do with political and scientific elegance than technical elegance.
3.3 The problem of what is in the BCH - i.e., whose databases are there - is not simply resolved by rules for completeness of data. The MOP must decide what information from non-Parties must be available to the BCH and whether non-Parties have "rights" of information refusal - and how non-Party information may be made available and its quality and completeness assured. This would of course include considerations of liability for incomplete or inaccurate information.
3.4 The MOP will also have to resolve the clash of public need for information with industrial need for CI, and this may best be approached through an ad hoc working group on the subject.
For the Pilot Phase, it was decided that no confidential information would be placed on the BCH. This should not be the case for the BCH when the CPB comes into force. If only information that is considered not confidential is placed on the BCH, it will be up to Parties to exchange information on a bilateral basis using other means of information exchange. Such bilateral arrangements, while protective of CI, may not lead to efficient or timely biosafety decision-making. In particular, molecular data for unique identifying purposes should be placed on the BCH and should not be classified as confidential information because this is a critical biosafety issue.
Which information is determined to be confidential will vary from country to country. If mechanisms of information verification (or warrantee) are created for the BCH, those mechanisms may not apply to CI. Without such verification (and the liability implied by it), Parties may have no way of verifying each other's CI.
3.5 In regard to information, another dimension must be considered: the ongoing need of Parties for up-to-date scientific information about biosafety. Particularly in countries where access to scientific journals is insufficient (for whatever reason - cost, language, etc.), this dimension must be addressed. Also, a link to the public DNA database, GenBank, should be provided for easy access to all the genome sequences.
Potential solution: The BCH can list all the relevant scientific journals, together with links to their websites. Further, the Secretariat can open discussions with journal editors (particularly peer-reviewed journals) with an eye to making all journals publicly available six months after they are published. If this were possible, it would greatly aid scientific capacity building within the context of the CPB, the CBD, and elsewhere.
4.1 Still another informational question presents itself, this for security reasons: information on the BCH must have backup.
Solution: The MOP must create fully secure mechanisms for backing up and storing material in the BCH, on a daily, ongoing basis. For a variety of reasons, not the least of which is the need to substantiate information useful to establish liability, the BCH should not be used until there is a backup mechanism that can provide time-dated information about any changes in the BCH.
4.2 In regard to security, the BCH (and its backup) must be unhackable (and those who create the BCH or its backup or work on them must be bonded) while at the same time those whose information appears on its pages must be able to update their pages. This is no small security matter, particularly with databases that are being made "interoperable". The MOP must decide procedures for altering data of any kind. The procedures should be revisited every time the MOP meets (or more often than that, if new developments warrant).
Security of the BCH is linked to security at all the sites to which the BCH links. Although it may not be easy in the future to change material on the BCH itself, it may be quite easy to change those pages to which the BCH links and from which the Parties gather information used in biosafety decision-making. The MOP may need to create policies and mechanisms to ensure that sites to which the BCH links remain, credible and verifiable.
Solution: The MOP - after soliciting and consulting papers from Parties, technicians, and NGOs - should set whatever level of security and whatever security protocols deemed necessary for BCH and linked sites.
5. Other details
5.1 It is very good that all the details of the roster of experts is made available on the BCH. More information should be made available so that BCH users may be better able to assess the quality of the expertise of those listed. The current information is still incomplete and not fully searchable.
Solution: No expert's name should be allowed to be listed without complete information about the expert's employment history, including sources of recent grants ("recent" being in the last 5-10 years, the exact period to be designated by the MOP), recent research, business partnerships and stock holdings. Without that information, BCH users may not fully be able to assess the quality of the expert or the "meaning" of his/her decisions. A declaration of conflict of interest should be required and this should be updated periodically.
Whatever is achieved by the Pilot Phase - however technically brilliant it may seem or however much hard work went into its construction - it should be understood that the Pilot Phase is only one step in a very long lesson about how to create a BCH that serves all the objectives of the CPB and the CBD. No detail of it should go unexamined by the MOP; no detail of it should be exempt from revision where such is deemed necessary by the MOP or any of its members. Indeed, the MOP should encourage people to complain about problems with the BCH -- such complaints will only serve to improve its operation and quality (and push the complainers to higher levels of biosafety sensitivity).
The BCH has the capacity to serve key needs of industry, providing-it is hoped-in the near future, a central database on national and international biosafety laws and regulations. Whether the BCH will be of equal use to countries and peoples remain to be seen. Hopefully, with sufficient funding, capacity, and political will, the BCH will usher in a new age of scientific openness and sharing.
This paper reflects many comments and concerns of the following individuals and organizations about the Biosafety Clearing House:
L. Bereano, Washington Biotechnology Action Council
1 Because the BCH is work in progress, it may be that some of the concerns expressed in this paper have already been addressed. Also, a paper of this sort is by its very nature critical in perspective, and as such, this paper focuses mainly on the deficiencies perceived by the authors.