Working Paper by Great Britain : Report of a Visit to a Pharmaceutical Research Facility (WP.258)
Click here to view the Disclaimer.
Click here to return to the list of Working Papers.
AD HOC GROUP OF THE STATES PARTIES
TO THE CONVENTION ON THE PROHIBITION
OF THE DEVELOPMENT, PRODUCTION AND
STOCKPILING OF BACTERIOLOGICAL
(BIOLOGICAL) AND TOXIN WEAPONS
AND ON THEIR DESTRUCTION
BWC/AD HOC GROUP/WP.258
9 January 1998
Original: ENGLISH
Ninth session
Geneva, 5-23 January 1998
Working paper submitted by the United Kingdom of Great Britain
and Northern Ireland
REPORT OF A VISIT TO A PHARMACEUTICAL RESEARCH FACILITY
INTRODUCTION, EXERCISE OBJECTIVES AND SCOPE
1. The United Kingdom recognizes that visits, such as random or clarification visits, have implications for the pharmaceutical industry. For this reason the United Kingdom and a major pharmaceutical company conducted a practice visit to explore the issues more thoroughly. The need to explore the Protocol's potential impact on a large site, of which only a small part carries out activities which might be declarable, was one of the company's main reasons for wanting to conduct a practice visit.
2. There were six exercise objectives:
(i) General objectives. To simulate as far as practicable a visit to a large multi-discipline industrial site of a type likely to be declared under the Protocol, and to evaluate the effectiveness of the visit in providing transparency and building confidence when managed access techniques are used. For exercise purposes we assumed that a pilot plant within a large research site was the triggered facility.
(ii) Declarations. To evaluate the process of reviewing a declaration made previously by the site, and to address problems arising from differences in interpreting either the declaration requirements or the declared data. To consider how a visit should go about evaluating the relationship of activities central to the declared data with other activities on the site.
(iii) Site preparations. To evaluate the role of government advisers in briefing management of a large industrial site about the concept of visits, and in advising and helping management to make specific preparations for a visit. To consider how site preparation can best assist the site management in protecting Commercial Proprietary Information (CPI) whilst addressing visit team's questions.
GE.98-60040
BWC/AD HOC GROUP/2P.258
page 2
(iv) Access to site activities. To determine how much access can be provided so as not to impinge on CPI:
(a) to activities relating to declarations, either relating to the triggering of the declaration or to information required in the declaration format; and
(b) to other site activities when this is useful to provide an overview of site functions.
(v) General organizational aspects. To identify any practical problems for industry and government advisers in handling such a visit, especially in regard to communication and organizational difficulties on a large site, and to keeping control of CPI at all times.
(vi) Sampling. Although no sampling was to be performed in the exercise, note would be taken of any occasions when the site management felt that in a real visit it would offer to analyse a sample, using host test procedures under the observation of the Visiting Team (VT).
SUMMARY OF EXERCISE ACTIVITIES
(a) Pre-visit activities
3. Two weeks before the visit, the government home team (GHT) met senior site representatives to simulate submission of the declaration and awareness-raising activities. The site was given a briefing package, including guidelines outlining procedures for preparation and for use of managed access techniques to protect CPI. It also included the procedural document which set out the rules for the visit and defined the VT's rights and obligations as well as those of the declared facility.
4. Following receipt of the "notification", the GHT went to the facility six days before the VT was due to arrive. The visit mandate was given to site representatives, and a review of the declaration took place. The GHT clarified questions about the declared information, including the scope of the associated scientific and technical functions and the range of materials produced in the pilot plant. The GHT briefed personnel likely to be affected by the visit on the background on: rules and procedures; the areas, documents and personnel likely to be of interest to the VT; and on managed access techniques.
5. Site representatives outlined their policies on interviewing, photography, sampling and analysis; and discussed the level of access that could be offered to documents and to non-declared areas. In particular, contingency plans were made to allow access to, for instance, culture collection lists without revealing CPI, and suggestions were made on possible ways to resolve difficulties in allowing access to contained facilities.
BWC/AD HOC GROUP/2P.258
page 3
6. On the day that the visit was due to commence, the GHT arrived at the facility a few hours before the VT to run through the opening presentation and advise the site representatives on their final preparations.
(b) Visit activities
7. The VT was limited to three people and the visit itself to two days. The visit began with a briefing covering three topics: the company's general structure and the functions of its various sites and associated companies; the triggered pilot plant; and a detailed description of the pilot plant itself. Published company documents were also provided.
8. VT questions covered inter alia the location and nature of animal holding facilities; the treatment of pilot plant wastes; identification of areas with vaccination requirements; content of research projects supported by pilot plant activities in the last year; distribution within each project of fermentation batches produced in the last year; the process by which requests for materials were submitted and how materials were transferred to other areas after downstream processing.
9. The VT accepted a tour of the pilot plant covering the fermentation rooms, the control room, downstream processing areas, assay laboratories, cold and raw material stores and the deactivation room. The VT and company personnel discussed pilot plant usage and containment features, and procedures for dealing with spills and leaks. The VT and senior site representatives spent 40 minutes discussing the declaration.
10. Interviews were conducted with five other individuals: scientists from the biomolecular structure and processing section, the animal house, a doctor from occupational health, and a pilot plant technician. These covered inter alia the relationship between research elsewhere on the site and the pilot plant; the use of animals with material made in the pilot plant; the nature of requests for production and the transfer of materials within the site; vaccination policy and how an individual's vaccination status is recorded; and whether there were batch records that could be examined.
11. The VT requested details of batch runs for one fermenter which were associated with a specified research project, and attempted to follow the audit trail pertaining to these requests. The company was able to show the code from a batch run which did not contain CPI, and volunteered this information to demonstrate the linkage between a request for material, its production and subsequent use in a research lab. Risk assessment documents were also made available to assist the VT. Site management saw no occasion where it would have been useful to analyse a sample in order to address a VT question.
BWC/AD HOC GROUP/2P.258
page 4
MAIN LESSONS
(a) Site preparations
12. Preparations by facility representatives and the GHT played a useful part in helping the company cope with the visit. Review of the declaration allowed the company to anticipate some of the issues raised by the VT and to consider responses. The detailed GHT briefing and advice given before the visit assisted the company in formulating policy on access to facilities and documents, interviewing and photography and sampling. The measures that were put in place to control information release, such as including a scribe in the home team to record details of all information passed to the VT, were successful in protecting CPI. For the company the pre-visit briefing was an essential element in the exercise.
13. As a result of the company's focussed preparatory work the GHT met no substantial problems during the visit. One lesson was the need for the GHT to scrutinise all documents before they are given to the VT. Time needs to be set aside for this. This may help to identify and clarify at an early stage any anomalies arising from documented information.
(b) The Declaration
14. Discussion of the completed declaration format did not reveal any substantive
shortcomings in either its design, or in interpretation of its questions. The only problems concerned questions which used terms from existing CBMs (detection, identification and diagnostic techniques, physical protection techniques and decontamination, and/or disinfection techniques). The company indicated YES to these questions. However, following closer scrutiny it emerged that the work done only referred to process development work on plant commissioning rather than to investigation or development of these techniques as a goal in itself. Such misunderstandings can be avoided by including guidance notes making clear that such activities are excluded from the scope of the declaration.
(c) Access
15. Although the visit focus was the declared pilot plant, the company was prepared to allow access to certain other areas in order to address specific questions that could have arisen from on-site activities in the pilot plant. However, the VT only had occasion to interview one scientist in his office which was located in another building. Contingency plans existed to offer controlled access to animal facilities, including highly contained laboratories, using video cameras and floor plans. The company would not have been able to prepare thoroughly had the entire site been open to the VT automatically. Access to all areas of the pilot plant itself presented no problems.
16. All requests for interviews were granted. The interviews that were conducted proved useful in clarifying and confirming facts. These played a critical role in helping the VT
BWC/AD HOC GROUP/2P.258
page 5
understand the overall operation of the site, and especially the relationships between the pilot plant and other research units elsewhere on the site.
17. Access to documentation was strictly controlled. All requests, such as those for batch records, were subject to managed access techniques. Information was either re-packaged to excise CPI, or CPI was blanked out in the versions handed over to the VT. There was usually enough information on the pages in this latter category to obtain an insight into the general nature of the work and confirm that it was consistent with other information acquired by the VT. It proved difficult however to track the details of requests for batches of material and their subsequent removal from the pilot plant because batch reference codes were blanked out to protect CPI. During discussion on vaccination policy the company said that it would be possible to have managed access to vaccination records which might have proved useful in confirming information given in the declaration and during the visit.
18. During discussions on the relationship between specific batches and specific elements in the overall research programme - the details of which were also confidential - it initially proved difficult to obtain a reassuring connection between requests for material and specific batches, and any subsequent work on the material. However, the company was able to identify one specific batch reference code which did not contain CPI. The VT could see the originator's initials on the production request and was able to trace the computer print-out to a specific laboratory notebook; a copy of the print-out had been pasted into the relevant scientist's laboratory notebook and had been initialled. Adjacent pages confirmed the nature of the request and some of the work done on the material. The company agreed that in fact the only CPI in any of their batch reference codes was the three letters referring to the protein: all other details on the fermenter number and originator's initials could be shown to a VT, enabling an audit trail to be established.
(d) Transparency
19. One of the arguments in favour of visits is the contribution that they might make to transparency of microbiological activities. On the one hand this visit demonstrated that caution must be exercised in drawing general conclusions about transparency values after a single visit given the differences there may be between sites within specific companies, both within countries and between countries.
20. On the other hand, the experience of the visit also suggests that if an inspectorate regularly visits a variety of facilities, including research ones such as the pilot plant, then they would be better placed to assess and interpret the evidence. They would know, for example, that the record keeping systems likely to be encountered in research facilities not operating on either Good Laboratory Practice or Good Manufacturing Practice would place limitations on their ability to conduct detailed audit trails. They would thus be less likely to see anything sinister in such situations.
BWC/AD HOC GROUP/2P.258
page 6
21. Moreover, the real benefits to transparency may only arise over a much longer time-frame - probably over 5 to 10 years - in which visiting teams become better versed in the type and nature of sites, and what they could expect to see at particular categories of facility. Furthermore, they would be able to observe directly how changes in technology and procedures as well as regulatory frameworks shape microbiological activities in different countries. Transparency would be improved through the cumulative effect of many visits across a representative range of facilities over a period of time.
(e) Confidentiality
22. The company protected CPI throughout the visit: at no time did it come close to losing control of CPI. There were six reasons for this. First, the procedures governing the visit made it clear that all on-site activities were subject to managed access. Second, the VT size was limited to three people and was not allowed to sub-divide. This meant that senior site representatives had a complete overview of all the information passed to the VT. Third, preparation and pre-briefing of company personnel helped ensure that those involved were better equipped to handle VT requests. Fourth, strict documentation control ensured that no CPI sensitive documents were inadvertently handed over: all non-published company documents were scrutinized by company personnel before they were released. Fifth, the visit's limited duration kept potential exposure to CPI to a minimum and reduced the risk that company personnel might lower their guard. Sixth, re-packaging information in summary form to exclude CPI enabled the company to discuss key areas of its activities.
23. Tight control of information made it difficult for the VT to acquire the data it needed. However, given the wealth of open literature about the facility's activities coupled with the comprehensive briefing, the control needed to protect CPI did not make the VT's task impossible. Summaries of research projects and the distribution of fermentation batches between them contained sufficient information to allow the VT to make the necessary connections between activities in various parts of the site. Such linkages are essential for building confidence since they indicate that what is being seen is a true reflection of activities underway at a declared facility.
24. During a real visit, the company's level of confidence in the ability to protect CPI might be different depending on the VT's behaviour and approach. However, the experience of the 12 OPCW (Organization for the Prohibition of Chemical Weapons) inspections held in the United Kingdom to date suggests that inspection teams do conduct themselves in a professional manner, and do not push for access to CPI or national security information. Similar modalities and training would be required for a future inspectorate to ensure that a uniform and measured approach applied in all visits.
BWC/AD HOC GROUP/2P.258
page 7
CONCLUSIONS
25. The main conclusions are:
(a) No CPI was divulged, as a result of careful preparatory work by the company, which was facilitated by the detailed briefing and advice given by the GHT a week before the visit, and due to the use of managed access techniques throughout;
(b) Given the substantial staff time (an estimated 35-50 staff days in this case - including some essential senior management time) taken up in preparing and hosting the visit, the company representatives would much prefer that there should be no on-site activities other than challenge inspections. However, they acknowledged that the burden should not be intolerable for such a site if visits were operated in the same manner as in the exercise, that is a two day visit by a three man team;
(c) The exercise confirmed the potential utility of visits as a means of increasing the understanding of information in a declaration and how it related to specific activities in a declared area and elsewhere on a site; and of resolving any ambiguity or anomaly. It was not designed to provide significant insights into site activities having no scientific or technical links with the declared activities, and did not do so;
(d) Although the VT eventually acquired some understanding of the way in which this largely research facility operated, it was nonetheless a time consuming and at times difficult task. This is likely to be the norm at facilities that do not operate quality systems such as ISO 9000, whereas at manufacturing facilities more formalized record keeping may be expected which would facilitate the VT's task. Even in this research context the VT did gain insights into the commercial and regulatory rationales reflected in the physical attributes and operating modalities of the declared activities. This would be useful in informing an inspectorate about operational trends in United Kingdom industrial facilities;
(e) There was a range of opinions among the company representatives on whether visits offered any advantage for industry. Some felt that the effort expended in preparing and hosting a visit outweighs any benefit that might accrue. Further work on visit modalities might be useful in this regard. However, there was an acceptance of the view that, should visits be included in a future regime under the BTWC, this would be because of a need to strengthen global security through increased transparency;
(f) The constraints chosen for this exercise (similar ones could well apply in a future Protocol) in respect of the small size of the visiting team and the short time allowed for the visit, and the resulting limitation in the possible degree of scrutiny of site activities as a whole, suggest that such procedures should only address transparency objectives and should not be used to determine the compliance of the facility or the site with the BWC;
BWC/AD HOC GROUP/2P.258
page 8
(g) Overall, the results suggest that although the transparency benefits from a single visit in isolation may be limited, the accumulation of experience over several years from visits to facilities in the various regions would represent a valuable additional level of information on relevant activities that would not otherwise exist.
---------