Decision on own-initiative inquiry OI/3/2014/FOR concerning the partial refusal of the European Medicines Agency to give public access to studies related to the approval of a medicinal product
Case OI/3/2014/FOR - Opened on Wednesday | 16 April 2014 - Decision on Wednesday | 08 June 2016 - Institution concerned European Medicines Agency (No further inquiries justified )
This inquiry is concerned with how the European Medicines Agency (EMA) should deal with requests for public access to documents containing information on the safety and efficacy of medicines. The specific focus is on the right of public access to three clinical study reports on Humira, a widely sold anti-inflammatory drug.
In 2013 EMA decided to grant public access to these reports. However, the pharmaceutical company that sells the drug (AbbVie) took court proceedings against EMA which had the effect of blocking the proposed release of the reports. In 2014, before the conclusion of the court proceedings, EMA and AbbVie made an out-of-court agreement by which EMA would grant public access to redacted versions of the reports. The Ombudsman contacted EMA to check whether the redactions were justified. Following this check, the Ombudsman was not convinced that all of the redactions were in fact justified. The Ombudsman then began an inquiry, on her own initiative and in the public interest, into EMA's approach to the granting of public access.
In the course of the inquiry it emerged that EMA, in response to other public access requests for the same reports, had released much fuller versions of them. Nevertheless, certain redactions remained.
The Ombudsman accepted that some of these redactions were justified (because of the need to protect personal data). But she was not convinced that other redactions, made to protect commercial interests, were justified. In these latter instances, the Ombudsman took the view that there was likely to be an overriding public interest in disclosure. As a general observation the Ombudsman noted that, where the information in a document has implications for the health of individuals (such as information on the efficacy of a medicine), the public interest in disclosure will generally defeat any claim of commercial sensitivity. Public health should always trump commercial interests.
In closing the inquiry, the Ombudsman recognised that EMA has now made very significant progress with its proactive transparency policy, effective since January 2015. However, in relation to some specific portions of the reports, the Ombudsman questioned EMA's continued reliance on the protection of commercial interests. With a view to promoting systemic improvements, the Ombudsman made several suggestions to EMA as to its future practice in this area.
The background to the inquiry
1. This own-initiative inquiry concerns the manner in which the European Medicines Agency (EMA) dealt with a request for public access to clinical studies relating to Humira, an anti-inflammatory drug used to treat illnesses such as rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, Crohn's disease, ulcerative colitis and psoriasis. The inquiry is of public interest, not only because it concerns the transparency of reports aimed at assessing the safety and efficacy of a widely-used medicine, but also for the approach it proposes to providing public access to all clinical studies held by EMA.
2. In 2012, EMA received a request for public access to three clinical study reports submitted to EMA in support of requests to amend the Humira marketing authorisation: studies CSR M02-404 and CSR M04-691, related to the request to extend the marketing authorisation to include treatment of Crohn's disease, and study CSR M05-769, submitted with a view to updating the summary of product characteristics and removing a recommendation of use.
3. EMA decided to grant access to the reports in January 2013. However, before that decision could be implemented, AbbVie, the pharmaceutical company that currently markets Humira, asked the EU General Court to annul EMA's decision to release the reports and to grant interim measures blocking the release of the reports until the General Court had ruled on that request for annulment.
4. When the Ombudsman became aware of the steps taken by AbbVie, she requested the General Court to allow her to intervene in the court case. The General Court agreed to the request in September 2013. The Ombudsman then submitted her intervention to the General Court in which she explained why she considered that the reports should be released.
5. In the meantime, on 25 April 2013, the President of the General Court had granted AbbVie's request for interim measures, which effectively blocked EMA from releasing the reports until the General Court ruled, in the main proceedings, on whether EMA was correct to consider that the reports should be made public. To justify its position, the General Court stated that the question of principle governing public access to clinical study reports cannot be ruled on for the first time by a judge hearing an application for interim measures, but rather requires an in-depth examination in the context of the main proceedings. It added that the case raises complex questions which deserve to be resolved through a thorough examination in the context of the main proceedings. The President added that the question of whether an overriding public interest might nevertheless justify disclosure would call for a weighing up of the AbbVie's commercial interest in not having the reports disclosed and the public interest in ensuring the broadest public access possible to documents held by the European Union. Such a weighing up of the various interests present would call for delicate assessments, which must be a matter for the Court adjudicating on the substance of the case.
6. EMA then submitted an appeal to the EU Court of Justice against the Order of the President of the General Court.
7. On 28 November 2013, the Vice-President of the European Court of Justice overturned the Order of the President of the General Court granting interim measures. He noted that interim measures serve to avoid "serious and irreparable damage" to the party seeking the interim measures. He added that, in order to establish the existence of "serious and irreparable damage", it is necessary to show that damage is foreseeable with a sufficient degree of probability. Thus, the party seeking an interim measure is required to at least prove the facts forming the basis of its claim that "serious and irreparable damage" is likely. The Vice-President of the European Court of Justice disagreed that such a standard of proof had been met. He stated that AbbVie should produce such proof as regards any specific information or documents that it considered confidential. He noted that the interim measures could be granted only in respect of that specific information or those specific documents. He concluded by stating that it was for the President of the General Court to examine whether it was possible to authorise partial access to the three clinical study reports, based on the proofs put forward by AbbVie.
8. After an exchange of views, EMA and AbbVie then reached an understanding on a version of the requested documents to be disclosed. EMA then withdraw its decision of January 2013 granting wide access to the requested documents and replaced it with a decision by which the three reports, redacted in a manner consistent with its understanding with AbbVie, were released.
9. As a result, the General Court did not have the opportunity to adjudicate 1) on whether AbbVie's request for interim measures should be granted and 2) on whether EMA's decision of January 2013 (which granted much broader access) should be annulled.
10. When the Ombudsman learned of these developments in April 2014, she requested EMA to provide her with the versions of the reports that had now been released. EMA sent the Ombudsman the redacted documents. It also provided her with a very brief explanation for the agreed redactions. The Ombudsman, after a careful examination of the reasons put forward for the redactions, was not convinced that all of the redactions were justified. She therefore decided to inquire into the matter using her powers of own initiative.
▪ The inquiry
11. On 16 April 2014, the Ombudsman opened the present inquiry and requested (i) all the correspondence between EMA and AbbVie relating to the new redactions, (ii) the new EMA decision, communicated to the citizen requesting access, setting out the justifications for the redactions and (iii) the non-redacted versions of the requested documents, namely the three clinical study reports: CSR M02-404, CSR M04-691 and CSR M05-769.
12. The Ombudsman analysed the extensive documentation provided by EMA. She took the view that certain redactions were justified, primarily those relating to the protection of personal data. However, she identified a series of redactions which were, in her view, problematic. She then asked EMA to answer 75 questions related to these redactions.
13. EMA provided a detailed response to all of these questions. It also informed the Ombudsman that it had, in the meantime, released much of the remaining information in response to subsequent access to documents requests by a researcher. The Ombudsman therefore asked EMA to provide her with the most recent public versions of the requested reports.
14. The Ombudsman then examined the response of EMA and the latest public versions of the reports.
Redactions of the requested clinical studies
Arguments presented to the Ombudsman
15. In its response to the Ombudsman's questions, EMA noted that the rules on public access to documents allow it to refuse to disclose documents or parts thereof if such disclosure would undermine certain interests, including commercial interests.
16. It went on to state that EMA enjoys a wide discretionary power in the scientific assessment underpinning claims of confidentiality for information contained in marketing authorisation dossiers.
17. EMA explained that, in the course of the judicial proceedings, AbbVie (which had up to that point insisted that the reports were, in their entirety, confidential) proposed redactions to the requested documents. EMA stated that, since the President of the General Court had explicitly invited EMA and AbbVie to state whether they had reached agreement on disclosure, EMA carefully examined AbbVie's proposals. While it rejected some proposed redactions, it accepted those which were closely related to the ongoing confidential commercial development of Humira.
18. EMA noted that its obligation to state why a redaction was justified was limited to referring to the category of information protected by the relevant exception. It stated that justifying the redaction in detail would run counter to the efficient use of resources and would slow down the release of documents. EMA thus concluded that the redactions were in line with Regulation 1049/2001 and with EMA's Policy on Access to Documents from 2010.
19. EMA further noted that the three clinical studies were, in the meantime, subject to further public access requests. Each time such a request was made, EMA reassessed the redactions. As a result, previously redacted information had, in the meantime, been disclosed by EMA (EMA identified these disclosures in its reply to the Ombudsman).
20. EMA went on to address the specific concerns put forward by the Ombudsman regarding certain remaining redactions.
21. In relation to the redactions on pages 81 and 82 of Clinical Study Report (CSR) M04-691, EMA stated that the redacted section describes the collection, storage and shipment of blood samples which were used for the study. The detailed description of this procedure was construed as commercially confidential.
22. In relation to Table 28 in CSR M04-691, EMA stated that it could not legitimately dismiss the claim, made by AbbVie, that this information was to be used in an ongoing development plan regarding improved dosing adjustments. It added that the value of the redacted information was minimal as regards understanding the scientific information. Similarly, the redacted text in Table 29 was going to be used for an ongoing development aimed at exploring the dose adjustment of Humira in different subgroups.
23. As for the tables on pages 258-261 in CSR M05-769, EMA explained that the redacted data would provide insights into AbbVie's current developments of a new dosing regimen. EMA added that AbbVie was discussing the new dosing regimen with a non-EU regulatory body. The data was therefore considered commercially confidential at this stage.
24. Finally, as regards redactions of lot numbers from CSR M02-404, CSR M04-691 and CSR M05-769, EMA explained that the lot numbers are often formulated in a way which discloses the manufacturing site, month of manufacture and number of specific batches in a calendar year. The disclosure of such data could, it stated, possibly undermine a company's legitimate commercial interests.
The Ombudsman's assessment
25. A clinical study report describes in detail the tests and trials carried out to determine if a medicine is safe and effective for treating specific illnesses. EMA grants EU-wide marketing authorisations for medicines based on its evaluation of such clinical study reports.
26. The public disclosure of clinical study reports serves various important purposes.
27. First, it allows the work of EMA to be made subject to independent review by third parties, such as researchers. By being willing to subject itself to independent review, EMA will generate and maintain the public's trust in its ability to fulfil its important, sensitive and complex mission, which is to protect the health of people.
28. It may also be the case that the independent review of information, already analysed by EMA in the context of its analysis of a market authorisation request, will, in time, identify additional information and insights of relevance to the safety and efficacy of medicines.
29. Medical practitioners need as much information as possible about the safety and efficacy of the medicines they prescribe. It is thus very important that EMA discloses as much information as possible relating to how these trials are conducted and the results of these trials.
30. The Ombudsman recognises and appreciates the significant steps taken by EMA to inject greater transparency into its work in recent years. These steps include EMA's new policy, effective as of 1 January 2015, of proactively publishing clinical reports.
31. However, given the importance of the role of EMA, the Ombudsman considers it necessary always to evaluate carefully if there is room for further improvement. This is all the more important since, under the EU's new Clinical Trials Regulation, EMA will have the task of determining to what extent all future clinical trials conducted in the EU (and not only those that are submitted to EMA for the purposes of a marketing authorisation requested) can be made publicly accessible.
32. Any restrictions on the public's right of access to documents held by EU public bodies must go no further than is strictly necessary to protect defined interests. These interests, which are set out in EU law on public access to documents (Regulation 1049/2001), include the need to protect the public interest regarding privacy and personal data, the need to protect the purpose of investigations and the need to protect commercial interests. The exceptions protecting certain of these interests, such as the protection of the purpose of investigations and the protection of commercial interests, do not apply where there is an overriding public interest in the disclosure of the document in question. It is the Ombudsman's view that there is always a public interest in the disclosure of documents containing information on the safety and efficacy of medicines. Whenever EMA gets a request for public access to a clinical study report, or considers whether to make such a report public proactively, it must ask: 1) does the report contain information which would, if disclosed to the public, undermine one of the protected interests, and 2), and, if so, does the public interest in the disclosure of the document outweigh that protected interest.
33. In the present the case, the Ombudsman welcomes the fact that EMA has now disclosed very significant parts of the Humira clinical study reports. This is a great step forward compared to the situation in 2013 and 2014. In 2013, AbbVie insisted that the requested documents should not be disclosed at all. In 2014, EMA reached an agreement with AbbVie which led to the publication of redacted versions of the report. However, those redactions were still significant. EMA has since informed the Ombudsman that most of the previously redacted information has now been made public. The Ombudsman notes that this includes information which appears to be of importance to clinicians and researchers, such as information on testing methods, information on the determination of sample size, information on the results of specific tests, information on protocol changes and information on statistical changes.
The Analysis of the Redactions
34. Article 4(2), first indent, of Regulation 1049/2001 allows EMA to:
"(...) refuse access to a document where disclosure would undermine the protection of commercial interests of a natural or legal person, including intellectual property, [...] unless there is an overriding public interest in disclosure."
35. When opening this inquiry, the Ombudsman noted that AbbVie had, when reaching an understanding with EMA (see paragraph 8 above), identified to EMA a broad series of redactions it was seeking. However, upon examining the reasons put forward for these extensive redactions, the Ombudsman was concerned to find that these reasons were often extremely general. She thus set out her concerns in 75 questions posed to EMA.
36. The principal reason put forward to justify redactions related to the alleged fact that the information was of "relevance to an ongoing development programme".
37. The Ombudsman agrees that it is possible that information contained in a clinical study report may relate to the development of new treatments and possible new medicines. Such information can be commercially sensitive (companies may not wish competitors to know of their future product development plans). The public also has, the Ombudsman notes, an interest in companies developing new medicines and new treatments.
38. However EMA should not rely on the mere assertion that information relates to an ongoing development programme. It should ensure that it has sufficient information from the company on that development programme so as to demonstrate its existence and the commercial sensitivity of the information relating to it. To do otherwise would be to permit a company to self-assess on whether information is genuinely commercially sensitive or, worse, to permit a company to make spurious unfounded claims as regards the commercial sensitivity of information. In this context, the Ombudsman notes that the core purpose of a clinical study report is to verify that a medicine is safe and that it is effective in treating specified illnesses in humans. It would be purely incidental for a clinical study report to contain information which does not relate to the safety and the efficacy of a medicine in treating those illnesses for which it is intended. Thus, information in a clinical study report dealing with the future development of new products, or possible additional uses of the product in question, would not be typical.
39. Of course, this does not mean that EMA, after obtaining information on a development programme, is necessarily required to disclose any information on that development programme in order to justify a redaction. Rather, it means that EMA should always put itself in a position to reassure the public that it has obtained all the necessary information to take an informed view as to whether a redaction is justified. It should also be aware that the Ombudsman will, if concerns are raised, check whether EMA has obtained sufficient information to be satisfied that such claims of commercial sensitivity are valid.
40. The Ombudsman also asked EMA to comment on the fact that many years had passed since the marketing authorisation procedure in question had ended (the documents date, at the latest, from 2007). Thus, she had certain doubts as to whether information, which may have related to ongoing developments when first submitted to EMA, could currently be classified as such.
41. As regards the judgement to be made on whether information does in fact relate to ongoing development (of new treatments and new medicines), the Ombudsman recognises that EMA is, as a highly competent technical body, very capable of taking a view on the validity of any assertions that information relates to an ongoing development. However, the Ombudsman insists that EMA must, in order to take an informed view on such assertions, ensure that it is provided with sufficiently detailed and up to date information to sustain such assertions.
42. Even if, exceptionally, a clinical study report is proven to contain commercially sensitive information, that information will still have to be released if there is an overriding public interest in disclosure.
43. On the issue of an overriding public interest, it is the Ombudsman's strong view that where the information in the document in question has clear health implications, as is the case with information on the efficacy of a medicine, the relative efficacy of a medicine, or the safety of a product, the overriding public interest provision is very likely to be engaged.
44. In this context, the Ombudsman asked EMA to state if the redacted information, relating to ongoing developments, was information which was relevant to the current clinical use of Humira, either on-label or off label. She also asked EMA whether researchers/practitioners could use this information to better understand the overall risk-benefit of Humira in treating patients, either on-label or off label. The reason these questions were asked is that, even if certain information did relate to an ongoing development (such as the development of a new treatment), it might well be the case that the development relates also to a known "possible" use of the product and it could even be the case that doctors are already using the product, off-label, for that use. In such circumstances, there would also be strong reasons to consider, if the product is being used off-label, that there is an overriding public interest in the disclosure of the information at issue. That overriding public interest would be based on the view that such information could be useful to allow for a better understanding of that off-label use.
45. It is the Ombudsman's view that any undermining of the protection of commercial interests is very likely to take second place to the overriding public interest in ensuring that important health information is made available to the public.
46. The Ombudsman also takes the view that it should have been for AbbVie not only to provide EMA with detailed and convincing reasons why the disclosure of specific information would undermine its commercial interests (as noted above, it simply made general assertions), but also to explain why that information did not relate to the safety and efficacy of Humira to treat illnesses, either on-label or off-label. It would then be for EMA to decide if those explanations were convincing.
47. Concerning the very limited redactions that remain in the three clinical study reports, the Ombudsman considers that most of them relate to the need to protect personal data. The Ombudsman has no specific comments to make on these redactions. However, she will make comments on four sets of redactions with respect to which concerns may exist.
48. The first set of redactions concern Section 9.5.4 of CSR M04-691. EMA stated that pages 81 and 82 contain a detailed description of the procedure used by AbbVie for collecting, storing and shipping blood samples, which was in part regarded as commercially confidential. When compared with the first version of the disclosed documents, EMA has now granted access to most of the previously redacted text. The Ombudsman recognises this step towards greater transparency. Two measurements, however, continue to be redacted.
49. The Ombudsman is not convinced by EMA's explanation as to why, while the majority of this information could be released, these measurements could not; EMA states that it could release most of the information as it had found that information to be already in the public domain. First, it is not clear to the Ombudsman why these two measurements would be deemed commercially confidential, especially when AbbVie seems to have released very similar measurements. The Ombudsman considers that it would have been more convincing if EMA had asked AbbVie for information on why these two measurements could be distinguished, substantively or contextually, from other measurements. Second, it is clear that if information is already in the public domain, there will be no reason why EMA should redact that information from clinical study reports. However, the opposite is not true. One cannot base an argument for redaction on the simple fact that information is not yet in the public domain. One of the main purposes of the EU's rules on public access to documents is to secure the release of information that is not already in the public domain. If the test applied by EMA is reduced to determining whether the information in question is already in the public domain, the rules become devoid of purpose. The only valid reason why a document may be withheld is if its disclosure would harm one of the public or private interests protected under the EU's rules on public access to documents.
50. The second set of redactions concern parts of Tables 28 and 29 in CSR M04-691, containing data on clinical responses to Humira in different subgroups. EMA argued that this redaction is justified on the grounds that the information was to be used in an ongoing development plan regarding improved dosing adjustments in different subgroups. It is, however, not evident that the data presented in the tables could be relevant for dosing adjustments. The data in fact presents clinical responses to the same dosage of the tested medicine in different subgroups. It does not measure clinical responses to different dosages of Humira. Other parts of this section of the report, which includes the two tables in question, were initially redacted on the same basis, namely the ongoing development plan regarding improved dosing adjustments in different subgroups. However, most of the section, including the accompanying text to the two tables, has since been disclosed in response to public access requests.
51. EMA's decision to disclose only some of the information pertaining to ongoing procedures, and the fact that it did not adequately explain why some information remains commercially sensitive, raises concerns. EMA argued that the value of the redacted information "for the understanding of the scientific information related to the approved indication of Humira is minimal and cannot reasonably generate any disadvantage to the requester".
52. The Ombudsman notes that the person requesting access "is not obliged to state reasons for the application." In fact, the reasons for the making an access request are not relevant to the decision as to whether the document sought may be made available publicly.
53. More importantly, the justification for a redaction must be based on a convincing argument that the disclosure of the information would cause specific harm (for example, to a commercial interest). If such harm is demonstrated, it then becomes relevant, as part of the assessment of whether there is an overriding interest in disclosure, to decide whether the disclosure of the information would also generate public benefits (for example, by revealing important information of relevance to the assessment of the safety and efficacy of the product).
54. This means that an institution cannot refuse access simply on the grounds that the document would be of minimal value to the person seeking access. Though the usefulness of the document, to the public more generally, would be relevant in applying the public interest test.
55. In any case, the Ombudsman is not convinced that the value of the redacted text is minimal. The text accompanying the two non-disclosed tables was disclosed in subsequent access to documents requests. That text does not refer to the numerical values presented in the tables and does not contain commentary on all the results included in the tables. As for Table 28, the text does not include a commentary as to the achievement of CR-70 for tobacco users or non-users, drinkers or non-drinkers and people who tested HACA Positive and those who did not. Additionally, there appear to be some inconsistencies between the text related to Table 29 and data presented in the same table. While the text states that there are no clinically important differences between the subgroups presented in the table, the numerical data presented would rather point to a different conclusion in at least some of the subgroups.
56. The clinical data contained in the tables is therefore of the utmost relevance. Even if the redacted data is commercially sensitive, it is regrettable that EMA did not assess whether there is an overriding public interest for the tables to be disclosed in their entirety.
57. The third redaction which might be problematic is the partial redaction of Tables 41, 42, 43 and 44 in CSR M05-769 in Section 18.104.22.168.1 (Histology). Virtually the entire section was redacted in the first public version of the document. EMA based this decision on the fact that the information was commercially sensitive since its release would provide an "insight [into AbbVie's] development to explore new endpoints in the context of a new dosing regimen, [discussed] with a non-EU regulatory body." As a preliminary point, it is not clear why this particular section would be related to a new dosing regimen. Furthermore, the redacted tables do not in fact contain any data related to different dosage regimens, but rather data related to clinical reactions at different points in time. If there is an explanation which clarifies this issue, EMA should provide it if an access request is made for this data.
58. The Ombudsman also notes most of the previously redacted information from these sections was disclosed in subsequent access requests. While the Ombudsman very much welcomes this decision, it is not clear why there is a continuing need to redact Tables 41, 42, 43 and 44. Specifically, in relation to Table 44, where some of the numerical values were subsequently disclosed and some were not, it is not clear how this partial redaction could protect the discussions with a non-EU regulatory body. Similarly, discussions with a non-EU regulatory body were also used as the basis to redact Sections 22.214.171.124.6 and 126.96.36.199.7 and Section 13.1 in CSR M05-769. These texts were, however, subsequently disclosed. This inconsistency in the treatment of data, previously redacted on the same basis, reinforces the Ombudsman's assessment that disclosing the remaining information would be unlikely to undermine commercial interests - and even if it did, EMA would in any case have to apply the overriding public interest test.
59. The Ombudsman also notes that the text related to the three redacted tables, which has been disclosed in subsequent access to documents requests, is descriptive and vague. It does not contain any numerical values, but merely describes results. The clinical data contained in the tables is therefore of the utmost relevance. Even if the redacted data is commercially sensitive, it is regrettable that the EMA did not assess whether there is an overriding public interest for the three tables to be disclosed in their entirety.
60. The fourth redaction concerns lot numbers. EMA stated that lot numbers are often formulated in a way which indicates the manufacturing site, month of manufacture and number of the specific batch in the calendar year and that this could possibly undermine a company’s commercial interests.
61. It is not clear how public knowledge of a lot number would undermine any commercial interest. Lot numbers appear on all medicines sold (they are printed on every box of medicine put on the market). It is not in any way evident how the disclosure of this information, made generally available for medicines sold every day, would in the case of a medicine the subject of a clinical trial, undermine any commercial interest.
62. The Ombudsman is therefore not convinced that the disclosure of lot numbers, relating to medicines used many years ago in a clinical trial, would undermine AbbVie's commercial interests.
63. The Ombudsman also notes that it is incumbent on EMA to weigh the interest to be protected through non-disclosure against the public interest in the document being made accessible. EMA's opinion does not state whether it actually carried out such an assessment for each of the redactions in question. The Ombudsman regrets this especially because the public interest in question is the protection of public health and, in the case at hand, verifying the safety and efficacy of a drug that is widely used to treat people. In this regard, the Ombudsman notes that lot numbers may in fact have clinical value. If a specific clinical effect, or the absence of specific clinical effect, occurs in relation to a given batch of drugs, but not in relation to other batches, it may well be the case that the effects are connected to a specificity of that batch (and not the medicine as such). It is thus important for clinicians and other researchers to know if the lot numbers of one test correspond to the lot numbers in other tests.
64. To sum up, the Ombudsman is not convinced that the redactions contained in the most recent public version of the requested documents are justified.
65. Nevertheless, the Ombudsman acknowledges EMA's point that it should not be necessary for it to justify every redacted word. However, the explanation put forward must include a statement of reasons from which it is possible to understand and ascertain whether the part of the document in question falls in fact within the sphere covered by the exception and, second, whether the need to protect the interest covered by that exception is genuine. In other words, the explanation should not leave room for doubt as to whether the redactions were in fact justified on the basis of one or more exceptions provided for in Regulation 1049/2001.
66. The Ombudsman also notes EMA's contention that the burden of providing too detailed an explanation would impair the efficient use of its resources. EMA argued that reference should therefore be made only to the "category of information protected by the relevant exception". Such a general approach is, however, justified solely where a request relates to a "manifestly unreasonable number of documents" which would "very substantially paralyse" the proper working of the institution. In such a case, the institution may refrain from carrying out a concrete and individual examination of each of the documents.
67. At the same time, the Ombudsman is sensitive to the pressures placed upon a public authority with limited resources. In her view, it is therefore important to explore systemic solutions so as to make the best use of public resources without in any way sacrificing the citizen’s fundamental right of public access to documents.
68. In this respect, the Ombudsman notes that even if EMA were justified in using such a general wording in response to an initial application for access to documents, the Ombudsman would expect that EMA would be more precise in the event that the applicant questioned any such redactions in a subsequent confirmatory application for the same documents. Thus, EMA should be willing, following detailed informed questions on any redaction, to provide more detailed responses.
69. The implementation of EMA's new proactive transparency policy, which took effect on 1 January 2015, is also of particular importance in this regard. The stated purpose of the policy is to ensure the proactive publication of clinical trial data, as well as access to full data sets to interested parties. In March 2016, EMA published detailed guidance for pharmaceutical companies on how to comply with this policy. This includes guidance on: (i) the procedural aspects related to submitting clinical reports; (ii) how to identify and redact commercially sensitive information in clinical reports; (iii) anonymising clinical reports. Companies must submit a proposed redacted version justifying any commercially sensitive information proposed for redaction. EMA may require further clarification or justification, which companies must address, before EMA publishes the final redacted version.
70. The Ombudsman is pleased to note that the onus has been placed on the pharmaceutical company seeking a marketing authorisation to make available to EMA a redacted version of the clinical study, with an objective justification for each redaction.
71. The Ombudsman is furthermore particularly pleased to note the guidance provided to pharmaceutical companies on redaction of commercially confidential information. She notes improvements regarding the release of information on secondary endpoints, protocol amendments, statistical methods and safety-related information such as adverse reactions. The Ombudsman is reassured by the fact that EMA has drawn important lessons from the series of Ombudsman inquiries conducted over the past decade resulting in what can only be described as a paradigm shift on public access to clinical study data.
72. The Ombudsman notes EMA's statement that it expects to make further updates to its guidance. In the spirit of constructive engagement, the Ombudsman makes the following suggestion on the basis of her handling of this own-initiative inquiry: in addition to requiring a pharmaceutical company seeking a marketing authorisation (or its amendment) to provide EMA with a redacted version of the clinical study, with an objective justification for each redaction, EMA could require the company to produce a public disclosure schedule, specifying the categories of information that can be disclosed at a later date and the event that will trigger such disclosure. Based on her analysis above of clinical study reports, the Ombudsman believes that the main, if not only, legitimate justification for redacting information from a clinical study report relates to the ongoing development of new treatments or of new medicines. A public disclosure schedule should certainly indicate that such information will be made public if and when the details of the development enter the public domain (for example, in scientific journals) or if the product begins to be used off-label for the illness which is the focus of the ongoing development.
73. The Ombudsman believes that the entirety of a clinical study report should ultimately be disclosed; with the sole absolute exception being any personal data of patients that might be contained in a report.
74. The Ombudsman suggests also, in relation to the overriding public interest test, that EMA should consider that there is always a compelling reason for information to be disclosed, where the information at issue has clinical value to clinicians and researchers (as regards understanding the safety and efficacy of a product for uses to which it is put, including off-label use). Such information should always be disclosed even where disclosure risks undermining a company’s commercial interests. In short, public health should always trump commercial interests.
75. The present inquiry is not complaint-based; the Ombudsman has not received a complaint concerning lack of public access to the Humira documents concerned. Rather, the present inquiry is focused on the general approach of EMA when dealing with requests for access to any clinical study reports. While the Ombudsman has some residual concerns regarding certain redactions to the Humira documents (which are not extensive in scope compared to the earlier redactions), she is broadly satisfied with the progress being made by EMA. On this basis, and having outlined her analysis of how EMA might act in relation to the disclosure of clinical study reports in future, the Ombudsman finds that further inquiries into the matter are not justified in the context of the present inquiry. However, the Ombudsman makes the further suggestion to EMA, in the event of receiving new access to documents requests for studies CSR M02-404, CSR M04-691 or CSR M05-769, that it reconsider the need to maintain the present outstanding redactions.
On the basis of the own-initiative inquiry, the Ombudsman closes it with the following conclusion:
The purpose of the inquiry has been achieved; no further inquiries are justified at this point in time.
The European Medicines Agency will be informed of this decision.
Suggestions for improvement
The Ombudsman invites the European Medicines Agency:
1. in order to ensure continuing systemic improvements, to consider requiring companies seeking a marketing authorisations to provide EMA with a public disclosure schedule, specifying the categories of information that can be disclosed at present and at a later date, and the event(s) that will trigger such disclosure;
2. also to ensure continuing systemic improvements, always to consider that there is a compelling overriding public interest for documents to be disclosed where the information they hold has clinical value to clinicians and researchers (as regards understanding the safety and efficacy of a product for uses to which it is put, including off-label use);
3. in the case of the clinical study reports at issue in the present inquiry, to reconsider the need to retain the remaining redactions, made for the purpose of protecting commercial interests, if it receives new requests for access to these reports.
 A clinical study presents the results of clinical trials conducted on humans intended to discover or verify the effects of a medicine, in particular as regards its efficacy in terms of treating specific illnesses and its safety.
 Humira was, according to press reports, the world's bestselling drug in 2014 in terms of turnover: http://www.theguardian.com/business/table/2014/mar/27/world-best-selling-prescription-drugs-pharmaceuticals-industry
 The EU-wide marketing authorization for Humira was granted in 2003 (for the treatment of rheumatoid arthritis). In the following years, EMA granted marketing authorisations covering additional illnesses.
 The European Ombudsman's intervention can be found here: http://www.ombudsman.europa.eu/en/cases/correspondence.faces/en/68005/html.bookmark
 Order by the President of the General Court of 25 April 2013, AbbVie v EMA, T‑44/13 R, EU:T:2013:221.
 Order of the Vice-President of the Court of 28 November 2013, EMA v AbbVie, C-389/13 P(R), EU:C:2013:794.
 The letter to EMA opening the present own-initiative inquiry can be found here: http://www.ombudsman.europa.eu/en/cases/correspondence.faces/en/54524/html.bookmark
 A complete list of questions to the EMA can be found here: http://www.ombudsman.europa.eu/en/cases/correspondence.faces/en/58319/html.bookmark
 Regulation (EC) No 1049/2001 of the European Parliament and of the Council of 30 May 2001 regarding public access to European Parliament, Council and Commission documents, OJ L 2001 145, p. 43.
 Many questions addressed to EMA were related to redactions which were omitted in subsequent publicly disclosed versions. The present decision focuses on the few remaining redactions and the present section includes only EMA's response on the remaining redactions. EMA's full response can be found here: http://www.ombudsman.europa.eu/en/cases/correspondence.faces/en/58946/html.bookmark
 See 'European Medicines Agency policy on publication of clinical data for medicinal products for human use', available at: http://www.ema.europa.eu/docs/en_GB/document_library/Other/2014/10/WC500174796.pdf
 See footnote 8 above.
 Off-label use refers to the use of a medicine for an indication (that is, an illness or an aspect of an illness) despite the fact that the medicine has not (yet) been authorised for use on that indication, or the use of a medicine for an unapproved age group, dosage, or route of administration. Off-label use can exist only where there is already an on-label use for the medicine (since it is only where the product has an on-label use that it will be sold on the market).
 This redaction was dealt with in Questions 7(a-e) in the Ombudsman's letter of 27 October 2014.
 According to EMA, commercially confidential information (CCI) shall mean any information contained in the clinical reports submitted to EMA by the applicant/MAH which is not in the public domain or publicly available and where disclosure may undermine the legitimate economic interests of the applicant/MAH.
 This redaction was dealt with in Questions 8(i) and 8(l) in the Ombudsman's letter.
 Other parts of the three clinical study reports that were initially redacted on the basis of an "ongoing development procedure" were later disclosed, namely, Section 9.4.5 in all three CSRs (dealt with in Question 1 in the Ombudsman's letter of 27 October 2014) and pages 25-28 and 40 in CSR M04-691 (dealt with in Question 6 in the Ombudsman's letter). It is not clear from EMA's opinion whether this ongoing procedure is in fact the development plan regarding improved dosing adjustments, invoked to redact Tables 28 and 29. If this is the case, it raises additional doubts as to the continued need for redaction.
 Article 6(1) of Regulation 1049/2001; see also Judgment of 7 June 2011, Toland v Parliament, T-471/08, EU:T:2011:252, paragraph 82.
 CR-70 shows a decrease in the CDAI score, CDAI standing for Crohn's Disease Activity Index. It is used to quantify symptoms in patients and is measured in absolute numbers. CR-70 means that the CDAI has lowered by 70 points.
HACA stands for human anti-chimeric antibody. This antibody can develop in patients undergoing autoimmune disease therapy with the drug Infliximab. Infliximab can cause the patient to develop antibodies (HACAs) to the drug itself.
 The redaction in this section was dealt with in Question 10(s) in the Ombudsman's letter.
 The redaction of these sections was dealt with in Questions 10(q,r) in the Ombudsman's letter.
 The redaction of this section was dealt with in Questions 15(a,b,d,i,j) in the Ombudsman's letter.
 The redaction of lot numbers was dealt with in Question 16 from the Ombudsman's letter.
 Judgment in Sweden and Turco v Council, Joined Cases C-39/05 P and C-52/05 P, EU:C:2008:374, paragraph 45.
 See 'External guidance on the implementation of the European Medicines Agency policy on the publication of clinical data for medicinal products for human use', available at: http://www.ema.europa.eu/docs/en_GB/document_library/Regulatory_and_procedural_guideline/2016/03/WC500202621.pdf
 In the present case, the initial position of AbbVie regarding the information that was deserving of protection was, in the Ombudsman's opinion, entirely unjustified. With each successive public access request, EMA has been charged with verifying what information can be disclosed.
 The Ombudsman notes, from her analysis of such reports, that the data contained therein is anonymised. She notes, however, that if the illness concerned by a report were a rare illness or if the patient groups were very small, the issue of anonymisation may be more complex.