Case Summary
Two complaints were received from anonymous, non-contactable, hospital health professionals about the conduct of the same AstraZeneca representative.
One health professional complained that the representative had recently discussed the unpublished Jupiter (Justification for the Use of Statins in Primary prevention: an Intervention Trial Evaluating Rosuvastatin) data. The representative admitted to the complainant that she 'should not strictly be discussing the data yet' but she had clearly initiated the discussion and facilitated further questioning regarding the data.
During the conversation it became apparent that her line manager knew that she was discussing the data despite the fact that to do so was a clear breach of the Code.
An anonymous consultant also complained that discussions were initiated by the representative regarding the unpublished Jupiter data.
The detailed response from AstraZeneca is given below.
The Panel examined the representatives' briefing material. The results of the study would be of interest to health professionals. The first briefing, a voicemail, was very positive and stated that 'This is great news for Crestor' and this would give customers the outcome data that many had been waiting for before positioning Crestor positively in their guidelines. It would give confidence for customers to use and recommend Crestor more widely. The voicemail concluded with a question 'What actions should I take?'. The answer made it clear that the study was completed in a group of patients who were outside the UK licence and 'so you must not proactively raise this study with customers. It is against the AstraZeneca Code of Conduct and the ABPI Code of Practice to promote any study that is outside of a product licence'.
All the briefing material was very clear that Crestor did not have a marketing authorization for reducing cardiovascular (CV) events or saving lives and therefore could not and must not be so promoted. Further guidance was given that sales calls must not be engineered to encourage customers to ask for further information on the use of Crestor to reduce CV events. The company had prepared a reactive statement for representatives to respond to unsolicited enquiries. Representatives had been instructed not to proactively raise the study with customers.
The representative's line manager had reissued the briefings by email with a reminder. The email also praised the account team and named two individuals (not the representative in these cases) for the high number of referrals they had generated '… through [the regional medical affairs executive] post Jupiter'. This was the highest in the UK. In the Panel's view this comment could be evidence that representatives were being encouraged to engineer discussions about the data and thus generate requests to be referred elsewhere for a response.
The Panel noted that a complainant had the burden of proving their complaint on the balance of probabilities. The Panel had some concerns about the material supplied to representatives but also noted the company's submission that the representative's line manager witnessed her responding correctly to a request for information about Jupiter on two occasions. The Panel considered that the allegation was a serious one but it did not consider that evidence had been provided by either complainant to show that on the balance of probabilities the representative in question had promoted an unlicensed indication as alleged and no breach was ruled.
CASES AUTH/2190/12/08 and AUTH/2194/12/08 ANONYMOUS HEALTH PROFESSIONALS v ASTRAZENECA
NO BREACH OF THE CODE
Conduct of representative
Two complaints were received from anonymous, non-contactable, hospital health professionals about the conduct of the same AstraZeneca representative.
One health professional complained that the representative had recently discussed the unpublished Jupiter (Justification for the Use of Statins in Primary prevention: an Intervention Trial Evaluating Rosuvastatin) data. The representative admitted to the complainant that she ‘should not strictly be discussing the data yet’ but she had clearly initiated the discussion and facilitated further questioning regarding the data.
During the conversation it became apparent that her line manager knew that she was discussing the data despite the fact that to do so was a clear breach of the Code.
An anonymous consultant also complained that discussions were initiated by the representative regarding the unpublished Jupiter data.
The detailed response from AstraZeneca is given below.
The Panel examined the representatives’ briefing material. The results of the study would be of interest to health professionals. The first briefing, a voicemail, was very positive and stated that ‘This is great news for Crestor’ and this would give customers the outcome data that many had been waiting for before positioning Crestor positively in their guidelines. It would give confidence for customers to use and recommend Crestor more widely. The voicemail concluded with a question ‘What actions should I take?’. The answer made it clear that the study was completed in a group of patients who were outside the UK licence and ‘so you must not proactively raise this study with customers. It is against the AstraZeneca Code of Conduct and the ABPI Code of Practice to promote any study that is outside of a product licence’.
All the briefing material was very clear that Crestor did not have a marketing authorization for reducing cardiovascular (CV) events or saving lives and therefore could not and must not be so promoted. Further guidance was given that sales calls must not be engineered to encourage customers to ask for further information on the use of Crestor to reduce CV events. The company had prepared a reactive statement for representatives to respond to unsolicited enquiries. Representatives had been instructed not to proactively raise the study with customers.
The representative’s line manager had reissued the briefings by email with a reminder. The email also praised the account team and named two individuals (not the representative in these cases) for the high number of referrals they had generated ‘… through [the regional medical affairs executive] post Jupiter’. This was the highest in the UK. In the Panel’s view this comment could be evidence that representatives were being encouraged to engineer discussions about the data and thus generate requests to be referred elsewhere for a response.
The Panel noted that a complainant had the burden of proving their complaint on the balance of probabilities. The Panel had some concerns about the material supplied to representatives but also noted the company’s submission that the representative’s line manager witnessed her responding correctly to a request for information about Jupiter on two occasions. The Panel considered that the allegation was a serious one but it did not consider that evidence had been provided by either complainant to show that on the balance of probabilities the representative in question had promoted an unlicensed indication as alleged and no breach was ruled.
Two separate complaints were received from anonymous non-contactable, hospital health professionals about the conduct of the same AstraZeneca UK Limited representative.
Case AUTH/2190/12/08
COMPLAINT
An anonymous health professional complained that named AstraZeneca representative had recently discussed the unpublished Jupiter (Justification for the Use of Statins in Primary prevention: an Intervention Trial Evaluating Rosuvastatin) data. The representative admitted to the complainant that she ‘should not strictly be discussing the data yet’ but she clearly initiated the discussion and facilitated further questioning regarding the data.
During the conversation it became apparent that her line manager knew that she was discussing the data despite the fact that to do so was a clear breach of the Code.
42 Code of Practice Review May 2009
After careful consideration the complainant considered that there was no alternative but to report the matter.
Case AUTH/2194/12/08
COMPLAINT
An anonymous consultant complained that during recent meetings the same representative had initiated discussions about the unpublished Jupiter data.
This was alleged to be a breach of the Code.
* * * * *
When writing to AstraZeneca the Authority asked it to respond in relation to Clauses 3.2 and 15.2 of the Code.
Cases AUTH/2190/12/08 and AUTH/2194/12/08
RESPONSE
AstraZeneca noted that both allegations were in essence identical and that the source of the allegations appeared to be very similar. Whilst it was always difficult to verify the independence and authenticity of anonymous complainants, AstraZeneca believed the spirit of the Code required it to deal with the complaints in good faith.
AstraZeneca took the complainants’ allegations very seriously and recognised the need for proper and thorough investigations. Regrettably, there was very little detail in either complaint, such as dates, which would have assisted in dealing with them. The first letter was from a ‘health professional’ in a specific hospital which allowed AstraZeneca to check the activity of the representative in this hospital. The second was simply from a ‘consultant’ in a same region of England.
The Jupiter study was presented at the American Heart Association congress and simultaneously published online in the New England Journal of Medicine on 9 November 2008. Jupiter was a placebo-controlled cardiovascular outcomes study using rosuvastatin (Crestor) in a primary prevention population, which was a group of patients not included in the currently licensed indications for Crestor. The unprecedented reductions in mortality and morbidity, which had resulted in the study being prematurely terminated in March 2008, suggested that the Jupiter study publication would achieve a very high level of media attention. AstraZeneca was thus particularly concerned to ensure that all employees were fully briefed on the requirements of the Code and that no one had any doubt about what they could and could not say about the study. Accordingly, a series of cascaded briefings took place using teleconferences and webex technology, emails, voicemails and face-to-face briefings for all relevant employees immediately after the online publication on Sunday, 9 November. Copies of these briefings were provided. Details of the briefings relevant to the named representative were given.
All briefings were signed off by two signatories as required by the Code. The briefings, inter alia, stated that the Jupiter study publication could not be raised proactively with any health professional and gave a brief, factual reactive statement which could be used by representatives if the study was mentioned by one of their customers. The representative confirmed that she had received all relevant briefings. The reactive statement was as below:
‘The Jupiter study was recently presented at the AHA and published in the New England Journal of Medicine. The Jupiter study showed that those subjects who received Crestor had a reduction in cardiovascular events vs placebo. As this study is in a population that is out of licence I cannot discuss the results – if you would like more information on the Jupiter study I can arrange a visit from one of the AZ Regional Medical Affairs team or request information be sent to you by Medical Information’.
Case AUTH/2190/12/08
AstraZeneca submitted that it was not against the Code to discuss unpublished material, although ‘data on file’ needed to be made available on request without delay and any unpublished data referred to in promotional material needed to be within the existing licensed indications for the medicine. Due to the results being potentially price sensitive, they were embargoed until the publication date. Therefore the representative did not know the results before Monday, 10 November (she was unaware that they had been published online at 2pm the previous day in the New England Journal of Medicine) and therefore no results could possibly have been discussed prior to that time. Therefore AstraZeneca had looked at all calls in the hospital in question until the date of the complaint letter. Eight different health professionals were seen in 10 separate calls between 10 November and 8 December. The representative received comprehensive briefings throughout this time period and was fully aware of her obligations in dealing with the Jupiter data. The representative had no recollection of any conversations taking place in any of the visits where the Jupiter study results were discussed, apart from using the agreed reactive statement in response to a question.
On two separate occasions the line manager had accompanied the representative when asked about Jupiter and on both occasions she responded correctly, using the short factual statement and offering referral to a regional medical affairs executive or medical information if the customer wanted more details.
The business manager, the line manager and the Head of medical affairs for primary care had all spent time discussing the allegations with the representative and had given her every opportunity to admit to ‘a genuine mistake’, if this had indeed occurred. Sufficient time had also been allowed during the investigation for the representative to recall anything that did not come to mind at initial interviews. The representative and all three separate individuals were consistent in their belief that no inappropriate discussions had taken place.
The representative’s line manager and the business manager were questioned about the allegations. Both denied any knowledge of the alleged discussions taking place. The line manager had in fact issued two additional briefing emails on 13 and 26 November to her team, reusing the signed-off briefing materials, and stressing the importance of adopting the correct approach to any queries around Jupiter. As elsewhere in AstraZeneca, the investigation concluded that there was a strong focus on governance and compliance issues in this region and all briefings stressed the importance of complying with the Code.
Case AUTH/2194/12/08
AstraZeneca submitted that the points made above in response to Case AUTH/2190/12/08 were relevant to this case. The representative was again contacted to discuss the second complaint and asked to try to recall any situation where a discussion about Jupiter might have taken place that could have been misinterpreted by the consultant as off-label promotion. The representative was consistent and adamant that no such discussions had taken place.
Conclusion
AstraZeneca took allegations about representative conduct extremely seriously and there would be serious repercussions for a representative who proactively discussed information about a medicine which was inconsistent with its marketing authorization.
AstraZeneca was confident that the various briefing materials that were issued both centrally and locally were timely, comprehensive and clear. The representative was of exemplary character and performance and the statements had been consistent and robust throughout the investigations. In addition, the evidence of the line manager (including as a witness to two calls on health professionals) and the business manager had also been consistent and robust in support of the representative. AstraZeneca could find no evidence to support the allegations and therefore it denied breaching Clauses 3.2 and 15.2 in respect of either complaint.
FURTHER INFORMATION
In response to a request for further information, AstraZeneca confirmed that its representatives were not given a copy of Jupiter nor were they instructed how to access the paper online. The paper was never distributed to the sales teams in any other format and they were not instructed as how to use it. The briefing material that representatives received, referred to above, contained clear instructions. AstraZeneca referred again to the precise wording to use in response to enquiries about Jupiter from health professionals and how these requests must be referred to AstraZeneca’s medical team or medical information.
To refer these enquiries the representatives had to generate a referral in the AstraZeneca database. This was then passed on to the appropriate regional medical affairs executive.
In the line manager’s email of 26 November 2008, the manager used content and language consistent with the clear instruction above and also internal jargon with reference to ‘the high number of referrals they have generated through [the regional medical affairs executive] post Jupiter’. It would be expected that a high number would be generated given the extensive media coverage of the study’s results in both the lay and medical press but only if the representatives followed the clear instruction above and generated the appropriate referral. The manager confirmed this was the case in her subsequent sentence in the email when she referred to this number providing ‘clear evidence that we are communicating with our customers through the right channels’. Clearly the line manager’s intention in this email was to reinforce the instruction on Jupiter communication and congratulate the team on appearing to diligently follow that instruction.
The reference to ‘code breaches’ in this email to the team referred to the fact that cases published by the PMCPA in the Code of Practice Review relevant to the sales team were discussed as part of their governance framework and such cases had been recently discussed at the manager’s local meeting.
This email was supplied to the PMCPA in good faith in response to the original complaints as relevant evidence in support of appropriate action taken by company representatives and their manager in response to Jupiter. AstraZeneca appreciated however that when this email was considered in isolation by an individual not familiar with AstraZeneca process and internal jargon, it could potentially be misunderstood. However, AstraZeneca hoped that its explanation had addressed any concerns that the Panel might originally have had.
PANEL RULING
The Panel noted that the complainants were anonymous and non-contactable. When an allegation had been made about what a representative had said to a health professional it was difficult to determine precisely what had occurred. The parties’ accounts often differed. In similar cases, before the Panel made its ruling, the company’s response had been sent to the complainant for comment. This was not possible here.
The Panel noted that both complainants referred to a discussion about unpublished data. It was not necessarily a breach of the Code to discuss unpublished data. It would be a breach of the Code to promote an unlicensed medicine or indication irrespective of whether data was published.
The Panel examined the briefing material provided by AstraZeneca. It considered that given the results of the study there would be interest from health professionals. The first briefing to all the CV salesforce was a voicemail dated 10 November. The Panel noted that the voicemail was very positive stating that ‘This is great news for Crestor’ and this would give customers the outcome data that many had been waiting for before positioning Crestor positively in their guidelines. It would give confidence for customers to use and recommend Crestor more widely. The voicemail concluded with a question ‘What actions should I take?’. The answer made it clear that the study was completed in a group of patients who were outside the UK licence and ‘so you must not proactively raise this study with customers. It is against the AstraZeneca Code of Conduct and the ABPI Code of Practice to promote any study that is outside of a product licence’.
All of the briefing material was very clear that Crestor did not have a marketing authorization for reducing CV events or saving lives and therefore could not and must not be so promoted. Further guidance was given that sales calls must not be engineered to encourage customers to ask for further information on the use of Crestor to reduce CV events. The company had prepared a reactive statement for representatives to respond to unsolicited enquiries. Representatives had been instructed not to proactively raise the study with customers. The line manager had reissued the briefings with a reminder. The Panel was concerned about the reference in an email dated 26 November to ‘… several code breaches across the UK’. The email also praised the account team and named two individuals (not the representative in these cases) for the high number of referrals they had generated ‘… through [the regional medical affairs executive] post JUPITER’. This was the highest in the UK. In the Panel’s view this comment could be evidence that representatives were being encouraged to engineer discussions about the data and thus generate requests to be referred elsewhere for a response.
The Panel noted that a complainant had the burden of proving their complaint on the balance of probabilities. The Panel had some concerns about the material supplied to representatives but noted the company’s submission that the representative’s line manager witnessed her responding correctly to a request for information about Jupiter on two occasions. The Panel considered that the allegation was a serious one but it did not consider that evidence had been provided by either complainant to show that on the balance of probabilities the representative in question had promoted an unlicensed indication as alleged and no breach of Clauses 3.2 and 15.2 was ruled in both cases.
AUTH/2190/12/08 received 10 December 2008 AUTH/2194/12/08 received 17 December 2008
Cases completed 20 January 2009
