Case Summary
A complainant who described him/herself as a concerned UK heath professional, complained about the email promotion of Feraccru capsules (ferric maltol) by Norgine Ltd. Feraccru was indicated in adults for the treatment of iron deficiency.
The email had been sent on behalf of Norgine from Guidelines in Practice and the subject line of the recipient’s inbox read ‘Managing patients with iron deficiency anaemia in hospital (promotional information from Norgine)’. The email itself was headed ‘Real World Economic Evaluation of Feraccru vs IV Iron’ and was referenced to Lovato et al (2018).
The complainant noted that below the heading of the email it was stated that researchers at a named NHS trust reported on ‘…the potential health economic benefit of using Feraccru in Inflammatory Bowel Disease (IBD) patients who would otherwise have been considered for IV iron therapy*’. The asterisk took the reader to much smaller text that‘Feraccru should not be used in patients with inflammatory bowel disease (IBD) flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL.’ The complainant stated that apart from this small wording, this special warning was not made clear and it was also not clear if the patients in the audit followed that warning.
The complainant submitted that the data appeared to be based on 28 patients in one small hospital and was conducted almost as a narrative with the absence of a primary endpoint, inclusion criteria, exclusion criteria, any attempt to quantify the values that were used, and a mix of prospective and retrospective data. The complainant stated that that was the only data upon which the whole piece was based, apart from further in-house extrapolation.
The complainant noted that the nursing time was estimated to be 75 minutes per patient although how that was estimated was not stated. In that regard the complainant submitted that two of the most widely used IV irons took 15 minutes to administer, that still left a further hour of time to be accounted for. Cannulation took minutes at most, and of course there was time to set up a giving set but still the time was unaccounted for. The patient might have to be observed for 30 minutes post-infusion, but this did not require a nurse or one-to-one care. The cost of both nursing time and the IV infusion were not specified (there were several treatment options and all had different costs).
The length of treatment was set at three months. The complainant noted that the studies mentioned in the Feraccru summary of product characteristics (SPC) stated that 90% or 68% of patients gained 1g of haemoglobin (Hb) in 16 weeks. If that was so there was a considerable percentage of the patients who would not have had sufficient treatment after 3 months - so again the costs of Feraccru were artificially low. The complainant did not know how low and noted that treatment could be required for perhaps months if the patient required an increase of more than 1g Hb. The complainant acknowledged that the same might be true of those treated with IV iron - there was not prespecified criteria of what ‘treatment’ was to be.
The complainant noted that the email concluded that according to the authors using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time. The complainant referred to the actual conclusion and that the key limitations had been removed - such as underestimating those that needed IV iron, patients might refuse treatment and of course finally patients might be excluded as having too low a Hb or active IBD or needed oral treatment over 3 months - which was highly likely.
The complainant further noted that extrapolated costing based on no evidence ended with ‘**Number of appointments for iron infusion could have been overestimated as some of the patients could have been admitted.’ The complainant alleged that this was misleading as well as being unsupported by any evidence.
The complainant alleged that when the data had been extrapolated to 100 patients there were several errors that had not been taken into account: what percentage would be unsuitable due to special warnings?; how many might refuse?; how many might require IV iron after treatment? In essence the data was insufficient to draw any meaningful conclusions and should not be used promotionally to clinicians.
The detailed response from Norgine is given below.
The Panel noted that the email was sent to prescribers, commissioners and relevant NHS decision makers and included two boxed sections detailing costs and nursing times. The first section was based on a small study of 28 patients over 9 months at one NHS trust (Lovato et al) available as an abstract from the European Haematology Association open access library. The authors had reported that of the 28 patients in the study, four (14%) had to be switched to IV iron. The authors calculated the cost if all patients had been treated with IV iron, although the IV iron was not specifiedand that using Feraccru could have led to savings of £8,955 and 3.75 days of nursing time.
The figure for nursing time saved in the email referred to a footnote directly below which stated that the number of appointments for iron infusion could have been overestimated as some patients could have been admitted. The Panel noted that this also appeared in the abstract as one of the study’s limitations. Included within the published abstract, although not in the email, the authors noted that limitations also included the possibility that the number of patients who needed IV iron could have been under-estimated. The authors also noted that not all patients with IBD would be able to have oral therapy and some might refuse it. The authors concluded that further data were needed to better estimate the real impact of using Feraccru.
Despite the limitations and conclusion set out by the authors, a second section of the email, labelled ‘Modelling/projected costs over 12 months’, was referenced to ‘Norgine Data on file’. On examination of that Data on file, although not obvious to the reader, it appeared that the company had simply extrapolated the data from 28 patients over 9 months to 100 patients over 12 months and noted that if all these IBD patients received IV therapy, treatments costs would have been £57,068 vs £25,031 if Feraccru had been used and still assuming 14% would need IV iron, with a potential saving of 13.4 days of nursing time.
Overall, the Panel did not consider that Lovato et al and the company’s Data on file was sufficiently robust as to support the claim of over 50% cost saving with a potential saving of 13.4 days of nursing time if 100 IBD patients over 12 months were started on Feraccru versus IV iron; in the Panel’s view, the email made too much of too little. Neither the preliminary nature of Lovato et al nor all of the study limitations had been communicated to the reader of the email and there was no indication as to how the estimated 75 minutes of nursing time for each infusion had been calculated by the authors or which IV iron was used. Further, readers would be unaware that ‘Norgine Data on file’ referred to a basic in-house calculation by marketing. Overall, the Panel considered that the email was not sufficiently complete to allow the recipient to fully understand the basis and significance of the data. A breach of the Code was ruled. The comparison with IV iron was misleading and a breach of the Code was ruled. The Panel considered that what appeared to be unequivocal claims regarding potential savings could not be substantiated and a breach of the Code was ruled.
The Panel further noted that Lovato et al was an evaluation of Feraccru vs IV iron only in terms of reduction in treatment cost and nursing time; there was no clinical evaluation of the comparative efficacy and safety of the two regimens. The email, however, included the claim ‘The authors concluded that using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time’ and while the Panel noted that the authors did indeed conclude that Feraccru had been cost effective reducing costs and saving nursing time, it considered that the term ‘cost-effective’ encompassed more than just a cost comparison; other factors such as relative efficacy and incidence of side effects needed to be taken into account. The Panel thus considered that the claim that, compared with IV iron, Feraccru was cost effective was misleading and could not be substantiated. Breaches of the Code were ruled.
The Panel noted thatit appeared from the email that for any IBD patient it was a simple choice between IV iron or Feraccru which was not so. An asterisk in the first paragraph led the reader to a footnote which stated that Feraccru should not be used in patients with inflammatory bowel disease flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL. The Panel noted that claims must be capable of standing alone and that in general, claims should not be qualified by the use of footnotes and the like. The Panel considered that the restrictions on the use of Feraccru in some IBD patients had not been clearly and adequately communicated and a breach of the Code was ruled.
Overall the Panel considered that in using data from a very small, preliminary study as the basis for bold promotional claims and not informing readers about all the relevant limitations, high standards had not been maintained. A breach of the Code was ruled.
Clause 2 was a sign of particular censure and reserved for such use. The Panel noted that it was not clear from the study, and therefore not clear from the promotional email, how the estimated 75 minutes of nursing time for each infusion had been calculated or which specific IV iron was used. The Panel was further concerned by the use of the term cost-effective in the email as noted above.
The Panel noted that it was crucial that health professionals and others could rely upon the industry to provide them with robust and accurate information to aid their decision making. The Panel noted its comments above and considered that the information in the email was not based on robust data and that it encouraged use of Feraccru in all IBD patients when not all of those patients would be suitable for such therapy. Overall, the Panel considered that the material brought discredit upon, and reduced confidence in, the industry. A breach of Clause 2 was ruled.
CASE AUTH/3365/7/20
COMPLAINANT v NORGINE
Promotion of Feraccru
A complainant who described him/herself as a concerned UK heath professional, complained about the email promotion of Feraccru capsules (ferric maltol) by Norgine Ltd. Feraccru was indicated in adults for the treatment of iron deficiency.
The email had been sent on behalf of Norgine from Guidelines in Practice and the subject line of the recipient’s inbox read ‘Managing patients with iron deficiency anaemia in hospital (promotional information from Norgine)’. The email itself was headed ‘Real World Economic Evaluation of Feraccru vs IV Iron’ and was referenced to Lovato et al (2018).
The complainant noted that below the heading of the email it was stated that researchers at a named NHS trust reported on ‘…the potential health economic benefit of using Feraccru in Inflammatory Bowel Disease (IBD) patients who would otherwise have been considered for IV iron therapy*’. The asterisk took the reader to much smaller text that‘Feraccru should not be used in patients with inflammatory bowel disease (IBD) flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL.’ The complainant stated that apart from this small wording, this special warning was not made clear and it was also not clear if the patients in the audit followed that warning.
The complainant submitted that the data appeared to be based on 28 patients in one small hospital and was conducted almost as a narrative with the absence of a primary endpoint, inclusion criteria, exclusion criteria, any attempt to quantify the values that were used, and a mix of prospective and retrospective data. The complainant stated that that was the only data upon which the whole piece was based, apart from further in-house extrapolation.
The complainant noted that the nursing time was estimated to be 75 minutes per patient although how that was estimated was not stated. In that regard the complainant submitted that two of the most widely used IV irons took 15 minutes to administer, that still left a further hour of time to be accounted for. Cannulation took minutes at most, and of course there was time to set up a giving set but still the time was unaccounted for. The patient might have to be observed for 30 minutes post-infusion, but this did not require a nurse or one-to-one care. The cost of both nursing time and the IV infusion were not specified (there were several treatment options and all had different costs).
The length of treatment was set at three months. The complainant noted that the studies mentioned in the Feraccru summary of product characteristics (SPC) stated that 90% or 68% of patients gained 1g of haemoglobin (Hb) in 16 weeks. If that was so there was a considerable percentage of the patients who would not have had sufficient treatment after 3 months - so again the costs of Feraccru were artificially low. The complainant did not know how low and noted that treatment could be required for perhaps months if the patient required an increase of more than 1g Hb. The complainant acknowledged that the same might be true of those treated with IV iron - there was not prespecified criteria of what ‘treatment’ was to be.
The complainant noted that the email concluded that according to the authors using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time. The complainant referred to the actual conclusion and that the key limitations had been removed - such as underestimating those that needed IV iron, patients might refuse treatment and of course finally patients might be excluded as having too low a Hb or active IBD or needed oral treatment over 3 months - which was highly likely.
The complainant further noted that extrapolated costing based on no evidence ended with ‘**Number of appointments for iron infusion could have been overestimated as some of the patients could have been admitted.’ The complainant alleged that this was misleading as well as being unsupported by any evidence.
The complainant alleged that when the data had been extrapolated to 100 patients there were several errors that had not been taken into account: what percentage would be unsuitable due to special warnings?; how many might refuse?; how many might require IV iron after treatment? In essence the data was insufficient to draw any meaningful conclusions and should not be used promotionally to clinicians.
The detailed response from Norgine is given below.
The Panel noted that the email was sent to prescribers, commissioners and relevant NHS decision makers and included two boxed sections detailing costs and nursing times. The first section was based on a small study of 28 patients over 9 months at one NHS trust (Lovato et al) available as an abstract from the European Haematology Association open access library. The authors had reported that of the 28 patients in the study, four (14%) had to be switched to IV iron. The authors calculated the cost if all patients had been treated with IV iron, although the IV iron was not specifiedand that using Feraccru could have led to savings of £8,955 and 3.75 days of nursing time.
The figure for nursing time saved in the email referred to a footnote directly below which stated that the number of appointments for iron infusion could have been overestimated as some patients could have been admitted. The Panel noted that this also appeared in the abstract as one of the study’s limitations. Included within the published abstract, although not in the email, the authors noted that limitations also included the possibility that the number of patients who needed IV iron could have been under-estimated. The authors also noted that not all patients with IBD would be able to have oral therapy and some might refuse it. The authors concluded that further data were needed to better estimate the real impact of using Feraccru.
Despite the limitations and conclusion set out by the authors, a second section of the email, labelled ‘Modelling/projected costs over 12 months’, was referenced to ‘Norgine Data on file’. On examination of that Data on file, although not obvious to the reader, it appeared that the company had simply extrapolated the data from 28 patients over 9 months to 100 patients over 12 months and noted that if all these IBD patients received IV therapy, treatments costs would have been £57,068 vs £25,031 if Feraccru had been used and still assuming 14% would need IV iron, with a potential saving of 13.4 days of nursing time.
Overall, the Panel did not consider that Lovato et al and the company’s Data on file was sufficiently robust as to support the claim of over 50% cost saving with a potential saving of 13.4 days of nursing time if 100 IBD patients over 12 months were started on Feraccru versus IV iron; in the Panel’s view, the email made too much of too little. Neither the preliminary nature of Lovato et al nor all of the study limitations had been communicated to the reader of the email and there was no indication as to how the estimated 75 minutes of nursing time for each infusion had been calculated by the authors or which IV iron was used. Further, readers would be unaware that ‘Norgine Data on file’ referred to a basic in-house calculation by marketing. Overall, the Panel considered that the email was not sufficiently complete to allow the recipient to fully understand the basis and significance of the data. A breach of the Code was ruled. The comparison with IV iron was misleading and a breach of the Code was ruled. The Panel considered that what appeared to be unequivocal claims regarding potential savings could not be substantiated and a breach of the Code was ruled.
The Panel further noted that Lovato et al was an evaluation of Feraccru vs IV iron only in terms of reduction in treatment cost and nursing time; there was no clinical evaluation of the comparative efficacy and safety of the two regimens. The email, however, included the claim ‘The authors concluded that using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time’ and while the Panel noted that the authors did indeed conclude that Feraccru had been cost effective reducing costs and saving nursing time, it considered that the term ‘cost-effective’ encompassed more than just a cost comparison; other factors such as relative efficacy and incidence of side effects needed to be taken into account. The Panel thus considered that the claim that, compared with IV iron, Feraccru was cost effective was misleading and could not be substantiated. Breaches of the Code were ruled.
The Panel noted thatit appeared from the email that for any IBD patient it was a simple choice between IV iron or Feraccru which was not so. An asterisk in the first paragraph led the reader to a footnote which stated that Feraccru should not be used in patients with inflammatory bowel disease flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL. The Panel noted that claims must be capable of standing alone and that in general, claims should not be qualified by the use of footnotes and the like. The Panel considered that the restrictions on the use of Feraccru in some IBD patients had not been clearly and adequately communicated and a breach of the Code was ruled.
Overall the Panel considered that in using data from a very small, preliminary study as the basis for bold promotional claims and not informing readers about all the relevant limitations, high standards had not been maintained. A breach of the Code was ruled.
Clause 2 was a sign of particular censure and reserved for such use. The Panel noted that it was not clear from the study, and therefore not clear from the promotional email, how the estimated 75 minutes of nursing time for each infusion had been calculated or which specific IV iron was used. The Panel was further concerned by the use of the term cost-effective in the email as noted above.
The Panel noted that it was crucial that health professionals and others could rely upon the industry to provide them with robust and accurate information to aid their decision making. The Panel noted its comments above and considered that the information in the email was not based on robust data and that it encouraged use of Feraccru in all IBD patients when not all of those patients would be suitable for such therapy. Overall, the Panel considered that the material brought discredit upon, and reduced confidence in, the industry. A breach of Clause 2 was ruled.
A complainant who described him/herself as a concerned UK heath professional, complained about the email promotion of Feraccru capsules (ferric maltol) by Norgine Ltd. Feraccru was indicated in adults for the treatment of iron deficiency.
The email in question (ref UK-HAE-FER-2000038) had been sent on behalf of Norgine from Guidelines in Practice and the subject line of the recipient’s inbox read ‘Managing patients with iron deficiency anaemia in hospital (promotional information from Norgine)’. The email itself was headed ‘Real World Economic Evaluation of Feraccru vs IV Iron’ and was referenced to Lovato et al (2018).
COMPLAINT
The complainant noted that below the heading of the email it was stated that researchers at a named NHS Trust reported on ‘… the potential health economic benefit of using Feraccru in Inflammatory Bowel Disease (IBD) patients who would otherwise have been considered for IV iron therapy*’. The complainant noted that the asterisk took the reader to much smaller text which read ‘Feraccru should not be used in patients with inflammatory bowel disease (IBD) flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL.’ The complainant stated that apart from this small wording, this special warning was not made clear and it was also not clear if the patients in the audit followed that warning.
The complainant submitted that the data appeared to be based on a poster (ie it had not been peer reviewed), based on one small hospital and was conducted almost as a narrative with the absence of a primary endpoint, inclusion criteria, exclusion criteria, any attempt to quantify the values that were used, and a mix of prospective and retrospective data. The complainant stated that that was the only data upon which the whole piece was based, apart from further in-house extrapolation.
The complainant noted that the nursing time was estimated to be 75 minutes per patient although how that was estimated was not stated. In that regard the complainant submitted that two of the most widely used IV irons took 15 minutes to administer, that still left a further hour of time to be accounted for. Cannulation took minutes at most, and of course there was time to set up a giving set but still the time was unaccounted for. The patient might have to be observed for 30 minutes post-infusion, but this did not require a nurse or one-to-one care.
The complainant further noted that the cost of both nursing time (was it a Band 5 or a specialist nurse?) as well as the cost of the IV infusion were not specified (there were several treatment options and all had different costs).
Then there was the length of treatment which was set at three months.
The complainant noted that the studies mentioned in the Feraccru summary of product characteristics (SPC) stated that 90% or 68% of patients gained 1g of haemoglobin (Hb) in 16 weeks. If that was so there was a considerable percentage of the patients who would not have had sufficient treatment after 3 months - so again the costs of Feraccru were artificially low. The complainant did not know how low and noted that treatment could be required for perhaps months if the patient required an increase of more than 1g Hb. The complainant acknowledged that the same might be true of those treated with IV iron - there was not prespecified criteria of what ‘treatment’ was to be.
The complainant noted that the conclusion provided in the email was ‘The authors concluded that using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time.’ The actual conclusion was:
‘Using ferric maltol in our NHS Trust has been cost effective reducing costs and saving nursing time. Only new patients with [iron deficiency anaemia] were prescribed ferric maltol, the patients who were already on IV iron maintenance continued with the parenteral therapy. The cost effectiveness will improve if we could switch to ferric maltol all the patients who were on parenteral iron maintenance and if ferric maltol was approved for treatment of any cause of iron deficiency. We recognise some limitation of this analysis. First the percentage of patients who had to stop the drug and needed parenteral iron could be underestimated because as patients prescribed ferric maltol towards the end of the study period had shorter follow up. Second, the number of appointments for iron infusion could have been overestimated as some of the patients could have been admitted. Last not all the patients, particularly those with active IBD could qualify for oral therapy and some could refuse it. Further data are needed to clarify these points and better estimate the real impact of this intervention.’
The complainant submitted that the key limitations had been removed - such as underestimating those that needed IV iron, patients might refuse treatment and of course finally patients might be excluded as having too low a Hb or active IBD or needed oral treatment over 3 months - which was highly likely.
The complainant noted that the entire promotional item was based on a sample size of 28 patients.
The complainant further noted that extrapolated costing based on no evidence ended with ‘**Number of appointments for iron infusion could have been overestimated as some of the patients could have been admitted.’ The complainant alleged that this was misleading as well as being unsupported by any evidence.
The complainant submitted that when the data had been extrapolated to 100 patients there were several errors that had not been taken into account: what percentage would be unsuitable due to special warnings?; how many might refuse?; how many might require IV iron after treatment? In essence the data was insufficient to draw any meaningful conclusions and should not be used promotionally to clinicians.
The complainant alleged that each point above was potentially in breach of Clauses 7.2, 7.3 and 7.4 and, as the evidence base was clearly very low, the complainant further alleged a breach of Clause 9.1. The complainant queried whether the poor quality of the material warranted a ruling of a breach of Clause 2.
RESPONSE
Norgine explained that the promotional email in question was based on an abstract, which described a real world economic evaluation of Feraccru in an NHS trust. The evaluation was presented as a poster at the European Haematology Association Congress in 2018. The abstract was available on the society's open access library and was the primary reference for the email.
Norgine stated that it had no role in designing the evaluation and, therefore, it confined its response to the use of the data in the promotional email.
With regard to the opening statement of the email, ‘Researchers at the London North West University Healthcare NHS Trust reported on the potential health economic benefit of using Feraccru in Inflammatory Bowel Disease (IBD) patients who would otherwise have been considered for IV iron therapy*’ and the asterisked statement, ‘Feraccru should not be used in patients with inflammatory bowel disease (IBD) flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL.’, Norgine noted that Feraccru was indicated in adults for the treatment of iron deficiency. The asterisked statement appeared in Section 4.4 of the SPC under special warnings and precautions of use; it was not a contraindication. Although the referenced abstract did not indicate whether any patients fell into that category, Norgine had added the statement to support the safe use of its medicine.
Norgine did not consider that the asterisked statement in smaller text was misleading and consequently it met the requirements of Clause 7.2. Norgine submitted that the statement could be substantiated and so also met the requirements of Clause 7.4 and that the statement about Feraccru was factual and did not compare it with another product, therefore the requirements of Clause 7.3 did not apply. Norgine denied breaches of Clauses 7.2 and 7.4.
Norgine noted that the reference cited in the email (Lovato et al) was an abstract, not a poster. Whilst originally published as a poster, the referenced study had been presented at the European Haematology Association Congress in 2018. Although not 'peer reviewed' in the sense that a full journal manuscript would be, it was usual for a conference organising committee to conduct internal quality control and review. The abstract of the poster was available from the European Haematology Association open access library, and therefore was capable of substantiation.
Norgine stated that it understood its responsibility to accurately reflect the published data, but it could not be responsible for the design of the economic analysis itself. The Code did not preclude the use of abstracts. Whilst abstracts never contained the granularity of detail available in full publications, as challenged by the complainant, the data from the abstract was accurately and clearly presented as required by Clause 7.2. With regard to the protocol, it was not possible for a real world service evaluation (observational) to conform to the principles of interventional study design, nor to justify that in an abbreviated publication form such as an abstract.
Norgine submitted that the data offered a unique real world perspective of economic outcomes related to the use of Feraccru originating from an NHS trust. The company did not consider that was unrepresentative of other NHS trusts. There was no other evidence of a similar nature involving Feraccru real world costing data to cite alongside Lovato et al.
Norgine considered that the presentation of the data in the email was sufficiently clear to enable health professionals and other relevant decision makers to critically review the claims in the context of their own units.
Norgine considered that the selection of Lovato et al for use in the email was valid and not misleading (Clause 7.2). The data quoted within the material was clearly referenced for substantiation (Clause 7.4). Norgine considered that as the complaint was about the quality and validity of the study, not about comparisons, the requirements for Clause 7.3 did not apply. Norgine denied breaches of Clause 7.2 and 7.4.
Norgine noted that the complainant disputed the authors' statements of nursing time required for administration of IV iron infusions, and patient safety observation post infusion. These figures were given in the publication and did not originate from Norgine. However, Norgine did not consider that the figures were inconsistent with the 2019 Royal College of Nursing guidelines on IV iron infusion. The guidelines highlighted a range of infusion times depending on the IV iron in question, but also stated that patients ‘should be observed for adverse effects for at least 30 minutes following each treatment’. These guidelines did not appear consistent with the complainant’s view that 30 minutes observation 'might' be necessary, and that it did not require a nurse.
In that regard Norgine considered that the information provided in the email was accurate to the content of the source data. In quoting the estimates of independent NHS clinicians Norgine did not intend to mislead or distort the interpretation of the reader. The company disagreed with the short time estimates suggested by the complainant, which were not referenced, nor cited as personal experience and it could not find supporting evidence in established guidance. In Norgine’s view the complaint was not related to any comparisons and thus the requirements for Clause 7.3 did not apply. Norgine denied breaches of Clause 7.2 and 7.4.
Norgine stated that the complainant’s comments on nursing costs linked closely to the points above regarding nursing time. The cost estimates provided in the abstract were linked to the trust and provided by independent NHS clinicians. The abstract did not state the IV infusion on which the data were based.
Norgine stated that the costs quoted in the abstract were accurately reflected in the email and referenced to the abstract and therefore could be substantiated. The complaint was not related to any comparisons and thus the requirements for Clause 7.3 did not apply. Norgine denied breaches of Clause 7.2 and 7.4.
Norgine noted that the complainant had raised concerns about the efficacy stated in the Feraccru SPC based on the Phase Ill studies. The SPC was clearly referenced on the promotional material. The complainant incorrectly quoted 16 week changes in haemoglobin (Hb). However, the IBD data contained within the SPC was for the 12 week endpoint. In addition, the endpoint cited by the complainant was a secondary endpoint which related to subgroups of the study population, not the primary endpoint. Norgine considered that this had led to an exaggerated impression of reduced efficacy on the part of the complainant.
Norgine stated that the 1g/dl change in Hb that underpinned the above misinterpretation was not relevant to the objectives of the service evaluation. The abstract did not state any criteria used for switching patients taking Feraccru to IV iron. Whilst potentially a shortcoming of the publication, mapping of the Phase Ill data to this cohort was therefore inappropriate. This again highlighted the value of considering real world costing data with clinical endpoints vs solely using the endpoints of academic or regulatory studies.
Norgine stated that whilst the Phase Ill data formed core parts of its promotional material, the objective of the material was to communicate real world data. The company therefore considered that the addition of a 1g/dl endpoint (non-primary), was inappropriate and potentially confusing in that context. The costs quoted in the abstract were accurately reflected in the promotional material and referenced to the abstract and could therefore be substantiated. The complaint was not related to any comparisons and thus the requirements for Clause 7.3 did not apply. Norgine denied breaches of Clauses 7.2 and 7.4.
Norgine noted that the conclusion of the abstract contained 3 statements:
1. Using Ferric maltol in our NHS Trust has been cost effective reducing costs and saving nursing time;
2. Only new patients with IDA [iron deficiency anaemia] were prescribed ferric maltol, the patients who were already on IV iron maintenance continued with the parenteral therapy;
3. The cost effectiveness will improve if we could switch to ferric maltol all the patients who were on parenteral iron maintenance and if ferric maltol was approved for treatment of any cause of iron deficiency.
Norgine submitted that the first statement had been paraphrased to support the general conclusion of the promotional material and that inclusion of the second statement was considered unlikely to alter the natural conclusions drawn by a reader. The third statement would have reinforced promotional messaging, however, as it could not be substantiated it was excluded from the promotional material.
Norgine stated that in the context of the email and its target audience, it was confident that the conclusions drawn by the reader would not be influenced or distorted by the omission of the authors’ second and third conclusions.
Since the complaint referred to incorporation of the authors' conclusions into the promotional material, Norgine considered that the requirements of Clause 7.3 did not apply. The company considered that the summary of the conclusions included in the promotional material met the requirements of Clauses 7.2 and 7.4 and it denied a breach of those clauses.
Norgine noted that the authors stated three limitations:
1. the percentage of patients who had to stop the medicine and needed parenteral iron could be underestimated because as patients prescribed ferric maltol towards the end of the study period had shorter follow up;
2. the number of appointments for iron infusion could have been overestimated as some of the patients could have been admitted;
3. not all the patients, particularly those with active IBD could qualify for oral therapy and some could refuse it. Further data were needed to clarify those points and better estimate the real impact of this intervention.
Norgine submitted that the first limitation was a common scenario for medium to long term treatments. The company’s view was that the intended audience could easily deduce this common limitation without prompting, and its omission therefore did not impact the overall interpretation of the data.
Norgine stated that the second limitation was included in the email. Confusingly, this was later mentioned in the complaint - its inclusion was claimed to be misleading and 'unsupported by the evidence'. The statement was not written as an absolute but was a carefully worded limitation to mirror the authors’ sentiments. It highlighted a variable the authors could not account for in their data set. In light of this contradictory complaint, Norgine trusted that this explanation would suffice to alleviate any concerns.
Norgine noted that the third limitation related to the authors' extrapolation of the data over 12 months. Norgine submitted that it did not include this extrapolation within the promotional material because it was irrelevant to the email.
In the context of this material and its target audience, Norgine was confident that the conclusions drawn by the reader would not be influenced or distorted by the lack of verbatim reproduction of the limitations summarised above. The conflicting complaint about the inclusion of the second limitation was also addressed below.
Norgine stated that as the complaint referred to incorporation of the study limitations into the promotional material, it considered that the requirements of Clause 7.3 did not apply. Norgine considered that the most important and relevant limitations had been included and that the promotional material met the requirements of Clauses 7.2 and 7.4. The company denied any breach of those clauses.
Norgine stated that the size of study was, in its view, immaterial; the service evaluation was the size deemed appropriate by the investigating physicians, presented at the European Haematology Association Congress 2018 and was not misrepresented in the promotional material. Whilst larger numbers increased the power of clinical studies, in this case there was no statistical analysis, so whilst a larger sample might be desirable, there was no direct impact on the conclusions. The email clearly stated the number of patients prior to any communication of results. The material was also clearly referenced and so could not be considered misleading.
Norgine noted that the complainant cited issue with the sample size of an independent study and thus in its view the requirements of Clauses 7.2, 7.3 and 7.4 did not apply. The company denied any breaches of the Code.
Norgine noted that the complaint’s comments with regard to extrapolated costing related to the inclusion of the second limitation as stated by the authors and was associated in the calculations based on the observed 28 patients within the trust (not the extrapolated costing). The complainant previously implied that all the conclusions and limitations should be included.
Norgine noted that the primary objection for the extrapolation to 100 patients appeared to be the complainant's impression of the audit data presented in section 1. That data was referenced as a data on file. Norgine noted that, up until the time that the complaint was submitted, no requests to see that data on file had been received. The company thus did not consider that the complainant had grounds to state that it was incapable of substantiation.
In summary, Norgine submitted that the model was clearly laid out for interpretation by the reader and clearly reflected the data from the referenced service evaluation (Clauses 7.2 and 7.3). The information within the email was clearly referenced to the audit design and data (audit and estimated costs section), Norgine data on file (modelling section) and the Feraccru SPC (supporting safety statement). These documents were all available for substantiation (Clause 7.4) (copies provided). Norgine denied any breach of Clauses 7.2, 7.3 and 7.4.
Norgine stated that it was committed to producing materials which were both of a high standard and complied with the Code. The company did not consider that the material at issue was in breach of the Code; it was based on and clearly referenced an abstract, and the information presented was sufficiently complete to allow recipients a balanced view on which to form their own opinion of the therapeutic value of Feraccru in NHS practice.
In summary, Norgine did not consider that there was any evidence to support any breach of Clauses 7.2, 7.3 and 7.4, and thereby no breach of Clauses 9.1 and 2.
PANEL RULING
The Panel noted that the promotional email in question was sent to prescribers, commissioners and relevant NHS decision makers. The email featured, inter alia, two boxed sections detailing costs and nursing times. The first section was based on a small study of 28 patients over 9 months at one NHS trust (Lovato et al) available as an abstract from the European Haematology Association open access library. The authors had reported that of the 28 patients in the study, four (14%) had to be switched to IV iron. The authors calculated that if all patients had been treated with IV iron, although the specific IV iron was not specified, the cost of treatment would have been £15,960 but using Feraccru could have led to savings of £8,955 and 3.75 days of nursing time.
In the email in question, the figure for nursing time saved referred to a footnote directly below which stated that the number of appointments for iron infusion could have been overestimated as some patients could have been admitted. The Panel noted that this sentence also appeared in the abstract as one of the study’s limitations. Included within the published abstract, although not in the email, the authors noted that limitations also included the possibility that the number of patients who needed IV iron could have been under-estimated. The authors also noted that not all patients with IBD would be able to have oral therapy and some might refuse it. The authors concluded that further data were needed to better estimate the real impact of using Feraccru.
Despite the limitations and conclusion set out by the authors, a second section of the promotional email, labelled ‘Modelling/projected costs over 12 months’, was referenced to ‘Norgine Data on file’. On examination of that Data on file, although not obvious to the reader, it appeared that the company had simply extrapolated the data from 28 patients over 9 months to 100 patients over 12 months and noted that if all these IBD patients received IV therapy, treatments costs would have been £57,068 vs £25,031 if Feraccru had been used and still assuming 14% would need IV iron, with a potential saving of 13.4 days of nursing time.
The Panel noted that the Code did not prohibit the use of claims substantiated by economic evaluations of real world observational data provided that the claims complied with the requirements of the Code including Clauses 7.2, 7.3 and 7.4. Context was important.
Overall, the Panel did not consider that the data from Lovato et al and the company’s Data on file was sufficiently robust as to support the claim of over 50% cost saving with a potential saving of 13.4 days of nursing time if 100 IBD patients over 12 months were started on Feraccru versus IV iron; in the Panel’s view, the email made too much of too little. Neither the preliminary nature of Lovato et al nor all of the study limitations had been communicated to the reader of the email and there was no indication as to how the estimated 75 minutes of nursing time for each infusion had been calculated by the authors or which IV iron was used. Further, readers would be unaware that ‘Norgine Data on file’ referred to a basic in-house calculation by marketing. Overall, the Panel considered that the email was not sufficiently complete to allow the recipient to fully understand the basis and significance of the data. A breach of Clause 7.2 was ruled. The Panel thus considered that the comparison with IV iron was misleading and a breach of Clause 7.3 was ruled. The Panel considered that what appeared to be unequivocal claims regarding potential savings could not be substantiated and a breach of Clause 7.4 was ruled.
The Panel further noted that Lovato et al was an evaluation of Feraccru vs IV iron only in terms of reduction in treatment cost and nursing time; there was no clinical evaluation of the comparative efficacy and safety of the two regimens. The email, however, included the claim that ‘The authors concluded that using Feraccru in their NHS trust was cost-effective, reduced costs and saved nursing time’ and while the Panel noted that the authors did indeed conclude that Feraccru had been cost effective reducing costs and saving nursing time, it considered that the term ‘cost-effective’ encompassed more than just a cost comparison; other factors such as relative efficacy and incidence of side effects needed to be taken into account. The Panel thus considered that the claim that, compared with IV iron, Feraccru was cost effective was misleading and could not be substantiated. Breaches of Clauses 7.2, 7.3 and 7.4 were ruled.
The Panel noted the reference in the introductory paragraph of the promotional email to IBD patients who would otherwise have been treated with IV iron, being treated with Feraccru. It appeared from the email that for any IBD patient it was a simple choice between IV iron or Feraccru which was not so. An asterisk in the first paragraph led the reader to a footnote which stated that Feraccru should not be used in patients with inflammatory bowel disease flare or in IBD patients with haemoglobin (Hb) levels <9.5g/dL. The Panel noted that claims in promotional material must be capable of standing alone and that in general, claims should not be qualified by the use of footnotes and the like. The Panel considered that the restrictions on the use of Feraccru in some IBD patients had not been clearly and adequately communicated and a breach of Clause 7.2 was ruled.
Overall the Panel considered that in using data from a very small, preliminary study as the basis for bold promotional claims and not informing readers about all the relevant limitations, high standards had not been maintained. A breach of Clause 9.1 was ruled.
Clause 2 was a sign of particular censure and reserved for such use. The Panel noted that it was not clear from the study, and therefore not clear from the promotional email, how the estimated 75 minutes of nursing time for each infusion had been calculated or which specific IV iron was used. The Panel was further concerned by the use of the term cost-effective in the promotional email as noted above.
The Panel noted that the small number of patients in Lovato et al (n=28) was stated in the first of the two boxed sections in large red font and in that regard the reader would not be misled as to the size of the study. The Panel noted that the second boxed section stated in smaller font in the top right hand corner that the cost calculation and potential nursing time saved for 100 patients was based on ‘Modelling/projected cost over 12 months’ referenced to ‘Norgine Data on file’. The Panel considered that the take home message to a busy health professional/commissioner would be the cost savings which were detailed in large bold red font in both boxed sections. The Panel considered that, based on a small preliminary study at one hospital trust, with a number of relevant clinical limitations as noted above and where the authors themselves concluded that further data were needed, the email had been constructed to encourage readers to consider more than halving costs by prescribing Feraccru instead of IV iron; in the Panel’s view, it was not possible to determine from the data whether the magnitude of the savings claimed in terms of costs and nursing time were accurate or realistic. Furthermore, the restrictions on the use of Feraccru in some IBD patients had not been prominently stated in the email.
The Panel noted that it was crucial that health professionals and others could rely upon the industry to provide them with robust and accurate information to aid their decision making. The Panel noted its comments above and considered that the information in the email was not based on robust data and that it encouraged use of Feraccru in all IBD patients when not all of those patients would be suitable for such therapy. Overall, the Panel considered that the material brought discredit upon, and reduced confidence in, the industry. A breach of Clause 2 was ruled.
Complaint received 14 July 2020
Case completed 15 March 2021
