International comparison of pharmaceutical industry payment disclosures in the UK and Japan: implications for self-regulation, public regulation, and transparency

Disclosure rules, practices, and data formed three distinct dimensions of transparency of self-regulated payment disclosure in the UK and Japan. Figure 1 summarises their constituent parts and relationships.

Fig. 1figure 1

Relationships between disclosure rules, practices, and data in pharmaceutical industry self-regulation in the UK and Japan

Disclosure rulesAim of disclosure

The ABPI (AC2019, Introduction) and JPMA (JTG2018, Chapter 1 – Purpose) identified transparency as the primary aim of disclosure, without, however, defining it.

According to the ABPI, transparency, or “openness”, was achieved by making payments publicly available [111, 121], and, as such, was instrumental in “building and maintaining confidence” in the industry- healthcare sector collaborations (AC2019, Introduction). Its additional benefits included demonstrating ethical behaviour, managing FCOIs in the NHS, and increasing public understanding of industry-healthcare sector collaborations [111, 121]. Likewise, the JPMA viewed transparency as key for addressing FCOIs and demonstrating commitment to ethical standards and scientific progress (JTG2018, Introduction).

Payment recipients Recipient categories

The ABPI and JPMA distinguished individual- and organisational-level recipients. Both trade groups defined organisational-level recipients similarly, using overlapping examples (Table 1). At the individual level, however, the JPMA (JTG2018, Chapter 3 – Disclosure Recipients) defined HCPs more broadly, referring to medical personnel, while the ABPI Code mentioned professions and roles in decision-making involving medicines (AC2019, Clause 1.4). Nevertheless, in 2018 at least some companies in the UK reported payments to professionals providing treatment or interventions without prescription medicines, such as occupational health specialists, speech and language therapists, and social workers. The JPMA mentioned an additional HCP category associated with “medical operations”, corresponding with ABPI’s non-HCP “other relevant decisions makers” (ORMDs), which companies reporting their payments in Disclosure UK in 2018 interpreted as NHS administrators, managers, and board members (AC2019, Clause 1.5).

Table 1 Payment recipients subject to disclosure under the JPMA Transparency Guidelines and the ABPI Code

Two other individual-level categories did not overlap. From 2023, the ABPI provisions will extend to patients and journalists as members of the public (AC2021, Introduction). Conversely, the JPMA requires the disclosure of payments to life science researchers (JTG2018, Chapter 3 – Disclosure Recipients). While not explicitly mentioned in the ABPI Code, payments reported in Disclosure UK in 2018 suggests that at least some companies voluntarily interpreted scientists, such as biochemists and microbiologists, as HCPs or ORDMs.

Recipient characteristics

The ABPI’s “disclosure template” (AC2019, Clause 24.1) required the reporting of recipients’ address information and “institution name” and, optionally, “location”. Some companies reporting payments in 2018 interpreted “location” as an organisational subunit (e.g., a department within a university). The ABPI Code did not explain the difference between “names” and “locations”, potentially leading to confusion in identifying recipients [31], which could not be addressed by optional – and not published – identification numbers [122]. Further optional characteristics were HCP “speciality” and “role”, but in the absence of a shared list of categories and their descriptors companies used them inconsistently, thereby hindering or preventing reliable aggregation [24, 31]. For example, in the 2018 Disclosure UK database anaesthesiology was referred to variously as “anaesth”, “anaesthesia”, “anesthesiology” [sic], "anaesthiesia" [sic], and “anaesthetics”.

While the JPMA demanded no organisational-level characteristics (besides names), the individual-level characteristics overlapped with those required by the ABPI: affiliation and, if available, organisational sub-units, such as clinics and departments (corresponding with ABPI’s “names” and “locations”) (JTG2018, Chapter 5 – Publication details). Companies also reported recipients’ roles but without specialty and work address. Like in the UK, recipient characteristics were often missing and unstandardised in disclosures.

Disclosed payments

The ABPI mandated the disclosure of payments made for “promotional purposes or otherwise, in connection with the development or sale of [prescription] medicines” (AC2019, Clause 1.10, 24.1–2). In describing disclosed payments, the JPMA did not refer to promotion specifically, instead mentioning collaborations related to research or other activities undertaken by payment recipients, not necessarily in academic settings (JTG2018, Chapter 1 Background of Transparency Guidelines Enforcement; Chapter 4 – Target Payments for public disclosure).

We follow the ABPI’s and JPMA’s distinction between payments unrelated to companies’ research and development (R&D) and payments – or expenditure more broadly – related to R&D. We also compare organisational- and individual-level payments corresponding with the basic recipient categories.

Non-R&D payments Organisational-level payments

The JPMA and ABPI mandated the disclosure of, respectively, one and four categories of non-R&D payments to HCOs (Table 2). While the JPMA category covered payments to HCOs exclusively, this was the case only for two of the four ABPI categories (donations and grants, and joint working).

Table 2 Non-R&D payments subject to disclosure under the JPMA Transparency Guidelines and the ABPI Code

The JPMA Category B covered payments for subsidising recipients’ academic activities unrelated to companies’ R&D, allowing recipients to specify the purpose of funding (JTG2018, Chapter 5, 3 – Publication Subject, B Academic research support expenses). Depending on the organisational context, this category was divided into “scholarship donations” to academic institutions, such as universities or hospitals (B1); “general donations” to foundations, and non-governmental organisations, for example, providing ballpoint pens and other items and donations to a university (B2); and “donations to academic or professional societies” made to support their activities, including meetings and lectures (B3). However, if a company event, such as a luncheon or seminar, was held at a conference organised by a professional society, the respective payments would fall under “conference co-sponsoring” (B4).

We link JPMA subcategories B1-B3 to the broader category of “donations, grants and benefits in kind” from the ABPI Code (AC2019, Clause 19.1–2, 24.2), subsequently broken down into “donations” ("physical items, services or benefits in-kind”) and grants (“funds”) (AC2021, Clause 1.5, 23.1). The ABPI category was more comprehensive as it was not limited to education or academic support, also relating to “supporting healthcare” (AC2021, Clause 23.2), and was not restricted to the types of recipients listed by the JPMA. Conversely, we interpret B4 (conference co-sponsoring) as belonging to the ABPI’s broader organisational-level category of “contributions towards the costs of meetings” (AC2019, Clause 22.1, 24.2), most recently renamed as “sponsorship agreements” (AC2021, Clause 1.4, 1.22, 10.11). The ABPI category covered meetings organised by drug companies, HCOs or “other independent organisations” more broadly (AC2021, Clause 1.22). It also included costs of subsistence for participating HCPs (AC2019, Clause 22.1, AC2021, Clause 1.4, 10.1). Contrastingly, the JPMA Code (JC2019, I-2 Medical products Promotional code, Chapter 8 Provision of goods) and the Fair Competition Code (FCC, Criteria for the Operation of Article 4 of the Code, Operational standards regarding examples of restricted offerings) banned payments for travel, attendance or entertainment related to non-promotional (e.g. academic) meetings organised by HCPs.

Unlike the JPMA, the ABPI Code included disclosure of “fees for service and consultancy” to organisations acting on behalf of their employees (AC2019, Clause 23.2, 23.4, 24.2) and contracts with organisations (AC2019, Clause 21, 24.2). In addition, the categories of “joint working” (AC2019, Clause 24.2 and 20), later subsumed under “collaborative working” (AC2021, Clause 20), involved projects combining resources provided by donors and recipients alike and, as such, were not mentioned by the JMPA.

Individual-level payments

The JPMA and ABPI each mandated the disclosure of two categories of non-R&D payments to individual-level recipients (Table 2).

The JPMA “Manuscript/writing fees” (category C) included two specific subcategories – “honorariums for lectures” (C1) and “manuscript writing or supervision” (C2) – and a more general one, “consulting and commissioning” (C3), covering other forms of consultancy, such as advisory boards (JTG, Chapter 5, 3 – Publication Subject, C, manuscript/writing fees, etc.). We consider C1-C3 as matching the ABPI’s individual-level “fees for service and consultancy” (AC2019, Clause 24.2, 23.2–4) or “contracted services” (AC2021, Clause 24), which mentioned “writing articles and/or publications” and “speaking at and chairing meetings” as examples (AC2021, Clause 24.1).

The JPMA also required the disclosure of expenses related to “information provision” (category D) (JTG2018, Chapter 5, 3 – Publication Subject, D, Information provision-related expenses). These payments covered, first, expenses for promotional meetings held by companies, including support for venue, HCPs’ attendance, travel, accommodation, and post-meeting receptions with food and drink (D1). This subcategory overlapped with ABPI’s individual-level “sponsorship” (AC2019, Clause 22.5), subsequently renamed as “support” (AC2021, Clause 1.23), of attendance at meetings/events. The overlap would only be partial because the ABPI category covered not only promotional but also scientific, educational and training meetings/events (AC2019, Clause 22.1).

Second, “explanation meetings” (D2) referred to payments for food and drink provided by company sales representatives during promotional meetings in clinical settings. Under the ABPI Code, these payments would either be excluded from disclosure, if the value of food and drink was below £75 ($100) per head, or prohibited, if the value was above that figure (AC2019, Clause 1.10, 22.1–2).

A final JPMA individual-level subcategory covered costs associated with the provision of materials to assist HCPs with their scientific research (D3). There was no clear corresponding ABPI Code category. Although donations to individuals were prohibited by the ABPI (AC2019, Clause 19.1), it was conceivable that HCPs or ORDMs could claim research materials as “expenses” associated with consultancy services (AC2019, Clause 23.4).

Aggregate disclosure

Following the interpretation of the UK Data Protection Act and European privacy laws by companies following the ABPI Code, all HCPs and ORDMs receiving payments could only be named in drug company disclosures depending on their consent. The value of payments lacking consent to disclosure was aggregated per category (e.g. fees for service and consultancy; support for events participation). However, since late 2021 the ABPI has recommended “legitimate interest” as the preferred lawful basis for disclosure [64]: “a company asserts their transparency commitments over the data rights of the individual HCP” [123], without seeking their explicit agreement before disclosing payments. Nevertheless, companies “must allow individuals to exercise their right to raise objections” [123, 124]. The only exception to this rule were contracted services delivered by members of the public, a category introduced by the newest ABPI Code, and lacking a corresponding JPMA category (AC2021, Clause 24). Unlike payments to HCPs and ORDMs, these payments were aggregated by default (AC2021, Clause 24.6).

In Japan, only the individual recipients of “Manuscript/writing fees” (category C) were named. The JPMA informally advised its member companies to seek consent regarding the publication of their names. However, this was not an explicit rule, and data on the percentage of individuals who refused disclosure was not published. In practice, many companies accepted that HCPs could not receive payments unless they agreed for their names to be disclosed.

Conversely, payments related to “information provision” (category D) were reported as a lump sum per company. Another aggregated category – and lacking a corresponding ABPI Code category – involved expenses for social courtesies (JTG2018, Chapter. 5. E – Other expenses), including expenses for congratulations and condolences, and provision of food and beverages. This was an annual total that was aggregated and did not differentiate costs for either HCPs or HCOs.

R&D payments and expenditure

The JPMA and ABPI disclosure requirements covered payments for clinical trials, such as consultancy fees, which, under the JMPA Transparency guidelines would fall under “clinical trials” and “specific clinical research (AC2019, Clause 23.2; JTG, Chapter 5, 3 – Publication Subject, A, R&D expenses, etc.) (Table 3). In addition, both trade groups’ requirements encompassed non-clinical studies (AC2019, Clause 23.2; JTG, Chapter 5, 3 – Publication Subject, A, R&D expenses, etc.). In practice, JPMA members and companies following the ABPI Code disclosed payments related to both company- and investigator-initiated or sponsored studies [46].

Table 3 R&D payments subject to disclosure under in the JPMA Transparency Guidelines and the ABPI Code

Finally, the ABPI Code category of prospective non-interventional studies would fall under the broader JPMA category of medical research involving human subjects, understood as any clinical research complying with the Ministry of Health, Labour and Welfare’s Ethical Guidelines for Medical Research Involving Human Subjects, which had to be followed in the conduct of virtually all medical research, whether interventional, observational, prospective, or retrospective [128].

The ABPI expected aggregate disclosure of R&D payments as a lump sum per company (AC2019, Clause 23.2). Nevertheless, reporting practices suggested that some of the key elements of what is commonly understood to be R&D expenditure [46] were excluded by some companies, for example, honoraria to contract research organisations and payments to study sites [46]. On the other hand, the ABPI allowed companies to decide which costs were considered “subsidiary” to R&D activities, and therefore reported as R&D payments (AC2019, Clause 23.2), or reported as non-R&D payments, and therefore potentially on a name basis.

Contrastingly, the JPMA specified which payments were to be reported as either R&D or non-R&D (categories B – academic research support expenses- and C – lecture, writing or consulting fees), with category A (R&D) covering only research specifically contracted to and conducted by medical institutions (JTG2018, Chapter 5, 3 – Publication Subject, A, Research and development expenses, etc.). The JPMA mandated the disclosure of organisational recipient names and clinical trial numbers. The investigators names were to be disclosed, too, but not their individual fees. Importantly, not only did the JPMA mandate the disclosure of R&D payments (e.g. consultancies paid to clinical trial investigators) but R&D expenditure including “research funds” and “expenses” paid to medical institutions for outsourced research. Minor exceptions from disclosure included loan of equipment and damages paid to clinical trial participants.

Payments explicitly exempted from disclosure

Both industry trade groups listed the same or corresponding payments as excluded from disclosure.

The ABPI and JPMA excluded payments related to drug samples (AC2019, Clause 1.10; JTG2018, Chapter 4 – Target Payments for public disclosure). The ABPI also excluded market research, understood as “collection and analysis of information [on medicines]” in instances “where the company does not know the identity of the participants” (AC2019, Clause 12.2, 23.3). While the JPMA did not use the term “market research”, it would ordinarily fall under subcategories C1-C3 (JTG2018, Chapter 5, 3 – Publication Subject, C, Manuscript/writing fees, etc.). Further, under the ABPI Code companies did not need to report payments regarding “items which are to be passed on to patients and which are part of a formal patient support programme” (AC2019, Clause 18.2), “inexpensive notebooks, pens and pencils for use at those meetings” (AC2019, Clause 18.3) and subsistence below £75 ($100) per head (AC2019, Clause 22.1). These payments would be covered by JPMA’s category D (JTG2018, Chapter 5, 3 – Publication Subject, D, Information provision-related expenses).

Some exemptions were only mentioned by either trade group. While companies following the ABPI Code were not expected to disclose payments regarding over-the-counter medicines and "ordinary course purchases and sales of medicines”, these payments were excluded implicitly by the JPMA by not being mentioned in the Transparency Guidelines (AC2019, Clause 1.10).

The JPMA excluded from disclosure investigational drugs, support for membership fees, advertising fees, and capital for exhibition fees at conferences (JTG2018, Chapter 4 – Target Payments for public disclosure). Although not regulated by the ABPI, some companies reported investigational drugs as R&D payments, while others excluded them from disclosure [46].

Disclosure practicesAvailability

In 2018, 122 companies published their payments in Disclosure UK, the ABPI’s online platform for payment disclosure [129]. Among the disclosing companies were 63 of the 67 ABPI members (94%) and 59 non-members disclosing payments voluntarily. Four of the 67 ABPI members (6.0%) – Adaptimmune, Daval International, Sintetica, and Stallergenes Greer – did not submit disclosure reports. As they joined the ABPI towards the end of 2018, it is likely that none of the payments they might have made fell under the disclosure requirements.

In 2018, the JPMA had 72 members, all of which published payment disclosures in accordance with to the JTG2018. However, the total number of disclosing companies was 86 as it included subsidiaries and affiliates. Some of the disclosures were published late, contrary to the JTG2018. As no comprehensive disclosure platform existed, we could not tell whether other companies complied with JTG2018 voluntarily.

An unknown number of pharmaceutical companies did not subscribe to the trade groups’ disclosure codes and therefore did not disclose payments. According to the ABPI [74] and JPMA [130], disclosing companies formed a vast majority, especially of larger ones, but important exceptions included Vertex in the UK and Gilead in Japan.

Accessibility

The ABPI and JPMA leave gathering disclosures to companies. The ABPI collates and integrates company disclosures [131], and subsequently publishes them as Disclosure UK, a centralised database searchable by company and recipient names [132]. Starting from 2023, aggregated payments for services rendered by members of the public will be disclosed on individual company websites, that is, separately from Disclosure UK [59].

JPMA members disclose all payments in disclosure reports published on their websites. The JPMA prohibits companies from requiring registration or entering additional information to access the disclosure data (JTG2018, Chapter 6 – Notes from the perspective of ensuring transparency). However, 28 of the 86 (30.2%) companies required registration to access the 2018 data, apparently breaching the Transparency Guidelines. One possible reason is that companies had insufficient time to comply with the new requirements prohibiting registration, which entered into force in September 21 2018, in time for the data release.

To increase disclosure data accessibility, two non-governmental organisations, Tansa and the Medical Governance Research Institute, collect disclosures made by individual companies in Japan, integrating them into the Money Database [38, 106]. It comprises data from 2016 to 2019 and is searchable by company and recipient names and payment categories [133].

Format

Disclosure UK is available online and downloadable for further analysis as an Excel file. By contrast, in 2018, none of JPMA members published their data in a directly analysable format, instead choosing PDF (portable document format) files or tables embedded in a webpage.

Evidence of underreporting

The PMCPA, the ABPI’s self-regulatory oversight body, publicised three cases of payment underreporting.

The Japanese company Astellas voluntarily admitted to failing to disclose payments to UK nurses manning patient support lines for a support program and to pharmacies in relation to patient enrolment [134]. This was part of a larger case involving a lack of oversight of, and materials produced for, patient support programs, leading to §2 rulings as the PMCPA considered Astellas had brought discredit upon and reduced confidence in the industry.

Following a complaint by an ex-employee, another Japanese company, Daiichi-Sankyo, was also ruled in breach of §2 and publicly reprimanded in 2019 by the PMCPA for grossly under-reporting its payments to at least 132 HCPs sponsored to attend conferences, including 98 who attended a five-day European Society of Cardiology (ESC) congress in Germany in 2018 [135]. In addition, at least 15 and 28 HCPs were sponsored to attend ESC congresses in 2016 and 2017. The company had paid for travel, accommodation, food, and registration fees for these 132 HCPs, but none of this was disclosed – a total of around $629,000 of which $531,600 was for 2018. As even more underreporting could have occurred, the company stated it did not have complete visibility of the total payments that had not been reported.

Finally, an ex-employee revealed that Indivior, a company that has ratified the ABPI Code, did not disclose any 2017 payments, allegedly because it did not know it had agreed to disclose them since becoming an official “non-member” in 2017 [136]. The PMCPA also ruled a breach of §2.

We could not establish whether any JPMA members had been caught underreporting or not. Notably, there was one company reporting the value of zero in category A (R&D), and one in both category C (Lecture/writing/consulting fees) and category D (information provision expenses). It is impossible to ascertain whether, in fact, no payments were made in these categories or whether payments were made but not reported.

Evidence of misreporting

We identified three PMCPA cases involving payments attributed to incorrect individuals in the UK, all reported by the affected HCPs. All these instances were consistent with the lack unique recipient identifiers, as described above.

Merck Sharp & Dohme reported having made payments to an individual that had received none. Apparently, the company had sponsored another individual with the same name [137]. Amgen made a similar mistake, explaining that it had sponsored another individual with the same name working in the same area [138]. Importantly, Merck Sharp & Dohme made an additional incorrect payment disclosure related to the same individual as in the above case, leading to a breach of §2 ruling by the PMCPA as the company had failed to address the underlying problems [139].

Without procedures for submitting complaints or publishing corrections, establishing whether misreporting had been identified in JMPA members’ disclosures was impossible.

Disclosure dataPayments to corresponding payment categories

In 2018, the 122 companies following the ABPI Code reported payments worth almost 4.5 times less than the 86 JPMA members, including subsidiaries (Table 5 in Appendix). Another key difference was the R&D shares, constituting 74.1% of the total in the UK but only 42.7% in Japan, even though the Japanese reporting also included R&D expenditure more broadly, as explained above.

Companies following the ABPI Code reported $632,985,775 (95.1%) using payment categories corresponding with the JPMA categories, including all individual-level non-R&D payments and R&D payments (Table 5 in Appendix). JMPA members reported $2,589,064,594 (86.6%) using payment categories corresponding with the ABPI categories, including $1,018,619,673.61 (39.3%) for individual-level payments and $1,570,444,920 (60.7%) for organisational-level payments.

Payments to aggregated recipients

Companies following the ABPI Code reported payments worth over $144.3 m to named recipients, representing 21.7% of the total, while the respective figures for JPMA members were $1.7bn and 59.9% (Table 4).

Table 4 Drug company R&D and non-R&D payments disclosed under the JPMA Transparency Guidelines and the ABPI Code (2018)

In the UK, the shares of payments to named recipients ranged across the payment categories – from all or nearly all (joint working and grants and donations) to none (R&D) (Table 4). Importantly, as R&D made up almost three-quarters of all reported payments, it constituted practically all aggregated payments (97.7%). The remaining aggregated payments were predominantly non-R&D individual-level payments, representing 42.7% of all non-R&D payments to individuals (Table 6 in Appendix).

In Japan, subcategory D1 (promotional meeting expenses) attracted the largest share (64.2%) of all aggregated payments. Of the 86 companies, 46 (53.5%) apparently breached the Transparency Guidelines by aggregating some payments in category A (R&D), with the aggregated amount representing 3.1% of the total payment value (Table 4). In contrast to the UK, aggregated non-R&D payments constituted 96.6% of all aggregated payments.

The values of aggregated payments per company ranged from $0 to over $48 m in the UK and from $0 to $96 m in Japan (Tables 7 and 8 in Appendix). In the UK, 71 of the 122 companies (58.2%) reported at least 50% of their payments in aggregate, but 21 companies (17.2%) reported more than 90% (Table 9 in Appendix). The company levels of aggregated payments were lower in Japan, with 20 of the 86 companies (23.3%) reporting over 50% in aggregate. The group of companies with very high levels of aggregated payments (over 90%) was not found in Japan, with only 2 (2.3%) reporting at least 75% (Table 10 in Appendix).

The UK and Japanese shares of aggregated R&D vs non-R&D payments reported by companies were highly contrasting. Only 8 of the 122 companies following the ABPI Code (6.6%) had over 50% of all their payments constituted by aggregated non-R&D payments (Table 9 in Appendix). However, for R&D payments the corresponding number of companies was 56, representing 45.9% of all companies following the ABPI Code.

In Japan, 66 of the 86 companies (76.7%) reported at least 50% of their non-R&D payments in aggregate (categories D information provision expenses – and E – Other expenses). However, for R&D payments the corresponding number of companies was 3, which constituted 3.5% of all the 86 reporting companies (category A – R&D) (Table 8 in Appendix).

In the UK, aggregated R&D payments were considerably larger than aggregated non-R&D payments, with the value of the third quartile over 11 times larger ($145,195.9 vs $1,615,243.1 – see Table 7 in Appendix). In Japan, the opposite was true; aggregated non-R&D payments were much higher than aggregated R&D payments, with the value of the third quartile almost 60 times higher ($16,872,972.9 vs $288,770.5 – see Table 8 in Appendix).

留言 (0)

沒有登入
gif