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

Abstract

Background

Self-regulation of payment disclosure by pharmaceutical industry trade groups is a major global approach to increasing transparency of financial relationships between drug companies and healthcare professionals and organisations. Nevertheless, little is known about the relative strengths and weaknesses of self-regulation across countries, especially beyond Europe. To address this gap in research and stimulate international policy learning, we compare the UK and Japan, the likely strongest cases of self-regulation of payment disclosure in Europe and Asia, across three dimensions of transparency: disclosure rules, practices, and data.

Results

The UK and Japanese self-regulation of payment disclosure had shared as well unique strengths and weaknesses. The UK and Japanese pharmaceutical industry trade groups declared transparency as the primary goal of payment disclosure, without, however, explaining the link between the two. The rules of payment disclosure in each country provided more insight into some payments but not others. Both trade groups did not reveal the recipients of certain payments by default, and the UK trade group also made the disclosure of some payments conditional on recipient consent. Drug company disclosure practices were more transparent in the UK, allowing for greater availability and accessibility of payment data and insight into underreporting or misreporting of payments by companies. Nevertheless, the share of payments made to named recipients was three times higher in Japan than in the UK, indicating higher transparency of disclosure data.

Conclusions

The UK and Japan performed differently across the three dimensions of transparency, suggesting that any comprehensive analysis of self-regulation of payment disclosure must triangulate analysis of disclosure rules, practices, and data. We found limited evidence to support key claims regarding the strengths of self-regulation, while often finding it inferior to public regulation of payment disclosure. We suggest how the self-regulation of payment disclosure in each country can be enhanced and, in the long run, replaced by public regulation to strengthen the industry’s accountability to the public.

Background

Collaborations between pharmaceutical companies and the healthcare sector facilitate drug development and commercialisation, disease awareness, and disseminating information on medicine use [1,2,3,4,5]. They have become ever more pronounced as new specialist drugs, including for genetic diseases and some cancers, proliferated, requiring advice from medical experts and researchers throughout the product life cycle [6,7,8,9,10]. Similarly, decreasing industry investment in in-house drug development has emphasised academic partnerships as sources of innovation [11,12,13,14].

However, these collaborations may generate financial conflicts of interest (FCOIs) biasing healthcare practice, research, education, and policy [15,16,17]. The main global response to these concerns is disclosure of drug company payments – including for consultancies, sponsorships, and grants – to healthcare professionals (HCPs) and organisations (HCOs) [18,19,20]. Payment disclosure involves different “transparency provisions”: (1) self-regulation, primarily voluntary codes of conduct developed and enforced by industry trade groups; (2) public regulation, typically sunshine legislation mandating drug companies to disclose certain payments to some recipient categories; or (3) their combination [18, 21,22,23,24,25].

With few exceptions [26], research to-date has utilised single-case studies to examine the strengths and weaknesses of public regulation in the US [23, 27, 28] and France [29]; and self-regulation in the UK [30,31,32], Germany [10, 33], Ireland [34], Australia [18, 35,36,37], and Japan [38,39,40,41,42,43,44]. There have been fewer direct comparisons of countries with public regulation and self-regulation [18, 21, 22, 24, 45, 46]. Compared with public regulation, self-regulation of payment disclosure is described by its proponents as less burdensome, more flexible and responsive to the needs of a globalised industry [10, 25, 45, 47,48,49]. Strengths of public regulation include mandatory disclosures and more healthcare industries and payments covered by the requirements [18, 22, 45]. Crucially, both are critiqued for low disclosure accessibility, quality, and user-friendliness [21, 24].

Against this background, except for Europe [24, 26, 46], the relative strengths and weaknesses of self-regulation of payment disclosure across countries are poorly understood. Nevertheless, the variability of European cases is low as nearly all national trade groups follow – but rarely exceed – the minimum requirements from the Code of Practice of the European Federation of Pharmaceutical Industries and Associations (EFPIA) [21, 24, 50, 51]. What also underscores the need for comparative analysis of self-regulation of payment disclosure is its global prevalence [36]. Indeed, it is the default approach in countries without public regulation [51] and only two of the largest pharmaceutical markets – the US and France – have comprehensive sunshine legislation, while other known examples have limited scope or exist alongside self-regulation [18, 21, 24]. Self-regulation is also an important policy tool, e.g. in benchmarking [18], with country rankings of disclosure statistics and code compliance becoming an argument against the introduction of sunshine legislation [47, 52, 53].

We seek to compare the strengths and weaknesses of self-regulation of payment disclosure using two crucial case studies: the UK and Japan. While each country has been analysed separately, the focus has been on select payments to HCPs and HCOs [30, 31, 38, 39, 41, 42, 54]. Disparate research designs have further constrained possibilities for systematic analysis of the uniqueness and commonalities relative to other cases, reducing generalisable international policy learning [30, 31, 38]. In addressing these limitations, we use a typology of dimensions of transparency developed in internationally comparative research on payments to patient organisations, an area of industry self-regulation separate from the scope of this article [55]. We examine three such dimensions: (1) the scope of payment disclosure rules established by the industry trade groups; (2) how payments are disclosed in practice; and (3) the insight into industry-healthcare sector financial relationships generated by the payment data.

Japan and the UK have the third and seventh largest pharmaceutical markets worldwide [56], with extensive drug manufacturing and research and development (R&D) capacity [57, 58]. They are likely the best cases of self-regulation of payment disclosure in Europe and Asia, and potentially globally, having been characterised as more comprehensive than another established example of self-regulation, Australia [36, 38].

Since 2015, the Code of Practice of the Association of the British Pharmaceutical Industry (henceforth ABPI Code) has included payments and recipients exceeding EFPIA’s minimum requirements [24, 31, 59]. Despite important shortcomings [30, 31], the transparency of disclosures by companies following the ABPI Code is the highest within the EFPIA ecosystem, especially following its major revision in 2021, which introduced the disclosure of additional payment and recipient categories [21, 24]. Further, while only around two-thirds of individuals receiving payments consent to disclosure [60], this reflects the pattern observed in most European countries where industry interprets payment data as falling under the European privacy laws [26, 30, 47, 52, 53].

As far as we know, since 2011 the Japan Pharmaceutical Manufacturers Association (JPMA) has developed Asia’s most comprehensive standards on interactions with the healthcare sector. Specifically, the “Transparency Guidelines for the Relation between Corporate Activities and Medical Institutions” (henceforth JPMA Transparency Guidelines) [61], focus on payment disclosure, while the “JPMA Code of Practice” (henceforth JPMA Code) [62] covers promotion.

Context of self-regulation of payment disclosure in the UK and Japan

The pharmaceutical industry globally faces similar political and regulatory pressures to enhance transparency [18, 21, 25, 36, 49], potentially resulting in similar approaches to disclosure. In the UK, key manifestations of these pressures have included a parliamentary investigation into the industry’s influence on healthcare provision [63] and a medical product safety review [64, 65]. Controversy has also been caused by the Japanese company Astellas’ illicit marketing, resulting in its temporary suspension from the ABPI [48, 66]. In Japan, several well-publicised scandals have involved kickbacks to increase prescription [41] and data fabrication in clinical trials [44, 67,68,69,70,71]. The ensuing criminal investigations have prompted policy discussions on transparency, leading to the development of the new Clinical Trials Act [44, 72, 73].

As members of the International Federation of Pharmaceutical Manufacturers and Associations (IFPMA) [74, 75], the ABPI and JPMA subscribe to its Code of Practice [62, 74] and transparency guidelines on fees for services [76], sponsorship of events and meetings [77], and continuing medical education [78]. They also share company members, including the UK-headquartered GSK and AstraZeneca and Japan-headquartered Takeda and Astellas [79,80,81,82]. Like other multinationals, their global codes of conduct [49, 51] make similar commitments to transparency [83,84,85,86] and payment disclosure [85, 86]. The ABPI and JPMA collaborate [87, 88], including via EFPIA Japan [89], an organisation representing European company interests which contributed to the development of the JPMA Transparency Guidelines [90].

Although the “broad parameters” of payment disclosure are shared internationally, detailed expectations vary across jurisdictions [18, 25]. Indeed, IFPMA expects its member associations to implement codes reflecting national circumstances [49]. Therefore, the ABPI implements EFPIA’s standards [74] that exceed IFPMA’s requirements regarding payment disclosure [51]. In comparison, the JPMA responds, among others, to the expectations of domestic fair trade and marketing organisations [91].

Payment disclosure is also affected by the trade groups’ varying governance structures. As part of EFPIA’s compliance monitoring, the ABPI publishes yearly evaluations of company compliance with its Code [47, 52, 53]. Complaints regarding alleged breaches of the ABPI Code are considered by the Prescription Medicines Code of Practice Authority (PMCPA), a quasi-autonomous body within the ABPI [92, 93]. The PMCPA publishes reports of all complaints and levies “administrative charges” on breaching companies, and in instances of particularly problematic company behaviour, it can rule a breach of §2 of the Code (promotion that “brings discredit to, and reduction of confidence”) and can issue a public reprimand [92]. Contrastingly, because the Transparency Guidelines are voluntary for its member companies, the JPMA does not monitor company adherence or sanction non-compliance [38, 94]. However, the JPMA’s executive board investigates complaints regarding alleged breaches of the JMPA Code, which is distinct to the Transparency Guidelines, and decides on penalties, but does not publish their details systematically [94].

Differences between each country’s “medical economy” are also significant [91, 95, 96]. Although both have publicly funded healthcare offering universal coverage to residents [57], Japan combines physician-owned hospitals and clinics with a sizeable sector of hospitals owned by public authorities and a few not-for-profit hospitals run by charities or foundations [58, 91]. Contrastingly, in the UK, healthcare is predominantly delivered by practice-owning physicians, who are patients’ first point of contact, and salaried staff in publicly-owned National Health Service (NHS) hospitals [97].

To some extent, these health-system level differences entail different patterns of the industry-healthcare sector interactions. Like in other major pharmaceutical markets [91], in the UK and Japan, industry marketing targets physicians, including medical Key Opinion Leaders (KOLs) [63, 94, 98, 99] and hospitals [41, 100]. The JPMA has also sought to regulate financial ties pertinent to Japan, such as extensive funding of physicians’ and university research [91]. Similarly, while attempts at establishing reciprocity and loyalty with physicians have been reported internationally [101,102,103], their role might be more pronounced in Japan, given the cultural importance of informal gifts, associated with evidence of bribes and kickbacks in medical procurement, prescription, and clinical trials [91].

Lastly, self-regulation belongs to wider mechanisms governing interactions between pharmaceutical companies and the healthcare sector [10, 51, 104]. In Japan, FCOIs are regulated chiefly via the Physicians’ Act, the Medical Service Act, and the Health Insurance Act, which criminalise bribes, kickbacks and gifts to publicly employed physicians [91]. In addition, the Fair Competition Code [105] determines how recipients can spend company payments [91, 105], while the Clinical Trials Act mandates financial disclosure by researchers and research institutions [44, 73]. Although the Clinical Trials Act does not mandate disclosure from drug companies, the study types from the Clinical Trials Act were voluntarily adopted in the JPMA Transparency Guidelines in 2018 in relation to the reporting of R&D payments and expenditure (no changes were introduced to the reporting of non-R&D payments). Nevertheless, FCOI regulation by professional organisations [91] and hospitals [106] has been historically weak, with reports of senior medics seeking to maintain close industry ties [42, 94].

In the UK, the overarching approach to undue influence is covered by the Bribery Act, making bribery a corporate offence [45, 93]. The NHS has voluntary guidance on FCOI disclosure, which is characterised by important loopholes [107, 108]. FCOIs are also self-regulated by the medical profession, but attempts at creating a centralised register for all clinical staff have been inconclusive [109, 110]. The ABPI maintains close contact with government and professional organisations [93], which support industry self-regulation of payment disclosure [64, 111], having issued joint guidelines for HCOs with the ABPI [112]. Finally, freedom of information legislation is increasingly important in gaining insights into FCOIs in the NHS [108].

Methods

We compared the transparency of disclosure rules, practices, and data [55].

We examined disclosure rules which the ABPI and JPMA described in codes, guidelines, and press releases. We identified these documents on trade group websites and by googling “Association of the British Pharmaceutical Industry”, “ABPI”, “Japan Pharmaceutical Manufacturers Association” (日本製薬工業協会 or Nihon Seiyaku Kogyo Kyokai), “JPMA”, together with the terms “transparency” (透明性 or Tomeisei), “disclosure” (公開 or Kokai), “payments” (支払い or Shiharai), and “transfers of value”. We also consulted the 2019 publication “Pharmaceutical Association Code of Practice”, comprising the JPMA Code of Practice, the Transparency Guidelines, and a glossary of relevant regulations.

We considered the JPMA Transparency Guidelines from 2018, both the English [61] and Japanese versions [113] (cited in-text as JTG2018), and the JPMA Code from October 2019, both the English [62] and Japanese [114] versions (cited in-text as JC2019). The Guidelines were further revised in 2022, without, however, introducing changes to payment disclosure. On the other hand, significant changes were being introduced to the ABPI Code at the time of writing and therefore we considered its 2019 version (cited in-text as AC2019) [74] and the newest revision implemented in 2021 (cited in-text as AC2021) [59]. We note any differences between the two, but if no change occurred, we only mention the 2019 version.

Drawing on earlier comparative research [24, 46] and single-case studies of ABPI and JPMA disclosures [30, 31, 38], we compared the disclosure rules in relation to the aim of disclosure, payment recipients, and payments either included or exempted from disclosure.

We examined disclosure practices suggested by to-date analyses of self-regulation [55, 115,116,117]. First, the availability of disclosures, understood as the share of companies publishing disclosures within companies committing to disclosure rules. Second, accessibility concerned where disclosures were published online. Third, we considered the electronic format of documents with drug company disclosures. Fourth, by evidence of underreporting we meant acknowledged cases of payments which should have been reported but were not. Fifth, evidence of misreporting involved attribution of payments to incorrect recipients.

We analysed disclosure availability, accessibility, and format in 2018 as this was the latest year for which we had collected the Japanese disclosures. In so doing, in the UK, we considered (1) Disclosure UK, the ABPI’s centralised payments database; (2) company “methodological notes” which describe their approaches to disclosure and are published alongside Disclosure UK [30, 31]; and (3) any relevant disclosure guides that we could find on the ABPI website [118]. In Japan, we evaluated disclosure availability, accessibility and format using payment disclosures published on individual company websites [38, 42, 43]. No changes to disclosure availability, accessibility and format have been made by either the ABPI or the JPMA since 2018.

We investigated the evidence of underreporting and misreporting by searching the PMCPA’s open archive for any cases involving breaches of clause 24 of the ABPI Code (“Transfers of Value to Health Professionals and Healthcare Organisations”); we considered the entire period since the current disclosure regulations were introduced in the UK (2015). There was no systematic evidence documenting potential underreporting and misreporting by JPMA members. Contrastingly, because JTG2018 are not mandatory (and involve no penalties for breaches), no platform existed for reporting potential cases of payment misreporting or underreporting.

In examining disclosure data, we summarised company disclosures using the original ABPI and JPMA payment categories, subsequently aggregating them into UK-only, Japan-only, and shared. We converted payment values from sterling and yen to US dollar using the average yearly exchange rates published by the UK Office of National Statistics [119] and the Bank of Japan [120] (1 USD = 0.75 pound sterling and 110.4 yen). The 2021 revision of the ABPI Code broadened the scope of one payment category (“joint working” replaced with “collaborative working”) and further specified the meaning of another (“fees for service and consultancy” replaced with “contracted services”) [59]. These changes did not alter the nature or structure of the disclosure data significantly, allowing for the same approach to data analysis to be applied in future research. The JPMA have made no changes to the reporting of payment categories since 2018. In practice, some companies delayed the reporting of R&D payments using the updated categories which the JPMA introduced voluntarily to JTG2018 to reflect the new Clinical Trials Act. Nevertheless, this delay did not affect the overall reported R&D volume of payments.

Two researchers collected and analysed the data separately for each country, with interpretations agreed by the entire team.

Results

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. 1

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

Disclosure rules

Aim 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 practices

Availability

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 data

Payments 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).

Discussion

In its current form, the self-regulation of payment disclosure has operated in Japan since 2011 and in the UK since 2015. In Japan, disclosure research has identified widespread FCOIs among clinical trialists [72, 140, 141] and authors of clinical practice guidelines [43, 142, 143], while in the UK – across hospitals [100], commissioning [144], expert advisory [145], and policymaking bodies [32]. However, given the overall volume of payments reported annually in each country – in 2018, close to $3bn in Japan and over $0.6bn in the UK – a significant share of FCOIs affecting healthcare actors is likely to go unnoticed. Our comparative case study of three dimensions of transparency – disclosure rules, practices, and data – sheds light into why this may be the case and how the transparency of payment disclosures could be improved.

The ABPI and JPMA disclosure rules covered a broad range of payment recipients. The ABPI mandated a greater scope of non-R&D payment disclosure. On the one hand, almost all JPMA non-R&D payments were encompassed by the ABPI categories. On the other hand, the ABPI categories often lacked corresponding JPMA categories. Conversely, the JPMA required the disclosure of not only R&D payments but also R&D expenditure. It also mandated disclosure without referring to recipient consent. Shared weaknesses included some payments being exempted from named disclosure, while others – not disclosed altogether.

Companies following the ABPI Code had more transparent disclosure practices. There was certainty about how many disclosures were available; they were more accessible and had an analysable format. The ABPI complaints mechanism provided some insights into possible underreporting and misreporting.

However, JPMA members generated more transparent disclosure data, with the overall share of payments to named recipients almost three times higher, and considerably higher company-level shares compared to the UK.

Implications for self-regulation and public regulation

We use our findings to evaluate claims regarding key advantages self-regulation as the industry’s preferred approach to governing drug promotion, stakeholder interactions and communication, and providing information on medicines [49, 104]. We also compare self-regulation with legally mandated transparency provisions identified in the US and eleven European countries [24].

Corresponding with IFPMA’s declared emphasis on increasing transparency [49, 51, 93], the JPMA and ABPI presented transparency as disclosure’s primary aim. Nevertheless, it was unexplained how transparency was to be achieved; for example, how disclosures were expected to be used, by whom, and to what purpose [20]. Limited attention given to “transparency relations” [146] is also common in European public regulation, as evidenced by a lack of granular and easily interpretable disclosures in many countries [24]. In comparison, public regulation does eliminate non-transparency resulting from the aggregation of certain payments by default [24]. Yet, the evidence from Japan (and Australia [36]) demonstrates that mandating disclosure may not require new sunshine legislation or clarifications of data privacy laws by public authorities, as it has been the case in Europe [18, 21, 24], and can be achieved within self-regulation.

In Japan, individual-level disclosure has arguably increased the industry’s accountability by revealing FCOIs underreported by KOLs [44, 99] and clinical triallists [72, 147] as well as examples of poor FCOI management [142, 143]. Contrastingly, the ABPI only “encourages” companies to observe its new non-mandatory guidance on “legitimate interests”, promising to “further increase the transparency” of non-R&D payments, but without committing explicitly to full disclosure, at least in the short run [64]. Yet, the comparison with Japan demonstrates that, contrary to the European industry’s emphasis on protecting commercial secrecy [46], it is possible to disclose at least organisational-level R&D payments without jeopardising companies’ R&D investments. It also shows the need for rebalancing the UK transparency debate – by focusing on encouraging the full disclosure of a relatively small share of non-R&D payments to individual HCPs [64] it has neglected systemic, non-individualised, and non-itemised disclosure of R&D payments to researchers and research institutions with key roles in drug development and marketing [46].

Reflecting IFPMA’s work on “synchronisation” of national codes [51], important similarities in the ABPI and JPMA disclosure rules were demonstrated by overlapping disclosure goals, recipient and payment categories. Nevertheless, as each trade group often referred to unique industry-supported activities, the extent of overlap was uncertain. Therefore, although the corresponding payment categories attracted around 90% of payments reported in each country, more granular descriptions of disclosures would likely produce lower figures. Furthermore, the emphasis of disclosure requirements differed, with the ABPI prioritising disclosure of non-R&D payments, while the JPMA – R&D payments and expenditure. Overall, while detailed internationally shared standards are a key advantage of self-regulation over national legislation in the EFPIA ecosystem [10, 22, 45], our findings suggest that the same cannot be said about the current state of self-regulation globally.

The assertion that self-regulation is more comprehensive than public regulation [10, 49, 51, 93, 148] received mixed support. The JPMA and ABPI disclosure rules covered prescribers as well as professions and roles shaping prescription through policy decisions [18, 31]. In this respect, self-regulation exceeded the scope of the US Sunshine Act, covering only physicians, nurses, and teaching hospitals [24], as well as disclosure legislation in several European countries, such as Estonia [149], Hungary [150] or Latvia [151], focusing on personnel involved in medicine prescription, supply, or distribution. Further, while the JPMA covered life science researchers, only Danish disclosure legislation included disclosure of payments to select researchers, i.e., bioanalysts and pharmacoeconomists [152].

Nevertheless, self-regulation missed some key areas of concern [153]. Unlike in some countries with public regulation, the ABPI and JPMA did not cover payments related to over-the-counter medicines [21], medical devices, or veterinary products [24, 36]. Also excluded from disclosure were organisations targeted by the industry’s lobbying [32, 103, 154,155,156], which are sometimes covered by public regulation, such as the Ministry of Health (Portugal) [157] and media organisations (France) [158].

The scope of payments covered by the ABPI and JPMA was broader than in some European countries with public regulation, such as Greece [159] and Hungary [150], concentrating on event delivery or participation. Notably, R&D payment disclosure mandated by the JPMA was significantly closer to the most comprehensive French and US disclosure requirements than self-regulation in Europe (R&D payments always aggregated) or Australia (R&D payments not disclosed) [18, 36, 46]. Contrastingly, payments for food and beverage were covered by public regulation in France and the US [18], but excluded from disclosure in the UK below the threshold of $100 (£75) and disclosed in the aggregate in Japan. This is an important gap in disclosure as data disclosed under the US Sunshine Act demonstrates that payments for food and beverage are widespread [160,161,162] and associated with increased prescription of promoted drugs [163, 164]. Finally, US disclosure requirements also include ownership or investment interests (e.g. royalties, licenses) held by physicians or their immediate family members – about 12% of reported payments amounts in the USA – but these are not covered by the ABPI and JPMA [165].

There was mixed evidence for the claim that self-regulation is more detailed than public regulation [51]. The disclosure categories used by the ABPI and JPMA were broad and the reporting did not distinguish their constituent recipients or drug company activities, making them difficult to interpret [166]. However, only two countries with public regulation – the US and France – use comprehensive sets of specific payment categories [18]; others either disclose few specific payments (Latvia [151], Lithuania [167, 168]) or use broad categories such as “professional affiliation” (Denmark [169]) or a “value, good or right” (Portugal [157]), without itemising them in the disclosure databases [24]. In addition, consistent with earlier research [30, 31], we identified examples of ambiguous or unstandardised reporting, potentially preventable by more detailed disclosure guidelines.

We found problematic the claim that self-regulation often involves higher levels of compliance than public regulation [49, 93, 170]. The cases of ABPI members not disclosing payments demonstrate that companies may start or stop following self-regulation at any point, creating potential gaps and uncertainties in reporting. Further, while the proponents of self-regulation would expect effective scrutiny of companies’ conduct from HCPs [93], the fact that only two HCPs complained about payment misreporting suggests, more than anything, that relatively few recipients check their disclosure records in sufficient detail, potentially making the PMCPA complaints-based system insufficient to reveal this practice systematically. This interpretation is supported by frequent reporting inaccuracies associated with the absence of recipient identifiers and categories [24, 31]. Similarly, the cases of underreporting considered by the PMCPA indicate that only current or ex-employees – and not, for example, competitor companies – may have sufficient knowledge of relevant internal practices, therefore limiting the pool of potential complainants [48, 92, 93, 171].

Key examples of noncompliance in Japan included delayed disclosure publication, some companies creating barriers to disclosure access, and reporting R&D payments in the aggregate [44, 106, 143]. However, consistent with earlier concerns about difficulties in measuring companies’ compliance with the JPMA standards [91], revealing underreporting and misreporting was practically impossible without publicly available disclosure standards (“methodological notes”) used by individual companies and a complaints mechanism overseen by the trade group. Just like in the UK, misreporting might be caused by ambiguities in disclosure data. Also reflecting earlier research [54, 99], underreporting could result from the relatively narrow JPMA payment categories, encompassing fewer industry activities than the ABPI categories. Notably, two of the three UK cases of underreporting involved Japanese companies. The PMCPA characterised one of them, Astellas, as displaying “multiple organisational and cultural failings” and a corporate culture that prioritised “the bottom line” over compliance obligations and ethical norms [48].

Contrastingly, the only available evaluation of the US Open Payments database that we were aware of demonstrated little evidence of non-compliance with the reporting standards [172], suggesting the superiority of legally binding requirements associated with potential penalties over self-regulatory standards [21, 24, 46].

Limitations

Our study has important limitations. As pharmaceutical industry self-regulation has not been mapped in Asia [51], we might have missed important comparators for the UK as the strongest case of self-regulation in Europe [24]. Further, because we compared only two cases, any strengths and weaknesses identified were only relative to the other case. Therefore, considering another comparator might have revealed disclosure aims, recipients or payments missing from disclosure. In addition, as the recipient and payment categories were unique, and often described only in general terms, we could only ascertain their “correspondence” but not “equivalence”. Consequently, the comparisons across the payment categories require cautious interpretation.

We analysed similarities and differences between the two cases descriptively, without explaining the underlying mechanisms. For example, we did not consider the company and trade group and cultures more broadly, which may be an important factor behind the varying disclosure practices [26, 51]. One key aspect of such cultures worth in-depth investigation concerns the mechanisms for collecting and checking the quality and accuracy of disclosure data.

Future research on varieties of self-regulation might employ mixed methods to examine the two-way relationship between self-regulation at the global, trade group, and company levels, as well as other transparency provisions and the “medical economy” more broadly [48, 66, 91]. Of particular importance would be to trace the development of the different forms of regulation over time [104].

Conclusions and policy recommendations

Our study indicates that measuring the transparency of payment disclosure associated with self-regulation should triangulate disclosure rules, practices, and data, given the possible inconsistences and important relationships between these three dimensions.

Overall, our comparative analysis of payment disclosure in the UK and Japan offered limited evidence to support key claims concerning the strengths of self-regulation. Despite its important advantages, self-regulation seemed inferior to public regulation, which was consistent with findings from Europe and Australia [24, 36, 46].

Our comparison highlights the priority areas for improvement for both the ABPI and JPMA.

• Eliminating exemptions from disclosure.

• Removing aggregate disclosure by default.

• Introducing more granular disclosure of payments and recipients.

• Increasing the richness of disclosure data by expanding recipient characteristics, following the example of the Australian pharmaceutical industry trade group [173].

• Introducing additional standards to prevent underreporting (mandatory declarations that no payments were made) and misreporting (mandatory recipient identifiers).

• Creating stronger compliance monitoring of disclosure practices and data [51], including regular independent audits to complement the reactive, complaints-based systems [48, 63, 66, 153].

• Introducing public rewards and sanctions specific to disclosure standards, e.g. highlighting top performers and underperformers across different disclosure practices and data [148].

We identified several priority areas for the JPMA.

• Extending the list of non-R&D payment categories (e.g. a broader range of consultancy-related payments, following the example of the ABPI Code) to enhance the scope of disclosure.

• Mandating companies to publish their internal disclosure standards using the example of “methodological notes” introduced by the EFPIA and ABPI.

• Creating a searchable and downloadable payments database for all companies subscribing to the JPMA Transparency Guidelines.

• Making the Transparency Guidelines mandatory for JPMA members, including sanctions for non-compliance.

• Increasing the transparency of the complaints process by establishing an arms-length body investigating and publishing details of alleged breaches of standards, potentially modelled on the PMCPA.

Finally, we propose the following priorities for the ABPI.

• Making the receipt of payments conditional on agreeing for their publication, which seems to be the established practice followed by the JPMA members.

• Expanding and standardising the scope of internal disclosure standards described in company methodological notes [46, 51].

Our findings reinforce earlier calls for replacing industry self-regulation with sunshine legislation made in the UK and Japan [38, 65] and correspond with the UK government’s ongoing work on expanding or replacing Disclosure UK with legislation [174]. Nevertheless, as such, public regulation not a panacea. Any legislation should be premised on a comprehensive international evaluation of disclosure rules, practices and data [24, 46, 55, 175], including their unintended consequences [23, 176, 177]. One structured way of embedding the perspectives of HCPs, HCOs, patients, and members of the public into the policy proposals would involve developing a “theory of change” [178] capturing multifaceted relationships between disclosure, transparency, as well as the economic power of the industry and the state’s regulatory power [18,19,20]. In so doing, it is important to consider a range of potential goals of transparency, such as increasing the industry’s accountability to the public revealing potential FCOIs [21, 24, 26, 55, 106] and protecting the integrity of healthcare research or policymaking [20, 146, 179].

Given the evidence of the limited impact of sunshine legislation on reducing FCOIs or encouraging their understanding by the public [23], any new legislation should be complemented by educational campaigns enhancing the understanding and scrutiny of disclosure rules, practices and data [180]. Any sunshine legislation should be integrated – including via shared databases – with FCOI evaluation undertaken by HCOs such as hospitals, commissioning bodies, regulators, and professional organisations. No less important are solutions addressing the underlying problem of financial dependency of healthcare actors on drug company funding [18, 117, 181]. Some of the available options include FCOI management policies such as limiting or banning certain forms of financial relationships [18, 20, 26, 182].

Appendix

Tables 5, 6, 7, 8, 9, 10  

Table 5 Distribution of payments across corresponding and non-corresponding JPMA and ABPI payment categories (2018)

Table 6 Payments to individual- and organisational-level recipients under the JPMA Transparency Guidelines and the ABPI Code (2018)

Table 7 Descriptive statistics for aggregated company-level payments reported by companies following the ABPI Code (2018)

Table 8 Descriptive statistics for aggregated company-level payments reported by JPMA member companies (2018)

Table 9 Aggregated payments – company-level data reported by companies following the ABPI Code (2018)

Table 10 Aggregated payments – company-level data reported by companies following the JPMA Code