Introduction

Patient groups have been set up in Canada to represent patients at multiple levels in the healthcare system and more widely. Typically, they are concerned with people suffering from a single condition, but they may also represent people with multiple related conditions, e.g., various forms of arthritis. They may lobby for Health Canada to approve new drugs and for particular products to be provided for their members. They also speak for patients with healthcare professionals and healthcare institutions such as hospitals and finally they are often the voice of patients in the media.

Since the Canadian federal government rolled back funding of patient groups in the mid 1990s [1] groups have had to seek new sources of revenue and many of them receive money from pharmaceutical companies. Some groups have lobbied provincial governments to include drugs on formularies made by companies from which they receive grants [2, 3].

There has not been any systematic analysis of patient group funding by pharmaceutical companies in Canada and whether there is an association between funding and the activities of groups. This study looks at this issue by examining the submissions that patient groups make to the Common Drug Review (CDR) and the panCanadian Oncology Drug Review (pCODR), the arms of the Canadian Agency for Drugs and Technology in Health that make recommendations about whether provincial governments should fund medicines in general (CDR) and oncology products in particular (pCODR). Specifically, this research examines whether there is an association between the positions that the groups take about the products under consideration and the receipt of money from companies.

Methods

Results

There was a total of 222 reports for drug-indications– 114 from CDR and 108 from pCODR. There were 372 submissions from 93 different patient groups, 230 submissions for products considered by CDR (range 1–6 submissions per drug-indication) and 142 for pCODR products (range 1–3 submissions per drug-indication). Individual patient groups made between 1 and 15 separate submissions (S1 Table).

Groups declared a total of 1896 conflicts with drug companies and organizations representing drug companies in 324 (87.1% of the total number of submissions) individual submissions to both CDR (208 submissions and 1493 conflicts) and pCODR (119 submissions and 403 conflicts). The median number of conflicts per submission for CDR was 7 (interquartile range (IQR) 4,10) and for pCODR it was 1 (IQR 1,5). Overall there were 279 submissions (86.1% of all cases where the submission declared a conflict and 75.0% of the total number of submissions) where there was a conflict with the company marketing the product for a total of 1557 conflicts–CDR 164 submissions and 1164 conflicts, pCODR 115 submissions and 393 conflicts. In 48 submissions there was no conflict with the company marketing the product although conflicts with other companies were declared–CDR 44 submissions and 329 conflicts, pCODR 4 submissions and 16 conflicts. In 30 submissions no conflict with any company was declared (CDR 15 submissions and pCODR 15 submissions). Whether conflicts existed could not be determined for the remaining 15 submissions–CDR 7 submissions, pCODR 8 submissions (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Number of conflicts per patient group submission.

https://doi.org/10.1371/journal.pone.0212399.t001

The degree of granularity in the declaration of conflicts varied considerably, especially for patient groups who made submissions to CDR. Some did not state the time period when the conflicts occurred, whereas others gave specific time periods but those time periods varied from 1 year to 12 years. In other cases, groups declared that they had a conflict with one company and a number of unnamed other companies. There were 20 submissions (8 different patient groups) where the conflict declaration stated how much of the group’s total budget came from donations from pharmaceutical companies, but many of the 8 groups were inconsistent in providing that information (S2 Table). In the case of Myeloma Canada, the Asthma Society of Canada and Tuberous Sclerosis Canada donations from companies were a substantial portion of the organization’s budget– 36%, 20% and <20%, respectively, but for Cystic Fibrosis Canada, Multiple Sclerosis Society of Canada and Foundation for Fighting Blindness, the amounts were very small– 1.5% - 2.0%, <2%, 0.9%, respectively. In two other submissions, patient groups declared their conflicts but did not seem to see any value in making those declarations and included the following statement: “We do not see the purpose of asking how much money has been contributed by any entity that may have an interest in this.”

(https://www.cadth.ca/sites/default/files/cdr/relatedinfo/SR0522_Galafold_Patient_Input.pdf) Occasionally groups gave vague statements about how the money that they received from companies was used, e.g., for research or educational events, but 265 of the 324 (81.7%) declarations only named the companies that had donated money to the groups. In addition, after September 1, 2017 21 submissions listed grants from companies in one of four brackets: $0-$5,000–18 companies, $5,000 - $10,000–12 companies, $10,001 - $50,000–20 companies, >$50,000–43 companies. (Some submissions named more than one company, some companies gave money to more than one group.)

There were 268 submissions to CDR and pCODR, excluding submissions involving subsequent entry biologics, where patient groups declared a conflict with the company making the product (216 submissions), the absence of a conflict with the company (29 submissions) or no conflict at all (23 submissions) and where the groups expressed a view about the drug-indication being considered. Inter-rater reliability (Kappa) in scoring the views of patient groups was 0.4662 (95% CI 0.3040, 0.6284). The relatively low agreement between the two reviewers was due to initial confusion about what constituted a “neutral” opinion (34 of 36 disagreements). Table 2 gives examples of positive, neutral and negative views from patient groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Examples of patient group positive, neutral and negative views about a drug-indication.

https://doi.org/10.1371/journal.pone.0212399.t002

Groups had positive views in 242 (90.3%) submissions, neutral in 24 (9.0%) and negative in 2 (0.7%) (Table 3). Table 3 also shows that views are consistently positive and that the distribution of views–positive, neutral, negative–is the same regardless of whether groups have a conflict with the company marketing the drug, a conflict with other companies or no conflict at all, p = 0.3117, Chi square.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Distribution of patient group views about drug-indications under consideration^*.

https://doi.org/10.1371/journal.pone.0212399.t003

The comparison between views of patient groups and CDR and/or pCODR recommendations was analyzed grouping neutral views from groups with negative ones. (Grouping neutral views with positive ones did not change the outcome of any of the comparisons, results not shown.) Fisher’s exact test shows no difference between the views of patient groups about drug-indications and the combination of CDR and pCODR recommendations (p = 0.78). Similarly, there was no difference when views of patient groups were compared separately with CDR or pCODR recommendations (p = 0.38 and p = 0.17, respectively, Fisher’s exact test) (Table 4). Finally, regardless of whether patient groups had a conflict with a company making the drug, a conflict with another company or no conflict, there was no statistical difference between their views and the combination of CDR and pCODR recommendations (p = 0.99, p = 0.20, p = 0.99, respectively, Fisher’s exact test) (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Comparison of views of patient groups with recommendations from Common Drug Review and panCanadian Oncology Drug Review^{*, †}.

https://doi.org/10.1371/journal.pone.0212399.t004

There were 22 submissions about subsequent entry biologics, but in one case it was not possible to determine if the group had a conflict and in a second case the view of the group was not available. In the other 20 cases the distribution of views did not show any clear pattern regardless of whether the groups had a conflict with the submitting company, the originator company, both types of companies or no conflict with any companies (Table 5). The small numbers precluded any meaningful statistical analysis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Subsequent entry biologics (SEB)–distribution of views of patient groups.

https://doi.org/10.1371/journal.pone.0212399.t005

Seventy (62.5%) submissions from patient groups (out of a total of 112 submissions where the group declared a conflict with the company marketing the drug) commented on pCODR preliminary recommendations. In 19 cases, pCODR did not recommend funding and in 17 of those 19 (89.4%), patient groups disagreed with the decision. Otherwise, patient groups either agreed or agreed in part with a decision to either fund or fund conditional on the drug-indication being cost-effective in 48 of 51 cases (Table 6). In the one case where a patient group disagreed with the recommendation to fund, it was because the group felt that the recommendation meant that certain groups of patients would not be eligible to receive the drug and this took away patients’ choice. The distribution of patient group responses in relation to the three funding recommendations from pCODR was statistically significantly different, p < 0.0001 (Fisher exact test). There were 10 submissions containing comments from groups with no conflicts with the company marketing the drug, too few to allow for any meaningful statistical analysis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Patient group response to preliminary recommendation from panCanadian Oncology Drug Review.

https://doi.org/10.1371/journal.pone.0212399.t006

Discussion

When patient groups made submissions to CDR and pCODR about drug-indications, 75% of the time these groups declared a conflict with the company marketing the drug. The total of 1896 conflicts is a minimum, as some groups only listed companies that had a direct or indirect interest in the drug under consideration, whereas other groups made more comprehensive declarations. Other times, groups declared conflicts with one company and said they had additional conflicts but did not name the companies. Finally, there was no independent search for additional conflicts that groups may have had.

Patient groups’ views about products were positive just under 90% of the time and whether or not groups had a conflict with the particular company making the product under consideration, a conflict with another company or had no conflict at all did not make a difference in the distribution of their positive, neutral or negative views. At the same time, the views of patient groups were not statistically different from the recommendations from CDR and/or pCODR regardless of the presence or absence of conflicts.

One explanation for the almost uniform positive recommendations from all patient groups is that they may be motivated to make a submission because they feel strongly that the drug-indication will have a positive benefit for the patients that they represent and having a conflict plays no role in coming to this conclusion. This view is supported by the lack of any statistical difference between views of patient groups and the recommendations from CDR and/or pCODR. pCODR’s refusal prior to September 1, 2017 to accept submissions from groups likely to be the most affected by conflicts–those receiving funding from only a single company and those where a single company contributed more than 50% of their budget–may have skewed the results. However, there is no publicly available data about whether any patient group was actually denied participation because of this provision. There were too few submissions regarding subsequent entry biologics to draw any conclusions about whether the presence or absence of conflicts was associated with the views of patient groups.

The presence of a conflict with a company making the drug under consideration may play a role in the views of patient groups in certain circumstances. When the preliminary recommendation from pCODR was not to fund a drug-indication, patient groups almost always (17 out of 19 cases) disagreed and all of the groups that disagreed had a conflict with the company making the drug. When the recommendation was to fund or fund conditional on the drug being cost-effective, groups were in agreement with the recommendation in 48 out of 51 cases. Patient groups may have disagreed with the preliminary recommendation not to fund a drug-indication because they felt that the recommendation did not fully reflect their views about treatment duration, side effects, adherence, independence, psychosocial quality of life and avoiding further disease [10].

This study does not prove that funding from pharmaceutical companies directs the views that patient groups have about whether drugs should receive public funding. However, industry funding does put patient groups in a conflict of interest situation, where their primary interest is in the welfare of the patients that they represent and a secondary interest is in the financial health of the companies that provide them with funding and that are marketing the drugs under consideration. The possibility that the position that patient groups adopt may be influenced by their funding is raised by the guide that Medicines Australia published for its membership. The guide says that companies that sponsor non-profit groups might find that such sponsorship increases their chances of getting their drugs publicly funded under Australia’s Pharmaceutical Benefits Scheme [11].

Batt has provided examples of how Canadian breast cancer patient groups altered their behaviour once they started to accept donations from drug companies [1]. After the National Institute for Excellence in Health and Social Services of Quebec recommended that the Ministry of Health not fund 4 oncology drugs due to cost-effectiveness concerns, the Coalition Priorité Cancer (CPC), a Quebec-based patient advocacy group, denounced the decision and lobbied extensively to reverse it and was successful for 3 of the 4 drugs. A subsequent investigation of the group concluded that its commitment to its patient-members did not appear to be optimal based on a number of observations, including the absence of a clear position or warning against the use of bevacizumab for breast cancer and the CPC’s focus on the issue of reimbursement of expensive, low-efficiency drugs [12].

Reports from other countries have documented an association between patient groups receiving industry funding and the positions that they take in areas of concern to drug companies. In the United States (US), the Centers for Medicare and Medicaid Services proposed a project aimed at lowering spending on the most costly treatments offered under Medicare Part B, the federal insurance plan that covers outpatient drug costs. The proposed reform was to study the effect of modifying reimbursement methods so as to decrease physicians’ incentives to administer the most expensive medications. In addition to drug companies and doctors’ groups, 147 patients’ groups signed letters opposing the project, 110 of which received funding from the pharmaceutical industry [13]. The Epilepsy Foundation which receives funding from 4 companies marketing medications for epilepsy and that has representatives from these 4 companies on its board, campaigned for bills introduced in US state legislatures that would make it harder for pharmacists to substitute generic drugs for brand name epilepsy medications [14]. A survey of patient and consumer groups in Europe found an association between receiving industry sponsorship and support for an expanded role for the pharmaceutical industry as an information provider to the public about the products that companies make [15].

There is no research into the number or percent of Canadian patient groups that receive funding from pharmaceutical companies, but two recent papers from the US have explored this issue. McCoy and colleagues [16] looked at 104 groups with annual revenue of at least $7.5 million (USD). Eighty-six (83%) reported receiving donations from the pharmaceutical industry and only 1 explicitly stated that it did not receive donations. A second survey of a random sample of 7,685 patient groups operating in the US found that just over two-thirds of the responding organizations reported receiving industry funding, with almost 12% getting more than half of their funding from industry [17]. The recently passed legislation in Ontario requiring the disclosure of all transfers of value to healthcare professionals, healthcare institutions and organizations and patient groups [18] should provide Canadian data about patient group funding, at least for groups based in Ontario, provided the newly elected Progressive Conservative government implements the legislation.

Many patient groups in North America accept industry funding, but the situation may be different in a couple of European countries. Batt’s interpretation (Sharon Batt, personal communication, August 18, 2018) of a report on health consumer and patient organizations in seven European Union countries [19] is that the relatively small number of health consumer and patient organizations in the Netherlands and Sweden (200–250 and 50–100, respectively) makes government funding there more feasible. The Ministry of Health in the Netherlands spends more than 40 million euros annually on these groups and only a limited number of groups in that country receive funding from pharmaceutical companies. In Sweden, almost all groups receive structural funding from the Ministry of Health and Social Affairs. Whether groups in these countries take different positions regarding industry interests compared to groups in countries that are more reliant on industry money has not been explored.

Conclusion

The large majority of patient groups that make submissions to CDR and pCODR have conflicts with the companies marketing the drugs and have positive views about the products that they are commenting on. Whether this is an association or a cause and effect or whether conflicts even play a role in the views of patient groups have important policy implications for patient groups, governments and health technology assessment bodies. If there is a cause and effect, then patient groups need to consider whether they are serving their membership by accepting industry funding, governments need to think about working with patient groups to develop new sources of unbiased support and health technology assessment agencies need to consider the weight that they give to patient group input. Further research to try and resolve this question is urgently needed. In the meantime, both CDR and pCODR should adopt a precautionary principle approach and require groups to disclose all of their conflicts, state the percent of their total funding that comes from pharmaceutical companies and organizations representing companies, extend the reporting period for conflicts to 5 years prior to the date of submission and revert back to the pre-September 2017 position of pCODR and exclude groups that are funded by only a single company or that derive more than 50% of their revenue from a single company.

Supporting information

Acknowledgments

The following people commented on an earlier version of this manuscript: Sharon Batt, Sarah Berglas, Alice Fabbri, Lisa Parker. Catherine Oliver did the duplicate coding. Statistical advice was provided by Alex Kiss and Kevin Brown.