Introduction

Conflict of interest (COI) has been defined as “a set of circumstances that creates a risk that professional judgment or actions regarding a primary interest will be unduly influenced by a secondary interest” [1,2]. Physicians’ clinical decisions and practice patterns can be influenced by industry relationships [3,4]. The potential for COI may be widespread; according to public registers in the United States, financial relationships with industry were reported by 45% of otolaryngologists, 50% of orthopedic surgeons, 55% of plastic surgeons, 60% of urologists and 78% of neurologists [5].

Various approaches have been implemented by funders, professional societies and academic institutions to manage COI. Disclosure policies are the most commonly applied strategy to address COI [6–8]. However, research shows that guideline authors [9] and professional association meeting participants [10] failed to disclose payments from industry. Analysis of COI policies at American [11] and Canadian [12] universities and research institutions found that, while the majority had COI policies, they varied in criteria for what to disclose and how to manage COI. Limitation is another strategy for managing COI whereby organizations such as hospitals apply policies that prohibit employees from accepting industry incentives, or restrict visits from industry representatives [13]. Despite the presence of such limitation policies, physicians who were interviewed said that they trusted and valued information from drug industry representatives, and felt that it benefited patients [14,15].

Physician relationships with device industry representatives have not been previously studied to identify whether COI is a concern or how it is managed. This is particularly relevant among physicians who use higher-risk cardiovascular or orthopedic devices that have been associated with adverse events [16,17], many of whom are involved in their development, or pre- or post-market evaluation [18,19]. Furthermore, interaction with representatives of the device industry differs from relations with drug industry representatives. Both types of representatives discuss and promote the uses and benefits of their products with physicians but, in contrast to drug representatives, device representatives are frequently present during instances of patient care when their products are being used. In part this occurs because hospitals seek to optimize the efficiency of supply management, standardize surgery and reduce costs through “just-in-time” delivery of medical devices by representatives [20]. Thus, physician and institutional factors may result in considerable interaction between physicians and device representatives. To better understand this phenomenon, including whether physician relations with device representatives creates opportunities for COI and how it is managed, this study explored the nature of interactions with device industry representatives among physicians who use implantable cardiovascular and orthopedic devices to identify opportunities for COI that may influence patient safety.

Methods

Results

Participants

Twenty-two physicians who implanted cardiovascular (n = 10) and orthopedic devices (n = 12) were interviewed (Table 1) from among 27 who consented of the 561 that were invited to participate. These included 8, 10 and 4 early, mid and late career physicians, respectively, from five provinces. Views were similar between those who implanted cardiovascular and orthopedic devices.

Download:

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

Table 1. Demographic characteristics of interview participants.

https://doi.org/10.1371/journal.pone.0174934.t001

Physicians reliant on representatives

Representatives served multiple roles (Table 2). Information and support from representatives was considered an important factor in making individual or group decisions about which devices to purchase. Apart from colleagues, representatives were most frequently mentioned as a source of information about devices. Representatives supported training by either providing or linking physicians with training opportunities, and providing support during the learning curve period. Most notably, representatives were needed to ensure that devices, many of which are complex and comprised of a large number of components, were properly selected, assembled and implanted. By providing support to the operating room nurses, participants said that they could instead focus on the procedure. This was thought to reduce the duration of surgery, enhance patient safety and outcomes, and potentially mitigate liability for an adverse outcome. This was said to be even more important in community settings with few operating room staff.

Download:

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

Table 2. Reported roles of medical device industry representatives.

https://doi.org/10.1371/journal.pone.0174934.t002

Discussion

Physicians who use implantable devices recognized that relationships with device industry representatives created the potential for COI to influence decision-making about devices, and for patient safety to be compromised if representatives did not share information about device warnings. However, they assumed responsibility for managing such relationships because representatives function as indispensable members of the operating room team with specialized and proprietary information that may be essential to achieving optimal patient outcomes. To manage COI, physicians said they avoided industry representatives, avoided taking money from them, consulted with colleagues for advice about devices, or stopped using a particular device if they had concerns about it. Hospital policies and business practices may encourage COI by facilitating representative presence in the operating room and, through purchasing contracts, mandating the use of particular devices that, in the view of physicians, were not safe or effective. Overall this research revealed a tension between physicians and device representatives that was characterized as “symbiotic”, but required physicians to be vigilant about COI and patient safety. There was a distinct concurrent tension between hospitals, whose policies and business practice were focused on cost-control, and physicians who were required to comply with those policies and practices, and use particular devices despite concerns about their safety and effectiveness.

Strengths of this study included the use of rigorous qualitative methods to capture detailed information about the physician-device representative relationship. Sampling was purposive to capture a variety of perspectives from individuals with differing characteristics, and achieved thematic saturation which signals that recruitment was sufficient to thoroughly identify relevant themes. However, several limitations should be mentioned. The study involved a small number of participants, and they were sampled from Canadian hospitals. Therefore the findings may not be transferrable to other settings, and should be confirmed through investigation of these themes in other settings.

Few comparable studies were identified, and largely focused on drug industry gifts or research funding. American physicians who took part in focus groups [14] and a survey [25] considered interactions with, and gifts and funding from drug representatives as positive and beneficial, disagreed that such relationships affected physician behaviour, and rationalized that they remained impartial. Similarly, physicians in this study said that device representatives offered many beneficial services and they could manage COI. Studies of academic institutions in the United States [11] and Canada [12] found that policies concerning COI were absent in hospitals and, as a result, the onus was on individual physicians to self-regulate. This also appeared to be true in our study. While there are similarities between representatives of drug companies and device companies there are also differences–drugs are not stocked in a just-in-time manner, may be much simpler to use, do not require hands-on instruction, and may not be subject to the same types of contractual restrictions compared with devices. Future research should explore whether these differences lead to different types of physician-industry interaction with differing implications for COI. Given that physicians may not have the capacity to objectively judge their own performance [26], ongoing research is warranted to assess the impact of behavioral interventions that influence cognitive dissonance, or awareness of actual versus desirable behavior, such as audit and feedback [27] on views about COI and relationships with industry representatives.

Participants in our study noted the variable reliability of representatives, which raises questions about the characteristics of such individuals. A study by Mueller et al. involving focus groups with 17 American cardiovascular device industry representatives reported that they had former experience in nursing, pharmacy, engineering, business, veterinary medicine and administration [28]. Notably, the same study revealed that device representatives experienced role conflicts and moral distress regarding their activities in the clinical setting and customer service obligations [28]. Representatives said that physicians and nurses lacked knowledge about devices, and asked them for assistance beyond their expected duties including making clinical decisions about how devices should be programmed and managed, deactivating devices, and informing patients and family about the deactivation process and outcomes. They said that their customer service responsibilities had grown tremendously in recent years, and referred to this as an unpaid service to remedy staffing shortages in the clinical setting. They were reluctant to decline clinician requests, fearing that this would affect their relationship with clinicians and impact their ability to sell their companies’ products. The study by Mueller et al. suggests the presence of a third tension between clinicians and device representatives that further contributes to the potential for COI. Ongoing research involving device representatives should further investigate the challenges they experience when interacting with physicians that may compromise patient safety. Such research could also involve nurses who may be witness to industry representative-physician interactions in the operating room.

These findings are novel because there is little research on the implications of representative presence in the operating room. In response to concern about their role and influence, some hospitals have instituted policies specifying that representatives must assist medical staff only verbally, acquire a distinct identification badge on each visit and, prior to their first visit, provide documentation of competency in the specific device, training in infection control, insurance certificate, proof of annual tuberculosis test and signed confidentiality agreement [29]. Nursing and surgeon professional societies have issued similar statements regarding the credentialing and appropriate role of representatives [30,31]. A survey of senior nurses in charge of 79 gynecology operating theatres in the United Kingdom found that 82% had no guidelines for representative presence in the operating room, and 42% obtained patient consent for visits [32]. In surveys of arthroplasty patients in Canada and the United States, most respondents reported that their physician had not mentioned COI in advance of surgery [33–36]. Those patients trusted physicians to self-regulate, and believed that oversight should be provided by professional associations. Current professional society standards for COI focus on how their own organization manages COI, for example, funding of continuing education and annual meetings, rather than how they can help individual members to mitigate COI in their daily interactions with industry representatives [37–39].

In conclusion, this exploratory study identified tensions between physicians, hospitals and device industry representatives that may contribute to COI. Physicians may be reliant on industry representatives for their knowledge and skill, and hospitals, by employing purchasing agreements and just-in-time operating room delivery of devices, promote physician reliance on device industry representatives. Further research is needed to identify the clinical implications of the characteristics and role of, and relationship with medical device industry representatives from the physician, nurse and representative perspective; how that relationship and its implications differ from physician interaction with drug industry representatives; whether and how hospitals do and should govern interaction with medical device representatives; and explore the views of leadership from professional societies on how they could help their members manage COI, and the effectiveness of behavioral interventions that influence physician views about COI.

Supporting information

Acknowledgments

The authors thank Julie Takata for assistance in organizing meetings for this study.