Introduction

Pharmaceutical clinical trials are increasingly being conducted in the private sector as part of a growing global contract research system [1]–[3]. In the United States, a major shift in the locus of pharmaceutical research has taken place in the last two decades. Whereas the majority of clinical trials were completed in academic medical centers in the 1980s, estimates reported in 2005 showed that more than 70% of US trials were conducted by nonacademic physicians [4]. Community physicians have largely replaced their academic counterparts as principal investigators (PIs) on pharmaceutical trials [1],[5].

US medicine has changed significantly in the past 30 years. As a direct response to the perception that US health care expenditures had reached a point of “crisis,” health care reforms in the 1970s and 1980s aimed to limit excesses in health care spending [17]–[19]. While the reality of this crisis is open to debate, it is a common perception that physicians' payments have decreased significantly due to managed care [20]–[22], whereas practice operational costs have increased [23]. Pham et al. report that US physicians feel “beleaguered” and forced into becoming more “business-oriented” [22]. Physicians have joined larger, multispecialty practices in order to defray costs and liability [24]. Additionally, many have referred patients to their privately owned ancillary service centers [25]. This focus on the business of medicine has raised concerns that medical professionalism in the US, particularly in terms of physicians' values and ethics, has been affected [26].

The pharmaceutical industry has provided an opportunity for US physicians in private practice to serve as PIs on clinical trials [11],[22],[27],[28]. The number of US private-sector physicians conducting pharmaceutical trials increased from 4,000 in 1990 to 20,250 in 2010 [29], with research contracts now worth more than US$11 billion annually [30]. This enormous growth in the number of private-sector physicians involved in contract research is due in part to their replacing academic physicians on pharmaceutical clinical trials, but it is also due to simultaneous exponential growth in the number of clinical trials initiated by industry during this same time period [31].

Pharmaceutical contract research is profitable to US physicians because of the fee-for-service payment system that structures companies' remuneration of PIs [32]. Specifically, beginning with informed consent, PIs get paid for each step in the screening process, and for each study procedure, history and physical, and blood draw. Reimbursement rates are significantly higher than what Medicare pays for the same services [1]. An American College of Physicians' report found that contract research provides individual physicians income of up to US$300,000 annually [33]. Although the potential revenue from contract research is substantial, this income is not without ethical risks. Nonetheless, there is little oversight of financial conflicts of interest in the private sector compared to academic settings, raising concern that if per capita payments are inappropriately high, private-practice physicians might make recruitment decisions that are not in their patients' best interest [34].

Bodenheimer has argued that certain conditions enabled the development of a private-sector clinical trials industry [5]. He explains that, originally, academic medical centers provided (1) trial design expertise, (2) a patient population from which to draw, and (3) prestige and publishing experience. However, the combination of private-sector companies, private-sector physicians, and high-level scientists and physicians employed directly by pharmaceutical companies and contract research organizations has proved a more advantageous drug development model [5]. This model is based on the premise that speeding up clinical development is an important way for companies to increase their profits on patented products and that private companies are more efficient than are academic medical centers [35]. Indeed, it has been reported that private companies complete trials up to three months faster than academic medical centers, resulting in about US$90 million in additional earnings on each new drug [36].

In spite of the dramatic shift in who serves as PIs on industry clinical trials in the US, there has been little research on the implications of this trend [1],[3],[5]. The purpose of our study was to explore the outsourcing of clinical trials to US private-sector research clinics, paying particular attention to the different groups involved in the research enterprise (e.g., pharmaceutical company representatives, physicians, study coordinators, and research participants) and to the everyday practices associated with conducting clinical trials (e.g., recruitment of participants, informed consent processes, and study compliance). The data were collected as part of an institutional ethnography that was conducted between October 2003 and September 2004 [1]. The study was designed in an open-ended way because little empirical research had been conducted on the clinical trials industry and there were few guides indicating specific areas in need of investigation. In June 2011, the authors completed a second-order analysis of the data to use the findings of the empirical project to generate insights about the growing scholarly literature that has cited evidence of an increasing array of unscientific research practices perpetrated by the pharmaceutical industry [3],[4],[6]–[16]. This paper reports on our findings that private-sector PIs view monetary incentives, clinical research, and ethics in unexpected ways.

Methods

Our study of the clinical trials industry used a qualitative research design, consisting of semi-structured interviews and observation that took place between October 2003 and September 2004 at 25 US private-sector research organizations, such as private practices, dedicated research sites, and large (nonacademic) hospitals. Two cities in the southwest US were selected for the study because of the preponderance of private-sector clinics in this region, according to information in an industry clinical trial clearinghouse database (CenterWatch; http://www.centerwatch.com). The database was also used to identify and contact individual sites. All sites in the two cities were contacted and invited to participate in the project. Although no incentives were offered to promote participation, 75% of sites in one city and 50% of sites in the other city agreed to take part in observation and interviews. According to information listed in the database, there was no discernable difference between sites that agreed or refused to participate, in terms of the types of studies conducted, size of the clinical trials operation, or experience of the site. Clinic specialties included family and internal medicine, gastroenterology, neurology, pediatrics, rheumatology, and urology. As we discuss in more detail below, all sites conducted clinical trials outside of their specialty area. The majority of sites conducted trials to test the efficacy of new products for conditions that already had safe and effective treatments on the market (e.g., allergies, asthma, hypercholesterolemia, and insomnia), as has been identified elsewhere [6]. Only one site consistently tested products for life-threatening conditions (e.g., AIDS or cancer). One site specialized in early phase studies with healthy volunteers.

Semi-structured interviews (Text S1) with 63 informants were clustered to get the perspective of multiple employees at each site, including 12 PIs, 18 coordinators, three recruiters, and seven non-physician site administrators, as well as 14 research participants and nine representatives of pharmaceutical companies. Interviewees were selected based on their availability during observational visits. For instance, during some research visits, the only person present in the clinic was a research coordinator; at other clinics, research visits generated interviews with multiple informants. The size of a clinic and its level of activity limited the number of interviews—and the variety of people available for interviews—at each site. No one who was asked to participate in an interview refused. Interviews were conducted until saturation was reached [37], meaning that additional interviews ceased to provide data that generated new insights or information about contract research. Interviews lasted an average of 40 min. This paper draws on our entire dataset, including themes and findings from observations at all 25 research clinics and all 63 interviews. Because of our focus on physicians in this paper, we quote most extensively from the interviews with 11 physician-investigators who served as PIs in contract research. Physicians in our sample were representative of the national demographics of industry-funded investigators [38]: ten men and one woman; ten white and one Hispanic. Most physicians were in their 40 s, with an age range from mid-30s to late-70s, with a broad range of experience conducting trials (1–20 y).

Observation in clinics was focused primarily on interactions between PIs and participants, as well as between coordinators and participants. Observation ranged from a single visit on one day to multiple visits spanning several months. Field notes were used to capture observational data [39]. In interviews, informants were asked questions about their job responsibilities as well as about their experiences working in clinical research, how research had changed over time, and what types of changes they would like to see in the future. Specific interview questions for PIs also solicited their perceptions of research ethics and of their responsibilities to participants, pharmaceutical companies, and science. All data were collected by one of the authors (J. A. F.), who is trained in qualitative sociological methods. This author also transcribed all the interviews in full and contacted interviewees to give them the opportunity to edit their transcripts. The study was reviewed and approved by the Institutional Review Board of Rensselaer Polytechnic Institute. Written informed consent was obtained from all participants. All research clinics and informants were given confidentiality as part of their participation in the study.

In keeping with a methodology of grounded theory [40], there were no predefined theoretical or conceptual frameworks or specific hypotheses identified for analysis prior to data collection. In addition, multiple conceptual frameworks were developed through the process of interpreting the findings from the project. As with most grounded theory projects, the data analysis relied on a multi-staged process of coding field notes and interviews in their entirety for core and emerging categories. All the coding was done by one of the authors (J. A. F.). Coding was multi-staged in order to revisit the data multiple times for depth of analysis and for the creation of cross-references among the data and the categories coded. Coding during the data collection period was done by creating detailed memos at the conclusion of site visits and by repeated, fine-grained reading of transcripts and field notes for additional themes that emerged as important. This process helped to guide subsequent observation and interview techniques by highlighting areas of inquiry that needed further exploration. In the final stage of coding after the data collection period, more subtle codes were added that aimed to create subcategories within emergent themes. The current analysis was validated in a 2-fold way by both authors: first, by reflecting on themes in light of the published literature on the topic and, second, by attending industry conferences and reading industry publications designed to provide advice and commentary on the clinical trials industry.

Results

Our research revealed trends in why US physicians become PIs on industry clinical trials and how they develop their professional identities as contract researchers. By “professional identities” we are referring to how physicians incorporate into their existing clinical roles their experience of the rights and responsibilities that accompany contract research. We discuss three aspects of their identities shaped by (1) their motivations to conduct clinical trials, (2) an orientation that privileges business over science, and (3) an industry-based sense of research ethics.

Discussion

The results from our study corroborate earlier research indicating that US private-sector physicians are motivated to participate as PIs in pharmaceutical clinical trials for financial reasons [5],[11],[22],[32]. We also discovered that while financial incentives might be the most salient, at least initially, physicians also report other advantages associated with contract research: intellectual reward, increased professional status, and benefit to their patients and communities.

Beyond private-sector physicians' motivations to participate in pharmaceutical trials, our findings reveal that the financial aspects of contract research are aligned with their professional identities. Instead of seeing themselves as scientists or researchers, they see themselves as trial practitioners and as businesspeople. At conferences the physicians in our sample attend sessions geared toward the business aspect of trials, suggesting that they might not consider additional knowledge about pharmaceutical science or medical practice to be of foremost importance for their professional development. This may help to explain the rise of industry conferences that are entirely dedicated to specific business aspects of trials, including patient recruitment [41]. Additionally, it was an unexpected finding that the financial incentive to participate in contract research propelled the private-sector physicians in our sample to serve as investigators on trials for investigational drugs in a broad range of therapeutic areas. That they often work on the margins of their specialized training provides one potential explanation for why the private-sector physicians did not see themselves as scientists and claimed little insight into the mechanisms by which particular drugs work.

Unlike traditional investigator-initiated research, such as studies funded by US National Institutes of Health, the prepackaged nature of contract research means that physicians do not need to have expertise in research design and methods because they have limited input into trial protocols [1]. As a result, private-sector investigators, who by self-report in our study are not interested in or trained to understand the science, potentially facilitate the pharmaceutical industry's ability to exert more control over information about their products, including any negative data generated from clinical trials [9],[12],[13],[42]–[47]. The literature has suggested that pharmaceutical companies are interested in controlling their trial datasets so that individual PIs do not have access to data collected at all participating investigative sites and any resulting publications can be prepared by ghostwriters [6],[10],[16]. Our findings indicate that this arrangement tends to be acceptable to private-sector physicians who rarely—if ever—participate in analyzing trial data or authoring publications.

Private-sector physicians incorporate ethics into their identities in unique ways as well. It was revealing that when asked about ethics, the physicians in our sample always answered the question in terms of responsible conduct of research. Instead of being concerned with the dominant domain of research ethics—the PI's responsibilities to research participants—they understood their primary “ethical” responsibility as providing accurate data to the companies that hired them. Within the context of their work, this makes sense. Contract researchers do not design the studies, nor are they given much opportunity to provide input on the protocols. Additionally, private-sector physicians often do not have institutional review boards associated with their clinics, and the review of clinical trial protocols is handled by centralized commercial companies [8]. These two factors imply that to a large extent PIs must trust that the sponsoring company has designed a scientifically valid and ethical trial and that the centralized institutional review board has established appropriate guidelines for the protection of enrolled research participants [27]. What this suggests, however, is that physicians do not perceive the potential ethical conflicts that emerge as a result of their dual roles as health care provider and clinical researcher when they are conducting clinical trials that enroll their own patients.

There are, of course, limitations to our study. Although we included a diverse sample of research organizations in our study and reached thematic saturation with our interviews, the number of physicians we interviewed was small. We limited our study to private-sector research in one geographical region of the US. While the published literature indicates that the trends we found in our research are likely generalizable to other regions of the US, the context differs dramatically in countries with national health care systems, so it is likely that private-sector PIs characterize their roles differently in other settings. We must also note that our data were collected eight years ago, so it is unclear how physicians' identities might have changed in response to subsequent changes in the funding of contract research, such as the rise in seeding trials, or the broader health care environment. In addition, our study was not designed to investigate physicians' professional identities as contract researchers. Instead, what we report here are unprompted themes identified during data analysis. To better understand private-sector PIs, future studies are needed to compare their identities with those of academic physicians and to investigate how differences in identity might influence the execution and results of contract research. Future studies could also compare cross-national differences in the perceptions and practices of pharmaceutical investigators. Indeed, contract research is a global enterprise [2], and our study highlights the need to understand better the changes that have occurred in the conduct of pharmaceutical clinical trials worldwide.

Supporting Information