Study design

We opted for a qualitative design to allow an in-depth exploration of a complex situation without missing out on any subtleties [16]. In fact, none of the questionnaires we had identified to run a quantitative survey addressed the issues we learned about as we were preparing for our project. The qualitative design allowed us to document such issues, including some types of incentives and monitoring strategies used.

We followed the Grounded Theory (GT) methodology based on Corbin and Strauss[16]. We used the systematic qualitative approach of GT for collecting data, analyzing and synthesizing constructs with the intent of inferring a middle range theory. This theory was meant to provide one framework linking between the important determinants governing the interaction, the nature of the relationships between the different players, and the eventual impact of those determinants on prescription practices.

We followed the consolidated criteria for reporting qualitative research (COREQ) in drafting this manuscript [17]. COREQ guidelines include a 32-item checklist that can help researchers to report important aspects of the research team, study methods, context of the study, findings, analysis and interpretations (S1 File).

The research team

The research team consisted of eight members. Six team members (RH, AB, MC, ST, RN, FH) were second year medical students and had no previous experience with qualitative research. Two team members (GHA and EAA) had repeated experience in conducting and publishing qualitative studies.

Participants

Our three target populations consisted of physicians, pharmacists, and pharmaceutical representatives practicing in Lebanon.

The corresponding sampling frames were:

1. Physicians practicing in clinics and/or hospitals in Lebanon;
2. Pharmacy staff, including pharmacists and pharmacy technicians, who handle drug prescriptions in an outpatient setting;
3. Pharmaceutical representatives working for either a local company or an international company based in Lebanon.

We excluded individuals who did not speak either Arabic or English, the two languages used for consenting and interviewing.

Recruitment and sampling

We conducted the interviews during the months of January and February of 2016. We used the known sponsor, purposive, and snowballing sampling strategies to identify interviewees from the three respective groups: physicians, pharmacists, and pharmaceutical representatives. We also aimed for maximum variation [16], by trying to recruit participants that varied by area, age, gender, and specialty.

We recruited physicians using the known sponsor approach. Each of the medical students involved in this study identified and invited, either in person, by email or by phone, an eligible physician already known to her/him. To avoid any undue influence, another team member conducted the interview.

We recruited pharmacists using a purposive sampling approach. A team member approached the staff of large pharmacies located in five major cities in Lebanon and invited them to participate. Those who agreed were interviewed by the same team member.

As for pharmaceutical representatives, one medical student identified a representative already known to her. This pharmaceutical representative identified and introduced us to other representatives. Using this snowballing strategy, we recruited the other representatives to participate in our study.

It is noteworthy to indicate that no dyad of interviewer and interviewee had previous relationships and no one approached refused to participate. We proceeded with recruitment until no new themes emerged (thematic saturation) [16].

Data collection

We conducted the interviews in private locations convenient to the interviewees. We audiotaped the interview when permitted by the participant, otherwise, we resorted to note taking. The interviews were conducted in either English or Arabic, based on the interviewee’s preference.

First, the interviewers administered a short questionnaire asking for demographic characteristics, and then interviewed the participant using a semi-structured interview guide (S1 Appendix). Questions addressed the nature and intensity of interactions and whether these interactions are impacting drug related practices. As a way to encourage participants to be more forthcoming with information, we asked them to describe interactions in general and not necessarily their own experience. We pilot tested the first interview with each professional to ensure clarity and relevance of questions. The interviews lasted approximately 30 minutes.

Data analysis

We did not include the data collected during pilot testing in the final analysis. We transcribed audio-recorded interviews conducted in English; while we first translated, interviews conducted in Arabic into English and then transcribed them. We anonymized all transcripts. Immediately following each interview, we proceeded with transcription, translation, and data analysis.

Two investigators worked independently on coding the transcripts line by line while maintaining a log of emerging codes, their definition, and exemplary quotes (open coding). Then they met to share, compare and contrast their independent data analyses (constant comparative technique). This debriefing helped achieve consensus on emerging codes, expand or merge codes (axial coding). A third investigator resolved disagreements. Following each discussion, we refined the interview guide to allow probing, amplifying or clarifying issues still uncovered. Investigators met to discuss the final findings and combine the themes into a conceptual framework (selective coding) [16].

Increasing rigor

While the two senior team members (GHA and EA) had experience in conducting and publishing qualitative studies, the remaining six members (RH, AB, MC, ST, RN, FH) had no such previous experience. Therefore, we conducted training on interviewing techniques. Also, with the first few interviews, we implemented peer observation and held debriefing sessions to reflect on the process and how it could be improved.

The team members made every effort to maintain a level of reflexivity to avoid the influence of personal views and biases during data collection and data analysis, and to prevent bias during the writing stage of the study.

Ethics

We obtained the approval of the Ethical Review Committee at the American University of Beirut prior to the initiation of the study. We obtained verbal consent from all participants. We anonymized all transcribed interviews and kept sensitive data in a separate document.

Demographics

We recruited 32 participants: 13 physicians, 13 pharmacists, and 6 pharmaceutical company representatives. Table 1 shows their characteristics. Participants were from four different regions (Bekaa, Beirut, Mount Lebanon and South) capturing urban, suburban and rural areas in Lebanon. The majority of participating pharmacists owned a pharmacy. Half of the pharmaceutical company representatives worked in international companies. The median number of years of experience was 15, 20, and 1 for physicians, pharmacists, and pharmaceutical company representatives respectively.

Download:

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

Table 1. Characteristics of participants.

https://doi.org/10.1371/journal.pone.0184662.t001

Emerging themes

The following themes emerged from the data: purpose and driver for the interactions, nature of the interactions, incentives, impact on prescription practices, ethical considerations, and suggestions for managing the interactions. When similar views were voiced we used the term “most participants”; when that was not the case, we used the terms “many” or few” as relevant. When including quotes along with our findings, we referred to the type of participants using the following abbreviations: Physicians (Phys), Pharmacists (Pharm), pharmaceutical representatives (PR).

Purposes and drivers of the interactions

The main purposes of the interactions were promotional and educational as indicated by most professionals: “Every time it’s for a different purpose, sometimes to invite me to a conference, sometimes to give me samples, sometimes to talk to me about the drug.” (Phys₀₄). However, monitoring physicians’ drug prescription practices was another essential purpose as revealed by the three professional groups. As a means of monitoring, some pharmaceutical representatives reported that pharmacists provided them with information on physicians’ prescription of their company’s drugs, and of drugs from competitor companies: “If you are close to the pharmacist he might tell you the name of the doctors who are prescribing or stopped prescribing or prescribing competitor products” (PR₀₁). Others directly asked physicians for an account of the number of prescriptions for the promoted drug. Some pharmacists indicated providing representatives with access to copies of physicians’ prescriptions. This practice was justified as being a sense of obligation by many physicians or pharmacists: “If the company that is paying money wants to know how many prescriptions a physician wrote, you tell them the number and some reps take a picture of the prescription to know if the physician is prescribing the drug” (Pharm₁₂).

Most pharmaceutical representatives said that market potential was the main driver to approach certain physicians and pharmacists and not others. They target physicians who are most likely to prescribe their drugs and those with a high turnover of clients. In addition, the socio-economic status of patient population covered by physicians and pharmacists dictated the type of drug being promoted: brand names with affluent communities, and generics with less affluent communities: “It depends on the location of the pharmacy; if it is a high-end, then people can afford to buy an expensive drug and so we tend to accept brand offers. If the region is low-end, we accept generics, same effectiveness but cheaper” (Pharm₀₄).

Nature of the interactions

All physicians and pharmacists’ participants had previous interactions with pharmaceutical representatives. Most of them indicated that pharmaceutical representatives’ visits are unscheduled. Very few physicians indicated having a set fixed times for visiting: “On a daily basis […] because I am here most of the day I don’t schedule and just see them when I am free” (Phys₀₁).

The main discussion during these interactions focused on either sharing new information about a certain new drug: “They bring you studies, they ask you if you have any questions so that they can get you the answers” (Phys₀₃), or providing evidence-based data on the benefits of their promoted drug compared to a competitor drug: “They show you studies, or head to head comparisons between 2 drugs, sometimes some of them they try to convince you based on the financial aspect, for example they would tell you that this medicine is cheaper” (Phys₀₉). However, some indicated that these interactions were of a social nature: “Once you sit down with her and she asks about your son and tells you about hers” (Phys₁₁).

When asked how they perceive the scientific merit of the information provided during these interactions, many physicians and pharmacists indicated that it is evidence-based. However, some suggested that the information is biased because it emphasizes the benefits of the drug while neglecting its side effects: “If there is a study that does not show the importance of that drug or if the results were different from what they were expecting, they may not discuss it with us. But be sure that the rival drug company will discuss it with us.” (Phys₀₃).

When we asked physicians and pharmacists about their perceptions of the pharmaceutical representatives’ role, some believed that they deserve sympathy: “We should be considerate with a med rep because he is working all day long and I believe it is his utter right to know if the doctor he is dealing with is actually prescribing this company’s drug or not “(Pharm₀₄). Others perceived them as a nuisance and at times disrespectful: “[they] are always following physicians around and hovering in the hallways” (Phys₀₅).

When asked how to best characterize the nature of the relationship between pharmaceutical representatives, physicians and pharmacists, most agreed that since the main purpose of this relation is to benefit patients, then there is no conflict of interest: “If it is a medicine that you are used to prescribing a lot and it is good I don’t consider it wrong if they offer conferences or journal reviews or things of the sort.” (Phys₀₇). However, the statements by many physicians and pharmacists revealed personal and financial conflict of interests: “Each pharmacist, wants to profit the most of his work, or else no matter how professional he may be, if he’s not a good businessman, he’s not going to stay in business.” (Pharm₀₃). And when not harmful to patients, some physicians reported they would modify their practices in order to maximize the benefits: In general if there are similar medication and they are all good, I try to use everything.” (Phys₀₈).

A few physicians admitted that the conflict of interest resulting from this relationship may be putting their patients in harm’s way “He’ll maybe prescribe drugs that are more expensive and this will harm the patient” (Phys₁₁). This view was seconded by one pharmaceutical representative: “Some doctors prescribe the same medication from two companies to keep receiving their deals, for example, two anti-inflammatory medications with different brand names, one pill in the morning and one pill at night” (PR₀₁)

Incentives

Pharmaceutical companies are prepared to offer a wide variety of incentives to pharmacists and physicians in return for more prescriptions. While they seem to be the primary initiators of gift giving, some physicians or pharmacists may be the initiators as well as voiced by one physician: “There is a doctor that has a sign outside his office “if you don’t have a sample don’t bother coming in” (Phys₀₁).

It appears that the value of the incentive is linked to the prescription load. Pharmaceutical companies offer greater incentives for physicians with a larger number of prescriptions: “Pharmaceutical companies classify physicians in three categories, according to which they decide on how much incentives they would provide: A (high turnover of patients), B (highest turnover in their respective region) and C (low turnover of patients).” (PR₀₄). The incentives provided to physicians and pharmacists are both monetary and non-monetary and vary between the two professional groups. Monetary incentives include receiving office equipment, electronics and monthly payments while non-monetary incentives include invitations to be guest speakers at high profile meetings, or opportunities to appear on television. Therefore, they “affect the popular reputation, patients would be impressed and say: Oh! They invited him to France to talk about this medicine” (Phys₀₅). Some participants interpreted this as pharmaceutical companies giving prominence to selected physicians within their communities. This was supported by reports of pharmaceutical companies giving prominent roles to “their physicians” during professional meetings that they sponsor.

A few physicians and pharmaceutical representatives reported cases of offering of sexual favors (Table 2), provided either by the representatives themselves or through nightclub invitations or room service during travel. “There are reps in a local company who sleep with the doctors for the sake of promotion” (PR₀₃).

Download:

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

Table 2. List of incentives according to the three groups of participants.

https://doi.org/10.1371/journal.pone.0184662.t002

Pharmaceutical company representatives were also incentivized. They received bonuses when they increase the sales and revenues of their company. Conversely, poor performance provided ground for warnings and potentially dismissal. Pharmaceutical representatives also benefit from free consultations from physicians and discounted medication from pharmacists.

Impact on prescription practices

The impact of the relationship between physicians, pharmacists and pharmaceutical representatives and its resulting conflicting interest varied by individuals. A few physicians insisted that their own prescription practices were not affected: “I don’t prescribe it if I am not convinced, some of the drug reps are actually my friends and I still don’t prescribe for them if I am not convinced.” (Phys₀₉). Some physicians said that when affronted by two similar drugs from competitor companies, they were inclined to prescribe the drug from the company that provided them the incentive. They indicated that it was a sense of obligation towards that company that drive their decision. “It is so difficult to say they supported me and I am not going to prescribe for them” (Phys03). Many physicians, however, observed that over-prescription of unnecessary and expensive drugs are common occurrences among physicians in their communities: “Take for example anti-osteoporosis medicines the doctor might tell people your bones are very good but this drug is preventive you should take it so they prescribe randomly” (Phys11).

Some physicians indicated that pharmacists tend to “sell the more expensive drugs because they benefit more” (Phys02). Several pharmacists reported that some pharmacies might be dispensing medication even without prescription: “Sometimes, someone would come to buy a medicine without prescription; we tell them no we can’t but they go to another pharmacy and they give it” (Pharm08). In addition, many pharmacists alluded to mutual referrals between physicians and pharmacists to maximize their benefits: “The physician sends [patient]) to the pharmacy and the pharmacy gives the physician a certain commission and recommends the physician.” (Pharm07).

Ethical considerations

There was an overall culture of acceptance of the gift giving. The boundaries between what is considered ‘ethical’ and ‘unethical’ in the gift giving were not consistent amongst participants. Some considered acceptance of gift of any kind to be unethical: “There shouldn’t be any promotional material such as gadgets, travels, covering trips, sponsoring of research” (Phys₀₃). Others believed that educational gifts or free samples for example, are acceptable provided these free samples are dispensed to patients and not sold: ‘There are some doctors who really do it (use the incentives/samples) in a good way to help unfortunate people” (Phys₁₁). Some participants noted that some physicians sell the free samples they collect from pharmaceutical representatives: “They [physicians] tell you that they are giving you a reduction on this medicine although they received it for free.” (Phys₁₀)

However, there was a consensus that anything that may harm the patient because of this interaction is unethical and should be a “red line”: “At the end of the day the aim is financial gain. The red line is whenever the health of the patient is put into jeopardy” (Pharm₀₅).

Suggestions for managing the interactions

Physicians and pharmacists indicated the need for laws to regulate interactions between physicians and pharmaceutical representatives. Some proposed adding caps to incentives “I think there should be rules and regulations and some things should be allowed and others not allowed especially promotional material offered to the doctors it should only be limited to scientific material” (Phys₀₃). A few physicians also proposed the prohibition of the monitoring of prescription practices by pharmaceutical companies: “I think it is unethical for them to know who is prescribing what but to know in general what their sales are, they should be allowed” (Phys₀₇). Others suggested implementing a reporting system that keeps track of physician’s prescription patterns, and their interactions with pharmaceutical companies: “I think it is important to monitor prescription patterns, especially if there is an accurate reporting system like gains or gifts that physicians get from pharmaceutical companies.” (Phys₀₅).

Some participants voiced skepticism about the efficacy of laws. A few indicated that in the absence of a mechanism for monitoring the interaction and prescription practices, these laws are useless: “Since it is not computerized, nobody will know, they (authorities) cannot catch hundreds and thousands of papers of what is sold and what is done every day” (Phys₀₂).

The problematic behaviors seem to be more common in interactions with representatives of local pharmaceutical companies, compared to representatives of international companies. Pharmaceutical representatives highlighted the need for local companies to adopt policies similar to those of international companies: “We could follow Big international pharmaceutical companies which have two ethical centers: a medical department that imposes selection criteria for congresses, and an audit office that checks and audits all expenses and activities. Unlike local companies no cash advances are made by multinational pharma” (PR₀₃).

Implications

Physicians, as individuals as well as a body, need to lead efforts to minimize any real or perceived unprofessional or unethical interactions with pharmaceutical representatives. They also need to minimize any influence of interactions deemed as acceptable on their prescription behavior. The government can also have a major role through forming an independent agency, whose role include supervision of all providers; monitoring of physician prescription similarly to what is done in Canada (Prescription Monitoring Program www.nspmp.ca); and establishing an independent advisory committee to reinforce and monitor the implementation of the code of ethics. Similarly, there is a need to identify alternative unbiased sources for continuous medical education such as practice guidelines, and peer-reviewed research. The government also needs to enhance the implementation of recently launched initiatives such as the generic drug substitution policy[30].

Future research needs to quantify the extent to which the different types of interactions and incentives are prevalent among physicians in Lebanon. It would also be interesting to explore whether the findings of this study are reproducible in other settings, particularly other LMICs. Another relevant area of research is the public as well as patients’ awareness and perceptions of interactions, and how they affect their confidence in the healthcare system.