Primary physicians’ attitudes toward their patients receiving a proposal to participate in a vaccine trial

ABSTRACT

A trustworthy relationship between primary physicians (PPs) and their patients is crucial for vaccine acceptance. Little is known about attitudes of PPs toward participation of their patients in a preventive vaccine trial (PVT) proposed by investigation sites.

A cross-sectional study was conducted in Auvergne-Rhône-Alpes region (France) including an anonymous questionnaire for general practitioners (GPs) and other specialists as well as face-to-face interviews. A scenario of a patient, with chronic medical conditions, invited to participate in a PVT and reporting this situation to his/her PP was drawn up. PPs’ attitudes were assessed in quantitative approach by a 5-point Likert scale and in qualitative approach by semi-directed individual interviews.

Among the 521 respondents to the questionnaire, 429 (82.3%) were GPs and 92 (17.7%) were other specialists. Only 7.5% (39/521) of respondents regularly practice clinical research. Confronted with the scenario, 312 respondents (59.8%) declared they would give their opinion spontaneously. Before giving their opinion, PPs would like more information about the trial (91.4%, n = 476). Whatever their attitude, 488 (93.7%) would be influenced by available safety data. Face-to-face interviews confirmed that PPs lack of knowledge about clinical research, and would like to obtain information from investigators, particularly about safety.

PPs seem to be concerned by the decision of their patients to participate or not in a PVT but would like more information about the trial and clinical research before giving their opinion. Getting PPs to be more involved in the enrollment of patients in PVT may improve recruitment.

KEYWORDS:

• primary physicians
• attitudes
• recruitment
• patient
• preventive vaccine
• vaccine trials
• clinical trials

Introduction

The development of new vaccines continues to be needed^1,2 in order to fight emerging infectious diseases and healthcare-associated infections.³ Prior to licensure of new vaccines, it is necessary to carry out clinical trials on healthy volunteers and the target population.⁴ The success of clinical trials depends on the recruitment of a sufficient number of volunteers. Real difficulties exist in recruiting, including and retaining volunteers or patients in clinical trials.^5,6 In a US study, people were asked if they were interested in medical research participation.⁷ Though as diagnosed volunteers – if diagnosed with the disease being studied – 70% of the responders agreed to participate in a medication clinical trial, only 59% agreed to participate if the evaluated product was a vaccine.⁷ Consequently, recruitment for trials evaluating candidate vaccines probably meets more difficulties. It is possible that vaccine hesitancy may affect the recruitment of patients or volunteers in preventive vaccine trials (PVT).⁸

Factors influencing participation in clinical trials have been studied and have shown the importance of the primary physician’s attitude.^9,10 The primary physician (PP), i.e., the physician following the patient on a regular basis, has a trustworthy relationship with the patient and is the most trusted source of information.⁹ Their recommendations have been found to be the most important factor in patients’ decisions about participation in clinical trials, specifically in oncology.^11–13 By contrast, the attitudes of PPs to the patient receiving a proposal to participate in a clinical trial have rarely been studied¹⁴ and no study has focused on vaccine trials. In order to identify such attitudes and determine the factors that could influence them, we conducted a mixed-methods study of GPs and other specialists to evaluate PP attitudes toward patient participation in vaccine clinical trials.

Results

Characteristics of the participants in the quantitative study and the participants in the qualitative study

A total of 521 PPs (429 GPs and 92 other specialists) answered the quantitative questionnaire, corresponding to a response rate of 10% for GPs and 23% for other specialists. The main characteristics of respondents are summarized in Table 1. Of the 521 PPs who answered to the question, only 39 (7.5%) practice clinical research regularly (daily, weekly). This regular activity was more frequent among other specialists (6/33 among GPs vs 33/69 among other specialists, p = .004). For qualitative study, all solicited physicians accepted to participate. Characteristics of the 15 physicians (9 GPs and 7 other specialists) who agreed to participate in the qualitative study are summarized in Table 2. The interviews lasted between 9 and 41 minutes.

Table 1. Demographical characteristics of the panel population (n = 521).^a

	All respondents
Characteristics	n	%
Specialty
General practitioners (GP)	429	(82.3)
Others specialties^a	92	(17.7)
Age
<30 y	5	(0.9)
30 to 39 y	156	(29.9)
40–49 y	131	(25.1)
50–59 y	130	(24.9)
>60 y	99	(19)
Gender
Sex ratio (male/female)		0.98
Male	258	(49.5)
Place of practice
Private practice	424	(81.4)
Hospital center	97	(18.6)
Clinical research activity
Yes	102	(19.6)
No	419	(80.4)

Infectiologists = 6, cardiologists = 3, pediatrics = 5, internal medicine = 4, surgery = 11, dermatologist = 1, endocrinologists = 2, gastroenterologists = 2, pneumologist = 1, anaesthetists = 6, geriatrics = 1, gynaecologists = 2, emergency physicians = 7, urologists = 3, rheumatologists = 6, oncologists = 3, ophthalmologists = 2, neurologists = 6, nephrologists = 1, vascular physicians = 5, oto-rhino-laryngologists = 1, reproductive physicians = 3, rehabilitation physician = 3, psychiatrist = 1, addictologist = 1, nuclear physicians = 1, not reported = 5.

Table 2. Qualitative case table (n = 15)

Doctor	Gender	Age (y)	Specialty	Exercise place	Clinical research activity	Length of interview (min)
A	M	65	GP	Private practice	No	25
B	M	49	GP	Private practice	No	9
C	F	32	GP	Private practice	No	17
D	F	49	GP	Private practice	No	12
E	F	54	GP	Private practice	No	41
F	M	60	GP	Private practice	No	27
G	M	57	GP	Private practice	No	19
H	F	59	GP	Private practice	Yes	26
I	F	38	GP	Private practice	No	27
J	M	49	Nephrologist	University hospital	Yes	15
K	M	46	Rheumatologist	University hospital	Yes	17
L	F	43	Cardiologist	Hospital	No	17
M	M	31	Pneumologist	Private practice	No	16
N	F	51	Gastroenterologist	Both	Yes	12
O	M	56	Rheumatologist	Both	Yes	31

The attitudes of PPs to their patients who were offered participation in a vaccine trial

Table 3 summarizes quantitative and qualitative data on enrolled physicians. Confronted with this situation, 312 physicians (59.8%) declared they gave an unsolicited opinion. This declaration was independently associated with male gender and age ≥50 y old (see Table 4). When the opinion was solicited by the patient, 186 physicians (35.7%) did not wish to give an opinion. This attitude tends to be more frequent in physicians who do not have any clinical research activity (see Table 4) and qualitative data confirmed this reticence to give an opinion among GPs (see Table 3). Among all respondents, 476 (91.4%) declared they wanted more information before giving their opinion; this attitude was independently associated with female gender (see Table 4). Among 424 responding GPs, 397 (93.6%) declared that they wanted more information before giving their opinion, this proportion was significantly lower in other responding specialists ((79 out of 92–85.7%), p = .012 in univariate analysis) (see Table 4). The qualitative data reflected the physicians’ unease concerning this lack of information about the vaccine trial and lack of knowledge about clinical research (see Table 3). Of the 521 physicians who answered to the question, 367 (72%) were in favor of having a prior contact with the investigator to receive information as well as the study protocol. In both qualitative and quantitative analyses, the great majority of physicians declared they respected the patient’s decision and left them to make up their mind independently (see Table 3).

Table 3. Attitudes of attending physicians when one of their patients is approached by clinical center of research to participate in a preventive vaccine trial

Quantitative responses (n= 521) Likert scale representation	Qualitative verbatim sentences related
Do you spontaneously give your opinion?	-
If the patient asked your opinion, you do not wish to give an opinion	“If patients ask me for my opinion, I can eventually determine myself.” (Doctor F) “You can’t give advice if you don’t know what you’re talking about at all” (Doctor I)
If the patient asked your opinion, you want more information before you go ahead	“In order to give an opinion in a field, it is necessary to be informed and you have to be initiated”(Doctor A) “To give an advice, maybe there’s a need for more information for GPs” (Doctor I)
If the patient asked your opinion, you give him your opinion but leave him free of his choice	“I push them to accept if they are really excited but on the other hand I let them make their decision alone”(Doctor E)
If the patient asked your opinion, you discourage him from participating	“Often I encourage them to follow these protocols, for me it is an interesting method of analysis that must be maintained” (Doctor G) « If they have been asked to be included it is because they fit the profile so there is no reason why I can advise them against it. » (Doctor N)
You would have preferred that your colleague contact you in advance to present the study and send you the protocol in question	“I should be able to contact the doctor who sets up the trial to find out what he is expecting, because I know absolutely nothing at all’ (Doctor I) “Indeed, there are always difficulties often because the inform consent form (ICF) for patients are often extremely poorly written with terms not always understandable to the non-specialist “ (Doctor J) “Very often they do not ask the questions to the investigators, they ask them to us (attending physicians).”(Doctor D)
What factors could influence your opinion?
Available safety data	“ What might come into play in my decision to recommend or not to recommend participation is the progress of the study. To know if there had already interesting data and lead to an improvement”. (Doctor B) “ I have no fear about the old vaccines that we know about or those that are being put on the market. But participating in tests and it seems really complicated to me. Vaccinations you will often have a fallout at 20, 30 or 40 years old and I think it is more dangerous so you have to be more careful.” (Doctor E)
Patient’s clinical condition	“ My patient’s health condition is a factor I will consider” (Doctor B) « My opinion will depend on my patient’s health condition » (Doctor I)
Patient’s expected benefit	“One element is the individual benefit that it could have a posteriori” (Doctor B). “The potential gain, the potential benefit for the patient will be an element that will guide my recommendation” (Doctor F)
Infection targeted	“First element will be the usefulness of the vaccine (depending on the pathogenicity of the virus or bacteria involved). If we are dealing with a pathology that does not kill, I would say to myself that it would be less important to offer it to my patients” (Doctor C)
Patient initial decision	“ I will check that the patient is really aware of what he is making as a decision and why he has made it. The rest, no, in my opinion it wouldn’t concern me, it’s the patient’s decision! » (Doctor E)
Difficulties for the patient	“If it doesn’t require too many invasive tests for them, I think it advances science”(Doctor D) “They had the right to an electrocardiogram, a chest x-ray and a biological assessment. That’s why I told them it was good to participate, because it came in addition to what we could do as a GP.”(Doctor D)
Placebo-controlled study	“The terms used in the ICF are anxious and you have to try to make him understand that there will be a random draw and that maybe he will end up in the placebo arm. So in the majority of cases, his participation in clinical research does not directly benefit him.” (Doctor J)
If the trial proposed was a drug trial, not a vaccine trial, would your opinion have been different ?	It seems to me that a vaccine is a drug!”(Doctor K) “Oh no, it’s not invasive for me. A sting does not hurt that much.” (Doctor L) “So I do not have the fear about old vaccines we know.”(Doctor E) “ I think the drug trial seems a little less dangerous to me than the vaccine trial because in a vaccine trial we’re going to have a very long-term impact”(Doctor E)

Table 4. Univariate and multivariate analysis of physicians’ attitudes when one of their patients is approached by clinical center of research to participate in a preventive vaccine trial – Odds-ratio [IC]

			If patient asked for your opinion,
	Spontaneous opinion		Do not wish give your opinion		Want more information		Patient free of choice
	Univariate	Multivariate	Univariate	Multivariate	Univariate	Multivariate	Univariate	Multivariate
Male gender	1.97 [1.29, 2.99]***	1.58 [1.02, 2.44]**	0.68 [0.47, 0.99]**	0.76 [0.51,1.11]	0.44 [0.22, 0.88]**	0.47 [0.23, 0.95]**	0.91 [0.42,1.99]	-
Age ≥ 50	3.26 [2.07, 5.11]***	2.90 [1.82,4.60]**	0.69 [0.47, 1.00]*	0.75 [0.51,1.10]	0.76 [0.40, 1.44]	-	1.65 [0.73, 3.76]	-
Hospital center practice	0.70 [0.43, 1.17]*	0.80 [0.47,1.34]	0.87 [0.53, 1.42]	-	0.45 [0.22, 0.94]**	2.2 [0.21,23.16]	0.74 [0.29, 1.90]	-
GP	1.27 [0.77, 2.10]		1.19 [0.74, 1.94]	-	2.42 [1.20, 4.90]**	3.34 [0.302, 36.98]	1.31 [0.51, 3.34]	-
Have a clinical research activity	1.16 [0.69, 1.93]		0.60 [0.37, 0.97]**	0.62 [0.38,1.01]	0.37 [0.19, 0.73]**	0.53 [0.21, 1.35]	1.06 [0.39, 2.86]	-
	If patient asked for your opinion,
	Discourage patient		Prior contact
	Univariate	Multivariate	Univariate	Multivariate
Male gender	1.99 [0.67, 5.92]	-	0.79 [0.52, 1.22]	-
Age ≥ 50	0.86 [0.30, 2.46]	-	1.67 [1.08, 2.60]**	1.67 [1.08, 2.60]**
Hospital center practice	0.65 [0.14, 2.96]	-	0.75 [0.44, 1.29]	-
GP	3.43 [0.44, 26.45]	-	1.37 [0.81, 2.32]	-
Have a clinical research activity	0.96 [0.27, 3.49]	-	0.72 [0.43, 1.20]	-

* p< 0.2.

** p< 0.05.

*** p≤ 0.001.

Factors influencing the attitudes of PPs

These factors based on the quantitative and qualitative results are shown in Table 3. The first factor that influenced the attitude of physicians was the availability of previous safety data, declared by 93.7% of respondents (n = 488). Qualitative data also brought insight into the fact that physicians considered “old” vaccines safer than “new” ones and that vaccine under study could have long-term side effects (see Table 3). In multivariate analysis, female physicians and physicians over 50 are more concerned with safety data (see Table 5). The availability of safety data tends to be more important for physicians who did not have clinical research activity (Table 5).

Table 5. Univariate and multivariate analysis of the opinion’s influencing factors when one of their patients is approached by clinical center of research to participate in a preventive vaccine trial – odds-ratio [IC]

	Patient initial decision		Patient clinical condition		Difficulties for patient		Expected benefit
	Univariate	Multivariate	Univariate	Multivariate	Univariate	Multivariate	Univariate	Multivariate
Male gender	0.76 [0.49, 1.20]	-	0.88 [0.49, 1.60]	-	0.83 [0.54, 1.25]	-	1.27 [0.70, 2.31]	-
Age ≥ 50	0.75 [0.48, 1.18]	-	0.91 [0.50, 1.65]	-	0.92 [0.60, 1.40]	-	0.76 [0.42, 1.37]	-
Hospital center practice	0.50 [0.29, 0.86]**	0.99[0.00, ∞]	0.76 [0.36, 1.60]	-	0.52 [0.31, 0.86]**	4.08 [0.38, 43.74]	0.60 [0.30, 1.20]*	0.42 [0.01, 13.32]
GP	1.77 [1.03, 3.04]**	0.99[0.00, ∞]	1.26 [0.60, 2.65]	-	2.11 [1.28, 3.49]**	7.53 [0.68, 83.43]	1.62 [0.81, 3.25]*	0.60 [0.02, 22.04]
Have a clinical research activity	0.54 [0.32, 0.91]**	0.69 [0.34, 1.38]	0.55 [0.28, 1.07]*	0.55 [0.28, 1.07]*	0.57 [0.35, 0.94]**	0.88 [0.45, 1.71]	0.81 [0.40, 1.65]	-
	Vaccine targeted		Available safety data		Placebo-controlled study
Male gender	0.92 [0.48, 1.78]	-	0.26 [0.08, 0.79]**	0.22 [0.07, 0.70]**	1.00 [0.67, 1.52]	-
Age ≥ 50	0.77 [0.40, 1.49]	-	2.09 [0.73, 5.98]*	3.03 [1.02, 9.00]**	0.96 [0.63, 1.44]	-
Hospital center practice	0.39[0.19, 0.79]**	0.43[0.09, 20.51]	0.39 [0.14, 1.07]*	0.51 [0.01, 38.7]	0.61 [0.36, 1.00]*	1.07 [0.14, 8.18]
GP	2.54 [1.24, 5.18]**	1.06 [0.02, 50.85]	2.47 [0.90, 6.77]*	0.6 [0.01, 44.6]	0.67 [1.01, 2.76]**	1.48 [0.19, 11.50]
Have a clinical research activity	0.51 [0.25, 1.06]*	0.93 [0.34, 2.51]	0.29 [0.10, 0.73]**	0.33 [0.10, 1.12]*	0.61 [0.38, 0.99]**	0.76 [0.39, 1.46]

*p< 0.1.

**p< 0.05.

The patient’s clinical condition was a factor influencing physicians for 459 of respondents (88%). Constraint for the patient (number of visits, vaccine schedule administration, study procedures …) was a factor that influenced 377 of them (72.4%) in their attitude. Qualitative data suggested that physicians made sure that the study design was not associated with too many invasive procedures for the patient. However, physicians considered these procedures as a good point associated with a more complete follow-up for their patients (see Table 3). Benefit expected for the patient was the third factor influencing their attitude for 87.7% of respondents (n = 457). All these factors were spontaneously declared in qualitative interviews (see Table 5). The infection targeted by the vaccine was also a frequent factor influencing their responses in 85.2% of cases (n = 444). Interviews also showed that for physicians, some infections seemed to be more serious than others (see Table 5). For 77% of respondents (n = 401), the patient’s initial decision influenced their attitude. It was all confirmed by qualitative data, which also revealed that physicians checked if their patient was well informed and able to understand what was being offered (see Table 3).

Among respondents, 21.2% (n = 110) did not consider the vaccine to be a drug like others since they stated their opinion would have been different for a medicine trial (Table 3). The fact that the study was a placebo-controlled study influenced 55.6% of physicians (n = 290). In qualitative data, physicians declared that it was difficult to explain what a placebo-controlled study is, as written in informed consent form (ICF) (Table 3). As reported in Tables 3 and 5, the importance of some factors varied significantly between GPs and other specialists.

Discussion

This study described for the first time attitudes of PPs toward their patients receiving a proposal to participate in a vaccine trial. Our mixed methodology allows us to explore their attitudes thoroughly. The role of the physician–patient relationship in clinical research in general has been previously described in literature^9–11 and has revealed how the PPs’ attitudes and opinions influence the participation of patients in therapeutic drug trials.^10,15 Concerning vaccine trials, data are lacking about the attitudes of PPs as well as their role in the patient’s decision process. Among older adults interviewed about their willingness to participate in a vaccine clinical trial, 55% considered that their PP should be comfortable with their participation in the trial.¹⁶ This observation highlights the potential role of PPs in whether a patient decides to participate in a vaccine trial. The present study presents data on the attitudes and factors influencing the reaction of PPs in this situation.

Firstly, we reported that PPs confronted with a scenario where one of their patients said he/she was approached to participate in a PVT have a mixed reaction. In fact, although a large majority of physicians would like to make a recommendation to their patients regarding their potential participation, they often lack knowledge on clinical research. This poor knowledge of PPs was previously observed in oncology, despite the fact that PPs showed interest in training on clinical trials.^17,18 Chen et al. recommended that all physicians receive an education about clinical trial design and learn how to advise patients as far as research participation is concerned. They also suggested that it could be part of courses on the patient–physician relationship.¹⁴ In our qualitative study, physicians recognized that research is an unknown domain for physicians who were not involved in this field (Doctor I, Table 3). It would be interesting to understand the nature and types of misconceptions about clinical research among physicians in order to accurately inform and re-educate them as suggested by Walsh et al. for potential participants.¹⁰ We demonstrated that the attitudes of PPs were not directly correlated with their specialty but tend to be correlated with the fact they have research activity or not. Therefore, it could be beneficial to create continuing medical education sessions about clinical research or to implement training in medical schools. In this sense, recommendations were issued to develop a real education for non-professionals in the clinical trial in France.¹⁹

Physicians also deplored that ICF given to the patients was the only source of information they had about the proposed clinical trial. As pointed out by Doctor J (Table 3) and highlighted by Pandiva et al., terms used in ICF are not always easy to understand for people not interested in clinical research such as patients²⁰ or even some physicians. They also regret the absence of previous communication with the investigator. Thus, it seems crucial to improve communication between investigational centers and PPs, and that the investigator who approached the patient should inform the PP directly. In the US, a patient’s reference guide suggests that patients may consult their PP to explore options about entering a clinical trial and to use this guide to discuss the decision to participate or not in a clinical trial with their PP.²¹ This better communication is probably beneficial to the patient, to the PP who is the trusted physician for the patient, and to the investigator, and this could increase enrollment in the PVT.

Secondly, we observed that the most important factor influencing their attitude is the availability of safety data of the trial. This concern was frequently reported in the general population who want to be informed of any medication adverse reaction, however rare it may be.²² It is notably the fact with vaccines, safety being the primary concern in Europe.²³ Among elderly people interviewed about their willingness to participate in a PVT, 75% of participants considered that safety precautions should be taken.¹⁶ Physicians shared this concern with their patients. This observation emphasizes that general population and physicians not involved in clinical research are not conscious that trials are stopped at the slightest safety problem.

Although the majority of respondents considered vaccines as a drug like any other, 20% of the physicians still declared they would think differently if the experimental product was not a vaccine. We cannot conclude from that whether they would be keener to recommend a medicine rather than a vaccine trial. However, in the qualitative study, some physicians reported concerns about long-term side effects of these new vaccines in development, as they have concerns about vaccines recently licensed. Vaccine hesitancy among physicians has been described.^24–26 We may hypothesize that hesitant physicians could discourage patients to participate in a PVT; however, the influence of vaccine hesitancy on their attitude cannot be measured here. In the same way, if PPs do not perceive in-development vaccines as of interest, they possibly deter their patients from entering a vaccine trial. We observed that the type of infection targeted by the vaccine would influence their opinion. This point needs to be studied further as some vaccines in development will target diseases rarely managed by PPs.³ If these vaccines seem less useful to them they could negatively influence their patients. Fortunately, the great majority of medical doctors remain favorable to vaccines²⁷, and most physicians recognize that new strategies need to be developed to protect or cure patients.²⁸

Our study has several limits. Firstly we have a limited sample of PPs and a low response rate for the questionnaire. However, the panel was representative of the physician population in France when compared to official data.²⁹ The responding physicians were probably more concerned about vaccination matters or clinical research either positively or negatively, but this could not be verified in the absence of data about the refusing PPs for quantitative data. The panel was limited to Auvergne-Rhône-Alpes region. However, this region is the second bigger administrative one of France and it is a region with a great activity around vaccine development³⁰ and two vaccinology clinical investigation centers. For the qualitative study, the sample was recruited according to guidelines³¹ and as interviews were all face-to-face there were conducted in Loire and Haute Loire counties as a matter of practicality. The number of participants was stopped at data saturation.³¹ Results from this first work could probably not be generalized but bring data in a field not previously well explored. Secondly, we do not explore attitudes of PPs toward vaccination in general. Vaccine hesitancy exists among physicians in France and may affect up to 10% of GPs.³² However, a great majority of them recommend the most of vaccines to their patients.²⁴ The impact of hypothetical vaccine hesitancy among the panel could not be explored. Thirdly it was a majority of GPs who participated in both the qualitative and quantitative studies. This could underestimate possible differences between GPs and other specialists who accounted for a smaller sample. However, in France, GPs continue to have a crucial role in the follow-up and vaccination of patients including those also managed by other specialists.³³ It was a GP (CT) who conducted the interviews. It undoubtedly allowed closer contact with the other GPs, our larger category in the study. Qualitative data shed light on the trends observed in the questionnaire results.

In conclusion, these results revealed that PPs would like to make a recommendation to their patients regarding their potential participation in vaccine trials but would like more information beforehand. We found that French physicians lacked knowledge of clinical trials and regretted not receiving enough information from investigators. Developing training and decision-making tools on vaccine trials to empower PPs (GPs or other specialists) may help investigators to enroll new patients. Decision aids (DA) featuring a clearly written aim of the study, available safety data, and graphical protocol presentations could clarify information and help PPs before they give their patients an opinion. DAs have proven to be an effective means of optimizing the informed consent process; they increase knowledge and reduce decisional regret about trial participation.³⁴ It would also be interesting to test these tools with PPs when they are confronted with patients asking for an opinion on a vaccine trial proposal. Further studies are also needed to better evaluate among patients the role of the PPs in their decision to participate or not in a PVT.

Methods

We have conducted a descriptive cross-sectional study based on both quantitative (online survey) and qualitative methods (face-to-face interviews) among PPs, i.e. General practitioners (GPs) and others specialists following patients with chronic medical conditions on a regular basis. GPs from the Auvergne-Rhône-Alpes region, France and other specialists at the University Hospital of Saint-Etienne were invited via email to answer an anonymous Web-based survey (GoogleDrive®). All questions were previously validated and tested by general medicine and infectious diseases specialists involved in the study. Between May and October 2017, the questionnaire was sent by the regional union of health-care professionals (URPS Auvergne-Rhône-Alpes) to their 4232 GP members. For the other specialists, the questionnaire was sent by the Hospital Board to their 400 physicians. The quantitative questionnaire featured 14 questions about the attitudes of PPs to their patients who were offered participation in a vaccine trial. The scenario presented to the physicians was as follows: “A colleague of yours from the clinical research center of vaccinology offered one of your patients’ participation in a PVT. The patient comes to you with a copy of the information consent form and informs you of the proposal made to him/her”. The PPs’ attitudes and the factors that influence their answers were determined based on a 5-category Likert scale (rating for each question was: strongly disagree, tend to disagree, do not know, tend to agree, strongly agree). The answers to each question were dichotomized into positive versus non-positive (including: don’t know or no answer).

For the qualitative study, the method followed the COn-solidated criteria for REporting Qualitative research (COREQ) Reporting Guidelines.³¹ Semi-directed individual interviews were held with GPs and other specialists settled in Loire or Haute-Loire -French counties located in Auvergne-Rhône-Alpes region- and working in private health-care centers or in public-sector hospitals. Recruitment was based on reasoned sampling, by seeking maximum variation according to the following criteria: age, gender, specialty, clinical or non-clinical research activity, and working in urban, rural or semi-rural areas.³¹ The study was carried out between December 2017 and June 2018. First contacted by phone, the PP who accepted received the interviewer for a face-to-face interview by appointment in their office. The meetings were conducted following on interview guidelines developed according to the researchers’ questions, literature data and the first results of the quantitative survey. The guide of face-to-face interviews is available as supplemental material. The interviews were recorded after consent by the respondents. Data collection was stopped when redundancy was observed (data saturation)³¹ . Interviews were transcribed word per word on an anonymous data computer file. Data were inductively analyzed using a non-blind open manual coding, descriptive then thematic also using a constructivist grounded theory approach.³⁵ Grounded theory allows to evaluate the interrelationship between meaning in the perception of the subjects and their action.³⁵ In fine, the results were reworked with a team in the clinical trial center of the University Hospital of Saint-Etienne, with validation proofreading by a dozen interviewees.

Statistics. We used Microsoft Excel for the collection of recorded quantitative data, SPSS® IBM SPSS Statistics 24.0 (NY, USA) software for statistical analysis, and chi2 or Fisher’s exact test to compare quantitative data. An univariate analysis was first conducted to determine factors associated with PPs attitudes. A p-value ≤0.05 was considered to be statistically significant. To adjust for confounding factors, we conducted a multivariate analysis to test associations between attitudes and the explanatory variables with a p-value ≤0.2 significance level obtained in the univariate analysis.

Ethics

Questionnaires were validated by the French National Commission for Data Protection and Liberties (Commission Nationale Informatique et Libertés). The study was received favorably by the local ethics committee (IRBN742017/CHUSTE).

Author’s contributions

MD, EBN designed the quantitative study and the questionnaire, and collected quantitative data. MD, AGB, and EBN carried out analysis of data. CT held the interviews, collected qualitative data and analyzed them with a non-blind open annual coding (CT, RC, MD, EBN). MD and CT drafted this manuscript. EBN, AGB, RC, FL critically reviewed this manuscript. All authors read and approved the final manuscript.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors wish to thank URPS Auvergne-Rhône-Alpes and direction department of CHU SAINT-ETIENNE for their help in sending the web-based survey. The authors also wish to thank Mr Philippe Michelucci and Ms Glyn Thoiron for providing English editing of the manuscript.