Introduction

Opiate dependence, whether it involves legal or illegal drugs, is a major public health issue worldwide. Opiate dependence affects about 15,479,000 people worldwide [1]. In Europe, the average prevalence of problem opiate users is estimated at between 3.6 and 4.4 per thousand inhabitants (15–64 years old). The effectiveness of Opiate substitution treatments (OST) such as methadone, buprenorphine or buprenorphine/naloxone has been widely demonstrated [2–4]). In France, buprenorphine plays a particular role in opiate dependence care provision. This role is rooted in history and in the consensus against opiate substitution that lasted well into the mid-90s in France.

Treatments for other chronic pathologies normally obtain Marketing Authorization (MA) after scientific appraisal, in which, inter alia, the risk/benefit relationship is assessed. Buprenorphine, however, obtained MA in an emergency situation where there was political and ideological upheaval, mainly linked to the HIV epidemic [5,6]. A unique system was then implemented: more than 30 years after Dole’s and Nyswander’s work [7], non-experimental methadone centers were set up, with limited access for patients. Since 1995, methadone has been prescribed only by specialized physicians in addictive disorders centers or in hospitals (since 2001). FPs can only prescribe methadone after referral from a specialized center. The key issue in this system was the implementation of buprenorphine prescription [2,4,6].

A unique feature of the French treatment system is that all registered physicians are allowed to prescribe buprenorphine without any special education or licensing. Once prescribed, buprenorphine can be delivered by any pharmacy chosen by the patient and named on the prescription.

This system rapidly proved to be efficient in terms of public health, access to care and risk reduction [2,6]. 170,000 patients are currently being treated for opiate dependence, 65% of whom are on BHD, and 35% on methadone [8]. Nevertheless the fact that buprenorphine is so accessible has facilitated misuse. Observing the types of misuse, and the associated damage, led the authorities to issue recommendations for good practice in substitution treatments. In 2004, the Heath Authority (HA) brought together experts on opiate substitution in order to issue national guidelines on therapeutic strategies for opiate dependent subjects [9].

These guidelines focus on stabilizing and improving the patients’ quality of life [9] They still apply today and are still validated by national experts. However, almost 10 years after these recommendations were issued, a large number of physicians are still reluctant to prescribe OSTs, or only consider prescribing them as a transitional measure before abstinence [10]. According to the 2009 Health Barometer in Family Medicine, 35.7% of interviewed FPs did not accept opiate dependent patients [11]. This reluctance is often explained by the fear of misuse or, to a greater extent, by the fear of abetting a type of behavior judged as morally reprehensible. Indeed, although in 85% of cases, FPs estimated that opiate dependence is primarily a disease, 90% of them are distrustful and think that opiate dependent patients are not compliant [12].

This diversity in interpretation and application of the guidelines generates inconsistent care provision that is detrimental to patients, especially in terms of OST access and follow-up.

We have reflected upon this diversity in the prescribing of buprenorphine within a legal context which still allows buprenorphine to be as widely available for prescription as any other treatment.

The study was conducted with physicians, whose characteristics varied in terms of education, location and type of practice, OST prescription practices, theoretical compliance with the statutory national guidelines (9), and their attitudes towards OST.

Material and Methods

Results

Descriptive analysis

542 physicians were contacted via email: 356 belonged to the FP group, 78 belonged to the Network group and 108 belonged to the Hospital group. The response rate was of 44%, with 193 questionnaires collected: 111 from the FP group (response rate 31.1%), 37 from the Network group (response rate 47%) and 45 from the Hospital group (response rate 41.6%). Socio-demographic and professional characteristics are reported in Table 1 below.

Download:

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

Table 1. Socio-demographic and professional characteristics of responding physicians.

https://doi.org/10.1371/journal.pone.0137708.t001

Ease of contact with opioid dependent patients.

The results regarding how easy the follow-up of opioid dependent patients was felt to be are reported in Table 2.

Download:

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

Table 2. Ease in opioid dependent patients follow-up.

https://doi.org/10.1371/journal.pone.0137708.t002

Analysis of the compliance with recommendations.

Fig 1 shows the distribution of average global scores for compliance (not weighted to take into account non-compliant answers) according to the 3 groups: FPs group, Network group and Hospital group. This distribution is congruent and the average distribution differs among the three groups.

Download:

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

Fig 1. Distribution of the congruent answers between professional categories(MG = Fps, Hopital = Hospital, Réseau = Network).

https://doi.org/10.1371/journal.pone.0137708.g001

The 3 samples present a normal distribution curve, with an average close to 5 exact-match answers for the « Network » and « Hospital » samples and an average of 4 for the « FPs » sample, and unequal variances between the samples: Var (FPs) > Var (Network) > Var (Hospital).

Table 3 presents the distribution within the compliance categories: partially compliant (PC), Mainly Compliant (MC), Compliant (C) according to the compliance scores weighted to take into account the non-compliant answers.

Download:

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

Table 3. Distribution of the FPs, Network and Hospital groups within weighted scores compliance categories.

https://doi.org/10.1371/journal.pone.0137708.t003

Comparative analysis

The comparison of the unweighted compliance averages within the control group FPs is presented in Table 4. Significant differences between the FPs’ averages and the Network and Hospital averages are observed. There were no significant differences between the averages of the Network and Hospital groups.

Download:

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

Table 4. Comparison of average compliance scores according to professional categories.

https://doi.org/10.1371/journal.pone.0137708.t004

Regarding the impact of variables on the extent to which therapeutic attitudes matched recommendations, belonging to a compliant group was independent of socio-demographic variables. Continuous training in self-reported addiction therapy had no impact on the compliant group (Table 5).

Download:

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

Table 5. Compliance in three groups of prescribers, according to number of patients followed up with BHD treatment.

https://doi.org/10.1371/journal.pone.0137708.t005

Regarding the Network and Hospital groups, the number of patients followed up by physicians does not seem to point towards an effect size. Regarding the FPs group, prescribing physicians who had prescribed BHD to more than 2 patients, were more efficient (61% complied closely or very closely with the guidelines) than non-prescribing physicians or physicians who prescribed BHD to fewer than 2 patients (34%).

Discussion

The main result of this study shows that physicians we interviewed rarely took into account guidelines regarding buprenorphine. These results were in line with what can be found in literature regarding the difficulties encountered by physicians in following guidelines on BHD or BHD/Naloxone[[4,11]. We are unable to know for certain whether this was due to difficulties in applying the guidelines or ignorance of the guidelines or rejection of these guidelines by the physicians. However, most countries (USA and Australia, among others) only recently introduced BHD or BHD/Naloxone, whereas in France this treatment has been available since 1996, and guidelines date back 10 years.

One of the main points of guidelines is to ease access to treatment with buprenorphine. Despite the fact that the regulatory prescription framework is extended to include any physician, it is noticeable that access to buprenorphine is not systematic. The population of physicians we looked at was representative in terms of the socio-demographic characteristics of the national population of FPs. Only 42% of independent FPs prescribed buprenorphine as a first-time prescription, and 40% of FPs do not follow up patients on buprenorphine.

These results were also in line with other studies [13,14]. However, their results differ from those of a previous study conducted in Family Medicine in 2009, where two thirds of the FPs interviewed declared that at least one opioid dependent patient had come in for a consultation over the previous year. These differences could be explained by methodological differences (FPs were interviewed over the year, and mostly by telephone) and age criteria (the population interviewed was older) but mainly it seemed possible to assume that some FPs followed up patients on BHD without getting involved in the prescription because they considered this treatment as different from others. Furthermore, as put forward in the literature, many physicians considered that the objective of managing opioid dependence is complete abstinence, via withdrawal, without OSTs [10], in spite of a bad prognosis on long term opiate withdrawal [14].

In terms of compliance with guidelines, 55% of FPs gave theoretical answers that only partially complied with guidelines. The differences in compliance with guidelines were noted according to different variables, and particularly according to affiliation to a network. The Network group has a 51% rate of answers which comply with recommendations, which is significant compared with the FPs or the Hospital group. The second variable that influenced the level of compliance to guidelines is the number of patients on BHD followed up by the physician. There was a clear effect after 6 patients, but that effect was already apparent after 2 patients. The more frequently physicians prescribe buprenorphine, the more comfortable they are prescribing it again which, in turn, increases their pharmacological and technical prescription skills. It is also possible to highlight that the physicians’ perceptions of their opioid dependent patients changed, based on the number of patients and on affiliation to a network [15]. In this study we address addictions, but other studies have shown that belonging to a care network for HIV, hepatitis or addictions, fosters the development of substitution practices and strengthen these practices [13].

Risk reduction and the definition of personalized care projects are among the fundamental values shared in these networks. Differences were found in perceived therapeutic objectives between the Network, Hospital and FPs groups. The physicians most involved in opioid dependent patients’ care were Network and Hospital physicians, and they clearly considered BHD a tool for personalized treatment, whereas FPs considered it as a social control tool [15]. Conversely, there was no difference in perception between the perception of BHD as a drug or the perception of BHD as a treatment. As was found amongst patients, the prescribers are ambivalent towards BHD, whether it be perceived as a drug or as a treatment [15,16].

On average, physicians declared a high rate of participation in continuous addiction therapy training. 38% of FPs and more than 80% of Network or Hospital physicians reported having attended continuous medical training (CMT) in addiction therapy. This prevalence is in contrast to previous studies which reported 33% CMT among BHD prescribing FPs [13]. On one hand, the declarations could be overestimated to enhance social desirability, on the other hand, some physicians may have attended training in an addiction field other than that specific to opiate dependence. But contrary to our assumptions, there was no statistically significant effect of self-reported CMTs on practices which complied with guidelines.

Regarding accessible and available medical information on opiate substitute treatments, FPs and Hospital and Network physicians have very different views on care. More than 60% of independent FPs considered that the type of medical information available on OSTs was insufficient and inadequate to their needs. Conversely, Network and Hospital physicians were mainly satisfied with available information Existing information materials seemed inadequate for Family Medicine, or were not displayed, handed out or accessible enough. This lack of access to information was identified as a limiting factor to OST access for patients, according to FPs [17]

This study had some limitations, which may have had an impact on the interpretation of the findings. First, the national guidelines were written 10 years ago and could lack clarity, leading to a possible information bias. Second, there are several debatable points in the guidelines, which limits their application in real-life situations, leading to a confusion bias. But our statistical analyses addressed this point and found that experienced physicians’ practices matched national guidelines. Third, the theoretical, clinical attitude which was expected did not take into account the adjustments that are sometimes necessary in reality, leading to another confusion bias.

Organizing patients’ care under OST was thus stratified. In this respect, it was possible to distinguish between the following types of physicians: non-prescribing physicians, who may receive patients on BHD or MTD in consultations but who did not get involved in their treatments; physicians who rarely prescribed BHD or MTD, whose practices only moderately complied with the recommendations and physicians who often prescribed BHD or MTD, who may or may not belong to a network or hospital. This stratification seemed organized, but certainly lacked transparency for patients.

This study highlights the limitations of the French system and the wide availability of buprenorphine treatments, which matches the findings of other recent studies [17]. Almost 20 years after marketing authorization, France is far from being in a situation where all physicians prescribe buprenorphine, with 40% of physicians not prescribing it at all. It is impossible to deny the benefits of such a system, in terms of public health, as has been highlighted in previous work, concerning mortality and morbidity rates [2,15] But the limitations are also substantial. A minority of physicians is responsible for the majority of the buprenorphine prescribed. FPs play a central part in treatment accessibility, but their profile seems specific: they are interested in patient-centered care which is a core competency of Family medicine according to Wonca [18]. That competency has recently been assessed in most French universities and results could be even higher in the next decades. We have to consider that physicians involvement in networks, and buprenorphine prescription experience, influence their ease and their compliance with the recommendations. They followed up several patients on BHD, given that ease and compliance with recommendations that are proportional to the physician’s experience and to his/her involvement in networks. Risks linked to non-compliance with prescription recommendations are many: under-dosing, ineffective treatment [19], « switching addiction », and misuse.

Buprenorphine misuse, by snorting or injecting, involved 10% to 30% of patients treated with buprenorphine [17,19,20,21] According to recent studies, 12.5% to 16.9% of buprenorphine reimbursement in France was destined for the black market [21]. 82% of physicians feel concerned about misuse and the diversion of medicines, according to the Benyamina study [17]. But the introduction of buprenorphine/naloxone in January 2012 may help reduce the diversion of buprenorphine [22]

Since undertaking this work, we find it deplorable that addictive disorders networks, previously under threat, have now been disbanded in France. Despite access to buprenorphine in primary care, the management of opioid dependence may well remain the concern of specialists in the field, and the development of studies into opiate dependence in Family Practice, involving primary care stakeholders, will help to re-establish opiate dependence care within Family Practice.

Conclusions

With the objective of improving opioid dependent patients’ care, and improving BHD use, it might be pertinent to formalize a continuum of care for patients on BHD. This could be achieved through BHD accreditation, as contemplated for methadone on the streets, or at least a treatment program, as is occasionally recommended for new treatment management of alcohol dependence. Strengthening the role of networks could be an alternative path, as could the selection of FPs trained in the competency of patient-centered care. Nevertheless the actual results, in mortality and morbidity rates, should be kept in mind to avoid “throwing the baby out with the bathwater” by introducing an inadequate reform.

Supporting Information