GPs, which are present throughout the course of patient care, are often the first to witness the emergence of SUDs and are uniquely positioned to identify associated SUDs. This study, which was a phone survey of a nationally representative sample of French GPs, confirmed the fact that GPs have integrated the treatment of SUDs and non-SUDs into their clinical practice. Ninety-six percent of GPs reported having patients with SUDs. In addition, GPs in France are now invested in the care of HUD patients, as 43 % of GPs reported treating HUDs and 82 % of these GPs reported prescribing OATs. GPs also reported that 83 % of HUD patients had an associated SUD. Lastly, GPs also reported having a significant number of patients with opiate analgesic use disorders, an area of SUDs that has not been studied extensively in France.

The first prevalence estimates of problem drug use in France were made in the mid-1990s. In 1995, the number of persons with HUDs was estimated to be at least 160,000, using a demographic method [4]. In 1999, the number of problem opiate or cocaine users was estimated between 146,000 and 180,000 [4]. In 2006, the estimate jumped to 230,000 users [4]. In 2012, the French Monitoring Center for Drugs and Drug Addiction (OFDT) estimated the number of problem drug users (heroin, cocaine, or amphetamines) to be 260,000, of whom 81,000 were using the intravenous route [5]. Although these data suggest a significant increase in the number of problem SUDs since 2006 [4], this impression is misleading [5] as the methods and study populations have changed. The definition of the group of interest, for example, has progressed from heroin users to problem opiate users (1995), to problem opiate or cocaine users (1999), and finally, to intravenous drug users or regular users of opiates, cocaine, or amphetamines (2006).

Distribution of GPs in the different regions and the characteristics of the physicians in our study are similar to those previously documented in the Cegedim Group registry. For example, the Cegedim Group reports that 73 % of GPs in France are male; in our sample 71 % of GPs were male. This comparison suggests that the algorithm that was used to select a nationally representative cohort of GPs was successful and that our data can be extrapolated to all of France, with little risk of introducing significant bias.

When we extrapolated our numbers to all French doctors, we found that there were an estimated 360,000 people suffering from opiate use disorder in France. This incidence of opiate use disorder is higher than expected. When a similar extrapolation was made with data from the “General Practitioners’ Health Barometer 2009” study, which surveyed 2083 physicians in France and reported similar socio-demographic profiles [6], the number of people suffering from HUD in France was estimated to be 110,000. This difference may in part be due to the fact that our survey did not make the distinction between experimentation, regular use, and dependence, as defined by the OFDT. In fact, it is likely that the estimated number of persons with an HUD described by the GPs in this study (500,000 people according to the OFDT database) included some experimental users of heroin [3]. Furthermore, these estimates should be interpreted with caution, as there is also the possibility that patients consulted several doctors. Medical nomadism, though in decline, has been confirmed by several studies. In 2007, patients saw on average two GPs per year, and 25 % of patients saw at least three physicians per year [7]. In the Loire region, 14 % of patients have been reported to have consulted multiple doctors [8].

Our data support previous data that show that the changes put in place at the end of the 1990s have increased the involvement of GPs in the treatment of SUDs. Since the introduction of high-dose buprenorphine (HDB) in private practice, a survey of 1186 GPs showed that while 53 % of doctors had seen an HUD patient in the previous year, only 24 % were involved in ongoing care, and only 13 % of doctors said that when asked for OAT, they would agree to prescribe buprenorphine [9]. In the 2009 “General Practitioners’ Health Barometer” study, about 50 % of GPs reported seeing at least one patient with an opiate use disorder per month, and 87 % of GPs who saw at least one patient with an opiate use disorder per month prescribed OAT [6]. Similarly, in our study, 43 % of GPs reported seeing at least one HUD patient per year, and 82 % of these GPs had prescribed OAT during the previous 12 months. These data are consistent with data that suggest that more than 60 % of patients with an opiate use disorder are treated [10] and that 70 % of heroin users received OAT [11].

In addition, our study shows that 59 % of prescriptions written by GPs for OAT were for buprenorphine and that 37 % were for methadone. These data suggest a shift compared to 2007 data, when HDB prescriptions accounted for 80 % of nonhospital prescriptions for agonisttherapies, and compared to 2010 data, which showed that HDB accounted for nearly 75 % of all nonhospital prescriptions [7, 12]. Data from the French national health insurance service support our findings, as the increase in reimbursement requests rose dramatically between 2004 and 2010 for methadone (+276 %) compared to HDB (+29.3 %) [12].

Better physician education about the pathology, treatment strategies, and modalities of OATs [13] are likely to have contributed to this positive progression in treatment practices. Fears about overdoses associated with OAT prescriptions [14] appear to be dissipating, possibly because they are offset by the fact that problems with aggression tend to decrease with OAT. In addition, the fact that patients suffering from SUDs consider the relationship with their physician as an important factor may also have shifted patients towards seeking help from their GPs who, by definition, provide a more rounded approach to patient care [1, 15]. However, the realities of clinical practice often impose time constraints, which limit physician availability, and various studies have shown the necessity of remaining cautious due to the risk of overdose, especially in combination with benzodiazepines [16] and alcohol [17].

Among physicians who do treat patients with an HUD or an opiate use disorder, the distribution of patients is uneven [8, 18]. Our data show that GPs can be divided into two broad groups: one group that sees a few, clearly identified, and regular patients, and one group that sees many users. We show that 36 % of GPs treat 10 or fewer HUD patients, whereas 2 % of GPs treat 30 or more HUD patients per year. These data are consistent with previously published data which have shown, for instance, that in the Loire region, a single doctor treated 112 patients with an opiate use disorder using OAT, whereas 127 doctors in the same region treated only 1 patient per practice [8]. In addition, a 2006 study showed that 26 % of prescribing doctors in France treat 75 % of all patients [10]. Better distribution of patients could alleviate some of the time burdens and generally improve care for SUDs.

This study found few differences in physician practices across regions (data not shown). These data differ from those reported by other studies, which have shown heterogeneity in OAT prescription practices among GPs [19, 20]. In the large OFDT study of the 13 Primary Sickness Insurance Funds, for example, 35 % of GPs prescribed OAT during the first half of 2002; but in Metz and Lille, the proportion was larger (50 %) [20]. A more recent study in the Nord-Pas-de-Calais region showed that GPs prescribed almost all the OAT for that region (98.2 % in 2009).