Factors influencing antidepressant prescribing in general practice: a national cross-sectional study

Technical Summary: Factors Influencing Antidepressant Prescribing in General Practice

Problem Statement
Depression is a prevalent condition in primary care, with general practitioners (GPs) responsible for approximately 90% of antidepressant prescriptions in France. Despite their central role, the quality of these prescriptions varies significantly. Concerns exist regarding insufficient mental health training among GPs, the increasing reliance on pharmacological interventions over non-pharmacological options, and the specific challenges faced by locum (temporary) physicians. These practitioners often operate under constraints such as shorter consultation times, limited access to patient medical records, and a lack of longitudinal follow-up, which may predispose them to inappropriate prescribing behaviors. The study aims to analyze how professional status (established vs. locum) and the level of mental health training influence the adherence to clinical guidelines when prescribing antidepressants for anxiety and depressive symptoms.

Methodology
The study employed a quantitative, cross-sectional design involving an anonymous questionnaire distributed to licensed GPs practicing in France between February and May 2025.

• Population: The target population included both established physicians and locum physicians. The final sample consisted of 471 GPs. Exclusion criteria applied to psychiatrists, residents, and those prescribing outside the context of anxiety/depression.
• Data Collection: The survey covered four domains: baseline characteristics, prescribing practices, mental health training levels, and daily practice challenges. Data collection occurred in two phases: an initial phase via institutional channels and a second phase via professional social media groups to ensure adequate representation of locum physicians.
• Primary Outcome: The appropriateness of antidepressant prescribing was the primary outcome, assessed against four criteria derived from French National Authority for Health (HAS) recommendations:
  1. Confirmed diagnosis of a Major Depressive Episode (MDE).
  2. Assessment of severity.
  3. Consideration of non-pharmacological treatment options.
  4. Planning of short- or medium-term medical follow-up.
     Prescriptions were deemed "appropriate" only if all four criteria were met; otherwise, they were "inappropriate."
• Statistical Analysis: Bivariate analyses (chi-square, Student's t-test) were used to compare groups. A multivariate logistic regression model was fitted to identify independent factors associated with inappropriate prescribing, adjusting for age, sex, and other covariates. Model performance was evaluated using AUC, Nagelkerke's pseudo-R², and the Hosmer-Lemeshow test.

Key Results

• Overall Appropriateness: Of the 471 prescriptions analyzed, 81.1% were considered appropriate, while 18.9% were inappropriate.
• Professional Status: Locum physicians were significantly overrepresented in the inappropriate prescribing group (52.8% of the inappropriate group vs. 33.5% of the appropriate group; p=0.001). However, in the multivariate analysis, professional status alone did not remain a statistically significant independent predictor after adjustment.
• Mental Health Training: Physicians who had received mental health training (regardless of duration) were significantly more likely to prescribe appropriately. In the appropriate group, 39.8% had received training compared to 21.3% in the inappropriate group (p=0.001). In the multivariate model, mental health training was an independent protective factor (aOR = 0.509; 95% CI: 0.274–0.910; p=0.02), reducing the probability of inappropriate prescribing by approximately 8.4%.
• Time Constraints: Time constraints during consultations were the strongest predictor of inappropriate prescribing. Physicians reporting time constraints had a significantly higher risk of inappropriate prescribing (aOR = 3.661; 95% CI: 1.748–7.604; p<0.001), increasing the adjusted probability of inappropriate prescribing by 22.4%.
• Other Factors:
  • Consultation Duration: Consultations lasting ≥20 minutes were associated with higher appropriateness (p=0.034).
  • Guideline Use: Regular use of clinical guidelines and standardized assessment scales was strongly associated with appropriate prescribing (p=0.001).
  • Contextual Pressures: The inappropriate group reported significantly higher frequencies of patient pressure for prescriptions without a clear diagnosis and lack of familiarity with the patient.
  • Demographics: Female gender was associated with more appropriate prescribing in bivariate analysis (p=0.001), though this was not the primary focus of the multivariate adjustment.

Significance and Claims
The study concludes that the quality of antidepressant prescribing in primary care is determined by a combination of physician training and structural practice conditions. The authors claim that:

1. Training is a Critical Lever: Mental health training, even if brief (seminars, university diplomas, or CME), acts as an independent protective factor against inappropriate prescribing. This highlights a gap in initial psychiatric education and the need for targeted continuing medical education (CME).
2. Structural Barriers: Time constraints are a major structural determinant of poor prescribing quality, independent of individual physician competence. This suggests that organizational factors, such as consultation duration and workload, significantly impact clinical decision-making.
3. Locum Physicians: While locum status was associated with inappropriate prescribing in univariate analysis, the study suggests this is likely mediated by factors such as lack of training, lack of patient continuity, and time pressures rather than the status itself. The authors note that recent regulatory changes (2019 decree) have excluded non-contracted locum physicians from CME funding, exacerbating training inequalities.
4. Policy Implications: The findings underscore the need to reintegrate locum physicians into CME systems and to strengthen mental health training in medical curricula. The authors advocate for a dual approach: improving educational resources and addressing organizational constraints to enhance the management of mental health in primary care.

The paper acknowledges limitations, including potential selection bias (engaged practitioners responding), self-reporting bias, and the subjective nature of the appropriateness criteria. It calls for further qualitative research to explore the subjective experiences of locum physicians and patients and to better understand the ethical dilemmas faced during prescribing.