[Ability of French General Practitioners to detect common mental disorders identified using the Patient Health Questionnaire: Has this changed with the introduction of gatekeeping and registration with a chosen doctor?]

OBJECTIVES: The general practitioner (GP) is the most frequently consulted health professional by patients with common mental disorders (CMD). Yet approximately half of cases are not detected by the GP. Many factors linked to the patient, the doctor and the health care system influence detection. For example, detection rates are higher when patients are better known to their GP. On the other hand, patients visiting a different GP for reasons of dissatisfaction with previous care are more likely to be detected on the survey-day. In France, a form of gatekeeping was introduced in 2005 to encourage patients to register with a doctor (most often a GP) of their choice (known as the Preferred Doctor), responsible for care coordination and referral if necessary to secondary care. Visiting a different GP, other than for non-avoidable reasons (for e.g. GP unreachable, patient on holiday), is still possible but financially sanctioned with lower reimbursement rates. We aimed to compare GP detection rates before and after the introduction of this gatekeeping scheme. Patient service use behaviour such as doctor-shopping and GP referral to secondary care were also compared.

METHODS: Two cross-sectional surveys using the same study methods were carried out 10 years apart. In 2003, 46 GPs and 1151 patients participated (approximately 25 patients per GP), with a 32.7% GP participation rate. In 2013, 38 GPs participated (of which 29 had participated in the previous study, with a 85.3% "recapture" rate) and 1133 patients (approximately 30 patients per GP). Patient participation rates were 89.8% and 67%, respectively. Patients completed self-report questionnaires in the waiting room of which the DSM-IV diagnostic criteria Patient Health Questionnaire (PHQ) and an adapted version of the Client Service Receipt Inventory (CSRI) on contacts with health care services in the previous six months. For each patient, the GP completed a questionnaire giving his rating of psychiatric illness on a five-point scale with his/her diagnosis for cases, and action undertaken.

RESULTS: Of the patients, 27% and 25.4% had a CMD according to the PHQ (defined as a diagnosis of minor or major depression, panic attack, anxiety or somatoform disorder) in 2003 and 2013 respectively. Corresponding detection rates were 51% and 52.6%. Rates were highest for threshold disorders: panic disorder (69.4% and 79.9% in 2003 and 2013, respectively), major depression (75% and 63.3% in 2003 and 2013, respectively) and other anxiety disorders (69.1% and 78.8% in 2003 and 2013, respectively). In 2003, the GPs declared seeing 15.5% for the first time on the survey-day, compared to 9.6% in 2013 (P=0.006). Doctor-shopping declined between the two studies, from 18.4% to 12.1% for practical and mostly unavoidable reasons, and from 9.8% to 4.2% for dissatisfaction reasons (P<0.0001). Referral to specialist doctors increased from 9.7% in 2003 to 14.7% in 2013 (P=0.014). In 2013, on the survey-day, 94.8% of patients had registered with a Preferred Doctor and 81.2% were seeing this Preferred Doctor. In 2003, 93.5% of patients declared having a usual GP and 79.9% were visiting this GP on the survey-day.

CONCLUSIONS: This is one of the first studies to report data from two repeated surveys carried out before and after a change in the health service organisation, with data collected from both the patient and the GP. We report relatively high GP detection rates for the two periods, with about 50% of CMDs, including subsyndromic conditions, detected by the GP. Rates are considerably higher for the threshold disorders. The overall detection rate did not increase as expected between the two studies. Detection is a complex topic, involving issues such as the suitability of applying categorical DSM-IV criteria diagnoses to primary care, the relevance of detecting subthreshold conditions and the ability of cross-sectional studies to correctly assess the ability of GPs to recognise cases. The introduction of gatekeeping with the choice of a Preferred Doctor has led to a decline in the frequency of doctor-shopping, whatever its reason, with patients no doubt being better known to the GP. Yet it appears most patients had already chosen a GP they were loyal to before the scheme, with a similar proportion of patients consulting their chosen GP or Preferred Doctor on both survey-days in 2003 and 2013, suggesting the scheme may to some extent only have officialised what already existed with respect to having a usual GP. The French reform still allows for doctor-shopping which can be considered as a positive aspect of the scheme: patients either dissatisfied with previous care or needing to change GP are thus able to "test" and choose the doctor that best suits their needs.