[Failure to draw up medical certificates increase judiciary releases]

In France, a systematic control of compulsory psychiatric admissions has existed since the enactment of the law of July 5th 2011. The Court of Cassation clarified that the liberty and custody judges (JLD) cannot supersede the medical opinion described in the medical certificates. In 2015, the JLD ordered the release of 8.4 % of all compulsory psychiatric admissions. The goal was to compare the quality of medical certificates derived from judicial release based on medical grounds with non-released witnesses from the cohort of compulsory psychiatric admissions ordered in the Groupe Hospitalier Universitaire Paris Psychiatrie & Neurosciences (GHU-Paris) between November 1, 2017 and October 31, 2018.

We included as cases all the medical certificates derived from judicial release based exclusively on medical grounds from the release cohort of the GHU-Paris from November 1, 2017 to October 31, 2018, concerning the systematic control 12 days after compulsory psychiatric admissions. A witness whose compulsory care had been maintained was matched according to the same judge, place and date of hearing, mode of compulsory care and site of hospitalization. Each certificate was analyzed according to a reading grid relating to the good decisions in matters of compulsory admission and medical certificates' redaction. An overall score, based on the description of the clinical and symptomatic evolution, the level of discernment, the capacity of consent and the mode of compulsory care was awarded to each certificate.

Seventeen release files were included in the comparative study. Globally, the clinical progression, psychiatric symptoms, level of consciousness and ability to consent did not differ in the two groups. The grade of quality of certificate was lower in case of withdrawal (2.92±1.08 VS 3.28±0.88, P=0.026). Psychiatric symptoms in "justifiable notice" (the last medical certificate prior to the judicial hearing) were less specified in case of withdrawal (58.8 % VS 94.1 %, P=0.015). Not describing any symptoms led to a 12.51 risk of withdrawal (95 % CI=[1.16; 135.19], P=0.038). Even with witness certificate, clinical progression was noticed in only 85.3 % of cases, in 89.3 % of psychiatric symptoms, in 68.0 % of level of consciousness and 80.0 % for the ability to consent.

Judiciary releases of compulsory psychiatric admissions exclusively based on medical grounds are not arbitrarily decided by the JLD but are based on a failure to draw up medical certificates. Doctors must comply with a careful drafting of all medical certificates: description of symptoms, clinical course, level of consciousness and ability to consent. It is necessary to be attentive to judiciary releases based on medical grounds to evaluate and improve medical practices concerning the drafting of medical certificates.