Intervention randomized controlled open study with 2 parallel arms . The objective of this study is to evaluate the feasibility of stopping the benzodiazepines treatment in elderly living in nursing homes.

It will check that the judgment of these molecules has no pejorative effect on sleep or behavior of residents and does not induce withdrawal syndrome. Two patient groups will be constituted. One will begin a gradual withdrawal of benzodiazepines in six weeks. The other group will continue his treatment and withdrawal will be proposed after 8 weeks by his general practitioner (delayed intervention).

Effects on sleep will be assessed by wrist actimetry for 10 weeks and a sleep diary . The repercussions of withdrawal on behavior will be rated by the NPI scale.

Introduction: In France, more than 30% of people over 65 consume benzodiazepines or related drugs (B & M) on a long term basis. These drugs are responsible for side effects like cognitive impairment, falls and fractures and the installation of an addiction, while their long-term efficacy is poorly documented. Recent best practice guidelines developed by the French Health Authority (HAS) focused on stopping the benzodiazepines and related drugs in elderly. Residents of geriatric institutions are particularly at risk of falls and cognitive impairment, it is possible to stop these drugs under facilitating conditions.

OBJECTIVES: To evaluate the feasibility and safety of stopping the benzodiazepines and related drugs in institutionalized elderly. Studying one year later to maintain the cessation of benzodiazepines and related drugs

METHODS:

DESIGN: Intervention Study randomized controlled open 2 parallel arms. Participating centers: 20 geriatric institutions (nursing homes)

Randomization: Residents will be included paired by geographical criteria (living unit). Residents of each pair will be randomized using a randomization list: a group in the Immediate and the other group in the delayed intervention.

Interventions: Immediate: Within 15 days after randomization, a medical investigator will meet the attending general practitioner to:

• remind good practice for B&M drugs prescription
• provide advices to stop the drug target with a concrete protocol and individualized for stopping in 6 weeks
• propose a supervision of the resident.

Delayed intervention: The resident will continue usual care without intervention from his general practitioner. Eight weeks after randomization, a letter will be sent to the general practitioner reminding the recommendation of prescription and proposing, if desired, protocols to stop the benzodiazepines and related drugs in the resident included.

OUTCOME MEASURES:

The measures will be carried out simultaneously in residents of a given pair, on schedule for the resident group Immediate:

• NPI-SCALE during the week preceding the cessation of benzodiazepines and related drugs (W2) during the corresponding week a 50% reduction of the initial dose (W6) During the week following full stop (W10).
• Sleep duration and number of awakenings evaluated by a sleep diary and wrist actimetry, obtained during the same week.
• Reports of symptoms of withdrawal symptoms during the same week.
• Reports of adverse events throughout the study.

To describe the included Residents :

• The current diseases and the autonomy Gerontology Groups Iso-Resources group ( AGGIR) measures group will be noticed

• The residents are evaluated by the cognitive test Mini Mental State Examination (MMSE) .

• If the general practitioner refuses to stop benzodiazepines and related drugs an anonymous questionnaire on the reasons will be presented.

  1. year follow-up will be done by studying the use of benzodiazepines and related drugs (long-term or occasional), The introduction of other psychotropic drugs, the occurrence of falls or fractures will be noticed.

The autonomy Gerontology Groups Iso-Resources group ( AGGIR measures ) will be noted and a MMSE will be realized.

Duration of the study :

The study duration for each resident is about 10 weeks after randomization. They will participate one day more for the year follow-up.

Residents come out of the study in the following cases:

• withdrawal by the patient or his representative consent,
• the investigator's decision to leave the patient in the study of the patient's interest, prolongation of the withdrawal of more than 2 weeks compared to the initial term due to the onset of withdrawal symptoms,
• the patient's death

Main Outcome Measures:

Feasibility:

• percentage of residents exposed to benzodiazepines and related drugs and eligible but refusing to participate in the trial
• included proportion of patients whose general practitioner refuses to stop benzodiazepines and related drugs

Tolerance:

• no negative impact on sleep: sleep disturbances severity assessed by the score of the NPI-scale sleep item
• not negative impact on behavior: the severity of agitation and aberrant motor behavior assessed by scores on items agitation / aggression and aberrant behavior of the same scale.

Secondary outcome measures:

• Average length of sleep
• Average number of nightime awakenings per week assessed using the sleep diary and wrist actimetry.
• Severity of behavioral problems assessed by scores on other items of NPI-scale.
• Proportion of patients with adverse events (including withdrawal symptoms).
• Proportion of patients discharged from study.
• Proportion of patients in the immediate intervention in which the benzodiazepines and related drugs are stopped on the 10th weeks and stopped after one year.

Statistical analysis:

The feasibility will be analyzed by descriptive statistics. The impact on sleep and behavior will be investigated by testing non-inferiority based on Student's distributions.

The comparison of proportions between the two groups (adverse events, withdrawals, falls and fractures) will be performed using the chi-square test.

EXPECTED RESULTS: This study will assess, in real conditions , medical practice , the feasibility of stopping benzodiazepines and related drugs in elderly living in geriatric institutions. The objective is to verify that the end of benzodiazepines and related drugs has no negative effect on sleep or behavior and does not induce withdrawal syndrome. Remote monitoring will show whether the withdrawal is maintained over time, if there is no carryover to other psychotropic drugs, The expected benefits of stopping benzodiazepines and related drugs will be noticed .

If the expected results are obtained, the protocol of withdrawal the benzodiazepines and related drugs will be widely proposed in geriatric institutions.

Inclusion Criteria:

• Residents aged 70 or older and institutionalized for over a month
• Receiving one or more benzodiazepines and related drugs for over 30 days;
• Corresponding to stop aspects of benzodiazepines and related drugs as recent best practice guidelines by the French health authority (HAS)
• Signature of informed consent by the patient attended or not his guardian, the guardian or a relative;
• General practitioner agreeing to adhere to recommendations.

Exclusion Criteria:

• psychiatric pathology,
• epilepsy
• end of life situation
• Refusal to participate (or representative if he is unable to consent).