In France, 40% of newly diagnosed HIV infections are concentrated in the Paris metropolitan area.

Two key populations are mainly concerned: persons born in a foreign country and men who have sex with men. The randomized trial ANRS DICI-VIH (2014-2015) showed that nurse-driven HIV screening for key populations, supported by research staff, in 8 emergency departments (EDs) of the region, was effective in addition to diagnostic testing. The strategy advocated by the WHO and the recent French recommendations support the proposal of screening for key populations in the EDs. Thus, it is important to evaluate the impact and the feasibility of the implementation of this strategy on a large scale.

The aim is to evaluate the impact of a wide implementation of nurse-driven HIV screening by rapid test in key populations combined with usual physician-directed diagnostic testing (intervention strategy) compared to diagnostic testing alone (control strategy) in the usual practice of the EDs.

The strategies will be compared during two periods in 18 EDs of Paris metropolitan area following a stepped-wedge cluster randomized trial.

During intervention period, nurses will suggest performing an HIV rapid test to patients belonging to key populations according to the answers to a self-administered questionnaire.

Background:

In France, 40% of newly diagnosed HIV infections are concentrated in the Paris metropolitan area.

Two key populations are mainly concerned: persons born in a foreign country and men who have sex with men. The randomized trial ANRS DICI-VIH (2014-2015) showed that nurse-driven HIV screening for key populations, supported by research staff, in 8 emergency departments (EDs) of the region, was effective in addition to diagnostic testing. The strategy advocated by the WHO and the recent French recommendations of the National Health Authority (HAS) and expert group support the proposal of screening for key populations in the EDs. Thus, it is important to evaluate the impact and the feasibility of the implementation of this strategy on a large scale in a context where local institutions are taking initiatives to curb the epidemic ("Pour une Ile-de-France sans sida", "Vers Paris sans sida").

Our hypothesis is that nurse-driven HIV screening for key populations can be integrated to the consultants' care pathway as usual practice of the ED and assured over the long term in the Paris metropolitan area; this strategy combined with physician-directed diagnostic testing could reduce undiagnosed infections and help control the epidemic.

Primary objective:

The primary objective is to evaluate the impact of a wide implementation of nurse-driven HIV screening by rapid test in key populations combined with usual physician-directed diagnostic testing (intervention strategy) compared to diagnostic testing alone (control strategy) in the usual practice of the EDs.

Patients aged 18 to 64 years and visiting a participating ED for reasons other than a potential exposure to HIV within less than 48 hours will be included.

The primary outcome is the proportion of new HIV diagnoses among included patients.

Secondary objectives are to :

• evaluate the feasibility of nurse-driven screening and its long-term implementation,

• evaluate the effectiveness of the screening process used by the caregivers, particularly the use of an electronic questionnaire, and by the patients as well as the acceptability,

• compare newly diagnosed HIV+ patients in the two groups in terms of :

  1. linkage to follow up care,
  2. proportions of patients who had a CD4 cell count greater or equal to 500 or 350 cells/µL and no HIV-related symptoms and proportions of patients with acute HIV infection.

• estimate through modelling the number of new HIV diagnoses for different levels of coverage of the screening strategy in the Paris metropolitan area.

Methods The strategies will be compared in 18 EDs of Paris metropolitan area during two periods. The start of the intervention strategy will be randomly allocated for each ED following a stepped-wedge cluster randomized trial.

EDs are selected according to the proportion of patients belonging to key populations in the patients they receive (data from two previous studies).

During intervention period, nurses will suggest performing an HIV rapid test to patients belonging to key populations according to the answers to a self-administered questionnaire.

The intervention period will last 3 to 11 months and the control period 0 to 8 months. Patients will participate up to 1 month if a follow-up is needed.

Inclusion Criteria:

- Patients aged 18 to 64 years and visiting a participating ED for reasons other than a potential exposure to HIV within less than 48 hours

Exclusion Criteria:

- Not applicable