Development and Implementation of an HIV-testing Intervention for Primary Care in Belgium

Using Mathematical Modelling to Determine the Number and Characteristics of People Living With Undiagnosed HIV to Inform Targeted and Innovative HIV-testing

An innovative HIV-testing strategy is developed based on mathematical modeling results on undiagnosed HIV and being implemented in primary care settings in Belgian's Flemish region. The systematically developed intervention aims at increasing the number of targeted HIV tests in primary care in order to identify people with undiagnosed HIV. The intervention tool is an HIV-screening advice targeting general practitioners (GPs), combining indicator-condition based screening and target-group based screening. A group-level training to apply this advice in routine practice is being delivered as part of the continuing medical education through GP-associations. The intervention is implemented across Flanders adopting a modified stepped wedged design: two different intervention levels (delivering written/online information on the HIV testing advice versus information plus group-level training) are being compared with a control condition (no intervention, i.e. standard practice) based on surveillance data. A mixed-method study evaluates the intervention's effectiveness, feasibility, and acceptability.

An innovative HIV-screening strategy is developed based on mathematical modelling on undiagnosed HIV and being implemented in primary care settings in Belgian's Flemish region. A multidisciplinary advisory board including stakeholders from the public sector, community-based organizations and primary care physicians (general practitionners; GPs) has been set up to advise on the different phases of the intervention's systematic development and implementation. Mathematical modelling using a back-calculation approach with national HIV-surveillance data delivered estimations of undiagnosed HIV and time-distribution between HIV-acquisition and diagnosis: 2805 (confidence interval: 2478 - 3186) people living with HIV remained undiagnosed in 2015, with highest rates among non-Belgian men having sex with men, and sub-Saharan African (SSA) women followed by SSA men (Marty et al., 2017) . These results informed the intervention objectives: to increase the numbers of HIV-tests performed, of new HIV-diagnoses among the priority groups with undiagnosed HIV, and of timely HIV diagnoses in these groups to link HIV positive individuals to care. A systematic literature review (Deblonde et al., 2018) indicated that many diagnostic opportunities were missed in primary care. Therefore, formative qualitative research was conducted with 122 purposively selected GPs to understand their perceived barriers and facilitators for HIV-testing. Based on these results, the following intervention determinants are addressed: HIV-knowledge, motivation to test and skills to proactively offer an HIV-test in a non-judgmental manner. Formative research and input from the advisory board also informed the delivery mode of the intervention. The intervention consists of an HIV-screening advice combining indicator-condition based screening (adapted for primary care) and target-group based screening. A multidisciplinary group-level training (one evening session) to apply this advice in routine practice is being delivered as part of the continuing medical education through GP-associations. The intervention is implemented across Flanders using a modified stepped wedged design t evaluate its effectiveness: two different intervention levels (1) delivering written/online information only on the HIV testing advice versus (2) information plus group-level training, are being compared with a control condition (standard of care or no specific intervention). regional HIV surveillance data are being used for the outcome evaluation. A mixed-method study evaluates the intervention's effectiveness, feasibility, and acceptability. Deblonde J, Van Beckhoven D, Loos J, Boffin N, Sasse A, Nöstlinger C, Supervie V. HIV testing within general practices in Europe: a mixed-methods systematic review. BMC public health. 2018 Dec;18(1):1191.

This mixed methods study adopts a modified stepped wedged design. The classical stepped wedged cluster design involves a random and sequential crossover of clusters from control to intervention until all clusters are exposed. It includes an initial period in which no clusters are exposed to the intervention. We adopt a modified version: the two different intervention levels are added stepwise, while a control condition is retained until the end of the study period.

General practitioners included at the first level receive the HIV-testing advice through a personal electronic mail by their local GP-organization coordinator containing an information message with the printer-friendly screening advice attached. The message also provides a link to the website of the Flemish umbrella organization for GPs (https://domusmedica.be), where the tool is available for download for all Flemish GPs. A reminder is sent out to all participants after 13 months.

At the second intervention level, GPs first receive intervention condition 1 and additionally the face-to-face group-level training session. These sessions are organized as part of regular 'continuous medical education' provided by the GP organizations ('quality circles') at their usual venues and are organized a few months after receiving intervention level 1. A reminder of the advice is sent out 13 months after the initiation of intervention level 1.

To increase GP's HIV-testing behavior a targeted HIV-testing advice is spread to participants through an electronic mail

To increase specific HIV-testing behavior among GPs a targeted HIV-testing advice is spread to participants through an electronic mail and an additional face-to-face group-level training is provided.

Change in the number of new HIV-diagnosis made by GPs (2016 data serve as baseline, and are compared to 2017 and 2018 data)

Change in the number of new HIV diagnoses made by GPs among the groups with high risk for undiagnosed HIV (2016 data serve as baseline, and are compared to 2017 and 2018 data)

Change in the number of HIV-tests performed by GPs in Flanders in 2018 (2016 data serve as baseline, and are compared to 2017 and 2018 data)

Fidelity in implementation is measured by assessing GPs' coherence of HIV-testing in routine GP care with the HIV-testing advice. This is assessed qualitatively during in-depth telephone interviews with participants, assessing if the intervention is implemented exactly as designed. This includes questions on the content of the intervention (e.g. proposed reasons for offering an HIV-test to patients), and frequency (i.e. which patients are proactively targeted for HIV-testing by the GPs).

Feasibility of implementing the HIV-testing advice/group-level training in GP's routine practice measured by a self-developed questionnaire and in-depth telephone interviews.

Feasibility of implementation is measured by assessing determinants (i.e. influencing factors) of routinely offering HIV-tests in GP practices. This is assessed with a mixed-methods approach: both by a self-developed online questionnaire for GPs (using a Likert scale from 1-10 expressing agreements to statements on feasibility indicators) and qualitatively during in-depth telephone interviews with GPs assessing their personal experiences with the HIV-testing advice.

Acceptability of the HIV-testing advice/group-level training in GP's routine practice measured by a self-developed questionnaire and in-depth telephone interviews.

Assessment of GPs' personal experiences with the implementation of the HIV-testing advice in the GPs practice and with their patients' reaction on it. This is assessed with a mixed-methods approach: both by a self-developed online questionnaire for GPs (assessing acceptability-indicators with a Likert scale from 1-10) and during in-depth telephone interviews with GPs of the HIV-testing advice in the GPs practice.

Inclusion Criteria: Flemish general practitioner associated with a GP-circle (local general practitioner-umbrella organization) Exclusion Criteria: None

Marty L, Van Beckhoven D, Ost C et al. Unraveling the geographic and population heterogeneity of the HIV epidemic in Belgium. Poster presented at the 9th IAS Conference on HIV Science, 23-26 July 2017, Paris, France.

Deblonde J, Van Beckhoven D, Loos J, Boffin N, Sasse A, Nostlinger C, Supervie V; HERMETIC Study Group. HIV testing within general practices in Europe: a mixed-methods systematic review. BMC Public Health. 2018 Oct 22;18(1):1191. doi: 10.1186/s12889-018-6107-0.

Apers H, Nostlinger C, Van Beckhoven D, Deblonde J, Apers L, Verheyen K, Loos J; HERMETIC Study Group. Identifying key elements to inform HIV-testing interventions for primary care in Belgium. Health Promot Int. 2020 Apr 1;35(2):301-311. doi: 10.1093/heapro/daz037.

Apers H, Vuylsteke B, Loos J, Smekens T, Deblonde J, Van Beckhoven D, Nostlinger C. Development and Evaluation of an HIV-Testing Intervention for Primary Care: Protocol for a Mixed Methods Study. JMIR Res Protoc. 2020 Aug 17;9(8):e16486. doi: 10.2196/16486.