Developing evidence-informed indicators to monitor HIV pre-exposure prophylaxis programmes across EU/EEA countries: a multi-stakeholder consensus

Several countries in the European Union (EU) and European Economic Area (EEA) established and/or scaled up HIV pre-exposure prophylaxis (PrEP) programmes between 2016 and 2023. Data on PrEP programmes’ performance and effectiveness in reaching those most in need will be needed to assess regional progress in the roll-out of PrEP. However, there is a lack of commonly defined indicators for routine monitoring to allow for minimum comparability. We propose a harmonised PrEP monitoring approach for the EU/EEA, based on a systematic and evidence-informed consensus-building process involving a broad and multidisciplinary expert panel. We present a set of indicators, structured along relevant steps of an adapted PrEP care continuum, and offer a prioritisation based on the degree of consensus among the expert panel. We distinguish between ‘core’ indicators deemed essential for any PrEP programme in the EU/EEA, vs ‘supplementary’ and ‘optional’ indicators that provide meaningful data, yet where experts evaluated their feasibility for data collection and reporting as very context-dependent. By combining a standardised approach with strategic opportunities for adaptation and complementary research, this monitoring framework will contribute to assess the impact of PrEP on the HIV epidemic in Europe.

This section aims to capture the latest updates regarding the status of PrEP implementation in your country. You will find the country case study that you submitted previously here, to assist you to fill in this section. If more options are possible, please tick all that apply.
A What is the current status of PrEP implementation in your country?
☐National program with co-payments ☐National program without co-payments ☐Private prescription and online purchase ☐Ongoing pilot or research project ☐Other: please specify B Are there any major changes in the main service delivery approach for PrEP [i.e. type of provider and delivery setting of PrEP], compared to the country case study? If yes, please specify. If no, please proceed to section 2.

DATA ON PrEP INITIATIONS
In this section we will focus on available data on people who initiate PrEP (i.e. first-time PrEP users or 'PrEP starters') in your country.
A How many people have started PrEP (i.e. have received PrEP for the first time in their lives) in the year 2019 1 in your country? Please specify how this number was produced. If no estimate on national level can be provided, please provide any other (more peripheral) available data on PrEP starters and specify the source(s). If no data at all are available, please proceed to question C.

B
Please specify for each level (i.e. national/regional, and facility-level), all possible available data sources that you are currently aware of and that could inform the number of PrEP starters in your country. If any (ongoing) research is available that could inform this number, please specify.
-Possible sources can include (but are not limited to): (sub-)national surveillance, clinical record data (i.e. patient data), pharmacy data (e.g. boxes of TDF/FTC sold), claims data (e.g. reimbursed prescriptions/consultations for PrEP), etc. -(Sub-)national or regional level concerns the availability of data covering the entire country (or county/region for sub-national level), either through central data collection systems, or through the reporting of peripherally collected data (i.e. at facility-level) to a (sub-)national focal point. -Facility-level concerns the availability of data at the specific facility or site where they are collected and/or registered (i.e. at pharmacy or at certain clinics).
Level (national/regional, facility) and research data Data sources (e.g. surveillance, clinical records, pharmacy data, claims data, other?). Please specify.
1. National or regional (if applicable) 2. Facility level 3. Research data (or ongoing research projects) Please specify. 1 We request data on the year 2019 as data on the reporting year of 2020 might not be available yet in some settings and/or COVID-19 might have caused considerable disruptions in data collection/registration compared to regular reporting years. However, if only data outside the requested time span are available, then please provide these data and specify the reporting period.

D
Please indicate whether the following information on PrEP starters is -to your knowledge -available in your country? If yes, can you explain briefly (e.g. data sources, national/regional and facility-level etc.) and provide an estimate if possible/available?

ADDITIONAL DATA RELATED TO PrEP
In this section we will focus on additional PrEP data (other than 'PrEP starters') currently available in your country.

A
To your knowledge, are any data on the following topics related to PrEP available in your country? If yes, specify the indicator used and its data source, and provide an estimate if possible/available. Supplementary material 3: list of candidate indicators and accompanying definitions, and a list of useful data sources for PrEP monitoring.
Indicator name Description

PrEP service availability
The number of PrEP providers in a given area or per 100.000 population.

PrEP awareness
The number of people who report being aware of the existence of PrEP as an HIV prevention option, regardless of whether or not it is available to them.

Willingness to use PrEP
The number of people who report being willing to use PrEP if it were offered/available to them.

Discussed PrEP with provider
The proportion of PrEP-eligible people that discussed PrEP with a provider, e.g. requested or received information on PrEP.

PrEP eligibility
The number of people that are (estimated to be) eligible for PrEP, according to local PrEP eligibility criteria.

Number of sexual health service attendees
The number of people that attended a sexual health service (SHS) in the reporting period (e.g. 12 months).

Current PrEP use
The number of people who used PrEP at least once during the reporting period (e.g. last 12 months).

New PrEP initiations
The number of people who initiated PrEP for the first time in their lives (i.e. who were previously naïve to the use of ARVs for the purpose of HIV prevention) during the reporting period (e.g. the last 12 months).

Profile of current PrEP users
The number of people who used PrEP at least once during the reporting period (e.g. last 12 months), disaggregated by some key characteristics of these users.

PrEP coverage
The number of people who used PrEP at least once during the reporting period (e.g. 12 months), divided by the size of the total PrEP-eligible population.

PrEP product use
The number of people who used PrEP at least once during the reporting period, stratified by specific PrEP product choice (i.e. oral TDF/FTC, oral TDF/3TC, oral TDF, or -in the future -injectable Cabotegravir or oral Islatavir).

PrEP dosing regimen
The proportion of people taking PrEP on a daily versus event-driven basis during the reporting period.

PrEP-to-need ratio
The number of people who used PrEP at least once during the reporting period in a given area (e.g. 12 months), divided by the number of new HIV diagnoses during the reporting period (e.g. 12 months).

Adherence
The proportion of PrEP users with sufficient or good adherence to the chosen PrEP dosing regimen (i.e. who take pills as directed to have protective drug levels in the blood during condomless intercourse).

PrEP continuation
The proportion of people who initiated PrEP and continued taking PrEP until a certain time point.

PrEP discontinuation
The number of people who are reported to have stopped PrEP OR who did not return for a follow-up visit during the reporting period (and who still received PrEP in the previous reporting period).

Retention in care
The number of HIV tests performed among all PrEP users during the reporting period (e.g. 12 months), divided by the number of HIV tests expected to have been performed among all PrEP users during the reporting period (e.g. 12 months).

HIV seroconversions among PrEP users
The number of new HIV diagnoses among people prescribed PrEP at least once during the reporting period and who had at least one follow-up HIV test.

PrEP safety
The number and type of (serious) adverse events that occur among PrEP users during the reporting period and are related to PrEP use.

STI diagnoses
The number of new diagnoses of chlamydia, gonorrhoea and syphilis among PrEP users during the reporting period.

ARV resistance
The proportion of blood specimens from PrEP users or people with a history of PrEP use and diagnosed with HIV that were tested and which show genetic mutations associated with (ineffective) PrEP use. 2) Readily available data 3) Population-level data 1) Databases not exhaustive (e.g., unavailability of data from prescriptions through closed health systems) 2) Possible lack of insight into the proportion of missing data or whether missingness is differential 3) Limited sociodemographic data (e.g., on sex, age and place of residence, but not KP)

Pharmacy dispensation database
Provides estimates on prescriptions filled Different sampling methods (e.g., venue-based or internet-based) may yield characteristically different population samples.

Medical claims database
Generally, internet-based surveys are more timely and lower cost than in-person surveys and have a broader geographic scope.
Supplementary material 4: methodological details of the Delphi process.

The Delphi technique
In order to facilitate the process of ultimately finding the most reliable consensus on indicators suitable to include in a practical monitoring tool for PrEP programmes in the EU/EEA, we used a modified Delphi method as an overall guiding approach 2 . The Delphi technique has been characterized as "a method for structuring a group communication process so that the process is effective in allowing a group of individuals, as a whole, to deal with a complex problem" 3 . Key advantages of using the technique are (1) the ability to build consensus in areas where scientific evidence is insufficient and/or conflicting, (2) the possibility to involve a wide range of experts from many countries at distance, with their indirect interaction being mediated by the study team, and (3) the ability to 'level-out' the influence of relatively dominant voices when using regular discussion groups. The modified Delphi method applied in this project existed of two phases, and combined a comprehensive review of existing evidence on the topic of PrEP monitoring (phase 1) with the collective judgement of a multidisciplinary expert panel (phase 2).
To this extent, an international expert panel was established with support of ECDC, consisting of 42 members. This panel included people from different backgrounds (clinical, research, and community), yet all with practice-based experience with the implementation and/or monitoring of PrEP or related programmes, and representing different EU/EEA Member States (see supplementary material 1 for the full list of experts). This expert panel committed to participating to the entire consensus-building process.

Consensus-building process
The data collection phase consisted of two online survey rounds, with an online meeting of the expert panel in-between survey rounds. Data collection took place between September and November 2021. We used Formsite as online survey platform, and Webex as online videoconference tool. The first Delphi survey was conducted in September 2021 and consisted of the 21 candidate indicators identified through the evidence review phase. Expert panelists were asked to rate the perceived importance and feasibility of implementing the candidate indicators in their respective contexts on a 7-point Likert scale (1=not important or not feasible at all; 7=very important or very feasible).
Experts were asked to base their ratings on the accompanying evidence summary tables (see Table 1 and Table 2), combined with their personal opinion and experience. Participating experts had the opportunity to provide additional qualitative comments for each indicator, to clarify their ratings or to express particular concerns they might have had. A link to the online survey can be found here.
In total, 30 out of 42 invited experts completed the first online survey. For each candidate indicator, we calculated the experts' median rating for importance and feasibility, and the percentage of experts rating on either extremes of the scale (i.e. 6-7 as top rating, and 1-3 as bottom rating).
Based on the ratings of the experts, the research team categorized the candidate indicators into their likelihood of being included in a future monitoring tool. The following algorithm was used to distinguish between indicators that could be (1) included without further discussion, (2) for which inclusion was uncertain, or (3) indicators that would be eligible for exclusion due to poor ratings of importance and feasibility: • INCLUDE WITH NO DISCUSSION: Median importance rating of 6-7 with NO DISAGREEMENT o Disagreement = 25% or more experts rated 1-3 on importance scale OR expressed substantial reservations in the qualitative comments • UNCERTAIN: Median importance rating of 3-5 (regardless of agreement) OR median importance rating of 6-7 WITH DISAGREEMENT o Disagreement = 25% or more experts rated 1-3 on importance scale OR expressed substantial reservations in qualitative comments • EXCLUDE: median importance rating of 1-2 (regardless of agreement) OR indicators from uncertain category that remain unresolved after discussion The online meeting with the expert panel was held on 9 November 2021. The aim of the meeting was to: 1. Agree to include those indicators in the monitoring tool that have a high median score of importance (6-7) with no disagreement.
2. Agree to discard indicators deemed of insufficient importance (median importance score of 1-2).
3. Discuss the appropriateness of including those indicators in the monitoring tool that fall under the 'uncertainty' category.
4. Discuss which indicators could be part of a minimum common 'core set' of indicators to be collected and reported across the EU/EEA.
After plenary discussion, during which indicators were refined to make them more accurate and adapted to the EU context, attending panellists were asked to vote for final inclusion or exclusion of the (refined) indicator in the monitoring tool, distinguishing between inclusion in a common 'core set' to be reported across EU/EEA MSs and relevant indicators to adapt on a country-level basis. The cutoff to include indicators in the final tool was set as follows: • INCLUDE: if 75% or more of panellists voted for either inclusion or inclusion of a modified indicator (and provided suggestions for modification).
• EXCLUDE: if >25% of panellists votes for exclusion of indicator.
See supplementary material 4 for a synthesis of expert panellists' ratings.
Pending issues were resolved in a steering group consisting of the research team, the ECDC coordinator and 5 expert panellists (see Supplementary Material 1). These panellists were selected based on their particular expertise with designing programmatic PrEP monitoring indicators (i.e. one expert led a previous publication on the topic), the status of PrEP implementation and monitoring in the country (including both representation from countries more advanced in PrEP monitoring as well as "starters"), and representation from different professional backgrounds (clinical, public health surveillance, and civil society) in addition to those of the research team (epidemiological, social sciences and M&E).

Supplementary material 5: synthesis of the expert panel's ratings for both Delphi survey rounds.
Indicator

New PrEP initiations
The number of people who initiated PrEP for the first time in their lives (i.e., who were previously naïve to the use of ARVs for the purpose of HIV prevention) during the reporting period (e.g. the last 12 months).

Number of sexual health service attendees
The number of people that attended a sexual health service (SHS) in the reporting period (e.g. 12 months) Outcome: Excluded as separate indicator; not discussed during Delphi session 2 -specified proxy for PrEP need perceived to be unsuitable in many (non-UK) contexts

PrEP coverage
The number of people who used PrEP at least once during the reporting period (e.g. 12 months), divided by the size of the total PrEPeligible population.

PrEP-to-need ratio
The number of people who used PrEP at least once during the reporting period in a given area (e.g. 12 months), divided by the number of new HIV diagnoses during the reporting period (e.g. 12 months).

Adherence
The proportion of PrEP users with sufficient or good adherence to the chosen PrEP dosing regimen (i.e. who take pills as directed to have protective drug levels in the blood during condomless intercourse). Outcome: refined indicator retained as core indicator.  suggestion included to integrate PrEP monitoring with STI monitoring systems.

PrEP safety
The number and type of (serious) adverse events that occur among PrEP users during the reporting period and are related to PrEP use.
Outcome: excluded as separate indicator; not discussed during Delphi session 2 -to be linked to general pharmacovigilance monitoring 15. ARV resistance The proportion of blood specimens from PrEP users or people with a history of PrEP use and diagnosed with HIV that were tested and which show genetic mutations associated with (ineffective) PrEP use.
Outcome: excluded as separate indicator; not discussed during Delphi session 2 -to be linked to general HIV monitoring

PrEP service availability
The number of PrEP providers in a given area or per 100.000 population.

PrEP awareness Via surveysresearch
The number of people who report being aware of the existence of PrEP as an HIV prevention option, regardless of whether or not it is available to them.

Eligibility
The number of people that are (estimated to be) eligible for PrEP, according to local PrEP eligibility criteria.
Include as defined 15.8 % Include modified* indicator 52.6 % Exclude 31.6 % *Devise a (nationally relevant) algorithm or proxy to predict the size of the PrEP-eligible population based on data of -among other possible predictors-STI incidence and HIV incidence.
Outcome: excluded as separate indicator; integrated in denominator of "PrEP coverage" indicator. N/A

Profile of current PrEP users
The number of people who used PrEP at least once during the reporting period (e.g. last 12 months), disaggregated by some key characteristics of these users. Outcome: excluded as separate indicator; no single best SES disaggregator, to be judged in context which SES indicator is most relevant to report.