Discussion

This review identified ten active national sentinel surveillance systems targeting illicit drug use, and, in some cases, the detection of novel psychoactive substances. Despite the systems sharing a similar process, no approach is completely the same. Identifying these different characteristics makes it possible to define specific aspects that could be included in the design of new sentinel systems directed toward illicit drug use monitoring and surveillance (Welty et al., Reference Welty, Harrison, Abram, Olson, Aaby, McCoy, Washburn and Teplin2016).

Descriptions of the characteristics of the sentinel population or the sentinel reporters are often limited in the literature despite their central role in shaping the type and quality of data obtained. Importantly, while the recruitment strategies of the sentinel population or the selection of the reporters introduces methodological considerations, the strategies do not undermine the sentinel value of the information gathered. In addition, surveillance of illicit drug use is challenging because its use is relatively uncommon in the general population and stigma may limit participation (Bathish et al., Reference Bathish, Madden, Duff and Ritter2024). The systems that carry out sentinel surveillance among users outside the hospital setting, such as IRDS and ERDS, typically recruit from syringe exchange centres, online platforms or nightlife venues (clubs, festivals), and may complement recruitment using a peer-referral or snowball technique. Focusing the sentinel population on persons under detention is an interesting approach, especially considering that the prevalence of drug consumption among them is higher than in the general population (Welty et al., Reference Welty, Harrison, Abram, Olson, Aaby, McCoy, Washburn and Teplin2016).

People with lived experience of drug use represent a highly diverse population, which makes defining the sentinel focus very challenging. Nevertheless, focusing on people who currently use drugs may offer early indications of new substances or changes in patterns of use (Soussan & Kjellgren, Reference Soussan and Kjellgren2016; Brunt et al., Reference Brunt, Nagy, Bucheli, Martins, Ugarte, Beduwe and Ventura Vilamala2017). For this reason, their contribution has been regarded as central in some sentinel models. At the same time, earlier studies highlighted methodological concerns regarding self-reported data in specific contexts (Sloan et al., Reference Sloan, Bodapati and Tucker2004). Some networks that use drug users as sentinel reporters provide incentives to participants. This aspect is controversial, especially in addiction research (Manriquez Roa & Biller-Andorno, Reference Manriquez Roa and Biller-Andorno2022).

In addition, information on drug-related behaviours or trends often relies on user self-report or that of their companions. However, in rapidly evolving illicit markets, substances consumed may differ from what individuals believe they have taken, which can introduce uncertainty in the absence of biological confirmation.

Purposeful sampling is another feature of sentinel surveillance. Some networks strategically position data collection in areas identified as high risk based on prior evidence. NDEWS, one such system, conducts rapid field assessments in locations characterised by high prevalence or rapidly changing patterns of drug use through its Rapid Street Reporting programme (Cottler et al., Reference Cottler, Goldberger, Nixon, Striley, Barenholtz, Fitzgerald, Taylor and Palamar2020; Palamar et al., Reference Palamar, Acosta, Cleland and Sherman2021). Similarly, DAWN incorporated components based on EDs in strategically selected areas to strengthen the early detection of emerging drug-related threats (Crane, Reference Crane2013; Substance Abuse and Mental Health Services Administration, 2023). In the context of purposeful sampling, many sentinel surveillance systems contribute to early warning functions by permitting the timely identification of emerging substances and unusual patterns of harm. Effective early warning approaches rely on collaboration between clinical services, forensic laboratories, and public health authorities, as well as on the triangulation of multiple data sources (Freestone et al., Reference Freestone, Siefried, Peacock, Fletcher, Barratt, Sutherland, Ezard, Doumany, MacDonald, Kypri, Acheson and Clifford2025). The systems identified in this review reflect varying degrees of this integrated model. However, the available literature rarely provides data on how quickly signals are generated, how they are interpreted, or whether they lead directly to documented public health responses.

The establishment of the sentinel component in nightlife venues could provide valuable information. This approach is implemented in SACENDU (Pluddemann et al., Reference Plüddemann, Parry, Myers and Bhana2004) and EDRS (Horyniak et al., Reference Horyniak, Dietze, Lenton, Alati, Bruno, Matthews, Breen and Burns2017). Another very interesting sentinel approach in leisure environments was conducted as repeated cross-sectional studies in clubs and music festivals, where the attendees are the sentinel population (Palamar et al., Reference Palamar, Le, Cleland and Keyes2023). It would be beneficial to integrate this approach into a sentinel illicit drug use surveillance system, allowing for a more complete and up-to-date understanding of the landscape regarding patterns of use and emerging trends among specific populations.

We identified other surveillance systems with a sentinel-like component worthy of discussion; however, they did not meet the inclusion criteria. In the case of Canada, the Canadian Hospitals Injury Reporting and Prevention Program was identified. This system collects self-reported data by means of a questionnaire distributed to patients who visit the EDs of sentinel hospitals (mainly paediatric) for treatment of trauma or other ailments (Campeau et al., Reference Campeau, Champagne and McFaull2022). Since it does not focus its activity directly on the surveillance of illicit drug use, this system was not included (Buxton et al., Reference Buxton, Spearn, Amlani, Kuo, Lysyshyn, Young, Purssell, Papamihali, Mill and Shapiro2019; McFaull et al., Reference McFaull, Champagne, Thompson and Bang2020). Other networks that have not been active in the last four years, such as the Swedish project STRIDA (Helander et al., Reference Helander, Bäckberg, Hultén, Al-Saffar and Beck2014), IONA (Identification Of Novel psychoActive substances) (Newcastle University, n.d.), and PSCAN (Psychoactive Surveillance Consortium and Analysis Network), had characteristics similar to those identified in the surveillance systems included in this review, and had EDs as a source of recruitment (Monte et al., Reference Monte, Hopkinson, Saben, Shelton, Thornton, Schneir, Pomerleau, Hendrickson, Arens, Cole, Chenoweth, Martin, Adams, Banister and Gerona2020). We have also identified some non-national systems that are not currently active but still contribute with interesting perspectives. This is the case of the Olympic surveillance system set up in Sydney in 2000 by the New South Wales Department of Health. This sentinel surveillance system is relevant as an historical example of an event-based sentinel surveillance (Indig et al., Reference Indig, Thackway, Jorm, Salmon and Owen2003). Establishing sentinel surveillance systems for illicit drug use and misuse during these types of events is of value not only from the perspective of health planning but also from the perspective of public health monitoring. It is also worth noting that other drug surveillance systems are wastewater analysis systems, which provide essential information for drug surveillance which lacks a population correlate. Finally, an institution that supports drug monitoring with a global perspective is the UNODC (United Nations Office on Drugs and Crime). This organisation collects information on drugs from member states and also runs the Global SMART (Synthetics Monitoring: Analyses, Reporting and Trends) Programme. This programme provides technical support to forensic laboratories, law enforcement agencies and researchers from partner governments in priority regions, with the aim of generating information on synthetic drugs (Laboratory and forensic science. United Nations Office on Drugs and Crime, n.d.).

This review has limitations. Surveillance systems that could have a clear sentinel component are not always classified as such, creating the possibility that published studies or surveillance systems might have been overlooked. Another limitation is the identification of sentinel systems whose main objective is not solely related to illicit drug use, but also captures other aspects, such as road accidents associated with illicit drug use (Dunn & Kelley-Baker, Reference Dunn and Kelley-Baker2021). Despite the fact that sentinel surveillance systems operate across various jurisdictional levels, we focused on the national level. Local systems are less representative of the national situation, mainly due 1) to their smaller population coverage; 2) the use of complementary sources such as surveys is limited, as these are rarely conducted at the this level; 3) the health areas covered by EDs tend to encompass larger areas, thus limiting the usefulness of the local systems; and 4) the resources available are fewer (Mounteney & Leirvåg, Reference Mounteney and Leirvåg2004). In addition, rapidly evolving surveillance initiatives may not always be captured in peer-reviewed literature and the design and results of sentinel systems may only be available through institutional sources, as is the case of the Epidemiological Surveillance System for Addictions (Sistema de Vigilancia Epidemiológica de las Adicciones-SISVEA) (Dirección General de Epidemiología, 2022). Although this system appears to have a sentinel component, this could not be confirmed due to the limited and difficult-to-access information available. In addition, access to some institutional webpages or online repositories was not consistent across locations or settings at the time of review, which may have limited the retrieval or verification of updated information for some systems. For this reason, direct contact with selected representatives or identified experts was used, where feasible, to clarify specific factual or operational aspects. Finally, the inclusion criteria of being considered ‘active’ should be interpreted in the context of the COVID-19 pandemic, which may have affected data collection, reporting continuity, or publication frequency in some surveillance systems.

This is the first review conducted with the aim of identifying and characterising sentinel surveillance focused on illicit drug use or drug misuse. These systems could be of value in the monitoring of illicit drug use by avoiding the limitations associated with traditional surveillance systems, especially the paucity of data and the inability to provide up-to-date information in a timely manner. The advantages of a sentinel system become more evident over time. As the length of time in which they operate increases, regular and systematic data collection help to identify when an apparent change becomes a meaningful trend or is an unsustained fluctuation (Crane, Reference Crane2013; Jouanjus et al., Reference Jouanjus, Gibaja, Kahn, Haramburu and Daveluy2015; Micallef & Mallaret, Reference Micallef and Mallaret2016; Substance Abuse and Mental Health Services Administration, 2023). The impact of such sentinel systems has the potential to reach beyond the monitoring of illicit drug use and could be used to monitor the impact of legislation and/or the control of other substances such as tobacco or alcohol. Sentinel reporting should be multidimensional and based on a variety of populations such as people who use drugs, health professionals and stakeholders at the community level. Investment of public funds to support sentinel surveillance appears critical, as their findings support the implementation of practices, programmes and policies at both the national and sub-national level.