Horizon Scan

Detecting Compositions of Unregulated Substances: A Horizon Scan of Emerging Drug-Checking Technologies

What Is the Issue?

• Canada continues to experience a toxic drug crisis driven by an increasingly complex and rapidly changing unregulated drug supply.

• Drug-checking services help identify substances and contaminants in unregulated drug samples, supporting harm reduction and public health responses.

• In response to the ongoing toxic drug crisis, new drug-checking technologies (DCTs) are emerging. Further research could help determine how these technologies might complement established options to address existing challenges — such as detecting low-concentration substances, novel psychoactive substances, and complex drug mixtures — and improve access to drug-checking services.

What Are the Technologies?

• Many emerging DCTs are portable, hand-held, or field-deployed, and are designed for use outside traditional laboratory settings while still offering the potential for high analytical performance.

• Emerging DCTs include portable Raman spectroscopy, surface-enhanced Raman spectroscopy, near-infrared spectroscopy, infrared spectroscopy combined with machine learning, portable mass spectrometry, sensor-based technologies, and multitechnology approaches.

• We searched key resources, including journal citation databases, and conducted a focused internet search for relevant evidence published since 2011.

What Is the Potential Impact?

• Initial studies suggest emerging DCTs may improve the detection of low-concentration substances, contaminants, novel psychoactive substances, and complex drug mixtures.

• Portable devices may expand access to drug-checking services in underserved communities as well as mobile, rural, remote, and event-based settings.

• Emerging DCTs may provide timely and detailed information about substance compositions, contaminants, and novel psychoactive substances. This information may support people who use drugs, harm-reduction services, health care providers, public health agencies, and surveillance programs in identifying trends in the unregulated drug supply.

• No single technology demonstrated advantages across all analytical and operational areas. Although emerging technologies demonstrate encouraging analytical performance, the technologies differ in sensitivity, portability, analysis time, training requirements, dependence on reference libraries, costs, and implementation needs.

• Combining emerging technologies with complementary established technologies may improve substance detection compared with relying on a single technology alone.

What Else Do We Need to Know?

• This Horizon Scan complements a 2026 Canada’s Drug Agency Health Technology Review by providing an overview of emerging DCTs and early evidence on their potential applications. The previous review evaluated established DCTs, while this Horizon Scan highlights technologies that are still evolving and where evidence remains limited. Emerging DCTs may complement, rather than replace, established DCTs and laboratory-based methods, which continue to play an important role in drug-checking services, including the identification and confirmation of novel or unexpected substances.

• Several emerging DCTs have been evaluated in community drug-checking settings, particularly in Canada; however, evidence is still limited for many technologies and consists largely of preliminary studies.

• More studies are required to better understand the accuracy, reliability, costs, and real-world implementation of these technologies. Further, evidence synthesis with critical appraisal will be necessary to provide conclusive statements about their performance.

• Decision-makers may need to balance analytical performance with feasibility, workforce capacity, infrastructure requirements, costs, and equitable access when considering implementation.

Purpose and Scope

We conducted this Horizon Scan to inform decision-making about emerging drug-checking technologies (DCTs) that may support harm-reduction services in Canada. The main objective was to identify and provide a high-level overview of emerging DCTs, including how they work, available evidence on their performance, and key implementation considerations. We also prioritized the following additional objectives:

• Explore the potential role of emerging DCTs within drug-checking and harm-reduction services in Canada.

• Summarize broader considerations related to the use of emerging DCTs, including cost; operational requirements; acceptability; privacy; regulatory context; and access in rural, remote, and underserved communities.

Members of the National Drug Checking Working Group reviewed the project plan and provided feedback on technologies of interest. They also identified considerations relevant to drug-checking services in Canada.

An information specialist conducted a customized literature search on May 1, 2026, balancing comprehensiveness with relevance of multiple sources and grey literature. Additional studies were identified through handsearching and reviewing the reference list of included studies. Further details are provided in the Methods section (Appendix 1).

What Is the Issue?

Canada continues to experience a severe and evolving toxic drug crisis.¹ Unregulated drugs include unexpected contaminants and new psychoactive substances (NPSs) such as synthetic opioids, benzodiazepines, veterinary sedatives, and stimulants, and can be lethal even in low quantities. The unregulated drug supply also changes rapidly, particularly within the unregulated opioid supply, where new contaminants and NPSs continue to emerge. Recent examples identified by Canadian drug-checking programs include medetomidine, etodezitramide, cychlorphine, and nefopam.² Exposure to unexpected substances can increase the risk of adverse effects, including sedation, respiratory depression, loss of consciousness, drug toxicity, and death, particularly when people are unaware of the substances present or their concentrations.³ This highlights the need for harm-reduction interventions to include approaches and technologies that can identify unexpected and emerging substances in the unregulated drug supply.^4,5 In Canada, DCTs are analytical tools used within drug-checking services to support harm reduction and public health activities.^6,7 They identify expected and unexpected substances in unregulated drug samples, helping people who use drugs and service providers to make more informed decisions, and supporting efforts to prevent drug-related harms and deaths related to drug toxicity.⁶ Aggregated drug-checking data also support surveillance by identifying trends in the unregulated drug supply, detecting contaminants and NPSs and informing public health responses.^6,8 These technologies vary by program and available resources, but commonly include Fourier-transform infrared (FTIR) spectroscopy, immunoassay test strips, and laboratory-based confirmatory testing.^6,9 They also differ in their analytical capabilities, infrastructure requirements, turnaround times, and costs.⁹

To date, the published evidence base primarily focuses on established DCTs, particularly immunoassay-based test strips and FTIR spectroscopy.⁹ Immunoassay test strips are relatively affordable, portable, and feasible to implement in community settings. They are designed to detect a targeted substance and are often used alongside other DCTs that provide information about the broader composition of a sample. For example, a fentanyl test strip can detect fentanyl but would not detect xylazine if it were present in the sample. FTIR spectroscopy can identify a wide range of substances in a mixture and is supported by comprehensive reference libraries. However, its performance may be reduced when target substances are present at low concentrations or within complex multicomponent samples.⁹ In this Horizon Scan, emerging DCTs refer to technologies that are newly applied or have limited adoption in drug-checking settings, with an evolving evidence base for this use.

A 2026 Canada’s Drug Agency (CDA-AMC) Health Technology Review⁹ aimed to evaluate evidence on test accuracy, limit of detection (LOD), repeatability, reproducibility, cost-effectiveness, and cost and operational considerations for DCTs. The review found that accuracy varied by substance and context, and that multitool approaches (i.e., combining multiple DCTs) can improve sensitivity and reduce false-negatives. The review highlighted that emerging technologies, such as portable gas chromatography–mass spectrometry (GC-MS) and surface-enhanced Raman spectroscopy (SERS), may improve accuracy. This review also identified an important research gap, wherein evidence on emerging DCTs remains limited and fragmented, including their feasibility and real-world applications. At the same time, there remains a need for technologies that provide broader substance detection, improved sensitivity for trace compounds, and faster turnaround times. There is also a need for technologies that are feasible for use in community, mobile, event-based, rural, and remote settings.^6,9 This Horizon Scan complements the previous review by focusing specifically on emerging and next-generation DCTs. This scan synthesizes peer-reviewed and grey literature to primarily describe emerging technologies, identify early evidence and implementation experiences, and summarize their potential role in addressing current limitations of established DCTs. The Horizon Scan was also undertaken in response to strong interest in emerging technologies beyond those evaluated in the previous review. This interest was expressed by the National Drug Checking Working Group, the Canadian Centre on Substance Use and Addiction (which co-chairs the Working Group), peer reviewers, and provincial jurisdictions.

In this rapidly evolving field, a range of emerging DCTs are being developed to address these limitations and aim to:

• improve the detection of low-concentration substances

• enhance characterization of complex mixtures

• reduce reliance on centralized laboratory infrastructure

• provide advanced drug-checking services outside traditional laboratory environments and therefore have greater feasibility for deployment in community, mobile, rural, and remote settings.

How Are DCTs Used in Canada?

Several jurisdictions in Canada offer drug-checking services as part of broader harm reduction and to prevent death related to toxicity strategies. Community-based organizations, public health agencies, and harm-reduction programs provide these services in a variety of settings. These include supervised consumption sites, community health organizations, outreach programs, mobile services, and community events (e.g., music festivals). The DCTs used in these settings have not been assessed by Health Canada to determine their safety, effectiveness, or quality for the purpose of drug checking.¹⁰

The potential benefits of emerging DCTs depend on how they are implemented within drug-checking services. For example, concerns about privacy, legal consequences, or interactions with law enforcement may prevent some people from accessing services. Programs such as Ontario’s Drug Checking Community operate through exemptions from Canada’s Controlled Drugs and Substances Act,¹¹ allowing the handling and transfer of controlled substances for drug-checking purposes. The program accepts anonymous sample submissions through participating community-based organizations. Results and samples are intended for public health and harm-reduction purposes and cannot be used for punitive, legal, regulatory, or compliance activities.¹² Additional factors, including transparent communication about technology limitations and uncertainty, quality assurance processes, and meaningful engagement with people who use drugs and community partners may influence trust, acceptability, and service uptake. When appropriately implemented, emerging DCTs may also reduce reliance on centralized laboratory infrastructure and support broader public health surveillance and harm-reduction efforts.^9,10 The broader ethical and governance considerations related to the implementation of drug-checking services are beyond the scope of this Horizon Scan.

Despite increasing adoption, access to drug-checking services remains uneven across Canada. Many rural, remote, and northern communities have limited access to advanced analytical technologies because of infrastructure, staffing, cost, and logistical challenges.^6,9 Emerging DCTs are gaining interest among harm-reduction organizations, public health agencies, community service providers, and policy-makers. They may offer improved portability, sensitivity for trace substances, turnaround times, and accessibility compared with currently used technologies. As these technologies continue to evolve, they may improve access to harm-reduction services in communities facing geographic, stigma-related, and health service barriers. They may also support timely monitoring of the unregulated drug supply.

How Are Emerging DCTs Used in Canada?

In Canada, community harm-reduction and drug-checking programs are piloting, implementing, and evaluating several emerging DCTs. Although some of these technologies are already being used, they are considered emerging because their adoption remains limited, implementation is still evolving, and evidence supporting their use continues to develop. British Columbia and Ontario have been particularly active in assessing technologies such as portable and miniature mass spectrometry systems, SERS, portable Raman spectroscopy, and machine-learning approaches integrated with infrared (IR) spectroscopy.^13-18 For example, a network of machine learning–enabled Raman spectrometers was deployed across 10 Ontario community organizations and generated real-time drug-checking results from more than 7,700 samples over 14 months.¹⁹ Community drug-checking programs have also independently evaluated technologies including Scatr's Series One (portable Raman), Spectra Plasmonics' Amplifi ID (SERS), and Waters' RADIAN ASAP mass spectrometry system.²⁰ These emerging DCTs aim to improve detection of low-concentration substances, contaminants and NPSs, and complex drug mixtures in real-world harm-reduction settings.^13-18 However, published information on the implementation of emerging DCTs across many regions and communities in Canada remains limited, representing an important evidence gap.^6,13

How Do Emerging DCTs Work?

This section summarizes the emerging DCTs identified in the literature. These include emerging spectroscopy technologies (portable Raman, SERS, near-infrared [NIR] spectroscopy, and IR spectroscopy combined with machine learning), portable mass spectrometry, sensor-based technologies, and approaches that combine multiple emerging DCTs. For each technology, we describe how it works and provide a high-level summary of the available evidence. We did not systematically synthesize or critically appraise the evidence.

Sensitivity, specificity, and LOD are commonly used measures to describe the performance of DCTs. Sensitivity refers to the ability of a technology to correctly identify the presence of a substance when it is actually present. A highly sensitive test will detect even small amounts of the target substance.²¹ Specificity refers to the ability of a technology to correctly identify when a substance is not present. A highly specific test will rarely give a false-positive for a substance that is not actually in the sample.²¹ LOD refers to the smallest amount or concentration of a substance that a technology can reliably detect.²² In this Horizon Scan, reported sensitivity, specificity, and LOD values are relative to the reference standard or comparator used in each individual study and should be interpreted within that study's context.

Refer to Table 2 and Table 3 in Appendix 2 for a detailed summary of studies that reported accuracy outcomes or LOD for emerging DCTs.

Emerging Spectroscopy Technologies

Portable Raman

How Does It Work?

Portable Raman spectroscopy uses laser light to measure the vibrational characteristics of molecules. When the laser interacts with a sample, a small portion of the scattered light shifts in wavelength according to the chemical composition of the substance. The instrument compares the resulting spectrum with reference libraries to identify compounds present in the sample.²³

Several portable Raman systems have been developed for drug-checking and forensic applications.²⁴ Traditional Raman instruments typically rely on spectral matching against onboard libraries, whereas newer systems incorporate machine-learning algorithms to improve identification of complex mixtures.¹⁹ For example, the Scatr Series One, a portable Raman device developed in Canada, combines Raman with cloud-based machine-learning analysis to identify substances commonly found in Canada’s unregulated drug supply.¹³

Evidence

We identified 2 studies, both evaluating the Scatr Series One in Canadian drug-checking settings.^19,20 Overall, results from these initial studies reported the technology correctly identified (sensitivity) most samples containing fentanyl, cocaine, methamphetamine, and 3,4-methylenedioxymethamphetamine (MDMA). In the largest study (1,083 samples), the device detected fentanyl, cocaine, methamphetamine, and MDMA in more than 90% of samples containing these substances (sensitivity) and correctly identified nearly all samples without them (specificity). However, performance was lower for benzodiazepines (72%) and xylazine (59%).¹⁹ Findings from Toronto’s Drug Checking Service showed that the technology correctly identified (sensitivity) fentanyl, cocaine, methamphetamine, or MDMA in up to 90% of samples. However, sensitivity ranged from less than 10% to 48% for some fentanyl analogues, benzodiazepines, veterinary tranquilizers, and some emerging substances (e.g., medetomidine and bromazolam).²⁰

Portable Raman technologies may have a reduced ability to detect substances present at low concentrations. One study reported a fentanyl LOD of approximately 1% in reference standard mixtures containing caffeine and/or mannitol.¹⁹ Toronto’s Drug Checking Service reported fentanyl detection at concentrations as low as 0.10%, although detection was inconsistent at low concentrations.²⁰

Surface-Enhanced Raman Spectroscopy

How Does It Work?

SERS is an emerging form of hand-held Raman spectroscopy designed to address Raman’s limitations, including difficulty detecting low-concentration substances or identifying components within mixtures.^25-27 It uses metallic nanostructures (typically gold, silver, or copper) to amplify Raman signals from drug molecules. The metallic nanostructures also reduce fluorescence interference that can affect standard Raman measurements.²⁸ Commercial SERS platforms, such as the Spectra Plasmonics Amplifi ID, use disposable nanoparticle-coated substrates that are mixed with a sample before Raman analysis to enhance detection of trace substances.²⁰

Evidence

These initial studies conducted in drug-checking settings reported varying performance across target substances. Sensitivity ranged from 79% to 96% for fluorofentanyl, bromazolam, xylazine, and etizolam, while specificity ranged from 86% to 96%.^18,29,30 Findings from Toronto's Drug Checking Service evaluating the Spectra Plasmonics Amplifi ID suggest that performance may vary substantially by substance. Correct identification rates (sensitivity) were 84% for fentanyl and 64% for para-fluorofentanyl, but lower for bromazolam (48%) and xylazine (35%).²⁰

Available studies suggest that SERS can detect fentanyl, fentanyl analogues, xylazine, and etizolam at low concentrations.^18,31 Reported LODs varied across substances, sample types, and experimental conditions. Several studies reported LOD in the ng/mL or nanomolar range, including 0.20 ng/mL for pure fentanyl, 0.35 ng/mL for 4-fluoroisobutyryl fentanyl, and 4.4 ng/mL for cyclopropyl fentanyl.^31,32 Fentanyl was also detected (LOD) in heroin and cocaine mixtures at concentrations as low as 0.05% and 0.10%, respectively.³¹

Portable NIR Spectroscopy

How Does It Work?

The literature describes portable NIR systems as rapid, nondestructive, and suitable for field deployment. NIR helps operators identify sample composition by producing a spectral “fingerprint.” The device produces the fingerprint by measuring how molecules in the sample absorb NIR light, typically across wavelengths from approximately 740 to 2,600 nm.

Recent literature explores the use of chemometric and machine-learning models to identify and sometimes quantify substances.³³ The studies applied machine-learning approaches to improve detection of complex mixtures.^33-35

Examples of available portable NIR devices in the literature include the MicroNIR OnSite-W 1700, Powder Puck, SCiO handheld spectrometer, and smartphone-integrated NIR sensors.^35,36

Evidence

From initial results of the identified studies, portable NIR technologies have demonstrated sensitivity and specificity above 90% for identifying cocaine, heroin, methamphetamine, MDMA, cannabis, and ketamine.^33-38 Across studies, more than 90% of samples were correctly identified (sensitivity), with some studies reporting sensitivity approaching 100% for cocaine, heroin, and cannabis.^33-38 These initial findings suggest that machine-learning and chemometric approaches (statistical and computational methods used to analyze chemical data) may improve the analysis of complex drug samples.³³ However, most studies evaluated seized drug samples in forensic settings rather than community drug-checking services. As a result, the performance, feasibility, and operational requirements of these technologies in real-world harm-reduction settings remain uncertain. Limited evidence exists on the ability of portable NIR technologies to detect low-concentration substances, coloured samples, and complex mixtures commonly encountered in community drug-checking settings.^33-38

IR Spectroscopy With Machine Learning

How Does It Work?

IR spectroscopy measures how a sample absorbs infrared light and produces a spectrum that reflects its chemical composition. Drug-checking services often use IR spectroscopy because it is rapid, portable, and nondestructive, and requires minimal sample preparation.

Interpreting spectra from complex drug mixtures can be challenging, particularly when substances are present at low concentrations.^15,39 Machine learning can help find patterns in IR spectra that may not be obvious through conventional library matching or manual interpretation. Machine-learning models are trained using spectra from samples with known compositions and can then identify substances or estimate their concentrations in unknown samples.^15,40

Researchers have evaluated several machine-learning approaches, including neural networks, random forests, k-nearest neighbours, and regression models. These approaches may improve detection of low-concentration substances and automate parts of the analysis process. They may also support more consistent interpretation across drug-checking services.^15,40

Evidence

Overall, these emerging studies suggest that machine learning may improve the detection and quantification of substances in complex drug mixtures.^15,39,40 In drug-checking settings, machine-learning models correctly identified (sensitivity) bromazolam and para-fluorofentanyl in 87% to 93% of samples, depending on the model used.⁴⁰ Another study reported that machine learning–assisted IR spectroscopy correctly identified (sensitivity) MDMA in 79% of samples and fluorofentanyl in 61% of samples, while correctly ruling out (specificity) these substances in 97% to 100% of samples that did not contain them.¹⁵ Machine-learning models have also been used to estimate fentanyl concentrations in drug mixtures, with 1 study reporting a fentanyl LOD of 0.35%.³⁹

Portable Mass Spectrometry

How Does It Work?

Portable mass spectrometry technologies include portable GC-MS, miniature mass spectrometry systems, and portable paper spray mass spectrometry (PS-MS). Although these technologies use different sampling and ionization approaches, they all rely on mass spectrometry to identify substances based on their unique chemical signatures. FTIR, Raman, and NIR spectroscopy estimate substance identity based on interactions with light. In contrast, mass spectrometry analyzes the chemical components directly, supporting more specific substance identification.^16,41

GC-MS separates a sample into its individual components before mass spectrometry analysis, which may improve detection of contaminants and NPSs in complex drug mixtures.¹⁶ In contrast, PS-MS analyzes samples directly from a paper substrate using a solvent and electrical voltage to generate ions for mass spectrometry analysis, reducing sample preparation and analysis time.^41,42

Recent advances have enabled smaller and portable mass spectrometry systems that can be used outside traditional laboratory settings. These technologies often include automated substance libraries and software-assisted analysis to support faster interpretation of results while maintaining portability.⁴¹

Evidence

Results from initial studies suggest portable mass spectrometry may detect low-concentration substances and characterize complex drug mixtures.^{14,17,20,42,43} In 1 study, portable GC-MS correctly identified (sensitivity) 100% of heroin and cocaine samples and detected fentanyl in 95% of samples.¹⁴ Another portable GC-MS system achieved a 90% detection rate (sensitivity) for synthetic opioids in screening-kit samples.⁴² However, performance varied by substance and technology. For example, Toronto’s Drug Checking Service reported high correct identification rates (sensitivity) for cocaine, methamphetamine, and MDMA (95% to 100%), but lower rates for fentanyl-related compounds (64% to 84%), benzodiazepines (48%), and xylazine (35%).²⁰

Evidence from community drug-checking settings showed that portable mass spectrometry identified substances that FTIR and test strips missed, particularly at low concentrations.⁴³ They also demonstrated the ability to detect fentanyl, fluorofentanyl, carfentanil, and etizolam at ng/mL concentrations, suggesting potential utility for identifying contaminants and NPS in drug samples.¹⁷

Sensor-Based Technologies

How Do They Work?

Sensor-based technologies are emerging portable analytical tools. They combine high analytical sensitivity with compact, field-deployable instrumentation and relatively low operating costs. Emerging sensor-based technologies increasingly incorporate nanomaterials; molecularly imprinted polymers; graphene derivatives; carbon nanotubes; MXenes; and metal-organic frameworks to enhance sensitivity, selectivity, and analytical performance.^44-46

Sensor-based technologies use different sensing mechanisms to detect specific target substances or classes of substances (e.g., heroin, fentanyl, MDMA). Electrochemical sensors detect substances by measuring electrical signals generated when a target analyte interacts with an electrode surface. When an electrical potential is applied, oxidation or reduction reactions occur, producing measurable changes in current, voltage, or impedance that can be used to identify and quantify the target substance.^45,47,48 DoseCheck is a drug-checking technology developed in Canada that applies voltammetry, an electrochemical sensing technique, in a portable, smartphone-connected platform to estimate the composition and concentration of substances in drug samples.¹³ Immunosensors are another type of sensor-based technology that use antibodies, aptamers, or other molecular recognition elements to selectively bind target analytes. They convert a binding event into an electrochemical, optical, or colourimetric signal to measure and interpret.^45,46 Aptamer-based biosensors are a related emerging technology that uses synthetic DNA or ribonucleic acid (RNA) molecules rather than antibodies and may offer advantages in stability, shelf life, reproducibility, and manufacturing scalability.⁴⁵

Although these technologies show promise for portable drug checking, most are at the proof-of-concept or early implementation stage. No eligible studies evaluating DoseCheck technologies were identified in this Horizon Scan.

Evidence

Results from 2 initial studies suggest electrochemical sensors may provide sensitive and targeted detection of priority substances, including opioids and MDMA, with low limits of detection and the ability to quantify analytes.^49,50 In 1 study, an electrochemical sensor correctly identified (sensitivity) 88% of heroin samples and correctly identified (specificity) 100% of samples that did not contain heroin.⁵⁰ In another study conducted in a festival drug-checking setting, the NarcoReader correctly identified (sensitivity) 100% of MDMA-positive samples and had a 70% specificity for correctly identifying samples that did not contain MDMA.⁴⁹

Both studies reported low LOD.^49,50

Portable Robotic High-Performance Liquid Chromatography

How Does It Work?

Portable robotic high-performance liquid chromatography (HPLC) systems are designed to automate sample preparation and chromatographic analysis in a compact format. Like conventional HPLC systems, they separate compounds before detection, which may improve the identification of complex mixtures and low-concentration substances while expanding access to chromatographic analysis outside traditional laboratories.¹³

Evidence

We did not identify any eligible studies evaluating this technology.

Combining Emerging DCTs

How Does It Work?

Multitechnology approaches combine 2 complementary analytical techniques in sequence to improve substance identification and address limitations of individual technologies. Both identified studies paired Raman spectroscopy with portable mass spectrometry, allowing samples to be analyzed using 2 independent methods. Raman spectroscopy provides rapid molecular fingerprinting and can be performed directly on samples, whereas mass spectrometry requires sample preparation but provides highly specific chemical identification and improved sensitivity for detecting low-concentration substances.^51,52

Evidence

Two identified studies combined Raman-based technologies with portable mass spectrometry to leverage the complementary strengths of each approach.^51,52 One study integrated SERS and paper spray ionization mass spectrometry (PSI-MS) on a single platform using a dual-purpose plasmonic paper substrate, allowing the same sample to be analyzed by both techniques.⁵² The other combined a hand-held Raman spectrometer with a transportable mass spectrometer, using Raman spectroscopy for initial screening and mass spectrometry for confirmatory analysis when contaminants or low-concentration substances reduced Raman performance.⁵¹

These initial studies suggested that combining complementary technologies may improve substance identification by pairing rapid screening with more sensitive confirmatory analysis.^51,52 Both studies reported high analytical performance in forensic evaluations.^51,52 In 1 study, the combined approach reported 100% overall accuracy. In the other study, the approach reported a 99.8% correct identification rate.^51,52 However, neither study provided a direct comparison of sensitivity and specificity against each individual technology alone.

Refer to Table 3 in Appendix 2 for a detailed summary of studies on combining emerging DCTs.

Considerations for Emerging Technologies

This section summarizes key considerations for emerging DCTs beyond analytical performance. Table 1 compares the technologies across implementation and operational factors from the identified evidence to help understand their potential advantages and limitations. The considerations are organized into the following categories: factors that may affect performance, portability and infrastructure requirements, analysis time, sample preparation requirements, sample destruction, dependence on libraries or software, artificial intelligence (AI), and other considerations. We extracted the information from the identified evidence, and not all emerging DCTs had information representing each consideration category. Whether these considerations represent an advantage or a limitation may vary depending on the characteristics, intended use, and implementation context of each technology.

Across emerging DCTs, common advantages included rapid analysis, portability, and reduced infrastructure requirements compared with laboratory-based methods.^{17,19,27,37,49} Among them, SERS, portable mass spectrometry, sensing technologies, and IR combined with machine learning may also improve LOD of low-concentration substances, contaminants and NPSs, and complex drug mixtures.^14,50,53 The ability to identify NPSs varied across technologies. Portable mass spectrometry and multitechnology approaches may identify NPSs.^20,51 Spectroscopy-based technologies and portable robotic HPLC may also identify NPSs when supported by updated reference libraries or machine-learning models.^{13,16,19,20,40} In contrast, sensor-based technologies are generally limited to detecting predefined substances or classes of substances.⁵⁰ Common limitations varied across technologies and included dependence on reference libraries, calibration models, or training datasets; challenges identifying novel substances and complex mixtures; and the need for ongoing updates to software, machine-learning models, or reference libraries.^{15,37,40,42,45} Although evidence is limited, implementation and sustainability may be influenced by maintenance, repair processes, technical support, software requirements, reference library updates, ongoing quality assurance, and training.⁵⁴ Additional data and evaluation of these data are needed to understand their implementation and operational requirements.^17,49,51,52

Table 1: Summary of Considerations for Emerging DCTs

AI = artificial intelligence; DCT = drug-checking technology; HPLC = high-performance liquid chromatography; IR= infrared; NIR = near-infrared; NPS = new psychoactive substance; SERS = surface-enhanced Raman spectroscopy.

Note: If a consideration category is not listed for a technology, that consideration was not described in the identified literature.

^aAnalysis time refers to the time required for a technology to analyze a prepared sample and generate a result.

What Does It Cost?

We identified limited cost information for emerging DCTs, and the costs vary across technologies. Software requirements, consumables (e.g., test strips, reagents), maintenance, training needs, reference library management, technical support, staffing, and administrative requirements may all influence implementation and operational costs.^6,9 Manufacturer survey data from the previous review by CDA-AMC⁹ showed that purchase agreements for spectroscopic technologies often included software updates, library management, technical support, online support, and warranties. Operational requirements and costs varied across technologies. Manufacturers also reported that training requirements ranged from minimal to moderate, depending on the technology and expertise required for spectral interpretation and data analysis.⁹

• Portable Raman devices have been reported to cost approximately US$10,000 to US$60,000, although additional expenses may include software licences, database subscriptions, library development, calibration, maintenance, cloud-based analytical support, staff training, and administrative activities related to quality assurance and library management.⁵⁹ Scatr Series One has been estimated to cost approximately CA$120,000 (approximately US$85,000) plus annual software and service fees.¹³

• SERS may be a lower-cost alternative to laboratory-based analytical methods such as GC-MS and LC-MS, although specific device costs were rarely reported. Spectra Plasmonics’ Amplifi ID, a portable SERS system, has been estimated to cost approximately CA$20,000 to CA$30,000 and requires ongoing purchases of disposable SERS substrates as well as staff time for training and operation.¹³

• NIR spectroscopy may offer cost advantages because hand-held devices are generally smaller and less expensive than many FTIR or Raman-based systems, although implementation requires development and maintenance of appropriate reference databases and chemometric models.³⁵ One report estimated the cost of the NIRLAB FIELDLAB at €10,800 (approximately CA$17,300) for the instrument, with an additional €1,200 (approximately CA$2,000) annual software subscription per application area or library.¹³ Ongoing costs may include software support, database maintenance, staff training, and administrative oversight.

• IR spectroscopy combined with machine learning may require lower implementation costs than some other emerging technologies because machine-learning models can be implemented using existing IR infrastructure, reducing the need for additional hardware investments.^15,40 However, there may be additional costs associated with model development, validation, software integration, database management, technical support, staff training, and model maintenance.^15,40

• Portable mass spectrometry systems are likely the most expensive among emerging technologies, with 1 estimate suggesting acquisition costs of approximately CA$200,000 to CA$300,000. Additional costs may include software development, calibration, maintenance, consumables, technical support, specialized staffing, and administrative requirements related to quality management and instrument operation.¹³

• Sensor-based technologies have been described as relatively low-cost alternatives to laboratory-based analytical methods, although specific device costs were not reported.^44,46,48 These technologies generally require less specialized infrastructure and may have lower maintenance, staffing, training, and administrative requirements than laboratory-based approaches.

The available cost estimates presented previously may not reflect current market conditions because the cost data were collected and reported at different times, including earlier years. Actual costs are likely to vary over time because of inflation, technological advancements that may affect production costs and product capabilities, software licensing fees, and changes in manufacturing and distribution costs. Organizations considering implementation should seek updated pricing from vendors.

Who Might Benefit?

People who use unregulated drugs are intended as the primary beneficiaries of drug-checking services. By providing information about the composition of unregulated substances, drug-checking services can help individuals identify unexpected or potent compounds, contaminants, and NPSs to support informed substance use. This may be important given the increasing prevalence of highly potent synthetic opioids, benzodiazepines, veterinary sedatives, and novel psychoactive substances in the Canadian unregulated drug supply.^6,9

Drug-checking information may also benefit harm-reduction workers, health care providers, public health authorities, and community organizations by improving awareness of changes in the local unregulated drug supply. This information may support efforts to prevent deaths related to toxicity. In addition, surveillance data generated through drug-checking programs can inform public health alerts, service planning, policy development, and resource allocation.⁶

Emerging DCTs may be particularly valuable for communities that currently have limited access to comprehensive drug-checking services — including rural, remote, and underserved regions — as well as mobile and event-based harm-reduction programs. Technologies that improve sensitivity, broaden substance detection, and reduce reliance on centralized laboratory infrastructure may expand access to timely drug-checking information across diverse settings.^9,14,60

More broadly, the topic aligns with jurisdictional priorities to strengthen harm-reduction efforts and address the health and social impacts of the toxic drug crisis, including pressures on emergency departments and other health services.

Looking Ahead

This Horizon Scan provides a high-level summary of emerging DCTs, including portable Raman, SERS, NIR spectroscopy, IR spectroscopy combined with machine learning, portable mass spectrometry, sensor-based technologies, and combining emerging DCTs. A common trend across these technologies is the development of smaller, more portable, and field-deployable devices that may expand access to drug-checking services in community, mobile, rural, and remote settings.

Overall, the available evidence from these initial studies suggests that some emerging DCTs may improve the detection and identification of trace substances, contaminants, NPSs, and complex drug mixtures. Key findings include the following:

• SERS, machine learning–assisted approaches, portable mass spectrometry, and combinations of emerging technologies showed positive initial findings in early evaluations.

• No single technology appeared capable of meeting all drug-checking needs.

• Each technology involved trade-offs related to analytical performance, portability, analysis time, operational complexity, training requirements, dependence on reference libraries or databases, and cost.

• Combining emerging technologies with complementary established technologies may help address some limitations of individual technologies and improve confidence in substance identification.

Future research may focus on:

• evaluating emerging DCTs in community drug-checking services and other real-world harm-reduction settings

• assessing implementation considerations, including sustainability, maintenance requirements, staffing needs, workforce training, and operational costs

• comparing multitechnology approaches with single-technology approaches

• exploring the social, ethical, economic, and regulatory implications of adopting emerging DCTs.

Drug-checking services should also consider how they communicate results to people who use these services when implementing emerging DCTs. The British Columbia Centre on Substance Use published guidance in August 2026 that recommends communicating drug-checking results and harm-reduction information clearly and in a person-centred manner, including setting expectations about the capabilities and limitations of DCTs.⁶¹

This Horizon Scan is not a comprehensive systematic review and did not include a formal critical appraisal of the identified evidence. The available evidence provides early insights into the performance and potential use of emerging technologies in real-world drug-checking programs. However, evidence remains limited and additional research is needed to evaluate their implementation, long-term performance, and use across diverse settings. Although we searched multiple databases and supplemented these searches with handsearching, citation searching, and expert input, some relevant studies may have been missed. We also identified limited evidence on implementation; economic considerations; acceptability among people who use drugs and service providers; and broader social, ethical, and regulatory considerations.

Considerations for Decision- or Policy-Making

In addition to analytical performance, decision-makers may wish to consider implementation factors, such as training requirements, maintenance and repair processes, software and library support, ongoing operational costs, and the ability of technologies to adapt to emerging substances.¹⁰

Understanding the limitations of individual technologies — including sensitivity, specificity, and LOD; reference library coverage; and their ability to detect low-concentration, novel, or unexpected substances — is important for accurate interpretation and communication of results. Emerging evidence also suggests that combining complementary technologies may improve confidence in substance identification and may have potential applications in the analysis of complex drug samples.^16,45 Future standards for drug-checking services could consider the use of complementary technologies. In forensic drug analysis, the American Society for Testing and Materials and the Scientific Working Group for the Analysis of Seized Drugs recommend 2 independent analytical techniques for substance identification, an approach that may be relevant given the increasing complexity of the unregulated drug supply.^62,63

AI and machine-learning approaches are increasingly being explored to support the analysis and interpretation of data generated by DCTs.^19,39,40 These approaches may improve the ability of some technologies to identify substances, interpret complex spectra, and detect patterns in large datasets.^19,40 However, integrating AI into drug-checking services may introduce additional considerations and underscore existing considerations, including:

• data privacy and management (e.g., appropriate use of collected information, safeguarding service user autonomy and confidentiality, preventing unauthorized access or misuse, and data sovereignty)

• technical security (e.g., protecting systems and data from unauthorized access or security breaches)

• transparency, explainability, and interpretability (i.e., the extent to which an AI-enabled system’s behaviour, logic, and outputs can be understood and meaningfully interpreted by end users)^64,65

• control and user autonomy (i.e., the extent to which individuals and organizations can meaningfully influence, monitor, and challenge the use of outputs from AI-enabled digital health technologies)⁶⁶

• algorithmic bias (when AI-enabled systems perform differently across various individuals or groups, leading to outcomes that are unfair, inaccurate, or inequitable)⁶⁷

• postdeployment monitoring.⁶⁸

These considerations may influence trust, accountability, and the responsible implementation of AI-supported DCTs. A detailed assessment of AI governance and digital health technology considerations was outside the scope of this Horizon Scan. Further research is needed to understand how machine learning–supported approaches can be implemented in ways that maintain trust and align with the goals of harm-reduction services. Forthcoming national guidance for AI in mental health and substance use health care may also provide relevant guidance on the safe, equitable, and ethical evaluation and implementation of AI-enabled health care services and solutions.⁶⁹

Broader considerations related to regulation and health technology assessment may also be relevant. Limited regulatory oversight and the absence of standardized evaluation processes may create challenges for organizations selecting technologies for harm-reduction services. Independent evaluation and transparent assessment processes may help support informed technology adoption and implementation decisions.^10,70

This Horizon Scan complements the existing CDA-AMC health technology review by focusing on emerging DCTs that have not yet been widely adopted or fully evaluated. While the health technology review assessed the evidence for currently available technologies, this Horizon Scan highlights innovations that may address existing limitations, such as broader substance detection, improved sensitivity for trace compounds, faster turnaround times, and greater feasibility for community and remote settings.¹⁰

As the unregulated drug supply continues to evolve, the ability of both established and emerging DCTs to identify a broad range of substances and adapt to newly emerging compounds remains important. Continued independent evaluation will be needed to determine their role in future drug-checking services.^10,70

Acknowledgements

The authors thank members of the National Drug Checking Working Group for providing input on the scope of this Horizon Scan, priority technologies and substances of interest, and practical considerations related to the implementation of DCTs in Canada. The authors also thank the individuals and organizations who provided feedback during the public review process. Their contributions helped inform the development and refinement of this report.

References

1.Public Health Agency of Canada. Opioid- and stimulant-related harms in Canada: key findings. 2026. Accessed 2026-07-10. https://health-infobase.canada.ca/substance-related-harms/opioids-stimulants/maps.html

2.Toronto's Drug Checking Service. 108 samples checked: May 30 – June 12, 2026. 2026. Accessed 2026-05-13. https://drugchecking.community/report/may-30-june-12-2026/

3.Schiller EY, Goyal A, Mechanic OJ. Opioid overdose. 2017.

4.Toronto's Drug Checking Service. Performance assessment: Medetomidine test strips. 2026. Accessed 2026-05-13. https://drugchecking.community/resource/medetomidine-test-strips/

5.Krotulski AJ, Mata DC, Smith CR, et al. Advances in analytical methodologies for detecting novel psychoactive substances: a review. J Anal Toxicol. 2025;49(3):152-169. PubMed

6.Toronto's Drug Checking Service. Drug checking technologies overview. 2023. Accessed 2026-05-12. https://drugchecking.community/resource/drug-checking-technologies-overview/

7.Canadian Centre on Substance Use and Addiction. Drug Checking. 2024. https://www.ccsa.ca/sites/default/files/2024-10/Drug-Checking-Evidence-Brief-en.pdf

8.Gray L. City overdose alert tied to loss of drug checking machine. 2026. 980 CJME. Accessed 2026-05-11. https://www.cjmecom/2026/04/15/city-overdose-alert-tied-to-loss-of-drug-checking-machine/

9.CDA-AMC. Drug-Checking Technologies to Detect Compositions of Unregulated Substance Samples. Canadian Journal of Health Technologies. 2026;6(1) doi:10.51731/cjht.2026.1328

10.Thompson H, McDonald K. Considerations for Purchasing Drug Checking Technologies: Perspectives from Toronto's Drug Checking Service. Research Support, Non-U.S. Gov't. Int J Environ Res Public Health. 07 31 2023;20(15):31. doi:10.3390/ijerph20156486

11.Government of Canada. Controlled Drugs and Substances Act. 1996. Accessed 2026-05-11. https://laws-lois.justice.gc.ca/eng/acts/C-38.8/

12.Ontario Drug Checking Community. Terms of Service (Version 5). 2026. May 2026. Accessed 2026-07-10. https://drugchecking.community/

13.British Columbia Centre on Substance Use. New Drug Checking Instruments in Canada: A Summary of Drug Checking Technology Developments. British Columbia Centre on Substance Use; 2024. https://drugcheckingbc.ca/wp-content/uploads/sites/4/2024/04/BCCSU_New_drug_checking_technologies2_2024.pdf

14.Gozdzialski L, Aasen J, Larnder A, et al. Portable gas chromatography-mass spectrometry in drug checking: Detection of carfentanil and etizolam in expected opioid samples. Research Support, Non-U.S. Gov't. Int J Drug Policy. 11 2021;97:103409. doi:10.1016/j.drugpo.2021.103409

15.Gozdzialski L, Hutchison A, Wallace B, Gill C, Hore D. Toward automated infrared spectral analysis in community drug checking. Drug Test Anal. Jan 2023;16(1):83-92. doi:10.1002/dta.3520 PubMed

16.Gozdzialski L, Wallace B, Hore D. Point-of-care community drug checking technologies: an insider look at the scientific principles and practical considerations. Review Research Support, Non-U.S. Gov't. Harm Reduction Journal. 03 25 2023;20(1):39. doi:10.1186/s12954-023-00764-3

17.Laxton J-C, Monaghan J, Wallace B, Hore D, Wang N, Gill CG. Evaluation and improvement of a miniature mass spectrometry system for quantitative harm reduction drug checking. International Journal of Mass Spectrometry. 2023;484:116976.

18.Martens RR, Gozdzialski L, Newman E, Gill C, Wallace B, Hore DK. Optimized machine learning approaches to combine surface-enhanced Raman scattering and infrared data for trace detection of xylazine in illicit opioids. Analyst. Feb 10 2025;150(4):700-711. doi:10.1039/d4an01496k PubMed

19.Aliyari E, Brown C, Van Boheemen M, et al. Street drug monitoring with networked spectrometers powered by machine learning: a pilot study in Ontario, Canada. Harm Reduction Journal. Feb 08 2026;23(1):08. doi:10.1186/s12954-026-01403-3

20.Toronto's Drug Checking Service. Performance assessment of Bruker’s Alpha II FTIR, Scatr’s Series One, Spectra Plasmonics’ Amplifi ID, and Waters’ RADIAN ASAP. 2026. Accessed 2026-05-12. https://drugchecking.community/resource/performance-assessment/

21.Shreffler J, Huecker MR. Diagnostic testing accuracy: Sensitivity, specificity, predictive values and likelihood ratios. StatPearls. StatPearls Publishing; 2023.

22.Guideline IHT. Validation of analytical procedures Q2 (R2). Vol. 1. 2022.

23.Gerace E, Seganti F, Luciano C, et al. On–site identification of psychoactive drugs by portable Raman spectroscopy during drug–checking service in electronic music events. 2019;38(1):50-56.

24.Wang Wt, Zhang H, Yuan Y. et al. Research Progress of Raman Spectroscopy in Drug Analysis. AAPS PharmSciTech. 2018;19:2921-2928. doi:10.1208/s12249-018-1135-8

25.Rana V, Canamares MV, Kubic T, Leona M, Lombardi JR. Surface-enhanced Raman Spectroscopy for Trace Identification of Controlled Substances: Morphine, Codeine, and Hydrocodone. JOFS. 2011;56:200-207. doi:10.1111/j.1556-4029.2010.01562.x

26.White P. SERRS Spectroscopy–a new technique for forensic science? Sci Justice. 2000;40(2):113-119.

27.Alonzo M, Alder R, Clancy L, Fu S. Portable testing techniques for the analysis of drug materials. WIREs Forensic Science. 2022/11/01 2022;4(6):e1461. doi:10.1002/wfs2.1461

28.Yu B, Ge M, Li P, Xie Q, Yang L. Development of surface-enhanced Raman spectroscopy application for determination of illicit drugs: Towards a practical sensor. Review. Talanta. Jan 01 2019;191:1-10. doi:10.1016/j.talanta.2018.08.032 PubMed

29.Gozdzialski L, Rowley A, Borden SA, et al. Rapid and accurate etizolam detection using surface-enhanced Raman spectroscopy for community drug checking. Research Support, Non-U.S. Gov't. Int J Drug Policy. 04 2022;102:103611. doi:10.1016/j.drugpo.2022.103611

30.Martens RR, Gozdzialski L, Newman E, Gill C, Wallace B, Hore DK. Trace Detection of Adulterants in Illicit Opioid Samples Using Surface-Enhanced Raman Scattering and Random Forest Classification. Anal Chem. Jul 30 2024;96(30):12277-12285. doi:10.1021/acs.analchem.4c01271 PubMed

31.Wang L, Vendrell-Dones MO, Deriu C, Dogruer S, de BHP, McCord B. Multivariate Analysis Aided Surface-Enhanced Raman Spectroscopy (MVA-SERS) Multiplex Quantitative Detection of Trace Fentanyl in Illicit Drug Mixtures Using a Handheld Raman Spectrometer. Appl Spectrosc. Oct 2021;75(10):1225-1236. doi:10.1177/00037028211032930 PubMed

32.Dogruer Erkok S, Hernandez E, Cruz J, Mebel AM, McCord B. Differentiating Structurally Similar Fentanyl Analogs by Comparing Density Functional Theory (DFT) Calculations and Surface-Enhanced Raman Spectroscopy (SERS) Results. Appl Spectrosc. Jul 2024;78(7):667-679. doi:10.1177/00037028241246010 PubMed

33.Saez Hernandez R, Soriano Hernandez S, Mazario-Garcia M, et al. A Novel Chemometric Local Approach for Qualitative and Quantitative Analysis of Cocaine, MDMA, and THC-Related Products: Method Application within the Mossos d'Esquadra (Catalan Regional Police). Anal Chem. Mar 31 2026;98(12):9250-9259. doi:10.1021/acs.analchem.5c07967 PubMed

34.Kranenburg RF, Verduin J, Weesepoel Y, et al. Rapid and robust on-scene detection of cocaine in street samples using a handheld near-infrared spectrometer and machine learning algorithms. Drug Test Anal. Oct 2020;12(10):1404-1418. doi:10.1002/dta.2895 PubMed

35.Fursman H, Morelato M, Chadwick S, et al. Development and evaluation of portable NIR technology for the identification and quantification of Australian illicit drugs. Forensic Sci Int. Sep 2024;362:112179. doi:10.1016/j.forsciint.2024.112179 PubMed

36.Kranenburg RF, Ou F, Sevo P, et al. On-site illicit-drug detection with an integrated near-infrared spectral sensor: A proof of concept. Talanta. Aug 01 2022;245:123441. doi:10.1016/j.talanta.2022.123441 PubMed

37.Coppey F, Becue A, Sacre PY, Ziemons EM, Hubert P, Esseiva P. Providing illicit drugs results in five seconds using ultra-portable NIR technology: An opportunity for forensic laboratories to cope with the trend toward the decentralization of forensic capabilities. Validation Study. Forensic Sci Int. Dec 2020;317:110498. doi:10.1016/j.forsciint.2020.110498 PubMed

38.Kranenburg RF, Ramaker HJ, Sap S, van Asten AC. A calibration friendly approach to identify drugs of abuse mixtures with a portable near-infrared analyzer. Drug Test Anal. Jun 2022;14(6):1089-1101. doi:10.1002/dta.3231 PubMed

39.Ramsay M, Gozdzialski L, Larnder A, Wallace B, Hore D. Fentanyl quantification using portable infrared absorption spectroscopy. A framework for community drug checking. J Vibrational Spectroscopy. 2021;114:103243.

40.Jai J, Gozdzialski L, Wallace B, Gill CG, Hore D. Neural Network-Based Detection of Adulterants in Opioid Samples Using IR Absorption Spectroscopy. Drug Test Anal. May 2026;18(5):619-624. doi:10.1002/dta.70050 PubMed

41.Xie Y, You L, Wang X, et al. Miniature mass spectrometry in drug and food analysis: Bridging laboratory and field applications. Review. TrAC - Trends in Analytical Chemistry. 01 Feb 2026;195:118594. doi:10.1016/j.trac.2025.118594

42.Leary PE, Kizzire KL, Chan Chao R, Niedziejko M, Martineau N, Kammrath BW. Evaluation of portable gas chromatography-mass spectrometry (GC-MS) for the analysis of fentanyl, fentanyl analogs, and other synthetic opioids. J Forensic Sci. Sep 2023;68(5):1601-1614. doi:10.1111/1556-4029.15340 PubMed

43.Karch L, Tobias S, Schmidt C, et al. Results from a mobile drug checking pilot program using three technologies in Chicago, IL, USA. Research Support, U.S. Gov't, Non-P.H.S. Research Support, U.S. Gov't, P.H.S. Drug Alcohol Depend. 11 01 2021;228:108976. doi:10.1016/j.drugalcdep.2021.108976

44.Hossain MI, Yi DK, Kim S. Recent Advances in Nanomaterial-Based and Colorimetric Technologies for Detecting Illicit Drugs and Environmental Toxins. Review. Applied Sciences (Switzerland). 01 Jan 2026;16(2):693. doi:10.3390/app16020693

45.Bojarska E, Zajaczkowski W, Furtak E, et al. Modern Methods for Detection of Fentanyl and Its Analogues: A Comprehensive Review of Technologies and Applications. Review. Molecules (Basel). Aug 31 2025;30(17):31. doi:10.3390/molecules30173577 PubMed

46.Anzar N, Suleman S, Singh Y, et al. The Evolution of Illicit-Drug Detection: From Conventional Approaches to Cutting-Edge Immunosensors-A Comprehensive Review. Review. Biosensors. Oct 03 2024;14(10):03. doi:10.3390/bios14100477

47.Truta FM, Cruz AG, Dragan AM, et al. Design of smart nanoparticles for the electrochemical detection of 3,4-methylenedioxymethamphetamine to allow in field screening by law enforcement officers. Drug Test Anal. Aug 2024;16(8):865-878. doi:10.1002/dta.3605 PubMed

48.Glasscott MW, Vannoy KJ, Iresh Fernando PUA, Kosgei GK, Moores LC, Dick JE. Electrochemical sensors for the detection of fentanyl and its analogs: Foundations and recent advances. Review. TrAC - Trends in Analytical Chemistry. 01 Nov 2020;132:116037. doi:10.1016/j.trac.2020.116037

49.Deconinck E, Polet MA, Canfyn M, et al. Evaluation of an electrochemical sensor and comparison with spectroscopic approaches as used today in practice for harm reduction in a festival setting - A case study: Analysis of 3,4-methylenedioxymethamphetamine samples. Drug Test Anal. Sep 2024;16(9):1031-1043. doi:10.1002/dta.3625 PubMed

50.Montiel NF, Mazurkow J, Van Echelpoel R, et al. Evaluation of an Innovative Portable Heroin Electrochemical Sensor for Empowering Forensic Laboratories. Evaluation Study. Drug Test Anal. Feb 2026;18(2):260-269. doi:10.1002/dta.70013 PubMed

51.Chang C, Monjardez G, Davidson JT. Assessment of a combined handheld Raman spectroscopy and transportable mass spectrometry approach for the analysis of seized drug mixtures. Forensic Sci Int. Jul 2025;372:112512. doi:10.1016/j.forsciint.2025.112512 PubMed

52.Burr DS, Fatigante WL, Lartey JA, et al. Integrating SERS and PSI-MS with Dual Purpose Plasmonic Paper Substrates for On-Site Illicit Drug Confirmation. Research Support, U.S. Gov't, Non-P.H.S. Anal Chem. 05 05 2020;92(9):6676-6683. doi:10.1021/acs.analchem.0c00562

53.Wilson NG, Raveendran J, Docoslis A. Portable identification of fentanyl analogues in drugs using surface-enhanced Raman scattering. Sensors and Actuators B: Chemical. 2021/03/01/ 2021;330:129303. doi:10.1016/j.snb.2020.129303

54.Thompson H, McDonald K. Considerations for Purchasing Drug Checking Technologies: Perspectives from Toronto’s Drug Checking Service. 2023;20(15):6486.

55.Smith E, Dent G. Modern Raman spectroscopy: a practical approach. John Wiley & Sons; 2019.

56.Dogruer Erkok S, Gallois R, Leegwater L, Gonzalez PC, van Asten A, McCord B. Combining surface-enhanced Raman spectroscopy (SERS) and paper spray mass spectrometry (PS-MS) for illicit drug detection. Talanta. Oct 01 2024;278:126414. doi:10.1016/j.talanta.2024.126414 PubMed

57.Kranenburg RF, Ramaker HJ, van Asten AC. On-site forensic analysis of colored seized materials: Detection of brown heroin and MDMA-tablets by a portable NIR spectrometer. Drug Test Anal. Oct 2022;14(10):1762-1772. doi:10.1002/dta.3356 PubMed

58.Foster SW, Xie X, Pham M, et al. Portable capillary liquid chromatography for pharmaceutical and illicit drug analysis. J Sep Sci. May 2020;43(9-10):1623-1627. doi:10.1002/jssc.201901276 PubMed

59.Harper L, Powell J, Pijl EM. An overview of forensic drug testing methods and their suitability for harm reduction point-of-care services. Review. Harm Reduction Journal. 07 31 2017;14(1):52. doi:10.1186/s12954-017-0179-5

60.Borden SA, Saatchi A, Vandergrift GW, Palaty J, Lysyshyn M, Gill CG. A new quantitative drug checking technology for harm reduction: Pilot study in Vancouver, Canada using paper spray mass spectrometry. Research Support, Non-U.S. Gov't. Drug Alcohol Rev. 02 2022;41(2):410-418. doi:10.1111/dar.13370

61.British Columbia Centre on Substance Use. Messaging: Sharing Drug Checking Results. Guidance on providing drug checking results with harm reduction information. BCCSU; 2026. https://drugcheckingbc.ca/wp-content/uploads/sites/4/2026/08/BCCSU_Drug-Checking-Messaging-Guide.pdf

62.ASTM International. ASTM E2329-17: Standard practice for identification of seized drugs. West Conshohocken (PA): ASTM International2017.

63.SWGDRUG. Scientific Working Group for the Analysis of Seized Drugs recommendations. Version 8.0. 2019.

64.Linardatos P, Papastefanopoulos V, Kotsiantis S. Explainable AI: A review of machine learning interpretability methods. Entropy. 2020;23(1):18. PubMed

65.Amann J, Blasimme A, Vayena E, Frey D, Madai VI, Consortium PQ. Explainability for artificial intelligence in healthcare: a multidisciplinary perspective. BMC medical informatics and decision making. 2020;20(1):310. PubMed

66.Canada Health Infoway. Toolkit for Implementers of Artificial Intelligence in Health Care. 2026. Accessed 2026-03-04. https://www.infoway-inforoute.ca/en/component/edocman/3998-toolkit-for-implementers-of-artificial-intelligence-in-health-care/

67.Hanna MG, Pantanowitz L, Jackson B, et al. Ethical and bias considerations in artificial intelligence/machine learning. Modern Pathology. 2025;38(3):100686. PubMed

68.World Health Organization. Ethics and governance of artificial intelligence for health: large multi-modal models. WHO guidance. World Health Organization; 2024.

69.Canadian Centre on Substance Use and Addiction. National Guidance for AI in Mental and Substance Use Health Care. CCSA. Accessed 15 August 2026, https://www.ccsa.ca/en/guidance-tools-resources/treatment-and-prevention/national-guidance-ai-mental-and-substance-use

70.New Zealand Government Ministry of Health Drug Checking. Regulation and legislation, Health providers and products the Ministry of Health regulates, and laws we administer. 2022. Accessed 2026-05-14.

71.Smith M, Logan M, Bazley M, et al. A Semi-quantitative method for the detection of fentanyl using surface-enhanced Raman scattering (SERS) with a handheld Raman instrument. J Forensic Sci. Mar 2021;66(2):505-519. doi:10.1111/1556-4029.14610 PubMed

Appendix 1: Methods

Please note that this appendix has not been copy-edited.

Project Planning

Prior to project initiation, CDA-AMC sought feedback from National Drug Checking Working Group members. We requested feedback on the project scope, priority technologies, implementation considerations, and evidence gaps. Feedback was received from individuals involved in drug-checking and harm-reduction services from multiple Canadian jurisdictions, including:

• a harm-reduction manager in Prince Edward Island

• a social worker with a mobile clinic in Quebec

• a drug-checking training coordinator in British Columbia

• someone who works in a leadership capacity at a drug-checking service in Ontario.

Their input provided insight into current drug-checking practices, implementation considerations, and evidence needs across Canada, helping to inform the scope and focus of this Horizon Scan.

Participants highlighted the importance of considering practical implementation factors in addition to analytical performance, including:

• device maintenance, repair processes, and service disruptions associated with equipment failures

• the distinction between training to operate a technology and training to deliver comprehensive drug-checking services

• the role of software, automated data interpretation, and reference libraries in supporting accurate results, including the ability to adapt to novel substances through library updates

• ongoing operational costs, including consumables (e.g., reagents, sample preparation materials), staffing, and administrative requirements, in addition to upfront equipment costs

Literature Search Strategy

An information specialist conducted a literature search on key resources including MEDLINE, Embase, the Cochrane Database of Systematic Reviews, the International HTA Database, the websites of health technology assessment agencies in Canada and major international HTA agencies, as well as a focused internet search. The search approach was customized to retrieve a limited set of results, balancing comprehensiveness with relevance. The search strategy comprised both controlled vocabulary, such as the US National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. Search concepts were developed based on the elements of the research questions and selection criteria. The search was completed on May 1, 2026, and limited to English-language documents published since January 1, 2011. The main search concepts were drug checking and emerging trends. Results from ClinicalTrials.gov were excluded and retrieval was limited to the human population. A second search using the same limiters was completed on May 6, 2026. The main search concepts were drug checking tests and illicit substances. The search strategies are available on request.

Authors also identified relevant studies through handsearching of reference lists, and suggestions from subject matter experts. They also conducted targeted web searching throughout the study selection process to find relevant publications that were not captured by database indexing or search strategies. They used the same eligibility criteria as records retrieved from database and grey literature searches.

Selection Criteria

One author screened search results and reviewed the full texts of potentially relevant publications. We included studies and reports that described an emerging DCT intended for detecting compositions of unregulated substances and contaminants. We also included primary studies, review articles, reports, and relevant grey literature that provided information on the performance, feasibility, implementation, costs, or potential impact of emerging DCTs.

The draft Horizon Scan underwent internal review by CDA-AMC staff, including Ethics and Scientific Advisors and Strategic Partner in Inclusion, Diversity, Equity, and Accessibility (IDEA). Following internal review, we posted the draft report for public feedback and shared with the National Drug Checking Working Group and other interested parties.

Appendix 2: Study Characteristics and Main Findings

Please note that this appendix has not been copy-edited.

Table 2: Summary of Characteristics and Findings of Studies on Emerging DCTs That Reported Accuracy Outcomes or Limit of Detection

FTIR = Fourier-transform infrared; FTS = fentanyl test strip; GC-MS = gas chromatography–mass spectrometry; GC-FID = gas chromatography with flame ionization detection; HCl = hydrochloride; HPLC-MS = high-performance liquid chromatography–mass spectrometry; HPMS = Hand-held high-pressure mass spectrometer; KNN = k-nearest neighbours; LC-HR-MS = Liquid Chromatography-High-Resolution Mass Spectrometry; LOD = limit of detection; LOQ = limit of quantification; MDA = 3,4-methylenedioxyamphetamine; MS = mass spectrometry; NA = not applicable; NIR = near infrared; NR = not reported; PS-MS = paper spray mass spectrometry; SERS = surface-enhanced spectroscopy; SHAP = shapely additive explanations.

^aThe study by Karch et al. (2021)⁴³ described the results of a mobile drug-checking service. It does not compare the accuracy of emerging DCTs to laboratory methods or gold standard tests.

Table 3: Summary of Characteristics and Findings of Studies on Combining Emerging DCTs That Reported Accuracy Outcomes or Limit of Detection

GC-MS = gas chromatography–mass spectrometry; LOD = limit of detection; MS = mass spectrometry; NA = not applicable; PSI-MS = paper spray ionization mass spectrometry; SERS = surface-enhanced Raman spectroscopy.