Toward a National System of Expanded Testing of Existing Urine Specimens: The Drug Outbreak Testing Service (DOTS)

Abstract

Background

NIDA’s National Drug Early Warning System (NDEWS) was established in 2014 with the mission of identifying and monitoring emerging drugs in the United States. Identification of emerging drugs has been complicated, however, by the rapid development of novel psychoactive substances such that users often cannot identify and report the drugs they have ingested. Biologic testing of urine, hair or blood is the only way to reliably identify the substances recently used. Unfortunately, the large number of up-to-date tests required is beyond the resources available to most organizations.

Methods

The DOTS study tested the feasibility of recruiting organizations to submit up to 25 de-identified urine specimens for testing for approximately 240 drugs, at no cost to them. The results were for epidemiologic purposes only and not for clinical use. Eleven sites who had questions about their patients or the results of their organization’s more limited urinalysis screens participated. These sites included drug treatment programs, medical examiners, hospitals and a criminal justice testing program.

Results

Extensive polydrug use and geographic differences in the drugs detected were found. All sites found the DOTS collaborating laboratory’s test results to be very useful for understanding the types of drugs being used recently and to assess the adequacy of their testing protocols.

Conclusions

The U.S. should consider establishing a program of expanded testing of already collected de-identified urine specimens in order to identify emerging drugs and track local patterns of use and availability.

Keywords:

• Drug surveillance
• drug testing
• toxicology
• urinalysis
• emerging drugs
• polydrug use

Introduction

The Coordinating Center for the National Institute on Drug Abuse’s (NIDA) National Drug Early Warning System (NDEWS) was launched in 2014 at the University of Maryland’s Center for Substance Abuse Research (CESAR) and maintained by CESAR until late 2020, when a new contractor assumed responsibility for the project. NDEWS was to be an updated and expanded version of the Community Epidemiology Workgroup started by NIDA in 1976. The mission of NDEWS is to help the United States to identify and monitor emerging drugs and drug trends and to facilitate information sharing. A variety of methods was used to achieve its mission, including the analysis of drug surveys and administrative data, monitoring drug reports in traditional and social media in 12 sentinel sites and conducting focused HotSpot studies to investigate local outbreaks.

Many of these data sources are limited because they rely on identifying drugs through self-reports and must assume that people can accurately recall and identify the drugs they have taken and are willing to disclose them. These assumptions may be less tenable because the number and types of drugs available has rapidly increased with the development of new psychoactive substances (NPS), such as synthetic opioids and cannabinoids, and because there are increasing reports that dealers may include other drugs and adulterants in their products without the knowledge of the user (Rosenblum et al., 2020).

Tests of biologic specimens such as hair, urine and blood can provide objective evidence of the specific drugs that a person has recently ingested. For this study, we analyzed only already collected urine specimens to assess recent drug use. Other biological specimens such as blood, plasma, oral fluid, or hair can also detect drug use but were unavailable for this study. Each of these measures has different windows of detection. The current study was based on the strategy of re-testing urine specimens that had already been collected by participating organizations.

The national Drug Use Forecasting (DUF) program and its subsequent Arrestee Drug Abuse Monitoring (ADAM) program (later redesigned and named ADAM II) were the first multi-state monitoring programs to track drug trends through urinalysis (DuPont & Wish, 1992; Office of National Drug Control Policy, 2014; Wish, 1990). Voluntary urine specimens were collected from samples of arrestees in urban booking facilities and tested for up to 10 drugs. DUF was based on the findings from the Washington, DC (DC) arrestee urine testing program that had demonstrated that the heroin epidemic in DC appeared in arrestees’ urinalysis results one to 1.5 years before it showed up in other community indicators. It was hypothesized that persons arrested for committing illegal acts might be those most likely to use newly available illegal drugs. The epidemics in DC of cocaine use in the 1980s and marijuana use among youths in the 90 s showed up early in the trends from the DC pretrial testing program of adults and juveniles.

After the ADAM II program was discontinued in 2014, researchers at CESAR developed a modified program for expanded testing of urine specimens already collected from probationers as part of routine drug testing programs. Instead of sending researchers into booking facilities to obtain urine specimens from arrestees, the new strategy collected de-identified urine specimens already collected by criminal justice drug testing programs from probationers and submitted them to a collaborating laboratory for expanded re-testing. Criminal justice drug testing programs routinely test specimens for a small number of drugs, often those most commonly involved in past epidemics. The Community Drug Early Warning System (CDEWS) that CESAR subsequently developed implemented a new strategy of sending already collected urine specimens to a collaborating laboratory for testing for over 240 drugs (CDEWS, 2020). CDEWS eventually collected specimens from over 15 sites, including persons tested by criminal justice programs, drug treatment programs, and hospitals. Most CDEWS sites submitted a one-time sample of 100–200 de-identified specimens in order to learn what drugs their population was using and to gauge the adequacy of their own more limited testing protocols.

The CDEWS results documented the extreme variety of metabolites found in the urines, as well as large geographic variations in the drugs detected. The wide variation in drugs found across the country suggested the value of obtaining local measures of drug use. National indicators would likely mask these regional differences. We had received limited funding from the Office of National Drug Control Policy (ONDCP) to conduct CDEWS, and we suspected there was a need to make available the extensive urine testing capabilities of our collaborating laboratory to more programs with limited testing resources for use as an epidemiologic tool to measure drugs being used locally. We subsequently designed the Drug Outbreak Testing Service (DOTS) study in 2017 as part of our NIDA-funded NDEWS Coordinating Center project. DOTS increased access to expanded testing to a broader spectrum of programs and enabled us to test the feasibility of obtaining a small number of de-identified urine specimens (up to 25) from interested treatment programs, medical examiner/coroner offices, and hospitals, and testing them for a large up-to-date panel of substances. DOTS would provide an objective measure of the drugs recently used by a small sample of persons tested by the participating organization. It was hoped that the results could be used to help the organization to detect the presence of drugs that their limited testing could not and to suggest research questions for future study. NDEWS had supported several research studies that followed up on urinalysis results (Hoffer et al., 2019; Lopez, 2019; Sorg et al., 2017).

Methods

Our prior research had suggested to us that having organizations submit up to 25 specimens would be manageable and give a useful picture of the types of drugs recently used. Recruitment materials were sent to prospective sites using several methods including, advertising the DOTS study on the NDEWS website, Network, and Twitter account, and distributing flyers and business cards at public health and substance use conferences. We had the most success with recruiting sites through the NDEWS virtual network of over 1600 participants.

As this was a feasibility study, only a small number of sites were recruited to participate. A total of 39 locations were identified as potential sites, of which 11 were determined to be eligible and participated. The primary reasons for excluding a site from the study included: 1) site was seeking clinically useful information about the person rather than de-identified information useful for epidemiologic purposes; 2) site sought analysis of a larger number of specimens than could be accommodated through DOTS; 3) site had an inadequate number of urine specimens available for testing from persons likely to be using drugs; 4) site had inadequate staffing to manage the tasks required for participation; and 5) site was unable to obtain administrative approval to carry out the study. To facilitate recruitment, a site participation agreement was developed to specify the details of study participation. Future efforts to implement this type of system can enhance participation by specifying the exact inclusion and exclusion criteria in the early stages of recruitment.

Site eligibility requirements included: 1) experiencing a drug-related incident or a desire to understand inconsistent or inconclusive drug test results; 2) able to submit up to 25 de-identified urine specimens that have already been routinely collected for epidemiologic, not clinical use; 3) can provide specific information on how the urine specimens would be selected and brief descriptive demographic information about the persons who provided the urine specimens; and 4) willing to discuss the implications of the DOTS test results with DOTS staff and to review drafts of their site’s summary for a special DOTS Bulletin. Once a site met the eligibility requirements and appeared likely to benefit from participating in DOTS, the site obtained Institutional Review Board (IRB) approval. CESAR received approval for the DOTS protocol from the University of Maryland’s IRB. A special exception to the above rules occurred for sites submitting specimens from deceased persons, as human subject and privacy regulations are less rigid when conducting research with decedents (Office for Human Research Protections, 2011).

CESAR staff instructed each site on how to collect, label, and ship urine specimens. Each site created a de-identified database or a summary of the demographic characteristics of the persons that had submitted the specimens so that the sample could be described in aggregate. Demographic and other characteristics available varied by site, but typically included specimen collection month/year, gender, year of birth/age, geographic location, race, medications administered, and toxicology results from the submitting program. Sites were instructed to transfer urine samples into de-identified specimen cups provided by DOTS, which were then mailed to our collaborating laboratory, the Division of Forensic Toxicology at the Armed Forces Medical Examiner System (AFMES). DOTS sometimes received test results several months after AFMES had processed them and CESAR reported the results to each site approximately 2 weeks after receiving them. This delay was not a problem because DOTS results were intended for use for epidemiological research, and not to provide rapid clinical feedback. CESAR staff reviewed the test results with each site and drafted a summary DOTS Bulletin, which contained site background information, methods, sample characteristics, and implications of the results for the site (CESAR, 2018a).

AFMES used liquid chromatography-tandem mass spectrometry (LC/MS/MS) to test specimens for approximately 240 drugs, including 46 synthetic cannabinoids, 21 fentanyl compounds/analogs, 95 other new psychoactive substances (including synthetic and other cathinones, phenethylamines, piperazines, tryptamines, and synthetic amphetamines), as well as 78 other licit and illicit substances. CESAR staff regularly track data on emerging drugs using national datasets such as the DEA’s National Forensic Laboratory Information System, DEA emerging threat reports, and the Center for Forensic Science Research and Education’s (CFSRE) NPS discoveries. These sources helped to identify the substances to discuss with AFMES staff for possible inclusion in their drug testing panels, and to ensure that the tests were as up-to-date as possible.

Results

The eleven sites

Table 1 displays the 11 sites that participated in DOTS, the number of specimens submitted, and the primary questions each site sought to answer. We have arbitrarily numbered each site to facilitate referring to them throughout this article. The 11 sites were located in 6 states and the District of Columbia. Five were from our own state of Maryland (1,3,8,9,11), and one each came from Washington, DC (7), Washington State (2), Florida (4), Vermont (5), Texas (6) and Oregon (10). The reasons sites requested the DOTS extended testing varied and some had multiple goals. One site (1) wanted to learn why people having an opioid overdose were less responsive to naloxone or more difficult to treat. Other sites wanted to test for drugs they could not test for, including fentanyl or its analogs (2,3,5,7,10), synthetic cannabinoids (3,7,8), or NPS (4,7,8,10). A number of sites just wanted to learn what drugs their test panel might be missing (6,9,11). It became clear from these requests that there was an important need for sites to gain knowledge of substances that went beyond the testing protocols that their organizations were routinely using. Much of this interest was prompted by the opioid crisis and the growth of analogs being developed.

Table 1. Summary of participating sites.

Site #	Site Name	Sample	Question to be answered through DOTS testing
1	Midtown Hospital Baltimore City, MD	8 patients treated for apparent opioid overdose. (Specimens collected May 2017)	To determine the substances that were involved in apparent opioid overdose cases during a 2-day period after a city-wide warning where patients required more naloxone than usual and were violent upon reviving.
2	King County Medical Examiner’s Office Seattle/King County, WA	20 deceased persons who died from opioid overdoses. (Decedent specimens included in DOTS tested positive for fentanyl, fentanyl analogs, or had inconclusive or discordant local toxicology results). (Specimens collected between October 2016 and November 2017)	To assess the presence of fentanyl analogs.
3	ACCESS to Behavioral Health Services Montgomery County, MD	20 uninsured or publicly insured residents of Montgomery County, MD seeking behavioral health and substance use services who screened positive for one or more drugs by their standard 12-drug panel. (Specimens collected between June and August 2017)	To determine whether drugs not screened for by the program’s onsite testing, particularly synthetic cannabinoids or fentanyl and its analogs, were present.
4	The Recovery Research Network Palm Beach, FL	23 uninsured persons who use opioids admitted to medication-assisted treatment (MAT) program. (Specimens collected between October 2017 and January 2018)	To obtain results for other synthetic drugs that are not detected by program’s onsite rapid 14-drug test.
5	Aspenti Health Burlington, VT	20 patients in drug treatment who had a positive fentanyl screen followed by a negative fentanyl confirmation test. (Specimens collected between November 2017 and January 2018)	To identify potential fentanyl analogs that are not included by the site’s drug test, to alert the provider community as to what fentanyl forms are being used.
6	Charlie’s Place Recovery Center Corpus Christi, TX	10 patients from a residential treatment program who tested positive for opiates. (Specimens collected between December 2017 and January 2018)	To determine the specific substances that were involved in opiate-positive cases. Results will provide insight into the substances being used by its patients and assist in its development of more effective protocols for future treatment.
7	Pretrial Services Agency District of Columbia	21 persons in the criminal justice system suspected of using new psychoactive substances. (Specimens collected between October 2015 and October 2017)	To determine whether specific synthetic cannabinoids, synthetic cathinones, and/or fentanyl analogs were involved in drug positive cases.
8	University of Maryland Medical Center Baltimore, MD	7 patients exhibiting atypical psychosis signs with suspected synthetic drug overdose.	To determine the specific synthetic drugs involved in atypical psychosis cases.
9	Stepping Stones Recovery Houses Annapolis, MD and Glen Burnie, MD	34 persons living in recovery. (Specimens collected March 2019)	To determine the specific drugs involved that could not be detected by program’s onsite drug test.
10	Oregon Health and Science University Hospital and Health System Portland, OR	17 patients admitted to the emergency department with suspected opioid overdose. (Specimens collected between October 2018 and March 2019)	To identify new psychoactive substances (NPS), in particular novel opioids, in emergency department patients at OHSU.
11	University of Maryland Prince George’s Hospital Center Prince George’s County, MD	10 specimens from patients admitted to the Department of Trauma Services for a gunshot wound at the University of Maryland Prince George’s Hospital Center. (Specimens collected between June and September 2019)	To identify drugs being used by persons being admitted to trauma services for a gunshot wound.

The populations from whom specimens had been obtained also varied, and included overdose patients (1,8,10), decedents (2), drug treatment program patients (3,4,5,6,9), criminal justice program supervisees (7), and shock trauma patients with a gunshot wound (11). Median ages of these persons ranged from 38.5 to 43.5 years.

Urinalysis results

Once the test results were obtained from AFMES, they were imported into a database for analysis. Because of the small number of specimens received from a site, we developed a matrix to present all of the drugs identified in each specimen. To protect the identity of persons who submitted specimens, results were not provided for drugs that were not routinely tested for by the submitting site that tested positive for a majority of the specimens. Instead, we placed asterisks in that drug’s column and typically reported only the total number of specimens positive for that drug.

Table 2 provides an example of the reporting matrix developed for each site. The Palm Beach Recovery Research Network (4) sent 23 specimens from persons admitted to an opioid medication assisted treatment (MAT) program. The patients were 22–61 years old and mainly white males. The program routinely tested patients for 14 drugs using an on-site test and wanted to obtain results on synthetic drugs not able to be detected by their testing. As Table 2 shows, all of the specimens contained multiple drugs. As would be expected of persons appearing for MAT, 19 of the specimens contained morphine and 19 contained fentanyl. THC (marijuana) and benzoylecgonine (cocaine) were each found in 10 specimens. In addition, many specimens contained non-fentanyl opioids, and fentanyl analogs. The drug matrix presents clear evidence of the predominance of multiple drugs in these specimens. We found that 91% of the specimens contained 5 or more substances and 35% contained 10 or more substances. Multiple drugs in urine specimens were the norm in all DOTS sites. Similar drug report matrices were prepared for each site and included in their DOTS Bulletin (CESAR, 2019a).

Table 2. Drugs or drug metabolites detected by NDEWS Drug Outbreak Testing Service (DOTS) laboratory urinalyses (N = 23 urine specimens submitted to DOTS by The Recovery Research Network, Palm Beach, Florida).

	Common drugs				Pharmaceutical nonopioid drugs																		Nonfentanyl opioids											Fentanyls						Other drugs
					Anti-histamines			Benzodiazepines					Anti-depressants
Specimen	THC (marijuana)	Benzoylecgonine (cocaine)	Methamphetamine	PCP	Diphenhydramine	Cetirizine	Hydroxyzine	Alprazolam/α-Hydroxyalprazolam	7-Aminoclonazepam	Lorazepam	Oxazepam	Temazepam	Desvenlafaxine/Desmethylvenlafaxine	Venlafaxine	Trazodone^	Quinine	Dextromethorphan	Loperamide	Gabapentin	Naloxone	Haloperidol	Cyclobenzaprine	Morphine	Codeine	6-Acetylmorphine (heroin)	Tramadol	Hydromorphone	Oxymorphone	Noscapine	Oxycodone	Hydrocodone	Buprenorphine/Norbuprenorphine	Methadone/EDDP	Fentanyl/Norfentanyl	Methoxyacetyl fentanyl	Cyclopropyl fentanyl	Carfentanil	Acetylfentanyl/Acetyl NorFentanyl	4-ANPP (Despropionyl fentanyl)	3,4,5-trimethoxycocaine	α-PVP	mCPP^
1	✓		✓	✓	*								✓	✓		✓		✓					✓	✓		✓								✓	✓	✓				✓
2		✓			*										✓	✓							✓	✓	✓		✓		✓				✓	✓	✓			✓			✓	✓
3		✓	✓		*			✓	✓							✓							✓	✓	✓	✓	✓					✓		✓
4	✓		✓		*	✓	✓			✓	✓	✓				✓	✓	✓														✓		✓
5	✓	✓			*											✓							✓	✓	✓		✓				✓		✓	✓	✓
6			✓		*			✓								✓							✓	✓	✓	✓						✓		✓
7		✓			*											✓							✓	✓	✓				✓					✓	✓		✓
8	✓				*				✓							✓			✓				✓	✓										✓	✓	✓
9	✓				*			✓								✓			✓														✓	✓				✓		✓
10			✓		*											✓							✓	✓			✓					✓		✓					✓
11	✓				*					✓										✓	✓					✓		✓		✓		✓
12		✓	✓	✓	*																		✓	✓										✓	✓
13			✓	✓	*																		✓	✓	✓	✓								✓
14		✓			*											✓							✓	✓		✓	✓							✓
15	✓	✓			*												✓						✓				✓					✓		✓
16	✓	✓			*															✓		✓					✓	✓				✓
17	✓	✓		✓	*											✓							✓											✓
18			✓		*													✓					✓			✓	✓							✓		✓
19					*											✓							✓	✓		✓								✓
20					*												✓						✓											✓			✓
21	✓	✓			*												✓						✓											✓
22					*																		✓					✓		✓			✓
23					*											✓							✓
Total positive	10	10	8	4	*	1	1	3	2	2	1	1	1	1	1	14	4	3	2	2	1	1	19	12	6	8	8	3	2	2	1	7	4	19	6	3	2	2	1	2	1	1

✓ Indicates positive test result for drug; blank indicates negative test result for drug.

* To protect the identity of persons who submitted specimens, results are not provided for drugs that are not routinely tested for by the submitting site and test positive by a high percentage of the specimens.

^ Trazodone is an antidepressant whose major active metabolite is mCPP. It is not possible to determine definitively whether the presence of mCPP was due to Trazodone use or whether mCPP was taken on its own.

Note: Specimens were collected by The Recovery Research Network from uninsured opioid users admitted to medication-assisted treatment (MAT) between October 2017 and January 2018.

Variation in drugs detected

Table 3 displays the top 6 substances detected in any of the nine sites that had submitted at least 10 specimens. In Table 3, the drugs not included in a site’s standard test panel are designated with an “*”. Cocaine and THC were detected in all 9 sites, and morphine and diphenhydramine were detected in 8 of the sites. Synthetic cannabinoids and PCP were much less prevalent, found in only 3 sites. As one might expect, the presence of stimulant drugs varied across the country. Methamphetamine was most common in opioid overdose decedents in Washington State (2) and residential treatment program patients in Texas (6). Cocaine was detected across most sites, but it was one of the top 2 drugs found in uninsured or publicly insured patients in Maryland (3) and criminal justice arrestees/probationers in Washington, DC (7). Our report on methamphetamine (CESAR, 2018c) described similar regional differences in methamphetamine use across the country.

Table 3. Top 6 substances detected in any of 9 NDEWS Drug Outbreak Testing Service (DOTS) sites (N = 175 urine specimens submitted to DOTS).

Substances among top 6 in any site	King County Medical Examiner’s Office˄King Co., WASite #2(N = 20)%	ACCESS to Behavioral Health Services Montgomery Co., MD Site #3 (N = 20) %	The Recovery Research Network Palm Beach, FL Site #4 (N = 23) %	Aspenti Health Burlington, VT Site #5 (N = 20) %	Charlie’s Place Recovery Center Corpus Christi, TX Site #6 (N = 10) %	Pretrial Services Agency (PSA) for the District of Columbia^˄ District of Columbia Site #7 (N = 21) %	Stepping Stones Recovery Center^˄ Annapolis and Glen Burnie, MD Site #9 (N = 34) %	Oregon Health and Science University Hospital and Health System^˄ Portland, OR Site #10 (N = 17) %	UM Prince George’s Hospital Center^˄ Cheverly, MD Site #11 (N = 10) %	# of sites detected
THC (Marijuana)	35	75	43	40	50	53*	3	69	80	9
Benzoylecgonine (Cocaine)	55	30	43	15	40	38	3	53	30	9
Morphine	60	5	83	40	100	24	0	73	30	8
Diphenhydramine	35	10*	87*	25*	10*	19*	12*	7*	0	8
Methamphetamine/Amphetamine	90	15	35	5*	100	29*	0	20	0	7
Hydromorphone	25	5	35	30	90	10*	0	0	60	7
Oxycodone/Oxymorphone	25	5	13	5	0	5*	0	6	40*	7
Any Fentanyl	74	0	83	60	0	24*	0	20*	50	6
Alprazolam/α-Hydroxyalprazolam	30	10	13	0	20	0	0	47	10	6
Gabapentin	20	5*	9*	5*	0	0	26*	27*	0	6
6-Acetylmorphine (Heroin)	15	0	26	0	30	10	0	53	0	5
Noscapine	5*	0	9*	0	50*	5*	0	40*	0	5
Tramadol	0	0	35*	5*	10*	0	3*	7*	0	5
Any synthetic cannabinoid	0	5*	0	0	10*	24	0	0	0	3
PCP	0	5	17	0	0	33	0	0	0	3
Number of substances positive (of 15)
0	0	10	0	25	0	14	62	12	0
1	0	55	0	15	0	14	29	12	0
2	10	10	9	20	0	24	9	6	40
3	20	15	4	15	10	19	0	18	30
4	25	0	17	5	30	19	0	6	20
5–9	45	10	70	20	60	10	0	46	10
Total	100%	100%	100%	100%	100%	100%	100%	100%	100%
Mean # of substances positive (of 15)	4.6	1.7	5.3	2.3	5.1	2.7	0.5	3.7	3.0

˄ Not all specimens for this site were tested for every drug due to insufficient urine quantity for a specimen. Given this, N’s vary for some drugs.

* Drugs not routinely tested for by the local program’s panel.

Note: The following drugs/drug classes are not shown in this table: α-hydroxymidazolam, amitriptyline, buprenorphine/norbuprenorphine, cetirizine, citalopram, codeine, haloperidol, lorazepam, methadone/EDDP, naloxone, naltrexone, promethazine, and quinine.

All sites had at least one drug in Table 3 that could not be identified by their standard test panel. Diphenhydramine, a cutting agent often mixed into heroin (particularly black tar heroin) and Tramadol, a sometimes misused opioid painkiller, were each not included in the test screens used by five or more of the sites (Mars et al., 2018).

Our prior studies have shown that we typically found multiple substances in the urine specimens obtained from high risk adult populations, including arrestees/probationers, overdose victims and hospital ED admissions (Billing et al., 2017; Wish et al., 2018). The bottom section of Table 3 presents the number of these 15 drugs detected in each specimen. In 4 sites, all of the specimens contained 2 or more drugs. These specimens came from overdose fatalities (2), drug treatment patients (4,6) and hospital ED patients (11). Eight of the nine sites had specimens that tested positive for 5 or more of the 15 drugs shown in Table 3. The average number of substances in each site’s specimens ranged from 0.5 to 5.3, with five sites averaging 3 or more drugs. In one site, there was a specimen that contained 9 of these drugs.

Site reactions to DOTS results

Each DOTS Bulletin contains a section describing the implications of the test results for the submitting organizations and their population. In some sites, the suspected drugs (often fentanyl analogs or NPS) were not detected and unexpected drugs were found to be prominent in multiple sites. Below are illustrative excerpts from DOTS Bulletins of the comments received from sites:

Multiple drugs and fentanyl but no NPS (1)

“The fact that most of these specimens contained eight or more drugs suggests that the number of drugs taken by these patients could be more important to their presentation and response to treatment than any one drug. After the precipitating event has been controlled, efforts should be made to diagnose and treat the patient’s multiple drug use disorders. The study results support our original suspicion of fentanyl involvement in these cases; however, little evidence was found for involvement of SC and none for synthetic cathinones. The prevalence of fentanyl in this sample of patients should prompt physicians to consider administering larger doses of naloxone to patients presenting with similar symptoms (Dezman et al., 2018).”

Current test panel underestimates drug use (3)

“Based on these results, it appears that while the standard 12-drug panel onsite rapid test cup currently used by ACCESS likely identifies the most commonly used substances (marijuana and cocaine), it may underestimate the extent of multiple drug use in its population (CESAR, 2018b).”

Many nonopioid drugs indicate complexity of MAT (4)

“The large number of nonopioid drugs detected underscores the complexity of treating these persons using opioid-focused MAT. These persons will likely need “ongoing, long-term education and interventions” that can address their extensive drug-using behaviors (CESAR, 2018f).”

Fentanyl analogs not detected (5)

“We were surprised to learn that none of these cases demonstrated fentanyl analogs. While the small number of cases may have precluded identification, this sample had been selected so as to increase the chance of fentanyl analog identification. This raises an important question as to how ubiquitous fentanyl analogs are in our state (CESAR, 2018e).”

Less fentanyl but unexpected methamphetamine (6)

“Although these results suggest that patients who screen positive for opiates in this treatment program were not likely to be using fentanyl or heroin cut with fentanyl, the significant use of methamphetamine is noteworthy. Charlie’s Place staff believe that the large number of methamphetamine positives may be due to the practice of speedballing (using an opiate and a stimulant at the same time), which is popular in their area. According to the site, these results will allow Charlie’s Place to provide a more accurate heroin treatment protocol (CESAR, 2018d).”

Misuse of gabapentin (9)

“The study also confirmed what we thought specifically of the increasing prescription and use of gabapentin. We see that gabapentin is clearly the most popular drug prescribed and is replacing opioids and being misused (CESAR, 2019c).”

Misuse of XANAX (10)

“Instead of the novel opioids that were expected on testing, the large doses of naloxone given in this series are likely due to the prevalence of the co-ingestion of a typical opioid with “pressed Xanax” or “pressed Xanie bars” that were reported in 7 of the 17 patients. This also explains the frequency of benzodiazepines (10/17), including alprazolam (7/17), in this study. This information will be used to develop treatment guidelines that can be circulated via the Oregon Poison Center to area emergency departments, police, and prehospital agencies to better inform their care of patients visiting the emergency department for an opioid overdose (CESAR, 2019b).”

Fentanyl screening in ED needed (11)

“While there was a substantial amount of fentanyl/norfentanyl detected, it is unclear if this was administered for analgesia in the emergency department (ED) or ingested as part of recreational use prior to arrival to the ED. Fentanyl testing is not routinely performed in the ED and typically escapes most hospital drug use screens. As clouded sensorium and bad judgment may have played a role in some of these violent events, additional screening for fentanyl is needed (CESAR, 2019d).”

It is clear from these comments that the extensive test results from the small number of urine specimens sent to DOTS provided these organizations with useful information about their testing protocols and the drugs that their populations were exposed to.

Discussion

CESAR staff designed the DOTS study to determine the feasibility and usefulness of enabling sites to have a small number of their urine specimens tested for a large number of drugs at no cost to them. We found that it was most feasible to recruit sites through the NDEWS dissemination vehicles we had developed. While the DOTS study has been completed, we still find that as people learn of the DOTS results, they send us requests to participate in DOTS extended testing.

The DOTS study supports the value of the strategy of retesting already collected urine specimens for a large number of drugs as an epidemiologic tool. The regional differences in the types of drugs being used found by DOTS are consistent with the findings from other studies using more traditional techniques of measurement. But DOTS has also uniquely uncovered the multiple substances that users are exposed to and the complexity involved in describing the nature of the drug problem. Persons appearing for treatment for an opioid related event had often been exposed to multiple drugs. This finding was expected, as some of the sites submitted specimens from persons that appeared to be resistant to the usual dose of naloxone. Nevertheless, the results provide clear evidence of the weakness of labeling drug events as having been precipitated by a single drug (Compton et al., 2021).

The large number of substances detected in some specimens, often more than 5, suggests to us the likelihood that the users may have unknowingly taken some of them because of intentional or unintentional adulteration at the point of manufacture or transportation. Self-report measures of drug use would therefore likely underestimate the true nature of the drugs ingested or available on the street. People can only report what they believe they ingested. Objective biologic tests are necessary to identify the nature of the drugs recently used.

The DOTS results also have important clinical implications. It is our experience that most organizations cannot conduct or afford the many types of tests that DOTS had available. Similar expanded testing of samples of specimens could be used to refine hospital and treatment programs’ testing protocols and to better inform patients and doctors regarding the drugs that their patients were exposed to. The inability of many of these organizations to test for the approximately 240 drugs on the DOTS panel could severely hamper the understanding of the nature of adverse drug events and the best ways to treat and prevent them. Several of the sites (5, 6, 10) learned that the NPS, typically fentanyl/fentanyl analogs or other novel opioids, they had suspected were being used by their population were not detected by the DOTS extended testing. Instead, these sites found traditional opioids in combination with other substances, such as benzodiazepines and methamphetamine. This information might help them to target these drugs for further attention. In addition, as new analogs are developed, it is evident that they will go largely undetected using standard limited urinalysis testing protocols or by relying solely on users’ self-reports.

Potential limitations

The DOTS results are derived from small convenience samples of already collected specimens that were submitted to answer specific problems that arose in each site and should not be generalized to an entire population or locality. Specimens often came from patients suspected of having used drugs or experiencing unusual health consequences that were resistant to treatment. The findings also depend heavily on the panel of drugs that AFMES tested for, and while considerable, their test panels could have omitted some substances. Future iterations of a DOTS type program might benefit from the use of liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QTOF-MS) for nontargeted data acquisition (Krotulski et al., 2020) that can detect many more substances, particularly NPS, as it is possible that some NPS may have gone undetected due to limitations in the substances included on the test panel. The lag time in the collection of the urines and their processing might have also contributed to our laboratory’s missing some substances because of degradation of metabolites.

The many drugs found in most specimens do not necessarily indicate intentional use. It is likely that drugs sold illicitly are intentionally or inadvertently mixed with other chemicals without informing the consumer (Rosenblum et al., 2020). Similarly, the user or the dealers might have unintentionally adulterated their drugs during their transport. Urinalysis results cannot determine whether the licit drugs identified were taken under a physician’s supervision or whether they had resulted from diversion.

The limited funding and timeline available for this study did not allow us to complete a comprehensive evaluation of the DOTS program and its implications for each site. However, each site was asked to provide feedback on the outcome of the testing for their site and some of this feedback is presented in the report, with more details contained in the respective DOTS bulletins that have been made available (CESAR, 2019a).

Final recommendation

We propose that the U.S. and other countries seeking to develop drug detection systems establish national programs to enable hospitals, physicians, and others to submit small samples of already collected, de-identified specimens to an independent laboratory for expanded testing. Biologic testing methods utilizing hair, blood, serum, and/or saliva may be considered depending on their availability. Several programs in the U.S. do test submitted biological specimens for emerging drugs including, the DEA TOX program (Diversion Control Division (DCD), n.d.), the Minnesota Drug Overdose and Substance Use Pilot Surveillance System (MNDOSA) (Moyer, 2019), and NPS Discovery (CFSRE, 2021). These programs differ from the proposed approach in important ways such as by testing limited types of specimens or focusing on small geographic areas or specific classes of drugs. As an example, epidemiologists at the Harvard School of Public Health are setting up the Global Immunological Observatory, a large surveillance system to monitor blood samples globally for the presence of antibodies to hundreds of viruses to alert scientists to immunological changes that could signal an emerging pandemic (Greenwood, 2021).

CESAR is currently pilot testing the ONDCP-funded Emergency Department Drug Surveillance (EDDS) system that conducts expanded testing of urine specimens obtained from patients appearing at emergency departments for a suspected drug-related event (CESAR, 2020). This process can be repeated once or twice a year, at minimal or no cost to the submitting organizations. It will allow for the tracking of traditional illicit drugs and prescription medications, as well as NPS.

Laboratories participating in a national system could collaborate to keep their test panels and methods uniform and up-to-date allowing for comparisons of trends across sties. Currently, it is left to individual labs and groups of scientists to reach out to each other to share recommendations for panels for specific types of drugs (e.g. Pacific Northwest National Laboratory ASTM Fentanyl Standards Working Group). An integrated data system could be developed to capture all of the results and prepare periodic reports that track the local trends in the drugs detected. Like DOTS, such a program could generate valuable results for epidemiological, rather than clinical purposes. The envisioned system could alert the country to the spread of new drugs as they emerge and help the local testing programs to update their test panels.

Acknowledgements

The authors would like to thank all of the sites that participated in this study. The authors would also like to thank Naheed Ahmed for her assistance with this report.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the article. The sponsor did not have any role in the study design, collection, analysis or interpretation of the data, or writing of this report.

Additional information

Funding

This work was supported by the National Institute on Drug Abuse (NIDA) as part of the National Drug Early Warning System (NDEWS) project under Grant number 3U01DA038360-04S2.