PEDIATRICS Vol. 119 No. 4 April 2007, pp. e843-e848 (doi:10.1542/10.1542/peds.2006-2278)
ARTICLE |
Results of Random Drug Testing in an Adolescent Substance Abuse Program
a Department of Pediatrics
b Division on Addictions, Harvard Medical School, Boston, Massachusetts
c Center for Adolescent Substance Abuse Research
d Divisions of General Pediatrics
f Adolescent/Young Adult Medicine
e Department of Psychiatry, Children's Hospital Boston, Boston, Massachusetts
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
OBJECTIVE. The objective of this study was to estimate from a random urine drug-testing program for adolescents the proportion of drug tests that are susceptible to interpretation errors.
METHODS. This was a secondary analysis of a clinical database and chart review from an adolescent outpatient substance abuse program at a large children's hospital. We analyzed from 110 adolescent patients (13–21 years of age) all 710 urine drug test results that were collected between December 2002 and July 2005 and 85 original laboratory reports for tests that were collected between December 2002 and May 2006 and were confirmed positive for opioids. We calculated the percentage of tests that were too dilute to interpret (potential false-negatives) and the percentage of confirmed positive tests for oxycodone that did not result in a positive initial screen (potential false-negatives). We also reviewed clinical information to determine whether confirmed positive tests resulted from legitimate use of prescription or over-the-counter medication (potential false-positives).
RESULTS. Of 710 drug tests, 40 negative tests were too dilute to interpret properly, and 45 of 217 positive tests resulted from prescription medication use for a total of 85 tests that were susceptible to error. Of the 85 confirmatory laboratory reports reviewed, 43 were positive for oxycodone, but only 16 of these had produced a positive opiate screen.
CONCLUSIONS. Unless proper procedures are used in collecting, analyzing, and interpreting laboratory testing for drugs, there is a substantial risk for error.
Key Words: adolescents • drug use/abuse • detection • screen
Abbreviations: GC/MS—gas chromatography/mass spectrometry • MRO—medical review officer • THC—delta-9-tetrahydrocannabinol
Drug-testing programs have been shown to be a useful therapeutic adjuvant for adults with cocaine disorders when used in combination with other treatments or as a sole therapy within treatment programs.1–6 Although few formal assessments have been conducted, many drug court programs for juvenile offenders rely on repeated drug testing to monitor adolescents,7 and judges have reported sustained improvements when combined with other services and close supervision.8,9 Random urine drug testing has been proposed as a screening and prevention method to reduce substance use by students through school-based drug testing programs. Proponents argue that testing will give students who are not yet drug involved a reason to "say no" and provide an opportunity for drug-involved youth to be identified early and receive a referral to appropriate treatment,10 although empiric evidence to support these claims is lacking.
Drug testing is a technically complex procedure that must be conducted under rigorously defined procedures for proper collection, screening, and confirmatory testing. The federal government has specified procedures that are known as Mandatory Guidelines for Federal Drug Testing Programs that must be used when federal employees are drug tested,11 and many workplace programs also have adopted these standards to minimize interpretation error. These guidelines were established for use with adults, and drug use by adolescents varies in important ways. For example, research has demonstrated that adolescents adeptly use information technology such as the Internet to modify their drug use behavior.12 These same technologies could be used to rapidly disseminate information on methods of defeating drug tests among peers. Many more adolescents than adults are prescribed stimulant medications, and at the same time, abuse of stimulant medications by teenagers is increasing. Adolescents use different drugs than adults. Inhalants are used more commonly by adolescents, and use of prescription medications by adolescents is increasing13; these classes of substances are not readily detectable on standard urine screens. These differences suggest that although drug-testing programs for adolescents must be as rigorous as workplace testing programs to avoid tampering with specimens, special adaptations may be required to avoid misinterpretation. Specifically, drug panels that are used for testing adolescents should be designed to detect the drugs that are used most commonly, and physicians or other professionals who are responsible for interpreting these tests must receive enough training in toxicology to interpret them correctly. There is potential for harm from false-negative tests, resulting in delayed diagnoses and reinforcement of behaviors by drug-involved youth who have learned how to defeat drug tests. False-positive tests could cause substantial harm to students who are not using drugs by placing them under a cloud of unjust suspicion that might be very difficult to dispel.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Areas for Analysis
On the basis of our experience in working with adolescents, we focused on the following areas in this analysis.
Specimen Dilution
A urine drug test may be negative even though an adolescent is using drugs if the specimen is diluted, either in vivo or in vitro, to drive the concentration of the illicit substance below the threshold detection level on an initial screening test. Proper collection technique can minimize in vitro adulteration and dilution, but in vivo dilution (consuming a large volume of fluid and/or a diuretic) will be detected only if each sample is checked for adequate concentration. Creatinine is a byproduct of muscle metabolism, and its concentration in urine is very sensitive to dilution. A urine creatinine level <25 mg/dL suggests a urine sample too dilute for proper interpretation of a negative drug test.14
Prescription and Over-the-Counter Medications
Laboratory testing for drugs is essentially a 2-stage procedure. The initial stage is a screening immunoassay for multiple drugs of abuse. Immunoassays are reasonably sensitive when the urine is normally concentrated but not very specific. Therefore, all positive screens must be followed by second-stage confirmatory testing with gas chromatography/mass spectrometry (GC/MS), which is highly specific.
Drug tests may be positive in the absence of illicit drug use when a prescription or over-the-counter medication or food substance cross-reacts with a screening immunoassay or contains a substance that is detected by the screening panel. False-positives from cross-reacting substances can be eliminated by confirming each positive screening test by GC/MS, but legitimate use of a prescription medication can be assessed only by a medical history that is supplied by the adolescent and/or a parent.
Limitations of Screening Panels
Screening immunoassays test for a limited number of drugs. Certain drugs that commonly are used by adolescents are not reliably detected by multipanel screens. Notably, multipanel tests include an opiate screen but are not designed to detect opioids. The term "opiates" refers to naturally occurring alkaloids that are obtained from the opium poppy (morphine and codeine) and their semisynthetic derivatives (heroin). "Opioids" are synthetic compounds (eg, oxycodone, hydrocodone, fentanyl) that have pharmacologic properties similar to morphine and affinity for the opioid receptor. "Opiate" assays are designed to detect morphine and codeine and will detect heroin, which is metabolized to morphine. There is variable cross-reactivity for detection of the synthetic opiates, which may be detected if taken in very high dosages. When use is suspected, testing for oxycodone or other opioids via specific immunoassay panel or GC/MS must be ordered separately. If this is not done, then the risk for a false-negative is high. Similarly, 3,4 methylenedioxymethamphetamine (ecstasy) is not detected routinely on amphetamine panels but may be detected by specially designed immunoassays or GC/MS, which must be ordered separately.
Drug-testing Program
The goal of this project was to analyze the drug-testing results from a random drug testing program that was administered by adolescent substance abuse experts using recommended procedures to determine (1) the proportion of tests that were too dilute to interpret; (2) the proportion of positive tests that were determined to result from a cause other than illicit drug use; and (3) the proportion of tests that were confirmed positive for oxycodone and were not detected on a standard drug screening panel that included an opiate assay. The results of this analysis can inform recommendations for adolescent drug-testing programs and may help to assess the feasibility of widespread random drug-testing programs in schools and other settings where access to expertise in medical toxicology may be limited.
We performed a secondary analysis of a clinical database that contained drug test results and of chart review to confirm positive drug tests. The clinical drug testing program that served as the source of the database is described in detail.
Patients
Adolescents and young adults who were 13 to 21 years of age and completed a substance abuse evaluation in a children's hospital–based substance abuse outpatient program were invited to enter a random drug-testing program when the supervising clinician believed that drug testing was clinically indicated. Generally, drug testing was offered to patients who had a history of illicit drug use other than inhalants or steroids and agreed to a trial of abstinence during the drug-testing period. A total of 110 patients entered the drug-testing program and completed at least 1 drug test between December 2002 and July 2005. Results from all drug tests were recorded in an administrative database, which formed the data set for this analysis. The team identified 89 GC/MS tests that were done between December 2002 and May 2006 and were recorded in the database as positive for any opiate. Of these, 4 original laboratory initial screening reports were not available (all were recorded as positive for morphine, which can be detected after use of any opiate that is metabolized to morphine, including heroin and codeine but not oxycodone). We reviewed all 85 reports that were available to estimate the false-negative rate.
Random Drug-testing Program
Patients who agreed to participate in the random drug-testing program selected a participating laboratory and presented on the day of entry into the testing program or first random call for a baseline sample and then were called on a random schedule with an average of 1 test per week. Patients were called by 10 AM and required to provide a sample the same day. All test results were reviewed by a certified medical review officer (MRO). The MRO is a physician who has completed special training in proper management of laboratory drug-testing procedures, urine toxicology, and interpretation of test results. Unanticipated results, which include dilute specimens or positive tests, were followed by a patient interview by his or her primary clinician. Patients and their parents were given the results of all confirmed positive or dilute urine tests after the patient–clinician interview. Parents determined consequences for positive drug tests for their children; staff consultation was provided on request. Typical restrictions included suspension of car privileges and grounding. Parents were advised to treat dilute tests as though they had been positive.
Laboratory Procedures
Drug tests were performed at a commercial laboratory using federally specified chain-of-custody procedures.11 After photographic identification, patients removed outer garments and emptied pockets. Specimens were collected in bathrooms with water turned off and toilet water stained blue to prevent adulteration. Specimen temperature was checked immediately, and samples that were below body temperature were not analyzed.
Each test included a standard drug screen immunoassay panel*, urinalysis, and creatinine level. The opiate assay of the urine screen that was used by the commercial laboratory is designed to detect morphine and codeine; oxycodone cross-reacts with the opiate assay when it is present in concentrations >300 ng/mL. Urine specific gravity and creatinine level were used to ensure specimen concentration and integrity. Confirmatory testing via GC/MS was performed on all positive screens, and all tests that were positive for D9-tetrahydrocannabinoid carboxylic acid (THC) were quantified. We added GC/MS tests for synthetic opioids to every order for each patient with a history of opioid use, because oxycodone is not detected reliably by standard screening immunoassays.
Analysis
Data were extracted from (1) an administrative database with the results of all drug tests conducted between December 2002 and June 2005 and (2) clinic notes to verify positive drug tests. In a second step, we reviewed the original laboratory reports of 85 drug tests that were confirmed positive for any opioid to determine the proportion of tests that were confirmed positive for oxycodone and were not detected by a routine drug screen. The Committee on Clinical Investigations (institutional review board) at Children's Hospital Boston approved this protocol.
Negative urine samples with a creatinine level <20 mg/dL were considered too dilute for proper interpretation. For all tests that were positive for opiates, amphetamines, or benzodiazepines, we reviewed the clinic note for the MRO visit immediately after the test date to determine whether the patient was taking a prescription medication that could account for the positive test or gave a history of (legitimate) use of an over-the-counter medication or food that could result in a positive test.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
A total of 110 patients completed 710 drug tests (mean: 6.5 tests per patient; range: 1–49). Forty negative tests (6% of total) were too dilute to interpret. Nineteen (17%) participants accounted for all 40 dilute drug tests; 15 participants had 1 or 2 dilute tests, and 4 participants had 4 or more (Table 1).
|
A total of 480 (68%) of 710 drug tests were negative; 125 (18%) were confirmed positive for THC, 46 (6%) for amphetamines, 10 (1%) for cocaine, 4 (<1%) for benzodiazepines, and 32 (5%) for opiates. In total, 177 (25%) tests were positive for a single substance, and 13 (1.8%) were positive for >1 substance, for a total of 217 positive individual tests.
After chart review, positive tests were categorized further as likely resulting from illicit drug use or legitimate prescription use. Of the 217 positive tests in the initial review, 45 (21%) were attributed to legitimate use of prescription or over-the-counter medications. Of the total 710 completed drug tests, 85 (12.0%) were susceptible to either positive or negative misinterpretation.
We reviewed 85 original laboratory reports for tests that were confirmed positive for opiates, including 43 (51%) for oxycodone. Of these 43, 16 (37%) were detected on a standard drug testing panel (ie, the opiate assay was positive; confirmatory testing was positive for oxycodone but not morphine) and 27 were detected only by GC/MS (ie, the opiate assay was negative, but GC/MS was positive for oxycodone). Of the 27 tests in which the opiate screen was negative and the accompanying GC/MS test was positive for oxycodone, 23 had completely negative drug screens and 4 had a positive screen for another drug (2 for cocaine and 2 for THC).
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This report demonstrates that a significant proportion of drug tests from a random drug-testing program for adolescents were susceptible to misinterpretation. We found that a small but substantial proportion (6%) of urine samples were negative but too dilute to interpret. This means that up to 6% of specimens, collected under rigorous conditions that are designed to prevent adulteration and dilution, still were too dilute to be interpreted and could have led to false-negative reports. Furthermore, 17% of the adolescents in this sample submitted at least 1 dilute sample, but only 4% submitted >2, suggesting that adolescents who were caught trying to defeat drug testing by dilution once were unlikely to attempt dilution again. Drug-testing programs that use less rigorous collection procedures (eg, staff member outside the bathroom) or no special procedures, as is common practice in most physicians' offices,15 likely would have much higher rates of adulterated specimens. We therefore recommend that all adolescent drug-testing programs use the same rigorous urine collection protocol as used with adults in workplace drug-testing programs. Furthermore, clinicians should recognize that even rigorous collection procedures cannot prevent a determined individual from tampering with a urine specimen, and, as in the federal protocol, directly observed specimens should be considered whenever urine tampering is suspected.
Twenty-one percent of all confirmed positive drug tests resulted from licit use of prescription or over-the-counter medications, including a large majority (91%) of amphetamines and a small minority (6%) of opioid-positive tests. We recommend that a clinician conduct a follow-up interview whenever an adolescent has a confirmed positive drug test. However, clinicians must know which medications can account for positive tests to make valid assessments. A recent survey found that few physicians have an adequate knowledge base; 99% misidentified 1 or more substances that can cause positive urine drug tests.15 Therefore, clinicians who administer adolescent drug-testing programs should obtain additional training and certification, as is required for those who administer workplace drug-testing programs. Furthermore, physicians who order urine drug tests for their adolescent patients should have frequent consultation with expert toxicologists to assist with proper drug test interpretation and to keep informed of changes that may occur in the laboratory, such as adoption of new reagents or new assays. We also recommend that physicians consider toxicology results in the context of all of the clinical information and call the laboratory toxicologist whenever there is a discrepancy, to review the laboratory procedures that were used and to question the results of the test.
Participants in this drug-testing program had a complete history before beginning drug testing, allowing the clinical team to order additional specific tests if the adolescent's drug of choice was not detected reliably on a routine panel. Had the clinical team relied on routine drug-screening tests only, two thirds of oxycodone use in this sample would have been missed, and more than half (53%) of the tests that were positive for oxycodone were negative for all other substances, suggesting that drug use would have been missed entirely in these teens. This is a significant finding, given that abuse of prescription medications such as OxyContin and benzodiazepines is on the rise,13 particularly among adolescents. Random school drug-testing programs could test all students with an expansive panel, although such a program likely would be prohibitively expensive. Alternatively, policy makers and program designers who implement programs with more limited testing should be mindful of this significant limitation and monitor the student body for increased use of drugs that are not detected on screening panels.
Limitations
The drug-testing program described here was conducted in an adolescent substance abuse program, and participants were identified as having used drugs before enrolling in the program. The percentage of positive drug tests, therefore, is likely to be higher than would be expected in a mandatory school-based program. If this is the case, then we would expect that the percentage of positive tests that result from legitimate use of prescription medication, over-the-counter medication, or food consumption would be higher in a school-based drug-testing program, and expertise in interpretation would be even more important to interpreting test results accurately. Furthermore, the participants from this program are likely to be more knowledgeable about drug use and methods of tampering with drug tests than adolescents in a general population, and the proportion of dilute tests may be higher than would be expected in a general population. However, random drug-testing programs are designed precisely to identify drug-involved teenagers, similar to those in our sample. The overall percentage of teens who try to tamper with specimens may be lower in a mandatory school drug-testing program; however, the same rigorous techniques must be used if the program is to identify successfully teens with drug problems, because a high percentage (up to 17% in this study) of drug-involved adolescents may try to tamper with drug tests by providing dilute specimens.
We extracted data from an administrative database that was not designed for research, and some information, such as urine specific gravity, was missing. It is possible that some of the dilute tests were incidental, and every patient with a dilute urine test should be interviewed to help determine whether intentional dilution was attempted to defeat drug testing. In our experience, many patients with dilute drug tests will acknowledge intentional dilution if asked. It also is possible that patients who are prescribed medications such as amphetamines for attention-deficit/hyperactivity also abuse them. This is an intrinsic limitation to drug testing and may limit its therapeutic or monitoring utility. A future study should address this.
Implications
This study demonstrates that drug-testing programs demand rigorous procedures and well-trained personnel to obtain accurate results. This finding has implications for the broad application of drug testing through school-based programs, which is currently advocated by the federal Office of National Drug Control Policy.10 Although thorough and rigorous drug-testing programs can be designed, they will be expensive to implement, potentially diverting resources away from much needed prevention and treatment programs. "Quick and dirty" drug-testing programs that use procedures of convenience are likely to result in unintended consequences, such as misidentifying some students as using illicit drugs when they are not and enabling others to continue illicit drug use by allowing them to easily evade detection. Adolescents who participate in drug-screening programs that provide little or no clinical interaction may change their drug of choice to a more dangerous substance that is not detected on routine screening panels, such as inhalants or synthetic opioids. Rapid dissemination of information could spread this behavior to an entire group of teens. More study is required to determine the efficacy of rigorous drug-testing programs in reducing drug use by adolescents. Because of the high error rate, we believe that there is no place for the widespread implementation of drug-testing programs that use substandard procedures.
|
ACKNOWLEDGMENTS |
|---|
We acknowledge support from grant K07 AA013280 from the National Institute on Alcohol Abuse and Alcoholism (Dr Knight), grants R01 DA018848 and R01 DA104553 from the National Institute on Drug Abuse, grant 45222 from the Robert Wood Johnson Foundation, grant 5T20MC000-11-06 from the Maternal and Child Health Bureau, and a philanthropic donation from Jonathan and Margot Davis.
We acknowledge Drs S. Jean Emans, John Kulig, and Elizabeth R. Woods for reviewing this manuscript.
|
FOOTNOTES |
|---|
Accepted Oct 26, 2006.
The authors have indicated they have no financial relationships relevant to this article to disclose.
* Standard panel includes THC, amphetamines, barbiturates, benzodiazepines, cocaine, methadone, methaqualone, opiates, phencyclidine, and propoxyphene but not synthetic opioids such as oxycodone or hydrocodone. 
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Carroll KM, Ball SA, Nich C, et al. Targeting behavioral therapies to enhance naltrexone treatment of opioid dependence: efficacy of contingency management and significant other involvement.
Arch Gen Psychiatry. 2001;58
:755
–761
[Abstract/Free Full Text] - Robles E, Silverman K, Preston KL, et al. The brief abstinence test: voucher-based reinforcement of cocaine abstinence. Drug Alcohol Depend. 2000;58 :205 –212[CrossRef][Web of Science][Medline]
- Silverman K, Chutuape MA, Bigelow GE, Stitzer ML. Voucher-based reinforcement of cocaine abstinence in treatment-resistant methadone patients: effects of reinforcement magnitude. Psychopharmacology. 1999;146 :128 –138[CrossRef][Medline]
- Katz EC, Gruber K, Chutuape MA, Stitzer ML. Reinforcement-based outpatient treatment for opiate and cocaine abusers. J Subst Abuse Treat. 2001;20 :93 –98[CrossRef][Web of Science][Medline]
- Bickel WK, Amass L, Higgins ST, Badger GJ, Esch RA. Effects of adding behavioral treatment to opioid detoxification with buprenorphine. J Consult Clin Psychol. 1997;65 :803 –810[CrossRef][Web of Science][Medline]
- Kirby KC, Marlowe DB, Festinger DS, Lamb RJ, Platt JJ. Schedule of voucher delivery influences initiation of cocaine abstinence. J Consult Clin Psychol. 1998;66 :761 –767[CrossRef][Web of Science][Medline]
- Christie T, Anderson JF. Drug treatment courts are popular but do they work and are they ethical and appropriate for Canada? Health Law Can. 2003;23 :70 –79[Medline]
- Lessenger JE, Roper GF. Drug courts: a primer for the family physician. J Am Board Fam Pract. 2002;15 :298 –303[Medline]
- Office of Justice Programs Clearinghouse. Juvenile Drug Courts: Preliminary Assessment of Activities Underway and Implementation Issues Being Addressed. Available at: www.american.edu/justice/publications/juvsum.html. Accessed October 2, 2003
- Office of National Drug Control Policy. What You Need to Know About Drug Testing in Schools. Washington, DC: White House Drug Control Policy Office; 2002
- Substance Abuse and Mental Health Services Administration. Mandatory Guidelines for Federal Workplace Testing Programs. Available at: http://workplace.samhsa.gov/ResourceCenter/r362.htm. Accessed January 30, 2007
- Boyer E, Shannon M, Hibberd P. The Internet and psychoactive substance use among innovative drug users.
Pediatrics. 2005;115
:302
–305
[Abstract/Free Full Text] - Friedman R. The changing face of teenage drug abuse: the trend toward prescription drugs.
N Engl J Med. 2006;354
:1448
–1450
[Free Full Text] - American College of Environmental and Occupational Medicine. Medical Review Officer Drug and Alcohol Testing Comprehensive Course Syllabus. Arlington Heights, IL: American College of Occupational and Environmental Medicine; 2003
- Levy S, Harris S, Sherritt L, Angulo M, Knight J. Drug testing of adolescents in ambulatory medicine: physician practices and knowledge.
Arch Pediatr Adolesc Med. 2006;160
:146
–150
[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  R. R. Brookman Unintended Consequences of Drug and Alcohol Testing of Student Athletes AAP Grand Rounds, February 1, 2008; 19(2): 15 - 16. [Full Text] [PDF]
|
|
eLetters:
Read all eLetters
- Random Drug Testing Can Be Effective
- Richard E. Struempler
- Pediatrics Online, 12 Apr 2007 [Full text]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
