All the authors except for Dr. Levy are affiliated with the McLean Hospital Alcohol and Drug Abuse Treatment Program, 115 Mill St., Belmont, MA 02478 (e-mail: firstname.lastname@example.org). Dr. Jaffee and Dr. Weiss are also with the Department of Psychiatry, Harvard Medical School, Boston. Dr. Teter is also with the School of Pharmacy, Northeastern University, Boston. Dr. Levy is with the Division on Addictions and the Department of Pediatrics, Harvard Medical School, Boston. Shelly F. Greenfield, M.D., M.P.H., served as editor of this column.
Methods for accurate urine drug testing have been available for several decades, and such methods are useful in assessing and identifying substance use (1). A number of current approaches employ urine testing as a means of determining and enhancing treatment efficacy. Urine testing is also an important deterrent in law enforcement, the workplace, professional and amateur sports, and more recently, schools. Regarding adolescent substance use treatment, urine testing is particularly salient; 95% of physicians who provide primary care to adolescents order urine drug tests (2). Despite widespread adoption of urine drug testing and delineation of workplace guidelines, studies have shown that medical students and residents receive inadequate training in these techniques (2).
Although seemingly straightforward, obtaining a valid test result is a complex process. Results are affected by several factors, including the substance of interest, test methodology, pharmacokinetics, chain-of-custody procedures, and intentional tampering.
The DSM-IV distinguishes 11 categories of abusable substances, including amphetamines; cannabis; cocaine; hallucinogens; inhalants; opioids; phencyclidine (PCP); sedatives, hypnotics, and anxiolytics; alcohol; caffeine; and nicotine. Despite this, many tests screen only for marijuana, cocaine, opioids, amphetamines, and PCP—the "NIDA five" (after the National Institute on Drug Abuse). Although these substances are the most frequently tested for, no industry standards exist. Terms like "comprehensive drug screen," "routine drug screen," and "standard drug screen" are not used consistently by different laboratories and test manufacturers. Additionally, it is important to note that several substances—notably methylenedioxymethamphetamine (MDMA, or "Ecstasy"), oxycodone, hydrocodone, buprenorphine, and clonazepam—are not included in many drug screens and must be ordered separately. Thus one of the most readily remedied sources of inaccuracy in result interpretation is the use of insufficiently broad tests.
Methods of urine drug analysis fall into two general categories—screening assays and confirmatory tests—although some methods of analysis are used for either purpose.
Two characteristics of the test are particularly important: sensitivity and specificity. Sensitivity refers to the lowest detectable concentration of drug, whereas specificity refers to how selective an assay is for a particular drug (3). Ideally, the standard procedure for urine drug testing should involve a highly sensitive screening technique, followed by the use of a highly specific confirmatory technique for samples identified as potential "positives" during screening.
Many screening procedures use immunoassay techniques that rely on competition between a drug chemically labeled with an enzyme, radioisotope, or fluorophore and the drug present in a biological sample (4,5). The labeled drug and the sampled drug compete for binding sites on drug-specific antibodies. The ratio between the two is used to determine the presence or absence of the drug in the biological sample (6). For example, by using radioimmunoassay methodology, the amount of labeled drug that is displaced from the antibody binding sites will determine the level of measured radioactivity, which in turn is proportional to the amount of drug in the sample. Immunoassays are known for having high sensitivity; they are thus useful for screening out "negative" samples that will not require confirmatory testing. However, immunoassays identify drugs with similar chemical structures, and therefore, they also have low specificity (6).
During confirmatory testing, drugs in the specimen are separated before detection. Separation is usually accomplished by using a method such as gas chromatography or high-performance liquid chromatography (HPLC) (4). After separation, analytes reach the detector at different rates (retention times). This information provides evidence as to whether the drugs in question are present in the specimen (6). Gas chromatography is often combined with mass spectrometry, which separates compounds into molecular fragments on the basis of the breakage of bonds within the compounds. The mass spectrometry measures both the mass and quantity of each fragment. Because each drug has a unique fragment pattern, a reliable confirmation of the presence of a drug is obtained (6). The combined use of gas chromatography and mass spectrometry is generally considered the most accurate and reliable method of confirmatory testing, although rare instances of false positives and false negatives have been reported (7).
Four possible outcomes exist for interpreting results: true positive, true negative, false positive, and false negative. However, it is important to note that when interpreting results, drug concentrations in the body may be lower than the designated cutoff. Thus an individual may be using drugs in such a manner (for example, intermittently at low doses) that a test result may appear to be negative when drugs are in fact present. It is also important to note that legitimate sources of false positives do exist, including pseudoephedrine (amphetamine screen), dextromethorphan (phencyclidine screen), and poppy seeds (morphine screen and confirmation) (8,9).
Variables affecting the results of urine testing
Cutoff selection. One variable influencing drug detection is the cutoff threshold. In other words, any sample having a drug concentration equal to or above a specified level is considered a "positive" result. This threshold may vary from context to context, as well as from screen to test. [A table showing commonly used cutoff concentrations for initial and confirmatory testing as well as average detection times for various drugs is available as an online supplement at ps.psychiatryonline.org.] A lower cutoff results in a longer detection time, although it also affects sensitivity and specificity. Lowering the cutoff increases sensitivity, although it also increases the potential for false-positive results by decreasing specificity (10).
Pharmacokinetics. Pharmacokinetic properties (for example, the mathematical description of absorption, distribution, metabolism, and excretion of substances) also influence detection times.
Absorption. Drugs are absorbed into the bloodstream at different rates, depending on factors such as route of administration. This in turn influences how quickly the drug will reach the urine (3), with faster absorption resulting in shorter detection periods.
Volume of distribution. Greater distribution of a substance throughout the body (with a corresponding decrease in the rate of elimination) results in longer detection times.
Metabolism. There is considerable variability in individual metabolism. Despite this variability, slower metabolism (therefore, lower total clearance) results in longer detection times.
Elimination. Final drug elimination from the body occurs primarily through the kidneys, which excrete drugs and their more water-soluble metabolites via urine. There is considerable variability in elimination rates because of factors such as urinary pH, with slower elimination resulting in longer detection times.
Serum half-life. The serum half-life is equivalent to the amount of time (usually in hours) required for drug concentration to decrease by one-half and is determined by pharmacokinetic parameters discussed above. Using the standard "five half-life rule" (4), 97% of a drug will be eliminated from the body within five half-lives. For example, a drug with a half-life of 24 hours may be detectable for approximately five days (depending on dose and cutoff), whereas a drug with a half-life of six hours may only be detectable for approximately 30 hours. Serum half-life tables are readily available and can be used to predict drug detection times (4).
As noted above, some individuals may be motivated to avoid detection of their drug use. Methods for avoiding detection of substance use ("tampering") include dilution of the sample with water or other liquids, substitution with a "clean" or synthetic urine specimen, and adulteration with other chemicals (11). Information on avoiding detection is readily available on the Internet.
To ensure integrity of testing results, mandatory guidelines for collecting and checking urine in federal workplace drug testing programs have been established (12). Although these guidelines are not mandatory in clinical contexts, they do represent a gold standard.
Several characteristics of urine may indicate whether the sample has been adulterated, including pH, temperature, creatinine, specific gravity, and human immunoglobulin (IgG) levels. In a healthy volunteer, urine temperature is expected to be 90°F to 100°F within several minutes of producing a sample. Temperatures outside this range suggest substitution. Similarly, the creatinine level is expected to be greater than or equal to 20 ppm. A creatinine level below this threshold suggests dilution, either directly or through excessive fluid ingestion. In a healthy volunteer, a urine specimen is expected to have a specific gravity level ≥1.003 and a pH level between 3 and 11. A pH or specific gravity level outside of these ranges may suggest chemical adulteration. Finally, IgG concentrations <.5 <g/ml suggest either substitution or dilution. Commonly available chemical adulterants include glutaraldehyde, pyridium chlorochromate, and nitrites (11). Tests for these chemicals may be performed by commercial laboratories, and "dipstick" devices to perform on-site screens are also commonly available.
In addition to the above integrity checks, other precautions, such as eliminating water sources, prohibiting outer garments and personal belongings from the collection room, having subjects wash their hands, and direct observation of the specimen collection, may also be useful (12).
Chain-of-custody procedures refer to the means by which a specimen is identified and associated with a specific individual throughout the collection, transport, storage, analysis, and reporting process. Chain-of-custody procedures may vary according to context and laboratory, but failure to implement and ensure accurate procedures poses a significant source of error. For details on implementing valid chain-of-custody procedures, the reader is referred to Substance Abuse and Mental Health Services Administration's Mandatory Guidelines for Federal Workplace Drug Testing Programs (12).
A number of factors that may contribute to a clinical false-negative result in urine drug testing include substance of interest; analytic technique used; interactions between time frame, substance, and use level; and violation of sample integrity and chain of custody. To ensure the accurate interpretation of urine drug testing, it is important to consider the following topics.
• Are the substances of interest included in the test? Because no standards for terms such as "routine" or "comprehensive" exist, it is important to ensure that the analyses performed will detect the specific substances of interest. Requesting a category of drug test is insufficient.
• Will the requested test provide the desired information? Screening procedures provide qualitative information, whereas confirmatory testing generally provides both qualitative and quantitative information. Confirmatory testing is generally more sensitive and specific, but it is also more expensive. Qualitative screening with confirmatory testing for positive results typically provides the desired information, although some clinical situations require quantitative information.
• Will the testing procedure detect the amount of use within the time frame of interest for a given substance? Urine testing procedures generally only detect relatively recent use. Additionally, the time frame for detection varies by substance (see table showing time frame for detection that is available as an online supplement at ps.psychiatryonline.org), as well as by the frequency and intensity of use. Regular and heavy use may be more easily detected, but intermittent and lighter drug use may be missed.
• Has the sample been diluted, substituted, or adulterated? These are methods used to falsify test results. Proper collection techniques, confirmation of pH, temperature, specific gravity, creatinine, and IgG levels as well as checks for adulterants, including glutaraldehyde, potassium chlorochromate, and nitrates, should be performed whenever tampering is suspected.
• Have accurate chain-of-custody procedures been implemented and followed? Ensuring that a specimen has not been mistakenly or intentionally switched with the specimen of another individual is fundamental to obtaining a valid result.
Opportunities for additional training exist. The Medical Review Officer Certification Council is a physician-based nonprofit board offering certification to physicians who complete a two-day training course and pass an examination (13). Certification is required for physicians who supervise government-mandated workplace drug-testing programs, but physicians in primary care are less likely to have completed this training. Workshops are also offered by professional societies. The degree to which the validity of urine testing is affected by procedural shortcomings is not known. However, research suggests that further dissemination of the means to ensure validity of these procedures is warranted, because it appears that some physicians are inconsistent when ordering tests, fail to use recommended collection techniques, and do not possess sufficient knowledge to accurately interpret these tests (2).
The process of obtaining a valid urine drug test result is deceptively complex and affected by several factors, including the substance of interest, test methodology, pharmacokinetics, chain-of-custody procedures, and intentional tampering. It is important to consider all of these factors when conducting urine drug tests.