 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 | |||||||||
 |
December 1999 Interface of Science & Law in Drug Testing By Mark P. Stevens; James R. Addison
James R. Addison is a physician in Marquette, Michigan. He specializes in emergency and internal medicine. In practice for 20 years, he earned his M.D. degree from Wayne State University and was a clinical assistant professor in the department of emergency medicine at Michigan State University Drug testing has become ubiquitous in the world of criminal law. Test results may be used to bolster a diminished capacity defense or undermine the credibility of a key witness. They may give rise to a probation violation hearing or serve as a basis for a pre-trial detention order. Yet defense lawyers, prosecutors, judges, and corrections officials often have little or no understanding of what drug tests actually measure, or the limitations of particular tests. Rather, they accept on faith that drug tests, in fact, answer the questions that they would like to have answered. There is an erroneous assumption that drug tests can provide data of the same character and quality as blood alcohol tests. Drug tests are sometimes faulted for supposed errors when, in reality, the fault lies not with the tests but in a legal system which fails to understand the inherent limitations of drug tests in the first instance. For example, the law may want a drug test to answer the question: “Was the defendant too high to safely drive an automobile?” The scientific community is not prepared to answer the question of how “high” is “too high.” Surprisingly little research has been done on the correlation between drug consumption and the degree of physical or mental impairment. Nor, unlike blood alcohol tests, do most drug tests answer the question as to whether a person is even “under the influence” at a given point in time. A drug test may only tend to show that a person had been exposed to a particular substance (or a chemically similar substance) within a period of days or weeks prior to the test. In short, there is a substantial gap between the questions that the legal community would like to have answered by drug testing and the answers that the scientific community is able to provide. The real danger lies in the legal community’s failure to “mind the gap” by drawing unwarranted inferences from drug test results. Science of Analytical Drug Testing Alcohol vs. drugs. There is a large body of legal literature regarding alcohol testing. Numerous published decisions, computerized models, treatises, as well as state police and manufacturers’ manuals, discuss alcohol tests (particularly breath analyzer tests) in exhaustive detail. There has been a great deal of qualitative and quantitative research conducted on the subject. It has been established that there is a direct correlation between alcohol intake and impairment1 and, based on the proof of such a fixed relationship, statutory thresholds have been established which separate the driver who has had a drink from the driver who is impaired. Any litigant wishing to challenge a test for alcohol consumption has a wealth of resources at his or her disposal and the luxury of a fixed and objective criteria for determining culpability. The exact opposite is true in drug testing. The vast majority of scientific research for analyzing drugs and their effects was completed in the mid-1970s. Since that time, there has been a great deal of research performed with regards to drug testing in the area of medicine. Medical research, however, is geared toward the needs of doctors in treating their patients and is often ill suited to adoption by other sectors of society such as law enforcement. An emergency room doctor, for example, may wish to conduct a screening test for the presence of alcohol or other drugs in order to quickly determine the proper mode of treatment; speed may be more important than absolute accuracy and oftentimes the results are not quantified. In the legal realm, accuracy is more essential than speed and a positive drug test result often means little unless a level of intoxication can be established. Much less research has been conducted on drug testing for forensic purposes in recent years. Thus, the literature available today on the subject of drugs and their intoxicating effects is largely unchanged from the information gained in the 1970s. Such resources may be wholly inadequate in coping with modern drug laws. Terminology. The reliability of a test depends upon how specific and sensitive the antibodies are to a given compound. Specificity is the ability of the test to distinguish one drug from another,2 whereas sensitivity refers to the smallest amount of the drug that can be detected in a sample.3 The term analyte merely means the substance being analyzed (identified or quantified) in a given test. After a drug is consumed, it is immediately absorbed into the blood steam. As the blood passes repeatedly through the liver and various other parts of the body, the drug is broken down by enzymes. Metabolites are the byproduct(s) of the drug that has been broken-down. Qualitative screening techniques are simply able to detect “some prior use of a drug,” whereas a quantitative test would indicate the amount consumed. Detection Methods. Various chemical techniques are used to test samples for the presence of drugs. Testing methods differ substantially as to cost, accuracy, time required for analysis, selectivity, and sensitivity.4 Generally, inexpensive and fairly rapid analytical tests [typically, either an immunoassay (EMIT, RIA, TDx) or thin-layer chromatography (TLC)] are used in the initial screening for the presence of drugs. Positive test results are often confirmed or questioned by a more expensive, more accurate testing procedure (usually by gas chromatography/mass spectrometry). The two most commonly used tests for the detection of drugs or drug metabolites are the immunoassay screening tests and the gas chromatography/mass spectrometry (GC/MS) tests. Thin-layer chromatography, in comparison to the immunoassay methods, is neither sensitive, selective, nor effective, and ought to be regarded as inadequate and obsolete.5 Blood, urine or hair samples can be utilized in any of these tests and there are advantages and disadvantages to each. The acquisition of a hair sample is the least intrusive collection method but hair testing is, by far, the most problematic method from a forensic standpoint.6 While urine testing is the preferred method of the National Institute on Drug Abuse, it is easier to take a blood sample from an unwilling subject than it is to collect a urine sample from that person. Because of the problems associated with hair testing, the practitioner will almost always be dealing with blood or urine tests. Immunoassays The immunoassays are drug-specific tests which are able to detect and identify the presence of a certain drug or its metabolite(s) in a person’s system by the binding of antibodies to the drug analytes being assayed in the test sample. Samples used in the immunoassay screening process are from urine, plasma, and/or blood serum. The antibody-drug analyte interaction can be labeled and measured by an enzyme link [enzyme multiplied immunoassay (EMIT) tests], a radioactive label [radioimmunoassay (RIA) tests], or florescent tag (TDx tests). The appeal of the immunoassay technique is based on the following characteristics: small sample size (microliters), direct detection, sensitivity (nanogram amounts of drug per milliliter of sample), instrument automation (and therefore precision and speed), and inexpense.7 The data received from immunoassay testing provides no useful evidence as to when the drug was consumed, the potency of the drug consumed, or even how much of the drug was consumed. The most controversial aspect of the immunoassay tests is their lack of specificity, because there are few antisera8 that are specific for a single compound,9 which leads to the problem of false positives. Further, immunoassays can only discriminate between the presence or absence of the suspected drug.10 Immunoassays, therefore, are at best semi-quantitative, meaning that the results yield a mere estimate of the quantity of the drug in a person’s system. Immunoassays are also deficient in that positive tests results may be the consequence of cross-reactivity; this also creates a problem of false positive test results and lower specificity rates. Cross-reactivity occurs when the chemical properties of over-the-counter drugs and/or prescription medications are mistaken as the chemical properties of illicit drugs. Due to the need for timely results and minimized costs, most companies, police and probation agencies, and even the federal and state governments, utilize the convenient immunoassay tests. However, forensic purposes in criminal and civil litigation demand precision and accuracy, and necessitate the ability of drug tests to quantify the amount of drugs or metabolites in a given sample, and further give evidence of the time, potency, and quantity of consumption. As has been demonstrated, the immunoassay tests fail to sufficiently satisfy these requisites. Manufacturers claim that the immunoassay tests are 95 to 99 percent accurate. These accuracy rates are based on results received when the testing is completed by the manufacturer’s technicians in closely monitored conditions. However, drug testing equipment is often-times sold to private laboratories and hospitals where unqualified and/or untrained technicians execute the tests and interpret the results. As a result, accuracy rates are decreased by means of human error. Thin-Layer Chromatography Thin-layer chromatography (TLC) is a method by which extracted urine samples are separated into components. The urine components are then placed on a glass plate and a solvent is added to force the urine components to travel to various places on the glass plate. The urine components and solvent are then sprayed with a solution that produces different colored spots; the different colors are representative of different drugs. Individual drugs will consistently travel to the same location on the glass plate; the colored spots at a specific location denote the presence of a particular drug or drug metabolite(s). There are, however, several problems with the TLC test. As previously mentioned, thin-layer chromatography ought to be regarded as obsolete because of its lack of sensitivity, selectivity, and efficacy. Cross-reactivity is also a problem; thus, certain prescribed or over-the-counter drugs will migrate to nearly the same location and have the same colored denotation as certain illegal drugs. Furthermore, TLC tests are merely qualitative, that is, they are only able to indicate either the presence or absence of a drug or drug metabolite(s). However, the TLC test’s qualitative ability is questioned by the test’s inability to detect low level concentrations of drugs or drug metabolites. Even in cases where there is a high enough concentration to be detected by the TLC test, thin-layer chromatography is unable to quantify any such positive result. Gas Chromatography/Mass Spectrometry The gas chromatography/mass spectrometry (GC/MS) tests are the most sensitive and specific screening techniques used in the identification of drugs and drug metabolites. GC/MS can be used to test for the presence of drugs and drug metabolites in urine and in blood. Gas chromatography (GC) separates molecules into different components. The components are then vaporized by a gas and are carried through a metal or glass packing tube or column. Identical components will travel through the tube at identical speeds. At the end of the column is a mass spectrometry (MS) detector that identifies, records, and quantifies the molecular components by breaking them into fragments. The mass spectrum of a drug consists of the manner in which that drug will break up into fragments. Each drug has a unique and specific mass spectrum which enables its identification. Unlike immunoassay screening, GC/MS tests are able to quantify drug concentrations in a fluid sample; they are also the least likely of the drug screen tests to encounter the problem of cross-reactivity. Both the qualitative and quantitative levels detected, however, vary among individuals due to factors such as diet, weight, health and drug use history. These testing techniques are very costly ($100 - $200 per test), and are therefore used primarily for the confirmation of preliminary test results. Manufacturers claim that the GC/MS tests are 100 percent accurate. Again, the facility and the individual who conducts the testing procedure and interprets the results may have an effect on the actual accuracy of the test. Forensic Toxicology. As has been previously stated, drugs or their metabolites can be detected through the use of various tests which screen for particular analytes. The chart demonstrates the relationship between consumption of a drug and the time period in which that particular drug is detectable in a person’s system. Several factors influence the length of time a drug or its metabolite(s) may be detected, such as the person’s drug use history, age, sex, weight, health, and even the actual physical and chemical properties of the drug itself. Different studies yield different results with regards to detection times for various drugs. Approximate detection times are listed. The conclusion to be drawn from the information illustrated in the chart is that drug tests, via urine or blood samples, are simply able to detect the presence of drugs in a person’s system for quite some time after consumption. The chart also demonstrates that the length of time a person has been consuming drugs (i.e., a single dose versus chronic use) affects the period during which metabolites or derivatives may be detected. However, such drug tests do not provide quantifiable evidence as to the amounts a person may have consumed at any given time. Furthermore, “Even if a drug or metabolite in urine is positively identified and precisely quantified, there is as yet no scientific basis for forming opinions as to when, how often, and how much drug was used — or on the past, present, or future effect of the drug on the performance, health, or safety of [the person tested].”13 Although the publication quoted here concentrates on the issues surrounding drug testing in the work place, the same conclusions can be applied to any drug testing situation. Thus, positive drug test results establish nothing more than some prior use or exposure, and fail to provide adequate evidence of intoxication or impairment. In its attempts to regulate drug testing procedures, the federal government has established sensitivity test levels for the detection of drugs. The recommended cutoff levels, however, are based not on any established correlation between consumption and behavior but rather on the analyzing capacities of the tests themselves. Identification of drugs or their metabolites in a person’s system has increasingly become an element of evidence in criminal and tort liability cases. Positive drug test results may be used as evidence for a defense, such as in cases involving a diminished capacity and/or involuntary intoxication defense. Conversely, in prosecutions for driving under the influence of drugs14, the drug test result may be the central piece of evidence in the prosecution’s case. Similarly, positive drug test results may be used as evidence of probation violation or drug possession. The problem with reliance by either the prosecution or defense upon such methods is that positive drug test results are often inaccurate. False Positives A false positive test result occurs when a drug-free sample is reported as having tested positive for the presence of a drug or its metabolite(s). In a study conducted in 1983 by the Center for Disease Control (CDC), labs reported 152 false positives in 106 of the 160 urine samples (or 66.5 percent).15 To the technicians conducting the tests, the discrepancy between the number of positive results and the number of “tainted” samples in the figures above would suggest that several of the subjects tested were poly-substance abusers, while in fact the test had detected several false positives within a single drug-free sample. There are several factors that may cause such a result: passive inhalation, improper laboratory procedures, contaminated laboratory equipment, mixed up samples, lost or mixed up paperwork, cross-reactivity with other legal drugs, tampering with samples, and various other unknown reasons. An additional danger of false positives has arisen in situations of periodic monitoring when the person being tested loses weight, as often happens during the rehabilitation period after drug usage has discontinued. Many drugs (especially THC, the main psychoactive chemical in marijuana) are highly fat-soluble and therefore tend to accumulate in stored body fat, allowing them to be detected for weeks after consumption has ceased. The more serious problem surfaces when a former drug user loses weight, since the drugs remaining in the fat cells are released as these cells shrink. This person will again test positively for drugs.16 False positive drug results are especially common where mass drug screening programs have been implemented. Quality control is often sacrificed in order to perform mass screenings in a timely fashion. Errors may arise at any stage in the process, including but not limited to the initial collection of the specimen, identification of the specimen, and the final recording of the results. Furthermore, there is no established government regulation of drug testing and screening. Administrators and examiners of the samples and testing procedures are often not sufficiently competent or capable of administering, analyzing, or reading drug tests and their results. Cross-Reactivity Cross-reactivity occurs when over-the-counter drugs or drugs obtained legally by prescription cause a drug test sample to test positive for the presence of illicit drugs or drug metabolites. For example, Ibuprofen (the main ingredient in Advil® and Nuprin®) has been found to produce a false positive test result when screening for marijuana, as have poppy seeds when screening for opiates. Likewise, quinine (an ingredient found in tonic water) may produce a false positive result for heroin use, and appedrine (the active ingredient found in many diet pills) has yielded false positive test results for amphetamine use. Even more astonishing, and conceivably troublesome for dark-complected individuals, is the fact that melanin, the skin-pigmentation substance found in all humans, may produce a false positive result for marijuana consumption. The table below 17 illustrates several examples of over-the- counter drugs and prescription drugs that may cross-react with various illicit drugs. Unknown or Involuntary Ingestion. Even in the absence of active drug consumption or inhalation, a drug or drug metabolites may be detectable in a person’s system. Exposure to secondhand smoke or the unknown ingestion of food or drinks containing illicit drugs are two factors that may result in positive drug test results. Depending on the amount and length of exposure, positive tests results may arise for several days after exposure. Further, drug testing methods are unable to differentiate between active or passive inhalation or ingestion. While a person may have unknowingly or involuntarily been exposed to or ingested a certain quantity of drugs, without ever experiencing the mentally or physically intoxicating and/or psychoactive effects of the drugs, sensitive testing methods will register the results in the same manner as if a person had actively consumed a drug. Effects on Performance. The chemical properties of various drugs complicate the issue of whether levels of impairment can be predicted. Drugs, once ingested or inhaled, are absorbed by the blood stream and circulated throughout the body. The drugs concentrated in the blood stream are quickly metabolized and/or redistributed. The metabolites are then slowly eliminated from a person’s system. This process may take several days or several weeks, depending on the amount consumed and previous drug use history. Drug metabolites may then remain detectable in a person’s system throughout and even after this elimination period, depending on the chemical makeup of the drug.18 Drugs or drug metabolites, therefore, are detectable long after the drug’s effects or impairment have worn off. “Except for alcohol there is a lack of published data on concentrations in blood, and none at all for drugs in urine upon which an expert may base an opinion of drug-related impairment or improvement.”19 While some drug testing methods can quantify the amount of drugs or drug metabolites in a person’s system, the quantity ingested, potency of the drug ingested, frequency of ingestion, time of ingestion, and the level of drug effect or intoxication are indeterminable. In cases where a person is accused of driving under the influence of drugs, establishing intoxication is a necessary element to prosecute such an offense. There is general agreement within the scientific community that urine concentrations for most substances cannot be correlated with performance, health, or safety, as there is no generally accepted scientific foundation for correlating urine concentrations with performance.20 Further-more, there is no scientific evidence that supports a fixed relationship between quantity detected and conduct or impairment. This is due, in part, to the fact that some drugs affect individuals differently21 and to the fact that the effects, or “high,” from a drug wear off in a matter of hours.22 This is not to say that it would be impossible to identify a fixed relationship between the quantity of a drug revealed in a test and specific mental or physical impairments but, to date, the research has not been done to establish such a relationship. Thus, at the present time, a positive drug test result merely reveals that a person has either actively unknowingly or involuntarily inhaled or ingested some quantity of drugs. It does not establish that a person has been physically or mentally affected or impaired by the drug. It is also entirely possible that the positive results may have been the consequence of errors in the testing procedure or in the test itself and that the test result is simply wrong. Admissibility of Drug Testing Evidence Arguments made in opposition to the admission of drug testing evidence will generally be centered on the standards set forth in Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharma-ceuticals, Inc.23 Although Daubert sets a more permissive and flexible standard for the admission of scientific evidence, it is not necessarily a boon to the prosecution. Under the old Frye24 test, drug test evidence would come in merely because a given test had achieved a somewhat amorphous acceptance in the scientific community; defects and linkage were questions generally going to weight rather than admissibility. The Daubert approach compels the trial court to focus on the questions of reliability and relevance. The focus on reliability makes it clear that the judge has some preliminary responsibility to insure that the tests and opinions derived from them are scientifically defensible and, as importantly, leaves the door firmly open for extensive cross-examination and rebuttal evidence. The focus on relevance compels the court to scrutinize what is arguably the weakest feature of drug test evidence: whether there is a firm connection between a fact at issue in the case and the fact that would tend to be established by the admission of a drug test result. Take, for example, a case in which the defendant is accused of having sexually assaulted the complainant in a date rape case. The defense is consent and there is no evidence of a struggle, screaming or prompt reporting. The prosecution seeks to explain this lack of corroborating evidence by admitting a drug test report showing that the complainant tested positive for marijuana the next day. He wants to argue that she was too high to resist and too high to consent. Under Daubert, a strong argument can be made that the test result is inadmissible. Presuming no other flaws, it merely shows that the complainant was exposed to marijuana within three or four weeks of the date of the incident. The relevance is slight; it does not rule out the possibility that the complainant was high on the evening in question but it certainly does not come close to proving it either. It merely shows that it was possible. Secondly, the drug test does not show that there was ever a point in time that the marijuana to which she had been exposed ever had any psychoactive effect on her. Thirdly, there is no known scientific link between marijuana consumption and the ability or inability to consent to sexual relations. The test results should be excluded here because the relevance is slight and it does not reliably prove what the proponent of the evidence wishes to establish. Conclusions In its efforts to win the “War on Drugs,” Congress and state legislatures have sought out ways to penalize individuals for drug use. Nevertheless, science has not yet caught up with what the legislators seeks to punish. Unlike the “war” on alcohol, where there was scientific evidence to prove the correlation between amounts of alcohol consumed and the effects it has on a person’s conduct and/or impairment, and thereby justify regulation, science and technology are at present merely able to detect the presence of drug metabolites. Moreover, unlike alcohol, there is no evidence to support the inference that there is a fixed relationship between the quantity of drugs or metabolites identified in, for example, an individual’s urine, and the degree of mental or physical impairment experienced by the user. There are also serious problems associated with the use of these tests for forensic purposes. For example, the EMIT immunoassay test is frequently used for detecting the presence of marijuana or its derivatives in a person’s urine. While the EMIT is able to detect and identify cannabinoids, it is unable to provide any indication of when the marijuana was consumed or absorbed. Marijuana derivatives may remain in an individual’s system for days or weeks after marijuana is consumed.25 The effects or “high” from the use of marijuana wear off in a matter of hours after consumption of the drug.26 Thus the derivatives remain in a person’s body long after any physical or psychological effects of the drug have worn off. Of the research that has been conducted for the purposes of forensic drug testing, much of it has been qualitative in nature and result. Scientists are able to chemically analyze bodily fluids for the presence and identification of a drug or its metabolites in a person’s system. However, scientific means are thus far unable to determine when the drug was consumed. Ultimately, the question which must be answered in the issue of drug testing is not only whether it is constitutionally permissible or scientifically possible, but also whether it makes for good public policy. Justice Marshall illustrated this point best in his dissent in Skinner v. Railway Labor Executives’ Association:27 History teaches that grave threats to liberty often come in times of urgency, when constitutional rights seem too extravagant to endure. The World War II relocation-camp cases, Hirabayashi v. United States, Korematsu v. United States, and the Red scare and McCarthy-era internal subversion cases, Schenck v. United States, Dennis v. United States, are only the most extreme reminders that when we allow fundamental freedoms to be sacrificed in the name of real or perceived exigency, we invariably come to regret it. [citations omitted] It is the role of defense counsel to ensure that the rules which emerge are not only constitutionally friendly, but also scientifically sound.28 Notes 1. A. J. McBay, Drug-Analysis Technology — Pitfalls and Problems of Drug Testing, Clinical Chemistry 33.11(B) (1987). 2. See Beverly Potter and Sebastian Orfal, Drug Testing At Work: A Guide For Employers and Employees (Ronin Publishing, 1995; ISBN: 0-914171-70-4) 46. 3. Id. at 46 4. Id. at 35 5.Bryan S. Finkle, Drug-Analysis Technology: Overview and State of the Art, Clinical Chemistry, 33.11(B) (1987). 6. Hair samples are particularly susceptible to environmental contamination; hair will become tainted by drug residue in the air from drugs which are smoked (marijuana, crack, heroin) or by physical, external contact with the drug. Thus a hairdresser or girlfriend, with residue on their hand, could unwittingly contaminate a test subject’s hair. Tests with a high threshold, designed to screen out environmental contamination, will also screen out many users. Washing the samples prior to testing may also wash out the substance which was absorbed into the hair through the subject’s body, thereby defeating the purpose of the test. Furthermore, samples may reveal a substance use history dating back months or years, defeating its use, for example, as a probation oversight tool. See generally Robert O. Bost, Hair Analysis –Perspectives and Limits of a Proposed Forensic Method of Proof: a Review, Forensic Science International 63 (1993), 31-42. 7. See Finkle, supra. 8.Antisera: Plural of antiserum, a serum containing antibodies. (Webster’s New Collegiate Dictionary, Springfield, MA: Merriam-Webster, 1979.) 9. John P. Morgan, M.D., Problems of Mass Urine Screening for Misused Drugs, Substance Abuse in the Workplace (San Francisco: Haight-Ashbury Publications, 1984) 21. 10. Potter and Orfal 35-6 11. Id. at 45. 12. M. A. Huestis, J. E. Henningfield, and E. J. Cone. Blood Cannabis I: Absorption of THC and Formation of 11-OH-THC and THCCOOH During and After Smoking Marijuana, Journal of Analytical Toxicology 16 (Sept/Oct 1992). 13. See McBay, supra. 14. Contrary to MCLA 257.625: MSA 9.2325 15. Center for Disease Control, The Results of Unregulated Testing, Journal of the American Medical Association (April 26, 1985). 16. Cynthia Kuhn, Scott Swartzwelder, and Wilkie Wilson, Buzzed: The Straight Facts about the Most Used and Abused Drugs from Alcohol to Ecstasy (New York: W. W. Norton & Co., 1998) 239. 17. The information contained in this chart is based on Potter and Orfal at 155, citing Abbie Hoffman with Jonathan Silvers, Steal This Urine Test (Penguin Books, 1987). 18. See drug detection rate chart for more specifics. 19. McBay 39(B). 20. Reese T. Jones, Drug of Abuse Profile: Cannabis, Clinical Chemistry 33.11(B) (1987), citing A.C. Moffat, Monitoring Urine for Inhaled Cannabinoids, Arch Toxicol 9 (Suppl.) (1986): 103-10. 21. Kuhn, et al. 22. See McBay, supra. 23. Daubert v. Merrell Dow Pharmaceuticals, Inc., 113 S .Ct. 2786, 509 U.S. 579 (1993) 24. Frye v. United States, 293 F. 1013 (1923) 25. Jones at 75(B). 26. Id.at 79(B). 27. Skinner v. Railway Labor Executives’ Association, 489 U.S. 602 at 635, 109 S. Ct. 1402 (1989). 28. A number of the standard scientific evidence treatises also address drug testing; these include Modern Scientific Evidence by Faigman, Kaye, Saks and Sanders (West, 1997), Scientific Evidence by Gainnelli and Imwinkerlried (Michie, 2nd ed. 1993) and Scientific Evidence in Civil and Criminal Cases by Moenssens, Starrs, Hendeson and Inhab (Foundation Press, 4th ed 1995). | |||||||||
 |
National Association of Criminal Defense Lawyers (NACDL) 1150 18th St., NW, Suite 950, Washington, DC 20036 (202) 872-8600 • Fax (202) 872-8690 • assist@nacdl.org | |||||||||
