Robert E. Gladd,
Thesis work-in-progress internet edition:

UNLV Institute for Ethics & Policy Studies

Chapter 2,
The epidemiology of illicit intoxication:
Estimating the extent and cost of workplace drug abuse

Quantification and assessment of the extent and impact of illicit drug use involve analyses of empirical data culled from a variety of sources including databases of hospital and clinic encounters, epidemiological studies conducted by institutions such as the U.S. Centers for Disease Control (CDC), the National Institutes of Health (NIH), the U.S. Substance Abuse and Mental Health Administration (SAMHSA, formerly NIDA, the National Institute for Drug Abuse), as well as studies undertaken by a host of university research centers (e.g., CASA, the Center for Alcohol and Substance Abuse research at Columbia University) and advocacy groups and private foundations such as the Partnership for a Drug-free America and the Robert Wood Johnson Foundation.

The scientific quality of these investigations runs the gamut, from the dispassionately professional and meticulous to the patently absurd and propagandistic. One encounters estimates of the “prevalence” (i.e. proportion, or rate) of drug-using employees that vary wildly, from a low of one or two percent to statements asserting that “Recent government statistics reveal that 1 out of 6 workers has a drug problem” (from a Psychemedics hair test marketing brochure) or that “Estimates of on-the-job cocaine use (including crack) range from 10% to 50% of all employees.” (Angela B. Miller, Working Dazed: Why Drugs Prevade the Workplace and What Can Be Done About it, Plenum Press, 1991, pg. 15. emphasis mine) .

A favorite tactic of some anti-drug advocacy groups seeking to inflate the apparent extent of the “problem” involves the aggregation of data covering prescription, over-the-counter, and illegal drugs, as well as alcohol and tobacco statistics, to be reported under the hazy rubric of “substance abuse” or “consumption of illegal and abused drugs.” For example, see “Alcohol and Other Drugs in the Workplace,” a page of statistical assertions proffered by NCADD, the National Council on Alcoholism and Drug Dependence, Inc. Their data overwhelmingly concern alcohol abuse, with the word drugs seemingly thrown in the mix for its marquee value. No instructive breakdowns are provided, i.e., alcohol vs. “drugs,” and within “drugs,” no stratification by type of drug (“licit” vs. “illicit,” by each substance?). Another example: Joseph Califano’s CASA reports that “92% of substance abuse-related health entitlement costs is spent to treat the consequences of tobacco, alcohol, and drug abuse. Only 8% is spent to treat alcohol, drug, or tobacco dependence.” (emphasis mine) Note again the generality. Visit the CASA website wherein these obfuscatory data reside. Observe in particular the artsy left-hand margin wallpaper montage, replete with totemic rolled-up $100 bill and powdered “cocaine” down the page aside the lengthy conglomerate litany of “tobacco, alcohol, and drug” statistics.

Where, one might rightfully wonder, are the Joe Camel and Johnny Walker Red renderings? A curious omission, given that a recent Califano article quotes the very same 1993 JAMA epidemiologic data cited by Ron Kotulak’s Inside the Brain. Recall my main page characterization of illicit drug mortality experience as a “relatively minor epidemiological concern” in the context of other, much more prevalent and severe sources of substance abuse harm.

Our leaders and citizens focus on the top killers: heart disease (720,000 deaths in 1990), cancer (505,000), stroke (144,000), accidents (92,000), emphysema (87,000), pneumonia and influenza (80,000), diabetes (48,000), suicide (31,000), chronic liver disease and cirrhosis (26,000), and AIDS (25,000). But they give scant attention to the causes of these killers, which, according to a 1993 Journal of the American Medical Association study, include tobacco (435,000 deaths), alcohol (100,000) and illicit drug use (20,000).

Joseph Califano, It's the Drugs, Alcohol, and Tobacco, Stupid

A last and fairly recent example of tobacco/drugs/alcohol conflation reveals the intractability of this addiction to data fog. The 1997 book Drug-impaired Professionals (Robert H. Coombs, The President and Fellows of Harvard College, publ, Cambridge, MA), opens its Preface with the assertion that [D]rug abuse is at least as prevalent among highly regarded professionals as among the general public.” This sentiment is shortly thereafter echoed in a chapter one section heading entitled Addiction: An Equal Opportunity Destroyer. The body of this work, however, once again provides the vague, mostly alcohol-referent rhetorical goulash heretofore surveyed:

Hickey (1990, p. 37) contends that many attorneys have difficulty admitting to themselves that they cannot manage their drinking. (Coombs, p. 13)

Airline pilots also “tend to see themselves as invincible.They see themselves as different from the average citizen because they are in a super-responsible position” an airline employee-assistance program (EAP) representative remarked. ”I’m a cracker-jack pilot,” reasoned a pilot (age 45). “It can’t happen to me, I'm not a skid-row character.”A treatment expert described the addicted pilot’s attitude like this: “How can I be an alcoholic when I'm the captain of a 747 aircraft.” (p.14)

The impact of substance abuse on professionals and their associates can be devastating. Obsession with alcohol and other drugs undermines physical and mental health; it also diminishes and destroys professional lives...An interview study of 86 pharmacists recovering from chemical dependency found that 44 had been arrested, and 24 had spent at least one night in jail. Forty five had experienced unemployment because of drinking or other drug use... (pp. 14-15)

Millions of Americans suffer and die from alcohol and drug abuse that often goes undiagnosed and untreated. About 43 percent of U.S. adults (76 million people) have beeen exposed to alcoholism in their families. They either grew up with, married, or had a blood relative who was an alcoholic or a problem drinker (Schoenborn 1991). (p. 26)

A national probability sample of U.S. households...found that 52 percent of Americans age 12 and older had used alcohol during the month preceding the survey. (p. 27)

Little is known about how many commercial airline pilots use marijuana and other illicit drugs, but a serious alcohol problem clearly exists. (p. 35)

A national study of 3,338 law students at 121 accredited U.S. law schools found that 14 percent had drunk alcoholic beverages 10 or more times during the previous month, and 3.8 percent admitted to daily use. (p. 33)

The foregoing is but a sampling of the murky substance abuse assertions that purport to sustain this work. Speaking of “sampling”—how, we might ask, did this author arrive at his conclusion that “[D]rug abuse is at least as prevalent among highly regarded professionals as among the general public”? Coombs describes his epidemiological methodology in the third paragraph of his Preface:

From 1992 through 1995 my assistants and I spoke with 91 addicted professionals (66 men and 25 women)—21 physicians and medical students, 11 dentists, 13 pharmacists, 12 nurses, 21 attorneys, and13 pilots; 10 experts (12 men and 7 women) who assist them in recovery; and 5 other people who felt that others might benefit from their experiences...

Under the Revival tent, such is called the leap of faith. Under the Big Top, it is known as working without a net.” In science it is simply called “anecdotal”—an insufficient “n,” with the corollary liability of built-in sampling bias: ungeneralizability, in a word.

Coombs offers up a curious conclusion in his Epilogue:

Tobacco and alcohol, the most widely used drugs, though legal cause more misery than all illicit drugs combined. (p. 281)

Interesting. Precisely a central point of this thesis.War on Drugs partisans are utterly disinterested in such an observation, however. They are certain to brandish this type of book (title prominently displayed for the cameras) as if it were one more (illicit) drug abuse Dead Sea Scroll justifying extreme countermeasures against everyone.

Valuing the drug abuse “losses”

Solid estimates of drug abuse prevalence are difficult enough to come by, but when we get to the appraisal of “economic losses” attributable to drug use, it often seems that policymakers just pick a large round number out of thin air with which to argue for public and political support. In the preamble to the 1988 federal “Drug-Free Workplace Act,” Congress summarily “finds” that drug abuse is “prevalent” and that it costs the U.S. $100 billion dollars per year in health, safety, and productivity “losses.” Where does such a figure originate? Scientific American writer John Horgan explains one way to derive it (see Your Analysis is Faulty, The New Republic, April 2, 1990):

Such empirical “reasoning” is all too common, the “correlation = causality” disease of the statistically credulous (many of whom are supposedly experts in their fields). Horgan wryly asked “ similar logic, should we conclude that Thunderbird wine hurts productivity but Chivas Regal scotch helps it?”

This type of inept inquiry and baseless calculation is nothing new. Recall from Chapter 1 Dan Baums account of impossible drug-related theft totals bandied about in the early 1970s. Another example of this type of data inflation is detailed in a 1971 monograph entitled The vitality of mythical numbers (see Judgment Under Uncertainty: Heuristics and Biases, Kahneman, Slovic, & Tversky, Ed.), in which author Max Singer evaluated a popular claim of the day alleging that “It is generally assumed that heroin addicts in New York City steal some two to five billion dollars worth of property a year...” A careful look through all pertinent sources of data led Singer to the conclusion that the “$2-5 billion” figure, while not inconsequential in absolute terms, was high by about a factor of ten.

NOTE:When confronted with an assertion such as “70% of all criminals used illegal drugs prior to arrest,” it is important to remember that 100% of them are also likely to have consumed water a short time prior to their arrests (and, all were “under the influence” of oxygen at the time they were detained). With respect to “gateway” substances, virtually all marijuana smokers started out on breast milk or Similac.

Moreover, it is, after all, a blinding glimpse of the obvious that those who engage in criminal activity will likewise have little regard for anti-drug laws. What is not clear, however, is that their drug use uniformly caused them to pillage and plunder.

The confusion of correlation and causation is probably the most frequently committed inductive error. Illicit drug use may indeed correlate significantly with all manner of workplace malaise, but it also correlates highly with alcohol use, tobacco use, poor diet, lack of exercise, and sleep disorders, to cite a handful of major factors. Such inter-correlations indicate a more global “factor”—call it “dysfunctional lifestyle”—that predicts poor performance even more effectively. Singling out one element of such a syndrome for coercive suppression will inevitably fail to lead to our indisputably laudable goals of improved health, safety, and productivity.

While no intellectually honest person can deny that drug abuse is a serious social problem, no one truly knows what the prevalence and costs of illicit workplace drug use are with any sort of precision. The most disinterested, objective, and comprehensive study to date was recently concluded by the National Academy of Science’s (NAS) National Research Council and the Institute of Medicine. Their 1994 hardcover report Under The Influence? Drugs and the American Workforce delivered the following among their principal findings and recommendations:

With respect to this last point, it is clear that some policy makers have not gotten the message. Witness the remarks made by Representative Gerald Solomon (R-NY) on the opening day of the first session of the 104th Congress, in a screed entitled “Redeclare the Drug War”:

On April 6, 1995 Mr. Solomon repeated his litany of drug casualty statistics in another House speech, with a couple of embellishments: instead of 11 percent of all live U.S. births being “drug-exposed,” he now claimed that “Today, 1 out of every 10 babies born in the United States is addicted to drugs” (emphasis mine) and asserted that “The nation’s health care system is straining from the war on drugs with nearly 500,000 drug-related hospital emergencies a year.” Given that U.S. Census Bureau figures on live births are 4,086,000 for 1994, Solomon implies an excess of 400,000 “drug-addicted” American newborns annually, an obstetrical disaster of major proportions that would be continuously blaring front-page news.

Were it in fact the case. Mr. Solomon acquires his “data” from The Partnership For a Drug-Free America and, no doubt, from its kindred and empirically incestuous advocacy brethren. A half-million druggie newborns a year? (the foregoing 10-11% figures.) Well, if we return for a moment to CASAand It's the Drugs, Alcohol, and Tobacco, Stupid, we find Mr. Califano claiming that “[T]he more than 500,000 newborns exposed each year to drugs and/or alcohol during pregnancy is a slaughter of innocents of biblical proportions.”

Once again: Drugs-and/or-Alcohol.

This is mantra, not measurement; allegory posing as analysis. Respectable empirical information is more likely to come from places like the CDC (U.S. Centers for Disease Control). On October 18, 1996, for example, CDC reported the findings of a 1994 Georgia Department of Human Resources pregnancy drug abuse study. Georgia health officials had anonymously tested every newborn in the state during a two-month period, and found 1 in 200 had been “exposed” to cocaine before birth. One half of one percent “exposed,” however arguably unrepresentative of the aggregate national obstetrical experience (and confined to assay for cocaine metabolite), is a very long leap from Solomon’s “10% addicted.”

The Georgia data are available in full online in the CDC Weekly Mortality and Morbidity Report, Volume 45, No. 41, October 18, 1996, located at, which requires an Adobe Acrobat reader for web viewing. Alternatively, view a one-paragraph CDC web page summary here.

All medical encounters are recorded and classified through standardized coding protocols. Everything diagnosed about or done to a patient is coded. Hospitals use the ICD-9-CM system, individual practitioners employ the CPT system. Computerized reimbursement systems rely on these codes for automated payment of charges, and “code gaming” (coding clinical episodes with an eye toward at once minimizing oversight and maximizing payment) has evolved into a fine art. Those who research clinical data repositories via computers face significant accuracy challenges.

To illustrate: When we probe hospital records for cases coded as “drug-related” encounters, in addition to illicit drug trauma, we find everything from acute alcohol intoxication to accidental poisonings to suicide attempts to “ADRs,” or “Adverse Drug Reactions,” which often means allergic reactions to therapeutic agents legitimately administered (the single largest category of “drug-related emergencies”). For example, a search for “drug-related” episodes in the 152,964 cases comprising the 1993 Nevada statewide hospital database turned up 4,619 cases, 3,730 of which turned out to be alcohol-related, and of the 889 remaining, only 237 could be legitimately classified as “acute (illicit) drug admissions.” Lacking national data at the moment, I could extrapolate linearly and multiply those 237 by the ratio of the U.S. population to Nevada’s (roughly 250/1.5 million, or 167), so 237 x 167 = 39,579. I’ll even throw in a comfortable pad and round up to 50,000, and, voila—one tenth of Representative Solomon’s estimate. Recall Disraeli’s lament: “Lies, damned lies, and statistics.”

Regarding “drug-addicted” neonates, the 1993 Nevada hospitalization data contain 19,997 records coded for “live birth.” Of those, 104 had “fetal/newborn” drug-related diagnostic codes, 30 of which were ICD-9-CM code 760.71, or Fetal Alcohol Syndrome. In light of the foregoing CDC Georgia data let us do a little quick math: 104 less 30 is 74, which when divided by 19,997 is 0.0037, or slightly less than one half of one percent.

    Tip: Want to inflate the apparent extent of drug abuse encounters? Hypothetical
    (and possibly representative?) case: A couple are out partying. They get quite drunk, smoke some pot, snort a bit of cocaine, and become involved in an injury automobile accident on the way home (the proximate cause of which was the alcohol). At the hospital (not to mention any and all follow-up clinical encounters) they generate a host of ICD-9 codes for every diagnosis and treatment. Two people involved in one incident have now contributed possibly dozens of “drug-related” codes that will become the grist for drug abuse prevalence “researchers” eager (or merely naive enough) to inflate the numbers concerning illicit drug morbidity by counting each code hit as a “drug-related encounter.”

In fairness, the Nevada Claims Database I used (Source: UNLV Center for Public Data Research) does not contain all possible diagnostic and procedural codes that could accompany a given encounter, so my queries may indeed undercount the phenomenon slightly. But the principal codes are captured (including those pertaining to emergency room episodes), so these numbers are probably not off by much. The point is that acquiring solid epidemiological data is no easy task, one utterly beyond the ken of the hapless Mr. Solomon and his “sources.”

An additional observation regarding the box just above; The interested reader can examine in detail the diagnostic codes of interest in recent federal national hospitalization data contained in the Detailed Diagnoses and Procedures, National Hospital Discharge Survey, 1994, published by the National Center for Health Statistics (NCHS). This report tabulates an annual estimate of acute-care hospital encounters, by “first-listed” ICD-9 diagnostic codes (a.k.a. the “principal dx”). A couple of summary observations: Legitimate drug-related encounters were estimated at 153,000. The alcohol-related estimate totaled 356,000. These represented 0.5% and 1.15% respectively of the total estimated 30,843,000 hospitalizations. Two conclusions should be evident after even brief consideration. First, there is essentially no such thing as an “acute tobacco-related admission.” (although we do in fact see a dx of 305.1, with 8,000 cases listed as “non-dependent drug abuse, tobacco,” and tagged with the caveat “to be used with caution.” These cases are most likely those of adolescents unhappily regurgitating their clandestine Marlboro adventures.) The damage caused by tobacco consumption, however, will be found sprinkled across a host of other codes identifying costly and severe maladies such as emphysema (dx 492), oral/ trachial/ esophageal/ lung cancers (numerous codes), COPD (dx 496, chronic obstructive pulmonary disease), and cardiovascular ailments (dozens of codes). Similarly, alcohol damage will be evident in liver, renal, gastrointestinal, and neurological ailments, to cite a few. Simple “principal dx” code counting, therefore, provides a misleading picture of the relative impacts of various lifestyle-related medical misfortunes.

Another important point: the authors of this report are careful to point out that their ICD-9 code frequency breakdowns reflects encounters, not individual patients. Since a relatively high proportion of medical services are rendered to chronic, repeat patients, it is easy to overestimate the prevalence of a condition if the databases are stripped of patient identifiers for confidentiality reasons, and code “hits” are summarily tabulated (recall that “prevalence” denotes the percentage of individual patients with a condition). The federal D.A.W.N. (Drug Abuse Warning Network) reports, which estimate “drug-related emergency encounters,” contain a similar disclaimer. The D.A.W.N. report also acknowledges that the medically indigent typically use emergency rooms as primary care facilities. If, for example, a welfare client or homeless person is treated at the E.R. for the flu or some other relatively minor ailment and suffers an “ADR” (Adverse Drug Reaction), or is simply detoxed for alcohol abuse, his or her codes may readily end up in a sloppily designed and executed “drug-related emergency” tabulation.

But surely, one might object, researchers are competent and vigilant against such naivete, right? Recall my main page assertion regarding estimation of the nature, extent, and cost of drug abuse: “Upon close critical examination, serious questions emerge concerning the reliability of the data, most of which are gathered and disseminated by “War On Drugs” partisians intent on buttressing foregone conclusions.” Consider the following recent news item:

Drug Statistics Questioned

When any kind of a revelation about drug abuse trends are made public, so too are the statistics backing up the statement. But some critics charge these numbers are just estimates that in no way should have an impact in making future drug policies, the New York Times reported April 20.

Drug statistics first became prominent in 1978, with the creation of the National Narcotics Intelligence Consumer Committtee. Mark A.R. Kleiman, a drug policy expert at the University of California at Los Angeles and chairman of the committee calls the process "estimation by negotiation." Kleiman says officials typically sit down and debate what the numbers should be. Peter Reuter, a drug policy specialist at the University of Maryland, says the numbers are irrelevant "because they play virtually no role in shaping the nation's drug policies."

And some contend the figures are merely a way to distract attention from the root cause of the problem. Just last year, an annual survey conducted by the University of Michigan reported that half of high school seniors admitted they have used drugs, up 20 percent from 1992. The study showed that the majority of young people are not hard drug users since most students said they experimented with marijuana. Of the students surveyed, less than 2 percent of seniors said they had used cocaine in the previous month and 0.6 percent said they had used heroin.

Eric D. Wish, director of the Center for Substance Abuse Research at the University of Maryland, is also highly critical of the statistics. Recently he wrote, "What is not so obvious is that the federal agencies that produce these statistics are also agents of the administration in power, and are not immune from pressures to interpret national drug statistics consistent with the ruling administration's view."

Date: 04/25/1997, Source: Join Together Online

Need more? No problem; consider the most recent examination of the drug data dubiety issue from The Washington Post:

Number Jumble Clouds Judgement of Drug War

Differing Surveys, Analyses Yield Unreliable Data

Washington Post Page: A01, Jeff Leen, Washington Post Staff Writer, Friday, 2 Jan 1998

...In spending a proposed $16 billion on the federal drug war in 1998—a 400 percent increase since 1986—lawmakers will rely on reams of data that often attempt to impose statistical order on a chaotic social problem that defies easy analysis. Extensive federally funded efforts to accurately assess the subterranean drug world have led to contradictory findings and occasional statistical curiosities, such as a 79-year-old female respondent whose avowed heroin usage in one survey resulted in a projection of 142,000 heroin users, 20 percent of the national total.

“It’s clear that these things are badly mismeasured and nobody cares about it,” said Peter Reuter, the former co-director of drug research for the non-profit RAND think tank and now a University of Maryland professor. “That’s because drug policy isn’t a very analytically serious business.”

Measuring the drug war with any precision is a daunting task. Hard-core drug users are hard to find, much less question, and people frequently lie on drug-use surveys—one study shows two-thirds of teenagers giving deceptive answers. Since surveys typically receive only a small number of positive responses, analysts risk making substantial errors in creating projections for the entire nation. Survey results sometimes include warnings acknowledging these obstacles, such as “subject to large sampling error” or “great caution should be taken.”

But the caveats often are downplayed or ignored, either by those issuing the data or by journalists and others promulgating the information. In reporting the apparent 1991 jump in habitual cocaine use, for example, the White House’s Office of Drug Control Policy noted that the statistics were both “cause for concern” and “highly unreliable.”

The difficulty in measuring and evaluating the nation’s illegal drug problem made it harder to set policy, stoked partisan rhetoric and confused the public, drug analysts say. Many experts, for example, believe cocaine and crack use are in decline, and the federal household survey indicates that overall drug use is down 49 percent from its peak of 25 million monthly users in 1979; yet many Americans still perceive the drug war as perennially lost.

“You really can’t tell from the big debate that goes on in public what the big picture is,” said David Musto, a Yale University medical historian who has studied drug trends for three decades. “When I tell people about it, they’re completely surprised by the fact there has been a decline since 1980.”

That big picture can be obscured by drug statistics that are “often incomplete, erratic and contradictory,” in the words of two RAND researchers funded by the government to measure cocaine consumption. The first problem of drug war analysis is the sheer number of measurements—there are more than 50 federal drug-related “data systems” with hundreds of “drug variables” produced by an array of federal agencies...

To review the full text of the above Post article click here.

With respect to drug testing specifically, the Post article notes problems with the Justice Department’s DUF (Drug Use Forecasting) program, wherein “voluntary” urine samples are collected annually from 30,000 inmates from 23 cities around the nation. Politicians invariably rush to extrapolate from such data to come up with highly suspect estimates of the prevalence of various types of drug use nationally. A similar flawed approach to workplace drug prevalence estimation is noted by John Gilliom in Surveillance, Privacy and the Law: Employee Drug Testing and the Politics of Social Control (1994, University of Michigan Press, Ann Arbor) who observes that:

As noted elswhere in this thesis, the American Management Association recently reported the aggregate national employee positive drug test rate at 1.9 percent (of all employees tested for whatever reason. Given that certain very low-prevalence employment strata have yet to buy into indiscriminate screening, the overall rate would have to be even lower than that).

Clearly something is amiss: either the continuing official workplace drug abuse prevalence assertion of slightly more than 8 percent (10 million workers) is grossly exaggerated, or the testing process is hemorraghing false negatives. Either way, we are wasting analytical resources.

In the foregoing we have seen how partisans have “cooked the books” to exacerbate public anxiety and justify extreme countermeasures, of which indiscriminate drug testing is but one. In the next chapter (Chapter 3), we examine the myriad complexities of the science and business of analytical chemistry. Can the labs manage the current and proposed specimen workloads accurately and cost-effectively?

Click here to return to the main page,
or just hit your Back Button.