Cosby Trial: Cold medicine in an alleged drugging case is … nothing to sniff at.

Bill Cosby is currently on trial for an alleged sexual assault on Andrea Constand in his home near Philadelphia in 2004. Last year, a previous court case ended in mistrial. In the interim, of course, the world learned of many more allegations against the entertainer over the decades.

At various times, Bill Cosby has claimed to have given women cold or allergy medicine. This was, I’m sure, intended to sound more benign than the Quaaludes (methaqualone) that he admitted purchasing, during a deposition in the civil case Constand brought against him in 2005.  He said then that he got Quaaludes to give to women he wanted to have sex with, and also admitted that he didn’t take that drug himself, as it made him sleepy.

But the “cold medicine” pivot doesn’t point to a lack of predatory intent. A little primer about the potentially big effects of diphenhydramine is order.

We commonly know diphenhydramine as an over-the-counter remedy, in the form of a little pink antihistamine; the most well-known brand name of this drug is Benadryl. It can make people drowsy, and in fact many people informally use it on themselves (or even their pets) to coax sleep on an otherwise agitated situation. This is a relatively safe practice if no alcohol is involved, and the dose is kept low. But at doses above the recommended 25-50 mg level, there have been reports of blackouts, feelings of heavy limbs or paralysis, and disorientation. There are also methaqualone and diphenhydramine combination pills; some formulations of these are blue tablets. Higher doses of diphenhydramine alone have been made in blue tablets, as well.

Recently, concern has emerged about nightlife welcoming the “Benadryl cocktail.” For the record, this is an ill-advised mixture. People have been mixing heavier drugs, like tranquilizers, with alcohol for a long time; in the 1950s, the Miltown Martini was much in vogue. I can remember a particularly notorious dorm party at college that involved NyQuil and vodka. Most people’s interest in these concoctions is also voluntary, recreational, and for self-dosing; nonetheless, diphenhydramine can and has been used in a predatory way. Nearly any substance can be pressed into predatory service under the right circumstances. Playing armchair toxicologist, by simply guessing on the basis of a victim’s symptoms, is not wise. To use legalistic language even outside a court setting: you’re piling on additional burdens of proof that are unnecessary if you simplimage of pills falling out of a bottley take a broader view of the total circumstances. The question really is: did someone suffer exploitation or assault, in part because they could not consent or resist?

The 2015 criminal complaint against Cosby takes diphenhydramine seriously – and it should. At a high enough dose, Benadryl absolutely could produce the symptoms that Constand described.  She said that he gave her three blue pills represented to her as herbal supplements, after she mentioned feeling out of sorts. Quickly, heavy symptoms emerged – dizziness, disorientation, and difficulty moving and staying conscious. She says she was also aware of a great deal of sexual touching during this time but was unable to resist or move away.

The prosecution this time around has a refreshing approach to the drugs aspects of the case — refreshing for being circumspect. It’s an approach, I think, with the practical goal of withstanding limitations the court might have (and did previously) place upon evidence about Cosby’s alleged pattern of behavior over the decades. But the role of drugs in the complaint keeps its sights on the underlying offenses. It takes a notably different tack than previous drugging allegation cases in some state courts and before college disciplinary boards. In my 2016 book Drink Spiking and Predatory Drugging: A Modern History, I identified a common, but often perilous, temptation to make drugs that central issue in assault cases, rather than the assault itself.

The key thing here, for establishing the non-consensual nature of the encounter, is the fact of Constand’s incapacity. The specific charge related to Cosby’s drugging behavior relates to facilitation of sexual assault, that he knowingly offered drugs that would lead to the incapacity and thus could not then argue that the encounter was consensual.

In this way, Cosby’s disassembling is the issue, not whether or not Constand took narcotics, high-dose diphenhydramine, or “herbal” supplements from Cosby. She took something from him, as they both agree, and became disoriented, weak, and only sporadically conscious afterward. This criminal complaint keeps it simple, in other words: Constand was in no state to consent, and the defendant’s behavior with whatever he may have given her strongly suggests consciousness of guilt and intent.

Contrast this approach with the one used in prosecuting and confirming convictions against Jeffery Marsalis, another Philadelphia-area case where the defendant was accused of multiple serial rapes and druggings:

But in trying to establish the likelihood of drugging, the press and the courts seemed at times, disturbingly, to rest their accusations of assault on it. It was a successful gambit for them, but risky. And it once again deferred the question of what right women had to bodily integrity when voluntarily intoxicated, as many of the women also were. So much emphasis was placed on the drugs that Marsalis’ violence—his decision to rape and exploit—seemed like some mechanistically simple and inevitable outcome of his drugging scheme. For instance, the courts belabored how he could have obtained drugs through his nursing and emergency medical technician (EMT) work. But by the time of the Marsalis allegations, in the early 2000s, obtaining drugs for such a purpose was hardly difficult. Benzodiazepines were everywhere. GHB was a popular club drug. Diphenhydramine (which was brought up as a possibility) is available over the counter. Basically, anyone who wanted to drug anyone else would not find many obstacles of a chemical sort. There basically are no barriers to means, nor have there been for a very long time. [p193]

The prosecution made the same mistake when they prosecuted Marsalis for a similar case in Idaho, developing a deep expert witness roster based on a theory of GHB drugging. Toxicology evidence was negative, and was misrepresented by a detective initially, although that wasn’t the only evidence that suggested drugging. But the drugs preoccupation put the conviction in jeopardy during appeal, where Marsalis’ lawyers argued, convincingly to at least some of the appellate judges, who dissented from affirmation of conviction, that the prosecution had made a particular drug scenario the centerpiece of its case, misrepresented the evidence, and then tried to argue that it wasn’t central. Once again, however, the appellate court majority did decide that the central issue was the complainant’s lack of capacity to consent, not an intricate and well-documented road map of chemical predation. A close call, and an example of how a drug-centric shaping of a case, can jeopardize the centrality of assault upon an incapacitated person.

The Cosby complaint takes its unknowns in stride, rather than running from them, headlong into phantom evidence that then weighs down the accuser with absences. It remains to be seen whether this broader-view approach makes a difference.

[back to the main page]


Get the e-book for $9.99 … sometime soon.

image of book cover Title is Drink Spiking and Predatory Drugging: A Modern History


Palgrave is featuring Drink Spiking and Predatory Drugging: A Modern History as a Daily Deal soon (it’s been postponed from April 3)

Springer said some sort of glitch appeared on the sale site, so …

In the meantime, previews are available at Palgrave, Amazon, and Google Books.

And related content is available here on my blog.

New Excerpt: What’s in your Halloween Cache?

Here’s an excerpt from the book, Drink Spiking and Predatory Drugging: A Modern History, about the attractions of drug scarelore, especially involuntary drug ingestion as a “problem solver” for frightened parents. It almost always rears its head around Halloween, but never really goes away completely. – PD

I am one and you are too? Narcissism, violence, lessons not learned, and the case of the Hot Chocolate rapist


On October 9, 2016, the Daily Mail (UK) reported the death of Harry Barkas, Australia’s so-called “hot chocolate rapist” who was convicted of drugging and assaulting women to whom he offered rides home from nightclubs. At the time, this offender’s exploits were covered as a serial rape case, rather than attempting to shoehorn it into the public drink spiking scare narrative. In Drink Spiking and Predatory Drugging: A Modern History (Palgrave Macmillan, 2016), I wrote about the failure of the press and other public health and safety communicators to learn about the nature of this crime from key American and Australian serial cases like Barkas’. Below, I provide an adapted excerpt from my book that refocuses on misplaced trust, context, and links to a chapter that discusses the dynamics of this crime.

From Chapter 6, Who and Where are the Druggers? 

(Other chapters are available as book previews at and Google Books)


[…] Chef John Xydias of Melbourne was accused of drugging, raping, and videotaping 13 women he had met through work. In some cases, he was introduced to his victims by a man named Harry Barkas, who was, at roughly the same time Xydias got caught, accused of being the “Hot Chocolate” rapist. Barkas, who worked in a medical clinic and had access to drugs, approached women as they left nightclubs and offered them a ride home. According to a 2008 Herald Sun article, he then offered them hot chocolate into which he had slipped tranquilizers and sleeping pills, including Rohypnol. Barkas was charged with a string of attacks between 1991 and 2005, and Xydias between 1995 and 2006. Both were in their mid-40s.

Xydias typically met women through the restaurant business, and Barkas sometimes did, too.  The Age, a newspaper in Melbourne, reported that one of Barkas’ victims worked at the same restaurant he did and regarded him as “an older brother or uncle.”   Xydias drugged and filmed many women while dressing them up and assaulting them. Upon sentencing in 2010, when he was convicted of 86 charges relating to 11 victims, he told the court that the women only lodged charges against him as vengeance for not continuing a relationship with them. Like other such serial offenders, he claimed that the acts were consensual and that the women were heavy drinkers and drug users.

While it may be the case that the reason nearly all of the men [accused of multiple attacks] say the acts were consensual is strictly a legal strategy—when the prosecution has a recording of your raping behavior, there are not that many defenses left—it is possible that pathological fantasy, on some level, has made this claim seem real to the offenders. They often see themselves as decadent nightlife hedonists who push the boundaries—so they think, well, why wouldn’t or shouldn’t their victims be? As in, I am one and you are, too. As edgeplay, they might even consider it relatively tame— “just” drugs. Xydias’ pre-sentence psychiatric report called him emotionally disconnected, as reported by the Sydney Morning Herald.  Barkas’ report suggested a narcissistic personality disorder; his ability to be accountable for his actions was limited; he was suspected of many more assaults than he could be charged with due to lack of evidence.

There was some question of Xydias’ relationship to Barkas. Though they were childhood friends and frequented the same venues, they did not, apparently, offend against the same women. Barkas appears to have been more sporadic with his assaults, with a suspected cluster taking place in the mid-1990s and another right before his arrest, leaving a gap of more than a decade, according to the Herald Sun. Although police suspected there must have been more victims in between, that is not entirely clear. Barkas was sentenced to 13 years and Xydias, 28.

The legendary public-place drink spiking, followed by a carry-off and assault, as we have seen, has too many moving parts to really be very common. What we find in its place, much more commonly, is voluntary intoxication followed by misplaced trust, or coevolving with it, and then victimization in a private setting. As with acquaintance rape generally, it is really the moment of misplaced trust that is exploited by the assailant. Intoxication helps, of course, in reducing the ability to resist unwanted sex, and as such, is simply another tool to facilitate an act of violence. Drugging can make detail retention hazy enough that the victim questions what really happened, and may be more reluctant to report it, though this amnesia is not guaranteed, any more than it is with large amounts of alcohol or voluntary drug ingestion. Culture then piles on by blaming women who drink for anything that happens after. On both individual and collective level, the gaslighting begins. Some gaslighters are better at this sort of thing than others. [I continue in the book by talking about the US case of Jeffery Marsalis]

[back to the main page]

The Opiate Overdose Crisis: An Experiment in Vancouver Suggests a New Approach

Across North America, first responders and hospitals report a deluge of overdoses among heroin users whose drugs are mixed with fentanyl, a strong synthetic opiate or the even more powerful carfentanil, which is used in veterinary anesthesia. Word from Vancouver is that public health workers have used a reagent test for the presence of fentanyl in other street drugs for the clients that they serve.

In addition to the dangers posed by heroin itself, these additives make street doses unpredictable in strength.  By volume, the synthetics are much more powerful, and street users depend upon illicit manufacturers and distributors to supply semi-predictable products. Although many street users now expect that the heroin they buy (and sometimes, the cocaine) will contain synthetics, they also have little or no way to know how carefully compounded these drugs are. Judging from the rapid rise in overdoses in recent weeks, the answer is that dosage chaos reigns.

Why Vancouver first? First, British Columbia is treating it as the emergency it is. It declared it as such in April 2016, with the CBC reporting that by June, 71 overdose deaths had taken place and 60% involved fentanyl.  But uniquely, the city has a government-funded organization called Insite, a supervised injection site for drug users. Nurses and other clinic workers staff the place, and it is the only program like it right now in North America. Insite also has some health care services, provides clean needles and can respond to overdoses on site.

Drug users bring in their own drugs, and they are free from the risk of arrest while they are there. This also provides an opportunity for drug testing. Consumers of the drugs can learn about the composition of the drugs they are about to inject, while the clinic can gain information about, and report on, trends in purity and composition of street drugs.

Which brings us to the potential of fentanyl test strips. Francine Diep reports in Pacific Standard magazine that fentanyl testing may – emphasis on may – provide some benefit to drug users by detecting its presence. Insite ran a pilot testing program this summer, using the strips for a month on as many samples as they could. 90% of the heroin samples contained fentanyl, and some of the cocaine samples did, as well.

There are two key issues here. First – how accurate the tests are from a strictly scientific viewpoint: what are the rates of type I and II errors, and what is the specificity to the problem of an adulterant like fentanyl as opposed to cross-reactions to other less deadly ingredients? The strips were originally designed to test urine, and in lab settings it has a 6% false positive rate and a 0% false negative rate. Most of the medical consultants on Insite’s project felt that it might do just as well in water-dissolved samples at the clinic. In the abstract, the low false negative rate is good news. The test strip doesn’t react to carfentanil, however.

It should be kept in mind that some users actively seek out synthetic opiate pills, either preferring them to injecting heroin or using them as a bridge at times when heroin is hard to come by.  People who buy hydrocodone illegally may be getting fentanyl instead, or a mixture, and again, the unexpected strength can be deadly. It appears that fentanyl in a hydrocodone preparation is what killed legendary musician Prince at age 57 in Minnesota this past spring.

Also, to what extent does a positive result for fentanyl discourage a user from injecting the drugs? Diep notes that few people at the Vancouver site seem willing to throw away contaminated drugs altogether. But staffers suggest they may be willing to lower their intake if they know the powerful synthetic is present.  The clinic reports that 86% of the drug samples they tested contained some fentanyl, which implies that at least for the population that uses that clinic, it may be hard to avoid.

Fentanyl is now everywhere. It can be purchased on cryptomarkets online, where either tablets or ingredients can be obtained. Making it is not that difficult, and mid-level distributors can move into manufacture. It can also be diverted from legitimate medical use, as it is often prescribed to people with cancer.

Similar ideas for other cities have been stymied by an insistence on abstinence and prohibition-based approaches and legislation which inhibits or quashes altogether such harm-reduction opportunities, in both Canada in the US. The only recent harm-reduction initiative that hasn’t been aggressively and routinely blocked is overdose-reversal medication like naloxone, which many communities distribute to first responders or make available to the public by prescription.

But as synthetic mixes hit the streets, naloxone is less effective. Reports from Indiana suggest that when carfentanil is present, even five or six doses of naloxone may not reverse the effect. Carfentanil is also powdery and light, posing risks to police and EMTs who respond to places where the drug is present just by touching or inhaling it, in a similar way that methamphetamine making can expose others to risk.

So, the way to think of the new synthetic adulterant problem is: a mess we don’t know how to handle, situated on top of a mess that we do know more about than we are willing to do.  You’d think decades of a failed War on Drugs – drugs won, in case you haven’t been keeping up – would demonstrate the need for greater humility and innovation in policy making. The pilot project in Vancouver can save lives and prevent injury, and similar programs could also tell us a great deal about drug adulteration, and how best to protect public health.

Like Many Things, the Promise of Light and Portable Drug Testing Depends Entirely on Context

images-pills-modern-cover-photo-fbI noted in my last blog entry on the dubious technology “drink spiking detection” devices — such as coasters, test cards, and other reagent-coated bar ware — that both false positives and false negatives are too high to make the technology useful. I also suggested that if such user-end technology — thus far deployed without much real consumer interest — were to become more popular, the imprecision associated with such products could instill a dangerous reassurance to people situationally suspicious enough to test their drinks, while creating undue fear, accusations, and other kinds of serious ramifications in the case of false positives.

But in another realm, there’s already a darker side to false positives in quick field tests. Nationwide concern in the US is beginning to pick up about the sometimes life-altering problems associated with inaccurate or inconclusive field tests used by law enforcement agencies. Field testing technology for drugs on the scene (say, at a vehicle stop) hasn’t developed much since its inception in the 1970s; it appears that what is new is the recognition that cross-reactions with any number of other ordinary chemicals and compounds produce unacceptably high levels of false positives. The New York Times Magazine reported in July that widely deployed reagent field tests have routinely produced false positives for methamphetamine and cocaine. But enter “field drug test” or “drink detector” into a search engine on line, and you’ll find little recognition of the cross-reaction problems that have been known for years.

Furthermore, the necessary visual interpretation of chemical testing results (it turned blue, it turned sort of blue, it turned light blue…) remains a key moment in decision-making in forensic detection and is fraught with the tendency to over-certainty and over-confidence.  This problem has been well-known since the most rudimentary lab techniques were developed in the early 1900s, and remains a problem for cheap, easy field deployables. The cobalt thiocyanate tests used to test for cocaine, for instance, will also turn blue with a number of over-the-counter medicines and cleaning products.

Technically, in order for a positive field test to lead to criminal conviction, it has to be confirmed in a lab by more precise tests. But circumstances rule. Most people arrested for drugs don’t have access to private lawyers or even readily available bail or bond money, and pleas before confirmation are the norm.  Some period of incarceration may ensue anyway, just enough to disrupt the stability of more vulnerable citizens. The Times article chronicles the downfall of a Louisiana woman arrested in Houston and saddled with a felony conviction for crack cocaine, which turned out to be a fragment of the over-the-counter aspirin and caffeine pill that she insisted it was. Having lost her steady home and job as a result, she didn’t even realize that she’d been part of a wave of similar exonerations until the Times informed her, years later.

But on-site drug testing can be used by consumers of drugs, as well. In a different context, the technology, even with its flaws, can prevent harm rather than inflict it. There is a well-functioning model out there – and it comes from the rave scene. In rave and electronic music settings, there’s actually a solid network of nonprofit drug testing that can analyze drugs on the spot. The DanceSafe outfit is probably the best known of these harm reduction groups, but the technology to provide such a service is widespread. MDMA, commonly known as Molly or Ecstasy, is one of the more widely adulterated illicit drugs. On August 3, Lauren Frayer reported on NPR’s All Things Considered radio show that on-site testing had been arranged for a larger music festival in the UK. Due in great part to advocacy on the part of researcher Fiona Measham of Durham University, a legal path has been cleared for the process, to encourage event goers to subject their drugs to testing. In the process, they can discuss the results with a trained counselor. In the US, DanceSafe sells testing kits to consumers, but can’t provide on-site testing and counsel due to risk of attendees’ arrests. (This organization and others are advocating for legal and enforcement-practice reform that would enable this service in the US.)

Online drug commerce has advanced the prospect of user-initiated testing, as well. A firm called Energy Control in Barcelona tests any drug sent to it and provides a report. It now receives government support in the interest of harm reduction, although it started out with a single doctor, Fernando Caudevilla, who posted to the now-defunct Silk Road darknet platform, offering testing and advice. The testing is highly precise but not immediate – with at least a few days’ turnaround time. The company takes Bitcoin payments to enhance anonymity.

A number of drug cryptomarkets have emerged in the wake of Silk Road’s closure in 2013, when it was seized by the FBI as part of the indictment of its founder, Ross Ulbricht. The market for online illicit drug transactions is growing rapidly, according to a newly-released RAND report.

So you can assume that there is a latent demand for user-end testing products, for people that use drugs voluntarily, and have the resources to buy them discreetly. People who purchase drugs in hand-to-hand transactions aren’t the same as those who buy on line. It’s unclear whether this sort of simple re-agent technology would be used widely by those who purchase street drugs.  Energy Control reported that cocaine was much purer in online samples than street ones, suggesting that a bifurcated market remains — starkly by economic resources — and perhaps also the interest in drug content verification. In any case, having users make contact with neutral testers and advice-givers seems all to the good, in terms of enhancing public health and at the same time adding to the research body of knowledge about illicit drug trends. Increasing awareness of the fallibility of light-and-portable tests — and resisting the “gee whiz” style of reporting on technology — also seems useful to everyone.