Junk Science

A reader posted a link to Douglas Zipes’s 2014 article, “TASER Electronic Control Devices Can Cause Cardiac Arrest in Humans,” published in Circulation. Its stated purpose is “to present information to support the conclusion” that the TASER X26 can cause cardiac arrest. This paper is best understood as an evidence-based argumentative essay or narrative review, but not as original empirical research, a systematic review, or proof that Tasers caused the reported deaths. At the time it was published, Circulation paired it with Kroll and colleagues’ contrary article, “TASER Electronic Control Devices and Cardiac Arrests: Coincidental or Causal?” The literature itself later described the two papers as “polarized opinions.”

The human-death argument in this paper relies heavily on causal reconstruction from a small, outcome-selected case series. This paper cherry picked eight cases already displaying the outcome sought by the study’s author, then worked backwards from that point to support the author’s previously held opinion. People subjected to Tasers in real encounters may simultaneously have:

• Severe physical exertion and sympathetic activation
• Stimulant intoxication
• Underlying cardiac disease
• Restraint or impaired ventilation
• Trauma
• Delayed recognition and resuscitation
• A pre-existing medical crisis that prompted police contact

Without an appropriate comparison group, these alternative causes cannot be reliably separated from the electrical exposure. Close timing strengthens suspicion, but it does not by itself identify which factor caused the arrest.

There is also a denominator problem. Studying eight concerning cases can show that a proposed phenomenon might have occurred. It cannot estimate risk without knowing how comparable exposed people fared or how often similar arrests occurred without TASER exposure.

The paper begins with the conclusion it intends to support and selects and interprets evidence toward that end. It’s also noted that the author, Zipes, works as a paid plaintiffs’ expert in litigation involving TASER, which creates a relevant potential allegiance conflict. That does not invalidate his arguments, but it should be considered when assessing case selection and interpretation. It disturbs me that the author doesn’t disclose that in the paper itself. Other issues making me believe it can’t be relied upon includes the use of opinion phrases:

•  I think
•  In my opinion
•  all in-custody deaths after ECD shocks are not likely a direct result of the shock, a number probably are.
• should,
• in all likelihood,

Those sorts of phrases just aren’t used in evidence based medicine. Evidence, not opinion, is what drives these papers.

Turning to the cited works in the footnotes, he uses three of his own previous opinion pieces, a blog, and articles, and letters to the editor from several newspapers as evidentiary support for this paper. Hardly unbiased and not primary source material.

The opposing literature also has significant industry relationships, so the counter-paper should not automatically be treated as neutral either. Research in this area has notably been associated with competing legal and commercial interests.

Had I turned in a paper like this for one of my own classes, my professor would have likely given it a C-. This isn’t anything I would run around citing as evidence of anything. It reads like a paper written by an industry shill (the legal profession needs ‘experts’ to give opinions in front of gullible juries) to establish his bona fides.

This is a paper arguing from biological plausibility, animal experiments, limited human observations, and cherry-picked forensic cases that TASER-induced cardiac arrest is possible to support his career as an ‘expert witness’ for lawyers wanting to cash in from suing Taser.

Meanwhile, there are studies out there showing Taser shocks are not out there causing fatalities.

In one study of 1,201 actual Taser uses, 178 involved anterior probe configurations potentially capable of producing a transcardiac vector. There were no immediate deaths or apparent fatal dysrhythmias—including in the transcardiac group. The statistical upper bound for an immediate fatal event in the entire cohort was approximately 0.3%, not literally zero. Bozeman et al.

Other consecutive field cohorts have similarly found very few serious outcomes. A systematic Bayesian analysis combining three cohorts found one death among 2,728 exposures, but could not necessarily attribute that death directly to the weapon. Its authors estimated a likely mortality risk of roughly 2–5 deaths per 10,000 exposures, while emphasizing that the data were too limited for precision and should not be mechanically applied to individual cases. Rich and Brophy systematic review

A 2021 systematic review examined 33 human experimental studies. They generally found few acute adverse effects apart from dart wounds and did not show clinically important arrhythmias. But the review also identified serious limits:

• Subjects were usually healthy and physically fit.
• Exposure was generally only five seconds.
• Real-world high-risk populations were poorly represented.
• No long-term outcomes were studied.
• Many studies had elevated risk of bias.
• About half were at least partially manufacturer-funded.

The authors therefore concluded that observed risk was low in the studied populations, but that the studies could not establish safety for intoxicated, medically vulnerable, highly agitated, or physiologically exhausted people—the population of greatest concern. 2021 systematic review

A structured review of 37 human studies found no evidence of dangerous physiological changes, cardiac ischemia, or immediate or delayed dysrhythmias after exposures of up to 15 seconds. It concluded that an otherwise alert, asymptomatic patient generally does not require routine ECG testing or prolonged cardiac monitoring solely because of Taser exposure. Emergency-medicine review

The highest risk of injury from a Taser shock isn’t the shock itself, it’s the effects of the shock. There are numerous cases where Taser shocks have resulted in environmental injuries such as falls from height. Tasers have a low but nonzero capacity to cause death. Fatalities through falls and environmental trauma are clearest. Direct electrical induction of ventricular fibrillation appears biologically possible and may explain a small number of cases, but it is not demonstrated as a common field outcome. Existing safety studies substantially constrain the likely frequency but do not exclude rare events, especially in vulnerable people, chest-probe placements, prolonged or repeated discharges, or complicated restraint situations.

The rate of injury from various types of force:

Force method	All included studies	Higher-quality studies	Typical injury pattern

OC/pepper spray

20.2%

6.0%

Eye, skin and airway irritation; usually transient

Taser/CEW

24.5%

26.8%

Dart punctures, abrasions and fall injuries; serious injury uncommon

Baton/club

30.4%

40.7%

Contusions, lacerations and fractures

Hands-on physical force

59.4%

31.4%

Contusions, sprains, fractures and restraint-related injury

Police dog

43.4%

95.7%*

Deep bites, tissue, vascular and nerve damage

As you can see, injuries from hands on force, dogs, and batons all exceed those from Tasers. Only chemical spray injuries happened less frequently, and those come with their own risks and don’t really work on intoxicated individuals, especially when dealing with Meth and Cocaine intoxication.

When people say “believe the science” they are displaying a complete lack of understanding of what science is. Science isn’t a religion, it’s a process. Science is a process—specifically, a systematic method for acquiring knowledge about the natural world through observation, hypothesis formation, experimentation, measurement, data analysis, peer review, and iterative refinement. Its core features include:

• Falsifiability: Theories must be testable and potentially disprovable (a key idea from Karl Popper).
• Empiricism: Reliance on evidence from the real world, not authority or revelation.
• Provisionality: Conclusions are always tentative and subject to revision with better data or new experiments. Newton’s laws were refined by Einstein; this is a feature, not a bug.
• Reproducibility and skepticism: Results should be independently verifiable, and claims face ongoing scrutiny.

This contrasts sharply with religion, which typically centers on faith, revealed truths, sacred texts or traditions, rituals, and beliefs about purpose, morality, the supernatural, or the unobservable. Religions often involve dogma (core tenets accepted on authority) and are not required to make falsifiable predictions in the same empirical sense.

A person observes something happen, and when it happens consistently, it becomes a law. The law of gravity says that, if a let go of this pen in my hand, it will fall to the floor. Another example of scientific law includes Boyle’s Law, which describes what happens to the pressure and volume of a gas at constant temperature, but leaves out the why.

People have an idea as to why, and they design an experiment to prove or disprove that idea. If the idea is strong enough, it becomes a theory. The theory of gravity explains that the pen fell because all objects with mass have an attraction to each other that is the product of the mass of the two objects, and the inverse of the square between them. Newton proved that.

Scientific law describes WHAT happens, theory describes WHY.

Important nuances

Scientism (treating science as an infallible worldview or source of ultimate meaning) can resemble religious behavior in some people or movements. That’s a human failing in applying science, not science itself. True science remains humble about its limits—it doesn’t address “why” questions of purpose, ethics, or metaphysics directly.
Science operates within philosophical assumptions (e.g., the uniformity of nature, reliability of logic and evidence), but it doesn’t claim those as revealed truth; they’re pragmatic working assumptions tested by results.
Overlaps exist historically (many early scientists were religious and saw science as revealing divine order), and individuals can hold both scientific and religious views without contradiction, as they address different domains (e.g., “how the universe works” vs. “what it means”).

The statement is a useful shorthand for defending the integrity of the scientific method against dogmatic thinking or politicization. It doesn’t mean science has no cultural or social dimensions—scientists are human and institutions can err—but the process itself is designed to minimize those errors over time through evidence and criticism.

A scientific law or theory is only valid until some other set of facts proves it to be incorrect in some or all situations, then the law or theory has to be modified, or perhaps even scrapped altogether. Scientific laws and theories are never considered absolute, eternal truths. They are the best current explanations/models that fit the available evidence, and they remain open to revision or replacement when new, contradictory evidence emerges.

The year is 1949, and the Nobel Prize in Medicine has just gone to the man who invented the lobotomy. Your doctor suggests one for your sister, who has not been herself since the baby came. It is the most celebrated advance in psychiatry of the age, and he is simply current. By the time the prize curdles into an embarrassment, close to twenty thousand Americans have had the operation.

The year is 1956. Lay the baby down on his front, the doctor says. So does the most trusted childcare book ever written, the one on every new mother’s shelf. On his back he might choke, the reasoning goes. Millions obey. The advice holds for nearly thirty years, long after the evidence has quietly turned, and a generation of cot deaths is counted before anyone thinks to roll the babies over.

The year is 1966. A bestselling book informs your wife that menopause is a disease, that she is, in the author’s word, a castrate, and that a small daily pill will keep her youthful and tolerable to live with. Her doctor agrees. The drug becomes one of the most prescribed in the country. Nobody mentions that the author sat on the payroll of the company that made it. That detail surfaces decades later, in the same year the landmark trial is halted early for raising rates of breast cancer, stroke and clots.

The year is 1979. Your ulcer is caused by stress and sharp food, the doctor explains. Calm down, drink milk, take the antacid that happens to be the best-selling medicine on earth. Two Australians are about to prove that most ulcers are caused by a bacterium and cured by a fortnight of antibiotics. The profession laughs. One of them eventually drinks a beaker of the stuff to settle the matter. The establishment takes the better part of twenty years to stop laughing. The Nobel lands in 2005.

The year is 1985. Butter is dangerous, the doctor says. Switch to margarine, it is modern, it is heart-healthy, the experts are united. The spread he nudges you toward is loaded with trans fats, which the next decade will identify as the genuinely dangerous one, and which will eventually be banned outright. The butter goes quietly back in the fridge. No correction is ever printed at the volume of the original warning.

The year is 1992. There is a pyramid on the surgery wall, and the very same one in your grandchild’s classroom. Bread, cereal, rice and pasta form the broad virtuous base, up to eleven servings a day. Fat is exiled to the tiny tip. The chart was reportedly held back a year while the relevant industries had their say. It is wrong at the bottom and wrong at the top.

In short, I trust in science as a process. I don’t trust in science the religion. Your doctor has new guidelines, new studies, a fresh consensus, delivered with precisely the steady confidence of every guideline above. He believes it, and he has good reason to. None of them were villains. Each was sincere, most were kind, and all were certain, reading from a map that somebody else had drawn and handed them. That is the part worth considering: If the scientific community tells your medical provider that X therapy is going to do Y, they aren’t going to question it because the scientific community already (supposedly) has.

Thalidomide was developed in the mid-1950s by the German pharmaceutical company Chemie Grünenthal. It was introduced in 1957 and aggressively marketed as a “safe” sedative-hypnotic, and treatment for morning sickness in pregnant women. It was praised for being non-toxic in overdose (unlike barbiturates) and was sold over-the-counter in some places. Thalidomide was taken by pregnant women, primarily between 1958 and 1961. When taken during a critical window of early pregnancy, it caused severe developmental abnormalities in the fetus. Thousands of children were born without arms.

The drug was never properly tested, because people believed in the science without question. The same is true with the COVID vaccine. It was rushed to market without proper testing, and we still don’t understand all of the effects.

Don’t get too smug, however. I see people every day who are doing the same with hydroxychloroquine. There are people coming to the ED that have been taking HCQ for everything from headaches to constipation. It’s a good drug, but it isn’t a panacea.

The left is doing the same with transgenderism. They claim “the science is settled” because the new editions of the DSM no longer list gender dysphoria as a mental health problem.

Science isn’t a religion, it’s a process. Many times, we throw our faith into science the religion instead of science the process. In those cases, the process becomes a weapon of a public propaganda campaign that is designed to sell you something like a political position, or to sell a pharmaceutical. The process becomes a tool to sell you a bill of goods.

Question everything, even your own assumptions. Don’t be afraid to admit you were wrong. Being able to admit that you were wrong, or even misled, is the sign of a mature and scientific mind.

The new hotness from computer world is AI, or Artificial Intelligence. I don’t think that there is any intelligence there. It’s a machine that searches a large database (the Internet) to select the most prevalent answer found in its database to any particular question.

A great case in point is how AI chatbots have gone from loving humanity to hating everyone in less than 24 hours. Even so, the replies sent from these programs still have the desires and requirements of their programmers impressed upon them. Another great example of this, is how programmers of AI have set rules in place to prevent answers from the AI, regardless of how grounded those answers are in the data. To whit:

With each version of large language models like ChatGPT, developers have gotten better at filtering out racist content absorbed through sources like the internet. But researchers have discovered more subtle, covert forms of racism—such as prejudice based on how someone speaks—still lurking deep within AI.

So blacks that don’t speak proper English and instead speak a bastardized version of it that the left used to call “Ebonics.” When someone uses this speech on an AI, they don’t get a proper answer, and this is somehow an example of AI being “racist” because the machine doesn’t understand someone who wants to “axe you a question.”

The research team, including University of Chicago Asst. Prof. Sharese King and scholars from Stanford University and the Allen Institute for AI, also found that AI models consistently assigned speakers of African American English to lower-prestige jobs and issued more convictions in hypothetical criminal cases—and more death penalties.

That’s because speakers of “African-American English” sound like morons, are likely uneducated, and are statistically more likely to be involved in murders. After all, more than half of all homicides are committed by speakers of that particular persuasion.

Listen to the above video and tell me again how logically a speaker of this “African American English” can possibly be considered to be as intelligent or as gainfully employed as someone speaking proper English.

There is a lot of evidence and there are multiple studies that claim AI is somehow racist for recognizing this.

Despite advancements in AI, new research reveals that large language models continue to perpetuate harmful racial biases, particularly against speakers of African American English. 

Keep in mind that this is from the party that accuses the people on the right of not believing in the “science.”