Lyman BioPharma Consulting LLC

Advice and Resources for the Biotech Industry

Advice and Resources for the Biotech Industry

Stacks Image 512

Stacks Image 805

Bias and Self-Delusion Are A Problem In Both Politics and Science

Coarse adjectives get tossed around the political arena as politicians and pundits try to score points with their supporters. This person is a moron, that one’s an idiot, and so and so is such an imbecile. These three adjectives are nowadays used as interchangeable insults, but what most people don’t understand is that these words used to have distinct scientific definitions in the (now discredited) field of eugenics. Eugenics, which was purported over a century ago to be a legitimate science, was later recognized as nothing more than a house of racial and ethnic bias constructed on a shaky foundation of non-rigorously collected “data” that didn’t stand up to careful scrutiny. “Data” were found and adjusted to fit this theory, rather than gathering up facts and then building a concept around them. Bias often lurks at the core of the thought process used by both politicians and scientists. Before going further, let me be crystal clear about one thing: while both groups suffer from biased thinking, it manifests itself in completely different ways.

Bias is Inherent in Politics
Bias is one of the hot topics dominating the news lately. I’m going to define bias as the selection, application, or interpretation of information that leads someone to a false conclusion. While some bias is deliberate, much of it is subconscious; people acting under its influence are unaware they’re doing so. For example, it’s easy for us to overlook the flaws in our political organizations and candidates and just focus on those aspects of them with which we agree. This affects both Republicans (who ignore Trump’s habit of lying on a regular basis) as well as Democrats (who gave Bill Clinton a pass by overlooking his sordid sexual behavior with women). In a recent NY Times
opinion piece, Daniel A. Effron of the London Business School illustrated the concept, “We use different standards of honesty to judge falsehoods we find politically appealing versus unappealing. When judging a falsehood that maligns a favored politician, we ask, “Was it true?” and then condemn it if the answer is no. In contrast, when judging a falsehood that makes a favored politician look good, we are willing to ask, “Could it have been true?” and then weaken our condemnation if we can imagine the answer is yes. By using a lower ethical standard for lies we like, we leave ourselves vulnerable to influence by pundits and spin doctors. In this time of “fake news” and “alternative facts,” commentators worry that people with different political orientations base their judgments of right and wrong on entirely different perceptions of reality. My research suggests an additional concern: Even when partisans agree on the facts, they can come to different moral conclusions about the dishonesty of deviating from those facts. The result is more disagreement in an already politically polarized world. Blame the human ability to imagine what might have been.”

Bias Exists in Science as Well
Data irreproducibility generate a lot of discussion in science these days, some of which arises from a type of bias. Irreproducibility is often caused by problems such as poor experimental design and execution, as well as weak statistical analysis. Bias allows scientists to fail to look at (or subconsciously downgrade) all of the available data, including information that could disprove their hypothesis. A classic example is the discarding of select data points because they are “outliers” (i.e. markedly different from the other pieces of data). Some type of rationale that supports the exclusion decision is generated, even if that rationale is neither written down nor verbally conveyed.

Scientists with good training are taught to envision and then conduct experiments that might disprove their hypothesis. Failure to disprove the hypothesis is supportive of their theory, but it cannot prove it. Many scientists fail to be as rigorous as they should be in their work. This usually stems from intellectual laziness, hope of financial rewards, poor training, bias, hubris, and/or dreams of scientific glory. Eradicating these problems is extremely difficult, and many don’t even recognize them when they crop up.

Bias Occurs at the Intersection of Politics and Science
As much as some people don’t like to admit it, there are many examples where bias in politics affects science. When politicians
refused for 22 years to allow the Centers for Disease Control to study gun violence (because of the Dickey amendment), this was a clear example of bias working against science. Politics has begun to work its way into the FDA’s drug approval process, as evidenced by the agency’s endorsement of Sarepta Therapeutics drug Exondys 51 (eteplirsen) for treating Duchenne muscular dystrophy. This approval was given despite a lack of compelling data and over the objections of in-house scientists who reviewed the application carefully. Politics allow for the existence of hundreds of dubious stem cell clinics in the U.S. These modern day purveyors of high-tech snake oil have operated for years with impunity, peddling bogus treatments for a wide variety of disorders for which they’ve never been tested. Reports have recently appeared indicating that patients in unregistered “clinical trials” have gone blind following injection of fat-derived stem cells into their eyes. Finally, medical clinics around the world offer untested treatments to terminally ill patients with the real intent to merely separate the sick and desperate from their money, while governments look the other way.

Bias When Politics Seek to Negate Science
The EPA recently angered environmentalists when it
announced it was putting in place a new rule that removes much of the science from its policy-making functions. This is the same type of rogue behavior that’s often seen in science denial campaigns, such as those ongoing in climate science and global warming. Sean Carroll, Professor of Genetics and Molecular Biology at UW-Madison, has talked about this process in an effort to help people identify scientific denialism. He’s outlined the following steps, “First, cast doubt on the science. Second, question the personal motives and integrity of the scientists. Third, magnify genuine disagreements among scientists, and cite non-experts with minority opinions as authorities. Fourth, exaggerate the potential harm caused by the issue. Fifth, frame the issue as a threat to personal freedom. And sixth, claim that acceptance would repudiate a key philosophy, religious belief, or practice of a group.

Differences Between Bias in Science and Politics
We know that human beings are, by our very nature, fundamentally flawed. Some of our leaders have many admirable qualities, along with a smattering of those we might loath. We readily accept this, even though we wish our friends, lovers, relatives, heroes, political representatives (and even ourselves) weren’t so darn imperfect. In science, failure to examine and integrate all of the relevant data leads to flawed hypotheses that eventually fall apart when others point out their inconsistencies. Failure to “look at all of the data” is actually quite common in politics as well, especially with candidates that are supported because of their views on a single issue (for example, abortion). Constituents will vote for such a candidate if they agree with him or her on this one core issue. They turn a blind eye to all of the other positions of their favored candidate, even though they might be illogical, contradictory, and even (at times) reprehensible.

Facts Are Important, But Avoiding Exploitation Is Too
In science, facts and expertise have traditionally been, and continue to be, highly valued commodities. In politics, these elements are not so treasured, especially now in the “post-truth” era of “fake news”, “alternative facts”, and other drivel that bears a closer resemblance to propaganda than reality. Science, however, is about identifying actual facts, uncovering the hidden details that explain how things operate in our universe. Having said that, moral and ethical considerations must enter into the scientific process when required. Too often these considerations have been ignored, with bias born of racism and a wanton disregard for the rights of the poor and underprivileged. This has led to the marginalization of these communities and their exploitation.

The infamous Tuskegee studies of end-stage syphilis in black men conducted by the U.S. Public Health Service were done to understand the natural course of the disease. This ultimately involved denying the men treatment with antibiotics that would have cured them. In a similar vein, dermatologists treated inmates at Holmesburg Prison in Philadelphia for nearly 25 years with a variety of dangerous agents (in non-therapeutic “studies”) simply because the incarcerated men were a ready source of “acres of skin.” Another well-known case was the deliberate infection of mentally disabled children at the Willowbrook State School in NY with live hepatitis virus, and then watching as they became sick. Ethics must always play a part in our thinking as we search for biological truths. Fudging data or promoting drugs for non-approved indications simply to make a buck only serves to impugn the reputations of the best and brightest scientists among us.

Facts constitute the warp and weft from which our theories are woven. These bend and change as new details are uncovered that force us to revise our thinking. Few would disagree with the idea that science should be extremely rigorous, demanding, and unbiased. When science gets lazy, we wind up with ideas built on an unstable platform of conjecture and magical thinking, not facts. This makes it difficult to proceed to the next experiment, and to change our thinking, because we don’t really know the truth about the foundations upon which our theories are built. This often leads to data irreproducibility, which is the kiss of death for science. There can be honest disputes about the interpretation of data. These disagreements lead to new experiments that are done to test which of two alternative interpretations of the data (or neither) is actually correct. However, in the absence of facts about which all can agree, performing these follow-up experiments becomes impossible. As Senator Daniel Patrick Moynihan once put it, “
Everyone is entitled to his own opinion, but not to his own facts.

People these days are demanding evidence-based medicine, not that which arises from conjecture and supposition. Clinical trials, unfortunately, are also subject to the same bias issues that bedevil basic science. This helps explain why we so often see post hoc analyses of failed trials that yield a “promising-looking” subgroup in which a drug or treatment seems to have performed well. These hints of success are merely a scientific mirage, conjured up by minds that want to believe so badly that the drug or treatment really worked. These folks will then devise a rationale for a new trial based on their biased interpretation of the subgroup analysis. When that “wish” is tested in a subsequent trial, the apparently positive outcome usually disappears, and the hypothesis once again crashes and burns (usually in an expensive way). The putative effect was indeed illusory. Bias in the selection and interpretation of elements of the data led to a second round of clinical failure. This bias may actually be conscious. Sometimes clinical trials are done not because the data are so compelling, but because investors in a company have bet on a particular outcome, and those running the trial believe that they have no choice but to move forward with less-than-stellar data. Companies that act this way often have no other drugs in their pipelines, thereby adding an element of desperation to the proceedings.

Biologist Sir Peter Medawar once wrote, "
The intensity of the conviction that a hypothesis is true has no bearing on whether it is true or not. The importance of the strength of our conviction is only to provide a proportionally strong incentive to find out if the hypothesis will stand up to critical examination." I don’t think there’s an equivalent phrase in politics, so I’ll coin one, “The intensity of the conviction that a politician will win an election and support legislation that you favor has no bearing on whether or not they will succeed. The importance of the strength of our conviction is only to provide a rationale to get others to vote for our candidate.

A key difference between science and politics: in politics, judgments are shaped by people’s needs, wants, and desires, whereas in science, decisions should be driven by facts i.e. solid data. Facts, of course, used to be something that everyone believed in. Nowadays, they are often used as political footballs where certitude is a distant virtue and “alternative facts” (previously known as falsehoods or lies) are paraded about in an effort to confuse the public. “Critical examination” has become an abandoned art in many quarters, rejected as “elitist” by folks who peddle nonsensical conspiracy theories that common sense should readily reject. We accept that science and politics are inherently distinct enterprises that operate quite differently. However, a failure to recognize the true value of facts as well as the pernicious effects of bias will doom both science and politics to abandonment at a time when the proper use of both of these processes is more important than ever.

Further reading:

Those looking to get a better understanding of eugenics should read Adam Cohen’s excellent historical treatise on the subject
Imbeciles: The Supreme Court, American Eugenics, and the Sterilization of Carrie Buck.

A historical account of the Tuskegee experiments on black men in the South is told with clarity in James Jones’
Bad Blood: The Tuskegee Syphilis Experiment.

The story of the ethically questionable studies done on prisoners are revealed in Allen Hornblum’s well researched
Acres of Skin: Experiments at Holmesburg Prison