Confirmation Bias In Emergency Medicine Essay

Question:

Discuss about the Confirmation Bias in Emergency Medicine.

Answer:

Introduction

Confirmation bias is affirming a body of belief based on preconceived notions that may have been unverified and based on internally represented perceptions. It vastly differs from measured judgment based on scientific analysis and inferences. People who portray this psychological trait tend to store and retrieve selective information. The effect of this type of subconscious information gathering process is deep entrenchment of existing beliefs often mixed with emotional information.

An example to this effect could be as simple as believing that a left-handed person has more cognitive power than a right-handed person. Whenever a person with this cognitive bias meets a creative and accomplished left-handed person, their cognitive bias is reinforced. This evidence is established as a valid truth and their belief system about this cognitive bias increases. To further establish their belief systems, they could seek out for more truths, and they may find thousands of such cases. Hence, their confirmation bias establishes up to the point of being a firm and infallible belief system.

Another example, which has a far-reaching consequence, is how confirmation bias influences people to choose leaders of nations. People with biased opinions only remember positive information about their subconsciously chosen candidates, disregarding the negative information. Instead of interpreting and analyzing objective facts, this blind-shot approach is deep-rooted (Cherry (2016)). Another fatal example of confirmation bias is misdiagnosis at the doctor’s office. A doctor could be predisposed to adjudging symptoms based on a several prior observations that confirm to a particular inference. Based on this lack of inductive reasoning, and more of hypothetical reasoning, the doctor could err on making a proper diagnosis. For example, a senior citizen experiencing dizziness could be diagnosed as having an internal ear-fluid imbalance. In fact, the senior citizen could only be having a gastric disorder and low blood pressure that may have led to such symptoms.

Confirmation bias exists in the investment and finance industry as well. This subconscious pre-mediation can deny a great investment opportunity. A probable investor may read rumors of an impending bankruptcy of a company. Without verifying the facts through the right channels, the investor already decides to sell their existing shares or choose not to buy new ones. Confirmation bias is how most overconfident investors invest and divest their money. It also explains the existence of bulls and bears. Also, the prevalence of ‘gut-feeling approach’ and ‘sixth-sense approach’ is high among investors. All of these are different manifestations of confirmation bias.

There are several ways to counter confirmation bias. One is to tap information from varied sources or heterogeneous sources of information. Another strong proposition to overcome this cognitive bias is to spend time with people who have opposing views. Giving a patient hearing to their views, and having an open mind to accept their views will lead to fresher perspectives and more awareness (Kemp (2011)). Although there is no proven cure for this habit, embracing all types and schools of thought without concluding is one way to start practicing inductive reasoning.

Literature Review

Cognitive bias and errors is a phenomenon that is highly witnessed in the medical industry. Morgenstern (2015) reviewed the common confirmation bias scenarios in an academic paper. According to the paper, cognitive bias has its roots in outcome bias, also known as the chagrin factor. It is the tendency of an individual to convince themselves of a pre-conceived notion instead of seeking alternatives to validate their belief. The next type of bias is based on selecting options with known probabilities, and this behavior is aptly titled Ambiguity Effect. For example, a patient could suffer influenza symptoms during a phase where chickungunya has been reported. But the ambiguity effect influences the practitioner to orient the treatment towards treating the patient for influenza.

Another dangerous fallacy in medical diagnosis is non-adaptation to changing patient status based on continuously updated medical information. This is called the Anchoring confirmation bias. Medical practitioners after making an initial diagnosis are unwilling to change diagnosis based on new medical information received of the patient. They are unable to interpret the patient with a fresh pair of eyes or diagnostic judgment, because of their anchoring tendency. Often the difference between a great medical practitioner and an ordinary medical practitioner is based on their tendency to adjudicate symptoms based on the availability or non-availability of relevant examples. It is common for doctors and physicians to diagnose based on Availability bias, which is neglecting the prevalence of rare conditions in lieu of common or recently available conditions in the general population.


McGee, (2015) was of the opinion that clinicians do not use mathematical models to aid their clinical decision making process. They rely on intuitive judgment based on probabilities and heuristics. These informal rules of the thumb methods often come heavily stacked with confirmation bias. Heuristics rely on recognizing patterns and integrating these patterns to form subconscious judgments. One such fallacy which is an outcome of this approach is called the Premature Closure confirmation bias. In this type of bias clinicians jump to inferences based on their unsubstantiated hypothetical arguments. For example, if a person has a long history of migraine, a recent headache in this person is generally attributed by clinicians as another bout of migraine. No consideration is given to the fact that this recent headache could be due to some other non-migraine factors.

Clinicians could also make the cardinal mistake of Attribution Errors or making judgments based on stereotypical information. For example, a drunken unconscious patient could be deemed as unconscious due to a high level of intoxication. But an underlying medical condition could have been the reason, which may have been completely ignored. McGee underlines specific strategies to minimize cognitive bias. Inserting reflection and factoring-in all possibilities instead of relying on heuristics is one approach. Awareness of the extent of medical danger a symptom can progress to is a technique to factor-in all possibilities, and seek for new clues. Expanding the knowledge domain of a particular symptom is the foundation for life-saving clinical interventions. Confirmation bias is often the short-cut that no clinician would want to consciously take.

An issue by Quick Safety, 2106, an advisory on safety & quality issues, discusses the cognitive biases in healthcare, based on information obtained from real events. Confirmation bias is categorized into personal factors, patient factors and system factors. Personal factors are cognitive bias, lethargy, tiredness and emotional angst. Patient factors are number of co-morbidities and incomplete medical history. System factors are the number of handoffs, complexity of tasks, insufficient time to analyze information and poor teamwork etc.

The academic paper identified more than 100 instances of confirmation biases. In one such incident, a patient with multiple medical conditions or co-morbidities was admitted to the ED. The co-morbidities were diabetes, obesity, renal problems and hypertension to name a few. The patient was admitted for chest pain. But due to the Framing Effect confirmation bias, a secondary complaint namely back pain was triaged. Since the patient’s medical history was known to this hospital/clinic, Ascertainment Bias made medical practitioners shape decisions based on previous medical records and incidences. They were predisposed to making decisions based on patient stereotypes.


The solution for such cognitive biases as proposed was a combination of inductive reasoning and meta-cognition. Critical evaluation methods such as Bayesian model that aid in probabilistic modeling of an issue provides alternative explanations and aids in well-rounded diagnosis. Framing effect can be reduced by including systematic and checklist-based methods that factor all probabilities. Another interesting solution is directed towards workflow design and ergonomic working requirements. This as per the authors reduced distractions and aided in clearer thought processes.

Pines, (2008) analyzes confirmation bias in emergency medicine and lists many potential solutions to tackle this subconscious psychological decision-taking medicine. One such solution is to practice cognitive forcing strategies. These strategies are categorized into universal, generic and specific respectively. A universal strategy is obtaining a generalized understanding of the extent and nature of errors in diagnosis. Post-this, application of meta-cognition is effected to reduce cognitive bias. A generic strategy is an overall understanding of the prevalence of heuristics in medical decision making. Blending this awareness with an awareness of the different confirmation biases that might arise leads to a potent cognitive forcing strategy. An example to this effect is stopping searches for coingestants when the primary toxic poisoning ingestant has been found. This approach uses heuristics, yet applies an awareness of probable confirmation biases. This approach helps medical practitioners make urgent clinical decisions in the ED room when time is at a premium.

Even in this academic paper, the prominence of Bayesian reasoning is established. Known tests on medical information are combined with heuristics to obtain diagnostic probabilities. This approach helps medical practitioners to converge upon a medical condition within pre-set tolerance levels.

Confirmation bias can be empirically and mathematically modeled. Some mathematical modeling techniques are based on the popular Bounded Confidence Model. This model is popular in the sphere of measuring opinion dynamics. Although this Model is generic to a large extent, it can be customized as per peer group requirements.

Confirmation runs high in the medical industry and amongst medical practitioners. Some common types of cognitive biases are Ascertainment Bias, Anchoring Bias, Stereotyping and Framing etc. The solutions proposed to counter confirmation biases are meta-cognition, awareness of the pitfalls of heuristics, Bayesian probabilistic models, ergonomic working environment and efficient medical systems.

Most academicians propose the idea of disapproving an immediately occurring belief and obtaining a well-rounded understanding of the nature, scale and scope of symptoms. This is an approach of novice medical practitioners, but fails to gain prominence in the experienced ones, who adjudicate based on previous case studies.

It is worthwhile for clinicians to adopt new clinical test strategies that focus on negative tests rather than the positive ones. Negative tests produce unknown or previously unrecorded outcomes. The effort is to re-emphasize the absence of non-confirmatory data and thus help in delivering a critically evaluated medical diagnosis.

I have personally been a victim of my own confirmation bias. My biased opinion about Chinese people led me to surprises. I was deployed in China for a project. Based on numerous homogenous sources of information, I was under the assumption that Chinese people are prolific eaters. I thought that they eat all types of meat on a daily basis. But I was pleasantly surprised when I saw numerous vegetarian restaurants even in up-market Chinese cities such as Shanghai. I couldn’t believe when I saw and interacted with umpteen numbers of vegetarians.

Going to China was like a nightmare for me because I was under the assumption that Chinese is a difficult language to pick, and most Chinese did not speak English. But on arrival and after a few months of stay, this assumption dissolved. Chinese is a fairly difficult language all right, but not impossible to learn to speak at least. There were so many Chinese who could speak English. In fact, there were Chinese who were doing college majors in the English language.

Conclusion

Had my confirmation bias been very strong, I would have not gone to China. In turn, I would have lost a golden opportunity to mingle with people of a beautiful land and culture. To minimize my confirmation bias, I do not read information that comes without the backing of sound research and statistical analysis. I apply this process to all aspects of work, education and life. Another way I minimize or totally negate confirmation bias is to not form opinions at all. I keep an open mind and assimilate experiences, concrete information and facts. This process helps me perform inductive reasoning to arrive at conclusions.

Another instance when I was overcome with confirmation bias was when interacting with people. I was selective of the people whom I spoke to. I made very few friends because of this. I was under the influence of stereotype confirmation bias. By using subconscious pattern matching, I tried to stereotype people based on their ethnicity, educational background, age, professional experience, personal habits, hobbies and social skills. This type of pattern matching did not allow me to freely interact with people. But I decided to drop prejudices and pre-mediated judgments. This was when I found stellar people who became my life-long friends.

A simple meta-cognition technique to assess my thoughts was enough for me to self-introspect. Although simple, it took time, effort and conscious de-learning. My interactions were heavily stacked towards Framing and Anchoring confirmation bias. I had to retract from such a route lest it hindered my social skills and progress in the social circuits. My approach and subsequent revelations of this approach inspired many others to take this path. To their surprise and my assurance, they were able to change their attitudes for the better. A positive attitude multiplies into a receptive attitude without the intervention of confirmation bias.

Now I don’t assume or hypothecate. I use facts more than fiction. Thanks to small adjustments to the way I processed thoughts, I am able to see through the glass and build better perspectives.

References

Cherry, K. (2016, June 22). What Is a Confirmation Bias? Examples and Observations. Retrieved from

Confirmation Bias. (n.d.) Retrieved from

Kemp, D. (2011, June 29). Overcoming Confirmation Bias. Retrieved from

Morgenstern, J. (2015, September 15). Cognitive Errors in medicine: The common errors. Retrieved from

McGee, L. D. (2015). Cognitive Errors in Clinical Decision Making. Retrieved from

Quick Safety. (2016). Cognitive biases in health care. Retrieved from

Pines, M. J. (2008). Profiles in Patient Safety: Confirmation Bias in Emergency Medicine. Academic Emergency Medicine. Volume 13, Issue 1. Doi: 10.1197/j.aem.2005.07.028

How to cite this essay: