/
A Comedy of Errors: Identifying, Classifying and Preventing A Comedy of Errors: Identifying, Classifying and Preventing

A Comedy of Errors: Identifying, Classifying and Preventing - PowerPoint Presentation

ellena-manuel
ellena-manuel . @ellena-manuel
Follow
391 views
Uploaded On 2017-05-19

A Comedy of Errors: Identifying, Classifying and Preventing - PPT Presentation

Michael Drabkin MD Margaret Steiner Isael Perez Steven Lev MD Nassau University Medical Center eEdE106 Purpose To review and classify causes of errors in diagnostic neuroradiology via a casebased approach with a focus on strategies for error prevention for the oncall resident ID: 549903

bias errors effect patient errors bias patient effect hemorrhage perceptual cognitive false basal ganglia background contrast diagnosis interpretation shapes

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "A Comedy of Errors: Identifying, Classif..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

A Comedy of Errors: Identifying, Classifying and Preventing Cognitive and Perceptual Errors in Neuroradiology

Michael Drabkin MD Margaret Steiner, Isael Perez, Steven Lev MDNassau University Medical Center

eEdE-106Slide2

Purpose

To review and classify causes of errors in diagnostic neuroradiology, via a case-based approach, with a focus on strategies for error prevention for the on-call residentSlide3

Approach/Methods

Various systems for the classification of radiologic diagnostic errors have been proposed, encompassing perception, interpretation, communication and technical

factors

. We retrospectively reviewed our database of resident- attending discrepancies of emergency and trauma cases at our level-1 trauma center over a span of 10 years. We then organized and classified the cases based on several criteria. Slide4

Approach/Methods

Errors were first divided into two groups, perceptual and cognitive. We further divided cognitive errors into false positives, false negatives, and misjudgments in which the abnormality was identified but misinterpreted.

Preliminary Report Error Identified

Perceptual

CognitiveSlide5

Interpretation

Radiologist looks for deviations from a mental template. Any variation from this template will raise concern for pathology. Slide6

Cognitive ErrorsSlide7

Bias

Despite identical levels of knowledge and training, two different radiologists may be inclined to different interpretations. This is dependent on several biases, stemming both from the radiologist’s temperament internally and also from the external circumstances.They may have a different bias/criterion

.

Raising one’s threshold to call an abnormality will decrease the false positive rate, but may also lead to more misses.

Likewise, lowering one’s threshold will decrease the false negative rate, but will also lead to more overcalls.Slide8

Cognitive Biases

AnchoringFramingSatisfaction of search

Multiple alternative bias

Availability Bias

Outcome biasSlide9

Anchoring

Locking onto a diagnosis from the beginningRelying too heavily on the first piece of information offeredUndervaluing data that would support an alternative diagnosis or that does not support the favored diagnosis (similar to confirmation bias). Slide10

Framing

Arriving at a conclusion based on how information is presented or “framed” Slide11

Satisfaction of Search

An eye-catching finding grabs the attention and diverts from consideration of other diagnoses.May lead to missed comorbidities, complications, or additional diagnoses. Slide12

Multiple Alternative Bias

Narrowing a list of diagnostic possibilities to a smaller subset with which a physician is most familiar.May fail to consider other possibilities. Slide13

Availability Bias

Considering a diagnosis more likely because it easily comes to mind.A notion that things which first come to must be important/relevant.Slide14

Outcome Bias

An optimistic radiologist might favor a diagnosis with a better outcome for the patient than one with a more harmful outcome. Slide15

Intraparenchymal Hemorrhage

Lesson

A unilateral high density in the basal ganglia should alert one to the possibility of hemorrhage and warrant follow-up examination. Bilateral basal ganglia high densities, on the other hand, often represent calcification. Density measurements are helpful, but can sometimes be misleading due to partial volume averaging.

Bilateral basal ganglia calcifications were overcalled as hemorrhage. Bilateral basal ganglia calcifications occur commonly as patients age.Slide16

Subdural Hematoma

Noncontrast Head CT demonstrating an overcall of an anterior interhemispheric fissure subdural hematoma due to “hematocrit effect” in neonates.Slide17

Satisfaction of search

Fixation on a large intraparecnchymal hemorrhage lead to overlooking the coexisting subarachniod hemorrhage.This large MCA aneurysm was an unusual cause of intraparenchymal bleed. Recognizing the SAH was important in prompting CTA.

Axial CT of 46 y/o female with headache shows large intraparenchymal bleed, with subarachnoid components and mild left to right shift.

Subarachnoid HemorrhageSlide18

Subacute Infarct

There is superficial enhancement of the brain parenchyma in distribution of the middle cerebral artery in a patient with recent infarct. This can be mistaken for subarachnoid hemmorhage.Slide19

Virchow-Robin spaces

A young patient with bilateral symetric well-defined hypodensities in the basal ganglia.These represent enlarged Virchow-Robin spaces.However, this can be mistaken for lacunar infarcts.Slide20

Cavernous Malformations

Initially interpreted as hemorrhage

.

Axial

T2WI in the same “trauma” patient reveals the right frontal (and left basal ganglia)

density

to be

a

caveroma

,

demonstrating mixed signal intensity and

popcorn

” appearance. The right frontal lesion shows a dark signal

hemosiderin

rim. Slide21

Perceptual ErrorsSlide22

Perceptual ErrorsSlide23

Contrast, Mach Bands, and the Background Effect

The Mach phenomenon exaggerates the contrast between edges of the slightly differing shades of gray where they meet. This enhances edge-detection by the human visual system.An interesting phenomenon is background contrast effect, where the perceived brightness of an object is affected by a neighboring object’s optical density. Slide24

Contrast, Mach Bands, and the Background Effect

In the figure below, the central grey squares all appear to have different densities, however, when the background is eliminated it is apparent that the small squares have the same gray levels.Slide25

Contrast, Mach Bands, and the Background Effect

In the figure below, the central grey squares all appear to have different densities, however, when the background is eliminated it is apparent that the small squares have the same gray levels.Slide26

Effect of Neighboring Density

There is diffuse cerebral edema in this 18 year-old male who experimented with cocaine. The vessels appear hyperdense against the dark parenchyma causing a “pseudosubarachnoid” hemorrhage effect.

Focal sparing of gray matter within an infarct may mimic hemorrhage as seen in the above block diagram and companion case.Slide27

Normal or Mass Lesion?

We try to complete shapes and objects. The line drawings below are created when we try to connect the dots. When four dots are present, some of us may see a square, others, perhaps more imaginative, see various shapes and forms emerge. The on-call resident questioned whether the patient to the right had a right parietal well circumscribed mass. This necessitated an overnight stay for the patient and follow-up imaging. Needless to say, the “mass” disappeared on follow-up? Was this a cure by repeat CT or an erroneous interpretation of sulcal lines that created the illusion of an object? Subjective contours

caused by the need for completeness create

pseudolesions

that can be mistaken for real ones.Slide28

Normal or Mass Lesion?

We try to complete shapes and objects. The line drawings below are created when we try to connect the dots. When four dots are present, some of us may see a square, others, perhaps more imaginative, see various shapes and forms emerge. The on-call resident questioned whether the patient to the right had a right parietal well circumscribed mass. This necessitated an overnight stay for the patient and follow-up imaging. Needless to say, the “mass” disappeared on follow-up? Was this a cure by repeat CT or an erroneous interpretation of sulcal lines that created the illusion of an object? Subjective contours

caused by the need for completeness create

pseudolesions

that can be mistaken for real ones.Slide29

Subjective

Contour Formation“Subjective contours” are the result of mental completion of partial lines to form familiar or expected shapes. These

contours are illusions formed by extending lines and boundaries. This can create the illusion of a pathologic lesion.Slide30

Perceptual Grouping

Perceptual grouping is a means by which the brain arranges objects according to certain available visual cues. In the Kanizsa triangle illusion, the positioning of black three-quarter circles in the “correct” orientation results in the appearance of a central white triangle, even though none really exists. The neural circuitry required involves the pathways connecting the lateral geniculate nucleus, V1, and V2. Slide31

Fogging Effect

Noncontrast Head CT demonstrating the fogging effect 2 days later, which can be missed depending on the timingSlide32

Discussion

It is imperative to both perceive and accurately interpret abnormal findings in order to facilitate appropriate and expeditious radiologic workup and clinical management. Slide33

Discussion

Perceptual errors are related to scanning, fixation, satisfaction of search, and the inherent characteristics of the human visual system, such as subjective contour formation and the Mach effect. Slide34

Discussion

A major contributor to cognitive false positive and false negative errors is misconstruing normal variants, especially pertaining to bony and vascular anatomy; errors in interpretation frequently involve infarcts and hemorrhages. These types of misjudgments may be caused by incomplete knowledge, misleading clinical history, and anchoring bias. Slide35

Discussion

The adverse consequences of misinterpretations include prolonged hospitalization, unnecessary testing and procedures, and needless patient anxiety.Slide36

Discussion

Strategies for improvement include maintaining a missed-case database which can be drawn from as an opportunity for learning. It is essential to communicate with referring physicians, assess for distinguishing radiologic features and consider all possible etiologies, however unlikely.Slide37

Summary/Conclusion

By utilizing such a system of error classification in conjunction with maintaining an organized case database, institutions may be able to provide improve quality in

neuroradiologic

diagnostic

imaging

.Slide38

Thank You