pediatrics grand rounds university of texas health science ... · university of texas health...

Pediatrics Grand Rounds 29 June 2012

University of Texas Health Science Center at San Antonio

1

Learning in Medicine

Jesus R. Guajardo

TWO TOPICS

1. Update on Diagnostic Reasoning Models

2. Share the results of a study: “the effects of patient’s name and image in educational

materials”

Objectives

• At the end of this activity the participant (you) will (should) be able to:

– Mention the names of at least two diagnostic reasoning models

– Describe the characteristics of at least one diagnostic reasoning model

– Describe the findings of an educational study recently conducted in our medical school

– Mention a few strategies that may foster learning in medical academic centers

Disclosures/COI

• UTHSCSA

• PhD through Mizzou (Univ of Missouri at Columbia)

• Financial Disclosures: Nothing Related

Medical Diagnostic Reasoning

~ clinical reasoning, medical problem solving, diagnostic reasoning, and

decision making

Medical Reasoning

Patel, VL 2004

• Cognitive process involved in making a medical diagnosis and the related care decisions

• Deductive? Inductive? Or Abductive ?

• Abstractionabductiondeductioninduction

• Forward or backward driven?

• Based on the level of expertise



2

Intermediate Effect May apply to reasoning, but apparently not to diagnostic accuracy.

Patel, VL 2001

What are the current conceptual models trying to explain diagnostic

reasoning?

Four Medical Diagnostic Models

• Decision-analytic approach

– Hypothetico-deductive (Elstein, 1978)

• Information-Processing models

– Prototype

– Instance-based

– Illness script (schema) or semantic network

Custers, EJFM 1996

Let’s briefly review each one

Hypothetic Deductive Model

• Generating hypothesis

• Testing hypothesis

• Deductive model, going from disease to patient’s signs and symptoms

• Doesn’t account for the differences between beginners and experts

Prototype Model

• Previous patients’ characteristics or data are abstracted

• Memory trace created with a representative model: prototype or exemplar

• New patients are categorized according to the archived prototypes

• Drawbacks: – Analytically decomposed (all factors are important)

– Initially made for simple clear-cut characteristics

– No time factor considered



3

Instance-Based Model

• No abstraction of characteristics

• Previous experiences with patients are stored as such (many examples)

• Abductive and forward driven (from patient’s sign & symptoms to the diagnosis)

• Does not provide a reason why certain characteristics are included

• Does not include explanations of the disease

Illness Script (Schema) Semantic Network

• Creation of nodes containing information • Mental links between those nodes • Lots of information: clinical and biomedical • Includes diagnostic reasoning • Links and nodes can be activated based on

multiple stimuli • Components:

– Enabling cause (i.e. tobacco smoking) – Faults (lung cancer) – Consequences (cough, hemoptysis, weight loss)

Illness Script/Semantic Network

Tree of Porphyry CCC AD

Illness Script/Semantic Network

NOW TO A DESCRIPTION OF A STUDY RECENTLY CONDUCTED HERE

Done with theory and models!

Effects of a Patient’s Name and Facial Image on Knowledge Acquisition

June 2011-March 2012

Study Collaborators: Glen Medellin, MD Jean Petershack, MD Jo Ann Lieberman John Schoolfield

PhD Committee Julie A. Caplow, PhD @ Mizzou John H. Littlefield, PhD @ UTHSCSA Michael Hosokawa, EdD @ Mizzou James Laffey, PhD @ Mizzou Joi Moore, PhD @ Mizzou



4

Title & Research Questions

• Title: Effects of a Patient’s Name and Facial Image on Knowledge Acquisition

• Research Question 1: – Does the use of a patient’s name and facial image in a

CLIPP VP case summary improve MS knowledge acquisition as measured by standardized medical examinations?

• Research Question 2: – Do MS have higher examination scores after being

provided with CLIPP VP cases as a learning and study resource compared to being provided with a summary of a CLIPP VP case as a learning and study resource?

Research Question 1 Effects of a Patient’s Name and

Facial Image on Knowledge Acquisition

Rationale

Background

• The illness script and related models stipulate that the semantic networks are created upon exposure to real cases (Charlin 2007)

• Therefore, it seems that the exposure to real patients is one of the main factors promoting the development of expertise in the medical field (Custers 1996)

Background

• Practicing Physicians: Exposure to real patients creation of semantic networks

• Educational materials for MS try to recreate and apply the practicing physicians’ context (i.e. realism)

• Realistic elements include the patient’s name and at least a facial image

Background

Literature on Realistic Materials

• Better scores from students who were provided VP versus text-only materials (Qayumi, 2004)

• Recommendations for showing patients’ images at the beginning of case presentations (Huwendiek, 2009)

• VPs improve knowledge acquisition (MS

comments only, not objectively evaluated) (Botezatu, 2010)

Background

• BUT: clinical medicine and biomedical science may be two different worlds (Patel 2004)

• Clinical Knowledge Complex taxonomy, pattern identification, categorization, etc.

• Biomedical Science General principles and causal mechanisms of disease



5

Background

Different Learning Objectives of MS vs MDs

Learning to explain

how a set of symptoms is consistent with a diagnosis (MDs)

MAY BE QUITE DIFFERENT THAN

how a disease is caused (MS)

Overly contextualized

educational materials

may reduce knowledge transfer (Patel 2004)

Are we really helping our MS by providing them with educational materials containing patients’ names and facial pictures?

Research Question 2 Comparison of a short summary versus a complete CLIPP VP Case

Rationale

Background

Length of Materials • Length concerns from students and faculty re

CLIPP VP cases • How do summaries compare to complete CLIPP

VP cases? • Is summarized medical information similar to a

much longer and contextualized complete CLIPP VP case regarding knowledge acquisition?

• Teaching/learning models indicate a positive relationship between study time and tests scores (to a point)

Effects of a Patient’s Name and Image on Knowledge Acquisition

Research Questions

Research Questions

Are we overly contextualizing medical students’ materials?

RQ1: Does the use of a patient’s name and facial image in a CLIPP VP case

summary improve MS knowledge acquisition as measured by standardized medical examinations?

Are we providing MS with very long cases?

RQ2: Do MS have higher examination scores after being provided with CLIPP VP

cases as a learning and study resource compared to being provided with a summary of a CLIPP VP case as a learning and study resource?



6


Method

CLIPP Materials and MS educational program

Brief Description

CLIPP VP Cases CLIPP VP Cases

CLIPP VP Case

#7

Week 1 Week 2 Week 3 Week 4 Week 5 Week 6

NBME Pediatric

EXAM “Shelf”

CLIPP Quiz 1 “CQ1”




CLIPP VP Case

#3

CLIPP VP Case

#2

CLIPP VP Case

#15

CLIPP VP Case

#18

CLIPP VP Case

#6

CLIPP VP Case

#9

CLIPP VP Case

#8 CLIPP

VP Case

#4

CLIPP VP Case

#30

CLIPP VP Case

#20

CLIPP VP Case

#17

CLIPP VP Case

#11

CLIPP VP Case

#28

Study Materials

Study Materials

Study Materials

Study Materials

Examination

Examination

Examination Examination Examination

MS Education (at time of study and very simplified)

Method ~ 200 MS per year



7

Method

• Recruited MS from four blocs on pediatrics • Provided with a summary of a CLIPP VP case

– Two blocs provided with a summary that had the name (29 times) and image of the patient

– Two blocs provided with similar summary but without the patient’s name or image

• Scores assessed looking for a difference on knowledge acquisition between groups – Summary with name/image vs. summary without

name/image – Summary vs. complete CLIPP VP cases

Summaries Name/Image vs no NI

Katie has had swelling around the eyes the past 10 days which is worst in the morning. There is no eye redness, but Katie has experienced fatigue and increasing girth. Katie’s physical exam reveals anasarca in face, abdomen and legs, ascites, weight gain and normal heart sounds. Lab tests performed on Katie confirm the diagnosis of nephrotic syndrome.

Nephrotic syndrome is characterized by swelling around eyes which is worst in the morning. There is no eye redness, but patients experience fatigue and increasing girth. Physical exam reveals anasarca in face, abdomen and legs, ascites, weight gain and normal heart sounds. Lab tests support the diagnosis of nephrotic syndrome.

VS No Image

Summary vs Complete VP Case

Five pages for the study case

20+ pages for the other CLIPP VP cases

CLIPP VP Case

#7


NBME Pediatric

EXAM “Shelf”



CLIPP Quiz Additional

“CQA”



CLIPP VP Case

#3

CLIPP VP Case

#2

CLIPP VP Case

#15

CLIPP VP Case

#18

CLIPP VP Case

#6

CLIPP VP Case

#9

CLIPP VP Case

#8 CLIPP

VP Case

#4

CLIPP VP Case

#30

CLIPP VP Case

#20

CLIPP VP Case

#17

CLIPP VP Case

#11

CLIPP VP Case

#28 CLIPP VP Case

SUMMARY

Study Materials

Study Materials

Study Materials

Study Materials

Examinations

Examination


Study Design

CLIPP VP Case

#7


NBME Pediatric

EXAM “Shelf”




“CQA”



CLIPP VP Case

#3

CLIPP VP Case

#2

CLIPP VP Case

#15

CLIPP VP Case

#18

CLIPP VP Case

#6

CLIPP VP Case

#9

CLIPP VP Case

#8 CLIPP

VP Case #4

CLIPP VP Case

#30

CLIPP VP Case

#20

CLIPP VP Case

#17

CLIPP VP Case

#11

CLIPP VP Case

#28 CLIPP VP Case

SUMMARY

Study Materials

Study Materials

Study Materials

Study Materials

Examinations

Examination


Study Design

NBME Pediatric

EXAM “Shelf”




“CQA”



(10 MCQs)

CQ1

CQA

(11 MCQs)

CQA_K (knowledge)

CQA_NI (name & image) (10 MCQs)

(12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4

(100 MCQs)

Shelf Quizzes



8


Results

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 1

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 2

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 3

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 4

Results

n=18

n=17

n=22

n=19

64.65 16.76

71.66 16.35

76.03 16.49

68.42 16.97

75.56 18.54

80.00 12.25

80.00 10.69

80.53 12.24

85.89 12.26

83.53 12.82

86.32 9.31

85.47 12.93

80.22 10.75

89.65 8.22

79.14 10.58

89.42 8.83

72.22 12.37

83.76 13.75

82.23 17.96

86.79 10.17

74.94 9.04

81.12 11.19

77.00 7.81

76.37 5.40

SUMMARY ------ COMPLETE CLIPP VP CASES-----

Test Correlation Table

Research Question 1

Are there any differences in knowledge acquisition in MS when they are provided with a

–summary that includes the patient’s name and facial image versus

–a summary without the patient’s name and facial image?

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 1

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 2

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 3

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 4

Comparisons

SUMMARY ------ COMPLETE CLIPP VP CASES-----



9

(11 MCQs)

CQA_K Bloc 1

(11 MCQs)

CQA_K Bloc 2

(11 MCQs)

CQA_K Bloc 3

(11 MCQs)

CQA_K Bloc 4

64.65 16.76

71.66 16.35

76.03 16.49

68.42 16.97

ANOVA

F (3,75) = 1.678 p = 0.179

n=18

n=17

n=22

n=19

Provided a summary with the patient’s name (Katie) and facial image

Provided a summary with NO patient’s name or facial image

Additionally: Bloc 1+3 (70.9) vs Bloc 2+4 ( 69.8) t test: p = 0.8

Name and Image

No statistically significant difference in scores, therefore it seems that there was NO difference

in knowledge acquisition (more a little later)

Research Question 2

Are there any differences in knowledge acquisition in MS when they are provided with a

–summary of a CLIPP VP case (5 pdf pages) versus

–complete CLIPP VP cases (>20 interactive) web pages?

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 1

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 2

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 3

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 (100 MCQs)

Shelf Bloc 4

Comparisons SUMMARY ------ COMPLETE CLIPP VP CASES-----

(10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 Bloc 1 (10 MCQs)

CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)


CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)


CQ1 (11 MCQs)

CQA_K (12 MCQs)

CQ2 (12 MCQs)

CQ3 (12 MCQs)

CQ4 Bloc 4

Comparisons

n=76 70.46 16.87

79.08 13.48

85.38 11.60

84.32 10.75

81.34 14.80

ANOVA F(4, 375) = 14.341, p < 0.0001 (pairwise comparisons showed a statistically significantly lower CQA_K score)

SUMMARY ------ COMPLETE CLIPP VP CASES----- Pairwise Comparisons



10

Summary vs. Complete CLIPP

The data here found seem to indicate that tests scores are statistically significantly lower when medical students are provided with a summary

of a case than when they are provided with complete CLIPP VP cases

Additional Results


Conclusion/Summary

Conclusion

1. The presence of a patient’s name and image on medical study materials seems not to make a difference on knowledge acquisition as shown by scores on standardized testing

2. Tests’ scores are significantly lower when medical students are provided with a summary of a case than when they are provided with a complete CLIPP VP case

Limitations?

• Oh, Many!

• One center, one class of MS

• Limited number of students

• Student variation (not cells or mice)

• Diminished MS interest on study materials and quiz

• Other limitations (i.e. comparing different cases w different tests at 1 week

only, no pre-test, lack of interactivity, unremarkable images, curricular changes, etc, etc, and

etc)

But keep in mind that these limitations are common in educational studies. This is not research on single cell populations or murine models

Importance & Recommendations

• First study showing the effects of a patient’s name and image in medical study materials on tests scores (supporting the overly contextualized educational materials effect)

• Findings are interesting, but should not alter the current design of study materials

• Are we trying to apply the practicing physicians’ contextual elements into the medical students’ educational materials?

• More studies are needed to corroborate these findings before we alter our approaches

General Recommendations

There are many recommendations on how to improve learning, for example:

– Adult Learning Theory (clear objectives, learner participation)

– Gagne’s instructional design (steps of instruction, feedback)

– Backward design (planning based on measurable outcomes)

– Importantly: stimulate the formation of schemes/nodes/scripts I will expand a little on

the last two



11

Foster Desired Outcome

• A little educational story about PFT’s reading

Readings & Lectures Methods Technique Instruments Description Uses

Practice &

Feedback

Examples and Non-Examples

Classic Approach

“This is a red circle”

Modern Approach

“This is a medium-sized red smooth circle”

“This is a large-sized red smooth circle”

“This is a small-sized red smooth circle”

“This is a medium-sized yellow smooth circle”

“This is a medium-sized red smooth triangle”

Green irregular…..

Wrap up

What Did We Just Review?

• Medical Reasoning

• Hypothetic-deductive, prototype, instance-based, and illness-script models

• Importance of Realistic Components

• Use of those elements on MS education

What Did We Just Review?

• Very specific/focused study • Effects of a patient’s name and facial image on

knowledge acquisition – Background and rationale for the study – Research questions – Methodology – Results – Discussion of Results

• Conclusion 1: Patient’s name and Image in study materials may not improve knowledge acquisition

• Conclusion 2: Longer and more interactive study materials may promote better knowledge acquisition (concerns about depth vs. width)

Study Acknowledgments

• Many!

• PhD Committee Members. Special Txs to Dr. Littlefield (Acad Ctr for Excellence in Teaching)

• UTHSCSA Peds Division and Pulm Department

• Drs. Medellin and Petershack

• Jo Ann Liebermann & John Schoolfield

• Jen and Jolie (Wife and Daughter)

pediatrics grand rounds university of texas health science ... · university of texas health...

Documents