Development of an Expert System for Classification of ... - ICMCC

icmcc.org

Development of an Expert System for Classification of ... - ICMCC

Development of an

Expert System

for

Classification of Medical

Errors

D. Kopec, K. Levy, M. Kabir,

D. Reinharth and G. Shagas


Findings

• The 1999 IOM Report

(Institute of Medicine)

indicated that between 44,000

and 98,000 unnecessary

deaths occurred in hospitals as

a result of errors committed by

medical professionals in the US


Importance

• The number of deaths

caused by such errors is

alarming and any useful

method which can aid in

their abatement would be

welcome


Research

• In previous research, (Kopec,

Kabir, Reinharth, Rothschild and

Castiglione, 2003):

– a new taxonomy of Medical

Errors was designed by

expanding the IOM

classification

– this model can be used as a

blue print for:

• Future design

Development

• Implementation


Research Goal

• To design an Expert System

utilizing our extended

taxonomy

• The system will:

– effectively classify medical errors

– serve as a testbed for healthcare

practitioners


Research Goal

System’s s goal would be to:

– perform convincingly as

advisory consultants

– exhibit knowledge and

expertise on par with and

beyond human experts in

specified domains


Types of Human Errors

• According to Reason ( (Human Error,

Cambridge: Cambridge University

Press; 1990), human error can be

subdivided as follows:

– Mistake: : made from

inadequate/incorrect info

– Lapse : correct information is

applied incorrectly

– Slip: : plan may be good but

execution is faulty


Types of Errors

• Reason also distinguishes

between two kinds of human

error:

– Active: : immediately

discernible

– Latent: : harder to detect; may

require considerable analysis

to discover


Taxonomy of Medical

Errors

• Five error types observed and

reported by U.S. family

physicians:

– Errors in prescribing

medications

– Errors in getting the right

laboratory test done for the

right patient at the right time

– Filing system errors

– Errors in dispensing

medications

– Errors in responding to

abnormal laboratory test

results


Taxonomy of Medical

Errors

• We found that errors related

to medication and clerical

procedures are abundant in

medical practice

• We altered the IOM

classification which resulted

in the following classification

of human errors in Medical

practice


Classification of Human

Errors

1. Errors in prescribing medication:

• Misuse by incorrect: medication,

route, dose, administration

• Overuse

• Underuse

2. Treatment Procedure(s)

3. Errors related to Clerical

Procedures

4. Errors related to Diagnosis

• Delayed Diagnosis

• Missed Diagnosis

• Wrong Diagnosis

5. Preventative Errors

6. Other

• Communication Failure,

Administration Problems


Proposed Classification

System


Conclusions drawn

• Errors in the medical field can occur

in many different ways, with

potentially diverse, wide-ranging

and hazardous effects.

• We are forced to ask…

Which expert will be able to

hold in his/her head all the

possible combinations of:

– Signs

– Symptoms

– Treatments

that can occur for all

possible medical conditions?


Definition

• An Expert System uses

knowledge specific to a problem

domain to provide “expert quality”

performance in that application

area

• Major Components:

– User Interface: : simplifies

communication between user

and system

– Knowledge Base: : contains

knowledge of an application

domain

– Inference Engine: : applies

knowledge to the solution of

actual problems


Architecture of an Expert

System


Need for an Expert

System

• With domain experts and a

substantial number of rules, a

system can be improved by

building an Expert System

• We intend to use the CLIPS Expert

System Shell to design this system

• The system will classify errors

based on a set of production or

decision rules


Mechanism of Action

Assume a patient was admitted to hospital due to

complications with influenza, however, after what

was considered an acceptable amount of recovery

time, the patient showed no signs of improvement,

and after assessment was found to be physically

worse than he had been upon admission. Some of

questions that the system might ask would be as

follows:

Q

A

Is the patient male or female?

male

Q What is the patient’s s age?

A 43

Q

A

Did the patient stay overnight in the hospital?

yes

Q How many days was the patient in hospital?

A 7

Q Number of medical staff exposed to patient

during his stay?

A 22

Q

A

Q

A

Did the patient come into contact with staff

through: medical devices, food trays,

medicine dispensation?

yes

Does the hospital have a hand-cleansing

protocol for staff?

no


Mechanism of Action

System might respond with the

following:

Patient may have contracted a

hospital acquired infection due to

absence of an effective hand-

washing protocol by staff.

Determination is based on a 75%

degree of certainty and on the

following elements:

– Patient was in hospital overnight

– Patient was exposed to more

than 5 members of staff

– Patient was exposed to staff

through medical devices, food

trays, etc.

– No hand-washing protocol for

staff exists


Further Research

• To reach a consensus on medical

error classification:

– necessary to develop a generally

accepted international medical error

taxonomy

– the major errors-types have to be

categorized and each type can be

associated with a specific mechanism

• May explain why and even predict

when and where an error will

occur which will:

– assist in the generation of

intervention strategies for each error

type

– aid in the reduction and abatement

of medical errors


THE END

More magazines by this user
Similar magazines