Clinical Measurement

Front 1

Key Properties of ClinicalMeasurements

Back 1

Key Properties of Clinical Measurements

(Very Reliable & Responsible Girl)

}Validity

}Reliability

}Responsiveness

}Generalizability

Front 2

Validity

Back 2

Validity:

The degree to which the results of a measurement correspond to the true state of the phenomenon being measured.

-Synonym is accuracy.

Front 3

Types of Validity

Back 3

Types of Validity

(Face Contest: Criteria Construct)

Face Validity: Does the measurement appear to test what it is supposed to?

Content Validity: Do the items sample from the universe of content that defines the variable?

Criterion Validity: Does the instrument predict a gold standard test?

◦Concurrent

◦Predictive

Construct Validity: Does the measurement correlate with other measurements of the variable/contstruct?

Front 4

Bias

Back 4

Bias

Aprocess at any stage of inference tending to produce results that departsystematically from the true (or valid) values (the opposite of a valid measure is a biasedmeasure).

Front 5

Reliability

Back 5

Reliability:

The degree to which a measurement yields the same results in repeated applications on an unchanged population or phenomenon.

Synonyms:reproducibility and precision

Front 6

Reliability

For repeated measures of a singlevariable in one person

Back 6

Reliability: For repeated measures of a single variable in one person

X = T + E

Where:

X =observed score

T =true score

E =random error

Front 7

Reliability:

For repeated measures across persons

Back 7

Reliability:For repeated measures across persons

VAR (X)= VAR (T) + VAR (E)

}Reliability is estimated as R where

R = VAR(T) / VAR (X)

= 1 – (VAR (E)/VAR (X))

Front 8

Why Reliability is often NOT high: The Sources of Variation in Clinical Measurement

The Examiner

Back 8

TheExaminer :(BI DIP)

B- Biologicvariation in the senses

I- Thetendency to record inferencerather than evidence

D- Ensnarementby diagnosticclassification schemes

I- Simple incompetence or inexperience

P- Entrapmentby priorexpectation

Front 9

Why Reliability is often NOT high: The Sources of Variation in Clinical Measurement

The Examined

Back 9

TheExamined(BIT M)

B- Biologicvariation in the system being examined

I- Effectsof illness and medications

T- Toss-ups

M- Memoryand rumination

Front 10

Why Reliability is often NOT high: The Sources of Variation in Clinical Measurement

The Examination

Back 10

TheExamination

E - Disruptive environments for the examination

I- Disruptive interactions between examiners and examined

D- Incorrect function or use of diagnostictools

Front 11

Strategies for Minimizing Observer Variation

Back 11

#Strategies for Minimizing Observer Variation (CST SP)

C=CLEARLY DEFINED CRITERIA

S=STANDARDIZED PROCESS }

T=OBSERVER TRAINING

S=SINGLE OBSERVER

P=PATIENT SELF ASSESSMENT

Front 12

Responsiveness

Back 12

Responsiveness:

The degree to which a measure detects clinically meaningful change, particularly the change due to a treatment effect.

Synonym is sensitivity to change.

Front 13

Defining Clinically MeaningfulChange

Back 13

Defining Clinically MeaningfulChange

}Expert Opinion }

}Patient Opinion }

}Benchmark Therapy }

}Statistical Estimate

Front 14

Effect Size

Back 14

Effect Size =

Posttreatment – Pretreatment Mean

--------------------------------------------------

Pretreatment StandardDeviation

Front 15

Different Measures ofResponsiveness

Back 15

Front 16

Generalizability

Back 16

Generalizability

The degree to which the measurementperforms in a similar fashion when applied to different populations orphenomena.

}Synonym is external validity.

Front 17

Dimensions of Generalizability

Back 17

Dimensions of Generalizability

}Demographics }}Diseases }}Doctors }Drugs

Front 18

Other Considerations in Choosing aMeasurement

Back 18

Other Considerations in Choosing a Measurement (ROS LI)

Relevance}}

Objectivity}}

Standardization}}

Interpretability}

Logistics