Transcript Insulin and Oral Hypoglycemic Toxicity

Diagnostic Testing
Ethan Cowan, MD, MS
Department of Emergency Medicine
Jacobi Medical Center
Department of Epidemiology and Population Health
Albert Einstein College of Medicine
The Provider Dilemma
A
26 year old pregnant female presents after
twisting her ankle. She has no abdominal or
urinary complaints. The nurse sends a UA
and uricult dipslide prior to you seeing the
patient. What should you do with the
results of these tests?
The Provider Dilemma

Should a provider give
antibiotics if either one
or both of these tests
come back positive?
Why Order a Diagnostic Test?
When the diagnosis is
uncertain
 Incorrect diagnosis
leads to clinically
significant morbidity
or mortality
 Diagnostic test result
changes management
 Test is cost effective

Clinician Thought Process
Clinician derives patient
prior prob. of disease:
H & P
 Literature
 Experience
 “Index of Suspicion”
 0% - 100%
 “Low, Med., High”

Threshold Approach to
Diagnostic Testing
Probability of Disease
0%
100%
Testing Zone
P(-)
P(+)
P < P(-)
Dx testing & therapy not indicated
 P(-) < P < P(+) Dx testing needed prior to therapy
 P > P(+)
Only intervention needed

Pauker and Kassirer, 1980, Gallagher, 1998
Threshold Approach to
Diagnostic Testing
Probability of Disease
0%
100%
Testing Zone
P(-)

P(+)
Width of testing zone depends on:
 Test properties
 Risk of excess morbidity/mortality attributable to the test
 Risk/benefit ratio of available therapies for the Dx
Pauker and Kassirer, 1980, Gallagher, 1998
Test Characteristics

Reliability
 Inter observer
 Intra observer
 Correlation
 B&A Plot
 Simple Agreement
 Kappa Statistics

Validity
 Sensitivity
 Specificity
 NPV
 PPV
 ROC Curves
Reliability

The extent to which
results obtained with a
test are reproducible.
Reliability
Not Reliable
Reliable
Intra rater reliability
 Extent
to which a
measure produces
the same result at
different times for
the same subjects
Inter rater reliability

Extent to which a
measure produces the
same result on each
subject regardless of
who makes the
observation
Correlation (r)
For continuous data
r=1
perfect
r=0
none

O1
O1 = O2
O2
Bland & Altman, 1986
Correlation (r)
Measures relation
strength, not
O1
agreement
 Problem: even near
perfect correlation
may indicate
significant differences
between observations

r = 0.8
O1 = O2
O2
Bland & Altman, 1986
Bland & Altman Plot
O1 – O 2
For continuous data
 Plot of observation
differences versus the
means
 Data that are evenly
distributed around 0
and are within 2 STDs
exhibit good
agreement

10
0
-10
[O1 + O2] / 2
Bland & Altman, 1986
Simple Agreement
Rater 1
+
total
 Extent
Rater 2
+
a
b
c
d
a+c b+d
total
a+b
c+d
N
to which two or more raters agree on the
classifications of all subjects
 % of concordance in the 2 x 2 table (a + d) / N
 Not ideal, subjects may fall on diagonal by chance
Kappa
Rater 1
+
total
 The
Rater 2
+
a
b
c
d
a+c b+d
total
a+b
c+d
N
proportion of the best possible improvement in
agreement beyond chance obtained by the observers
 K = (pa – p0)/(1-p0)
 Pa = (a+d)/N (prop. of subjects along the main diagonal)
 Po = [(a + b)(a+c) + (c+d)(b+d)]/N2 (expected prop.)
Interpreting Kappa Values
K=1
K > 0.80
0.60 < K < 0.80
0.40 < K < 0.60
0 < K < 0.40
K=0
K<0
Perfect
Excellent
Good
Fair
Poor
Chance (pa = p0)
Less than chance
Weighted Kappa
Rater 1
1
2
.
.
C
total
Rater 2
1
2
n11
n12
n21
n22
.
.
.
.
nC1
nC2
n.1
n.2
...
...
...
...
...
...
...
C
n1C
n2C
.
.
nCC
n.C
total
n1.
n2.
.
.
nC.
N
Used for more than 2 observers or categories
 Perfect agreement on the main diagonal weighted
more than partial agreement off of it.

Validity
The degree to which a
test correctly diagnoses
people as having or not
having a condition
 Internal Validity
 External Validity

Validity
Valid, not reliable
Reliable and Valid
Internal Validity
Performance Characteristics
 Sensitivity
 Specificity
 NPV
 PPV
 ROC Curves

2 x 2 Table
Disease Status
Test
Result
+
total
cases
noncases
TP
FN
FP
TN
cases
noncases
TP = True Positives
FP = False Positives
total
positives
negatives
N
TN = True Negatives
FN = False Negatives
Gold Standard
 Definitive
test used
to identify cases
 Example:
traditional agar
culture
 The dipstick and
dipslide are
measured against
the gold standard
Sensitivity (SN)
Disease Status
Test
Result
+
total
cases
noncases
TP
FN
FP
TN
cases
noncases
total
positives
negatives
N
 Probability
of correctly identifying a true case
 TP/(TP + FN) = TP/ cases
 High SN, Negative test result rules out Dx (SnNout)
Sackett & Straus, 1998
Specificity (SP)
Disease Status
Test
Result
+
total
cases
noncases
TP
FN
FP
TN
cases
noncases
total
positives
negatives
N
 Probability
of correctly identifying a true noncase
 TN/(TN + FP) = TN/ noncases
 High SP, Positive test result rules in Dx (SpPin)
Sackett & Straus, 1998
Problems with
Sensitivity and Specificity
 Remain
constant over patient populations
 But, SN and SP convey how likely a test
result is positive or negative given the
patient does or does not have disease
 Paradoxical inversion of clinical logic
 Prior knowledge of disease status obviates
need of the diagnostic test
Gallagher, 1998
Positive Predictive Value (PPV)
Disease Status
Test
Result
+
total
cases
noncases
TP
FN
FP
TN
cases
noncases
total
positives
negatives
N
 Probability
that a labeled (+) is a true case
 TP/(TP + FP) = TP/ total positives
 High SP corresponds to very high PPV (SpPin)
Sackett & Straus, 1998
Negative Predictive Value (NPV)
Disease Status
Test
Result
+
total
cases
noncases
TP
FN
FP
TN
cases
noncases
total
positives
negatives
N
 Probability
that a labeled (-) is a true noncase
 TN/(TN + FN) = TP/ total negatives
 High SN corresponds to very high NPV (SnNout)
Sackett & Straus, 1998
Predictive Value Problems
 Vulnerable
to Disease Prevalence (P) Shifts
 Do not remain constant over patient populations
 As P
PPV
NPV
 As P
PPV
NPV
Gallagher, 1998
Flipping a Coin to Dx AMI for
People with Chest Pain
ED AMI Prevalence 6%
AMI
No AMI
Heads (+) 3
47
50
Tails (-) 3
47
50
6
94
100
SN = 3 / 6 = 50%
SP = 47 / 94 = 50%
PPV= 3 / 50 = 6%
NPV = 47 / 50 = 94%
Worster, 2002
Flipping a Coin to Dx AMI for
People with Chest Pain
CCU AMI Prevalence 90%
AMI
No AMI
Heads (+) 45
5
50
Tails (-) 45
5
50
10
100
90
SN = 45 / 90 = 50%
SP = 5 / 10 = 50%
PPV= 45 / 50 = 90%
NPV = 5 / 50 = 10%
Worster, 2002
Receiver Operator Curve
1.0
Sensitivity
(TPR)
0.0
0.0 1-Specificity (FPR) 1.0
Allows consideration of test performance across a
range of threshold values
 Well suited for continuous variable Dx Tests

Receiver Operator Curve
 Avoids
the “single
cutoff trap”
Sepsis
No Effect Effect
WBC Count
Gallagher, 1998
Area Under the Curve (θ)
1.0
Sensitivity
(TPR)
0.0
0.0 1-Specificity (FPR) 1.0
 Measure
of test accuracy
 (θ) 0.5 – 0.7 no to low discriminatory power
 (θ) 0.7 – 0.9 moderate discriminatory power
 (θ) > 0.9
high discriminatory power
Gryzybowski, 1997
Problem with ROC curves
 Same
problems as SN and SP “Reverse
Logic”
 Mainly used to describe Dx test
performance
Appendicitis Example
Study design:
 Prospective cohort
 Gold standard:
 Pathology report from
appendectomy or CT
finding (negatives)
 Diagnostic Test:
 Total WBC

Physical Exam
+
OR
+
Appy
CT Scan
-
No Appy
Cardall, 2004
Appendicitis Example
WBC
Appy
Not Appy Total
> 10,000
66
89
155
< 10,000
21
98
119
Total
87
187
274
SN 76% (65%-84%)
SP 52% (45%-60%)
PPV 42% (35%-51%)
NPV 82% (74%-89%)
Cardall, 2004
Appendicitis Example
Patient WBC:
 13,000
 Management:
 Get CT with PO & IV
Contrast

Physical Exam
+
OR
+
Appy
CT Scan
-
No Appy
Cardall, 2004
Abdominal CT
Follow UP
 CT
result: acute
appendicitis
 Patient taken to
OR for
appendectomy
But, was WBC necessary?
Answer given in talk on Likelihood Ratios