Skip to main content

Table 3 Comparison of methods for diagnostic accuracy of diabetic foot ulcers (N = 518)

From: Diagnostic accuracy of existing methods for identifying diabetic foot ulcers from inpatient and outpatient datasets

Method

Chart review*

Accuracy and agreement measures (95% CI)

  

Yes

No

Sensitivity

Specificity

PPV

NPV

Kappa

Holzer

Yes

148

59

94.9

83.7

71.5

98.4

0.69

 

No

8

303

(90.1-97.8)

(79.5-87.4)

(64.8-77.5)

(97.9-98.7)

(0.66-0.72)

Mayfield

Yes

152

95

97.4

73.8

61.5

98.5

0.64

 

No

4

267

(93.6-99.3)

(68.9-78.2)

(55.2-67.6)

(98.0-98.8)

(0.61-0.67)

Ramsey

Yes

144

31

92.3

91.4

82.3

98.3

0.73

 

No

12

331

(86.9-96.0)

(88.1-94.1)

(75.8-87.6)

(97.8-98.7)

(0.70-0.76)

Harrington

Yes

147

36

94.2

90.1

80.3

98.4

0.73

 

No

9

326

(89.3-97.3)

(86.5-92.9)

(73.8-85.8)

(97.9-98.7)

(0.70-0.76)

New

Yes

145

33

92.9

90.9

81.5

98.3

0.73

 

No

11

329

(87.7-96.4)

(87.4-93.6)

(75.0-86.9)

(97.9-98.7)

(0.70-0.76)

  1. * Chart reviews identified whether there was any indication of a diabetic foot ulcer in the electronic medical records during October 1, 2002-September 30, 2003.
  2. PPV refers to positive predictive values and NPV, negative predictive values. PPV, NPV, and kappa coefficients were weighted.