Statistical analysis was performed using SPSS. Between-groups differences in mean values of con­tinuous variables were tested by independent samples t tests or by nonparametric Mann-Whitney Rank Sum tests when the distributions were not normal. The differences in fre­quencies of categorical variables were tested by % test with Yates’ correction for continuity or by Fisher’s exact test when the expected number of observations per cell was less than five. Associations between continuous variables were assessed by Pearson correlation coefficient.

Logistic regression analysis was used to analyze the asso­ciation of dichotomous outcome variable (malignant versus benign) with continuous and categorical predictor variables. The statistical inferences were based on the level of signifi­cance P<0.05. Receiver operating characteristic (ROC) curves were constructed for the biochemical variables.


TABLE 1 Patient demographics and final diagnosis

Patient

Sex

Age
(years)

Diagnosis

1

F

84

Carcinoma of the pancreas

2

F

61

Carcinoma of the pancreas

3

F

83

Carcinoma of the pancreas

4

F

71

Carcinoma of the pancreas

5

F

71

Carcinoma of the pancreas

6

F

78

Carcinoma of the gallbladder

7

F

78

Carcinoma of the pancreas

8

M

79

Metastatic obstruction from gastric

carcinoma

9

M

87

Metastatic obstruction from gastric

carcinoma

10

M

62

Carcinoma of the pancreas

11

M

78

Carcinoma of the pancreas

12

M

70

Carcinoma of the pancreas

13

M

75

Cholangiocarcinoma

14

M

83

Carcinoma of the pancreas

15

M

48

Carcinoma of the pancreas

16

F

56

Chronic pancreatitis

17

F

64

Benign intrahepatic duct stricture due

to operative trauma

18

F

25

Benign intrahepatic duct stricture due

to operative trauma

19

F

87

Papillary stenosis

20

F

41

Stricture at anastomotic site post-OLT

21

F

84

Idiopathic distal benign stricture

22

F

43

Primary sclerosing cholangitis

23

M

44

Primary sclerosing cholangitis

24

M

43

Ischemic stricture post-OLT

25

M

56

Ischemic stricture post-OLT

26

M

38

Chronic pancreatitis

27

M

52

Chronic pancreatitis

28

M

64

Papillary stenosis

29

M

57

Chronic pancreatitis

To determine optimal threshold levels for each diagnostic parameter, ROC plots were constructed using the observed true and false positive rates at each potential threshold level. A best fit ROC curve was constructed according to methods published elsewhere. The threshold value providing the best compromise between true and false positive rates was estimated from the ROC plot. Likelihood ratios were calculated from the fitted ROC curve.
Save on your pharmacy bills. Buy online