Table 2.

Measures of diagnostic accuracy of US examination, FNAC and core needle biopsy

	Results, n (%)
	FNAC (n = 58)	US examination (n = 58)a	CNB (n = 10)
True positive	0	2 (3)	0
Tumour
Breast cancer	n/a	1 (50)	n/a
Melanoma	n/a	1 (50)	n/a
SN location
Axilla	n/a	1 (50)	n/a
Groin	n/a	1 (50)	n/a
False positive	1 (2)	6 (10)	0
Tumour
Breast cancer	0	2 (33)	n/a
Melanoma	1 (100)	4 (66)	n/a
SN location
Axilla	1 (100)	6 (100)	n/a
Groin	0	0	n/a
True negative	43 (74)	38 (66)	9 (90)
Tumour
Breast cancer	17 (40)	15 (40)	9 (100)
Melanoma	26 (61)	23 (61)	0
SN location
Axilla	36 (80)b	29 (76)	9 (100)
Groin	9 (20)	9 (24)	0
False negative	14 (24)	12 (21)	1 (10)
Tumour
Breast cancer	4 (29)	3 (25)	1 (100)
Melanoma	10 (71)	9 (75)	0
SN location
Axilla	6 (43)	5 (42)	1 (100)
Groin	8 (57)	7 (58)	0
Measures, % (95% CI)
Sensitivity	0	14 (3–38)	0
Specificity	98 (90–100)	86 (74–94)	100
Positive predictive value	0	25 (5 – 59)	0
Negative predictive value	75 (63–85)	76 (63–86)	90 (63–99)

CI confidence interval, CNB core needle biopsy, FNAC fine needle aspiration cytology, n/a not applicable, US ultrasound

^aAccording to the Berlin criteria, a malignant/suspicious sentinel node on US examination was recorded as positive

^bTwo patients had SNs identified in the groin and axilla