Intestinal biopsy is not always required to diagnose celiac disease: a retrospective analysis of combined antibody tests

Table 3 Performance of single antibody tests and selected combinations n = 268, 149 CD patients and 119 controls

	fn	fp	nc	sens	spec	ppv	npv	effic	lr+	lr-
Single tests				%	%	%	%	%
IgA anti-dpgli	33	4	0	78	97	97	78	86	23	0.23
IgG anti-dpgli	22	10	0	85	92	93	83	88	10	0.16
IgA anti-tTG	5	16	0	97	87	90	95	92	7	0.04
EMA	3	18	0	98	85	89	97	98	6	0.02
Combinations of 2 tests IgG anti-dpgli +
IgA anti-tTG	2	5	39	83	82	96	98	83	20	0.04
IgG anti-dpgli + EMA	1	6	39	84	82	95	99	83	17	0.01
IgA anti-dpgli + IgG anti-dpgli	15	1	37	73	89	99	88	80	87	0.11
Combinations of 3 tests IgA anti-dpgli +
IgG anti-dpgli + EMA	1	1	62	72	81	99	99	76	86	0.01
IgA anti-dpgli + IgG anti-dpgli + IgA anti-tTG *	2	1	60	73	81	99	98	76	87	0.01
IgG anti-dpgli + EMA + IgA anti-tTG	0	5	45	83	80	96	100	81	20	0.00
Combination of 4 tests IgG anti-dpgli +
IgA anti-dpgli + EMA + IgA anti-tTG	0	1	65	72	79	99	100	75	86	0.00

*This test combination is identical to the test in Table 2.
Positive, above the cut-off in all tests; negative, below the cut-off in all tests; fn, number of false-negative patients; fp, number of false-positive patients; nc, number of patients not classified because of discordant antibody results; sens, sensitivity; spec,specificity; ppv, positive predictive value; npv, negative predictive value; effic, efficiency; lr+, likelihood ratio positive; lr-, likelihood ratio negative.

ISSN: 1471-230X