Comparison of nasopharyngeal and guttural pouch specimens to determine the optimal sampling site to detect Streptococcus equi subsp equi carriers by DNA amplification

Table 1 Sensitivity and specificity comparing 1) 123 eqbE LAMP results to the sample type (guttural pouch lavage (GPL) versus not [nasopharyngeal wash, or nasopharyngeal flocked swab]); 2) eqbE LAMP results from 44 GPL samples to abnormal guttural pouch endoscopic findings; 3) eqbE LAMP from 44 GPL samples to empyema found on guttural pouch endoscopy; 4) eqbE LAMP to seeI PCR from 44 GPL samples

		eqbE LAMP		Total	Sensitivity	Specificity	PPV^b	NPV^c	Correctly Classified	ROC^d
		Positive	Negative
GPL^a Sampling Method	Yes	24	7	31	77%	78%	55%	91%	78%	0.83
	No	20	72	92
	Total	44	79	123
Abnormal Guttural Pouch Endoscopy	Yes	20	7	27	74%	70%	80%	63%	72%	0.80
	No	5	12	17
	Total	25	19	44
Guttural Pouch Empyema	Yes	12	1	13	92%	61%	50%	95%	70%	0.85
	No	12	19	31
	Total	24	20	44
seeI PCR	Positive	15	3	18	83%	65%	63%	85%	73%	0.78
	Negative	9	17	26
	Total	24	20	44

The cutoff probability for sample type is ≥ 0.5
^aGuttural Pouch Lavage
^bPositive predictive value (based on the detection of S. equi DNA and not on the presence of live organisms)
^cNegative predictive value
^dReceiver operator curve
All analyses are adjusted for antibiotics usage and the time from last antibiotic treatment by determining if they were significant confounders within the logistic regression via STATA

ISSN: 1746-6148