. 2020 Jan 8;15(1):e0224359. doi: 10.1371/journal.pone.0224359

Table 2. Likelihood ratios (LR) estimating the value of a specific cervicovaginal infection (CVI) to predict a new incident CVI^a at the next visit based on 2-sided z-score test with p<0.05 significance level.

	Predicted Incident CVI Outcome
	Incident Vaginal Dysbiosis			Incident Sexually Transmitted Infection
Predictive Marker: CVI at the visit preceding the incidence of the CVI listed on the right	Nugent 4–6 (n = 450)	BV (n = 484)	Candida (n = 278)	NG (n = 72)	TV (n = 82)	CT (n = 58)	HSV-2 (n = 37)
Nugent 4–6		5.08^**	1.01	0.73	1.61	0.55	0.94
Nugent 7–10 (BV)	3.70^**		1.23	1.14	0.97	1.67	2.54^*
Candida	1.07	1.70^*		1.85	1.13	0.91^b	1.75^b
N. gonorrhoeae (NG)	0.85	0.86	1.43		0.88^b	2.01^b	2.o8^b
T. vaginalis (TV)	1.20	1.27	1.28	0^b		0.97^b	0^b
C. trachomatis (CT)	0.90	0.80	0.71^b	3.34^*^b	1.3^b		3.07^b
HSV-2	1.08	1.00	1.20	1.07	0.98	1.18

LR = 1 indicates a random occurrence of a CVI incident e.g. vaginal dysbiosis or sexually transmitted infection. LRs > 1 indicate higher and <1 lower likelihood of incident CVI.

*p<0.05

**p<0.01.

^aIncident CVI visit is defined as a CVI-positive visit diagnosed following a non-missing quarterly visit, which was negative for the same CVI.

^bThe Fisher exact test was used instead of the z-score test due to the small number of cases (<5) in the category examined.

Table 2. Likelihood ratios (LR) estimating the value of a specific cervicovaginal infection (CVI) to predict a new incident CVIa at the next visit based on 2-sided z-score test with p<0.05 significance level.

Table 2. Likelihood ratios (LR) estimating the value of a specific cervicovaginal infection (CVI) to predict a new incident CVI^a at the next visit based on 2-sided z-score test with p<0.05 significance level.