Table 2. Comparative Characteristics of Mixed Feeders and Non-Mixed Feeders (n=91).

Factor	Mixed Feeding (n=16)	Not Mixed Feeding (n=75)	OR (95% CI)	p-value
Feels pressure from family	12 (75.0%)	39 (52.0%)	2.8 (0.8-9.4)	0.11
Social Stigma	8 (53.3%)	35 (47.3%)	1.3 (0.4-3.9)	0.78
Non-disclosure to partner of HIV status	8 (50.0%)	55 (75.3%)	3.1 (1-9.3)	0.07
Partner does not support feeding decision	9 (60.0%)	63 (86.3%)	4.2 (1.2-14.4)	0.03
Financial Difficulties	6 (66.7%)	39 (68.4%)	0.9 (0.2-4.1)	1.00
Age ≤ 24	7 (43.8%)	20 (26.7%)	2.1 (0.7-6.5)	0.23
Not high school graduate	9 (56.3%)	35 (46.7%)	1.5 (0.5-4.4)	0.59
Presented to clinic in third trimester	12 (75.0%)	56 (74.7%)	1.0 (0.3-3.5)	1.00
Not a wage-earner	11 (68.8%)	36 (48.0%)	0.4 (0.1-1.3)	0.17
Christian	14 (87.5%)	67 (89.3%)	0.8 (0.2-4.4)	1.00
Lack of unprompted Knowledge of MTCT	2 (12.5%)	28 (37.3%)	4.2 (0.9-19.7)	0.08
Intention to Breastfeed	11 (68.8%)	21 (28.0%)	5.7 (1.8-18.2)	0.003
CD4+ T Cell Count ≤ 200	9 (56.3%)	50 (67.6%)	0.6 (0.2-1.9)	0.40
Husband has multiple wives	7 (43.8%)	34 (45.3%)	0.9 (0.3-2.8)	1.00
Non-disclosure to family/others of HIV status	10 (33.3%)	42 (56.0%)	2.5 (0.8-8.2)	0.16
Danger of being thrown out of home if HIV+	8 (50.0%)	36 (48.0%)	1.1 (0.4-3.2)	1.00
Desire to keep HIV within family a secret	16 (100.0%)	62 (83.8%)	--* --*	0.12

Italicized factors are the main variables from the semi-structured questionaire. Other factors are from the larger Jos PMTCT cohort study database. Since we used Fisher's exact test, some p-values were exactly equal to 1 (which is impossible for parametric tests like chi-square).

^*Owing to the presence of a cell of 0 in the 2×2 table, the OR estimate for this variable failed to converge.