Disclosure of HIV-serodiscordant relationships and association with viral suppression: results from the Positive Plus One study

Figures & data

Table 1. Socio-demographic characteristics comparing HIV-positive and HIV-negative partners in current relationships (n = 540 partners in current HIV-serodiscordant relationships).

Factor	HIV-positive N (%)	HIV-negative N (%)	p-value
All	312	228
Age, median years (IQR)	43 (34, 52)	42 (32, 52.2)	p = 0.703^a
Relationship duration, median years (IQR)	4.5 (2, 11.6)	5 (2, 13)	p = 0.637^a
Study dyad (both partners in study)	153 (49%)	153 (67.1%)	p <0.001
Gender
Male	215 (68.9)	177 (77.6)	p = 0.0579^b
Female	89 (28.5)	47 (20.6)
Other (includes 1 missing value)	8 (2.6)	4 (1.8)
Gender / sexual orientation
Men who have sex with men	144 (46.2)	116 (50.9)	p = 0.0439
Heterosexual female	70 (22.4)	39 (17.1)
Heterosexual male	48 (15.4)	49 (21.5)
Other	50 (16)	24 (10.5)
Race / ethnicity
White	189 (62)	164 (72.6)	p = 0.0249
Black	23 (7.5)	20 (8.8)
Hispanic	30 (9.8)	10 (4.4)
Aboriginal/Indigenous	31 (10.2)	13 (5.8)
Other	32 (10.5)	19 (8.4)
Not stated	7 (2.2)	2 (0.9)
Education
Less than secondary school diploma	42 (13.5)	17 (7.5)	p = 0.0591
Secondary school diploma	65 (21)	44 (19.4)
Beyond secondary school	203 (65.5)	166 (73.1)
Not stated	2 (0.6)	1 (0.4)
Income
< $20,000	131 (43.1)	62 (27.8)	p <0.0001
$20,000 – $50,000	102 (33.6)	71 (31.8)
>$50,000	71 (23.4)	90 (40.4)
Not stated	8 (2.5)	5 (2.2)
Region
Ontario	179 (57.7)	130 (57.3)	p = 0.8549
British Columbia	35 (11.3)	30 (13.2)
Prairie Provinces (Saskatchewan, Manitoba, Alberta)	40 (12.9)	32 (14.1)
Quebec	35 (11.3)	20 (8.8)
Atlantic Provinces (New Brunswick, Prince Edward Island, Nova Scotia, Newfoundland and Labrador)	21 (6.8)	15 (6.6)
Not stated	2 (0.6)	1 (0.4)
Reside in large city (>500,000 population)	193 (61.9)	148 (64.9)	p = 0.5247
Married or cohabiting	235 (75.8)	176 (77.9)	p = 0.6483
Relationship satisfaction (Range 1–5), mean (SD)	4.2 (0.8)	4.3 (0.7)	p = 0.3327^a
HIV diagnosis relative to relationship
Diagnosis before relationship start	209 (67)	34 (14.9)	p <0.0001
Diagnosis at same time of relationship start	24 (7.7)	124 (54.4)
Diagnosis after relationship start	79 (25.3)	70 (30.7)
Disclosed relationship to own friends or family‡ (children, parents, siblings, other)	249 (83.3)	160 (72.4)	p = 0.0039
Disclosed relationship to healthcare professional‡ (healthcare providers e.g., doctors/nurses, therapists, counselors, or other support services)	244 (81.9)	165 (76)	p = 0.1313
Other sexual partners during current relationship	136 (44.4)	85 (37.8)	p = 0.1468
Sexual agreement (re. consensual non-monogamy)‡‡	218 (69.9)	146 (64)	p = 0.1815
Discussed HIV testing with partner	227 (73.2)	185 (81.9)	p = 0.0021
Relationship power dynamic relative to partner on condom use
More say	49 (19.4)	32 (17.4)	p = 0.1016
Equal say	44 (17.4)	19 (10.3)
Less say	160 (63.2)	133 (72.3)
NA/DK/Not stated	59 (18.9)	44 (19.3)
Relationship power dynamic‡‡‡
More say	0 (0)	1 (0.4)	p = 0.484^a
Equal say	269 (86.2)	193 (84.6)
Less say	43 (13.8)	34 (14.9)
Relationship turmoil because of HIV	64 (20.8)	46 (20.4)	p = 0.9906

Values are numbers and proportions unless otherwise stated; IQR = interquartile range.

p-values are Pearson’s chi-square tests unless otherwise stated; ^a T-test; ^b Fisher’s exact test.

‡ Disclosure of serodiscordant status was assessed by asking: “Below is a list of different groups of people. How many people in each group know, or have you told, that you are in a serodiscordant relationship?” Response options included “none,” “some,” or “all” (“some” and “all” were grouped for analysis).

‡‡ Constructed from three questions: (i) any agreements allowing sex with people outside of the relationship, (ii) discussions about making such agreements, and (iii) a list of specific sexual agreements permitting/proscribing certain activities.

‡‡‡ “More say” or “less say” was coded if each topic area were answered the same way. Topic areas included: whose friends to go out with; decision to have sex; decision on general activities done together; timing of serious discussions; decision to use condoms; choice of sexual behaviors; financial decisions; and overall assessment. “Equal say” was coded for all other response combinations.

Table 2. Reported viral suppression by HIV-positive and HIV-negative partners (n = 529‡).

Group	Total, n (100%)	Self-reported viral suppression‡‡, n (%)	χ^2 p-value
HIV-positive	303	263 (86.8)	p = 0.738
HIV-negative	226	193 (85.4)
All	529	456 (86.2)

‡ Nine HIV-positive and two HIV-negative participants with incomplete data were excluded; responses of “don’t know” (12 HIV+ and 14 HIV-) were included in “unsuppressed”.

‡‡ Response options consisted of <40 copies/mL (undetectable); 40–10,000 copies/mL (detectable); >10,000 copies/mL (detectable); and detectable but don’t know copies/mL.

Table 3. Agreement on reported viral suppression within dyads (n = 149‡).

Viral suppression	HIV-negative partner
HIV-positive partner	Suppressed	Unsuppressed (including DK)
Suppressed	124	8
Unsuppressed (including “do not know”)	6	11
Cohen’s Kappa measuring agreement	K = 0.558
McNemar’s test for direction of difference	p = 0.7893

‡ 3 HIV-positive and 1 HIV-negative participant with incomplete data were excluded.

Table 4. Multivariable models estimating association of factors with self-reported viral suppression by HIV-positive partners in current HIV-serodiscordant relationships in Canada (n = 303^‡).

Factor	Viral suppression N (%)	Crude odds ratio (95% CI)	p-value	Adjusted odds ratio (95% CI)^b	p-value
Block 1
Age (in 5 years, centered)^a		1.37 (1.16, 1.63)	0.0001	1.28 (1.07, 1.55)	0.0051
Gender / sexual orientation
MSM	139 (93.5)	1	0.0098
Heterosexual female	70 (82.9)	0.34 (0.14, 0.830)
Heterosexual male	46 (78.3)	0.25 (0.1, 0.66)
Other	48 (81.2)	0.3 (0.11, 0.8)
Education
Up to secondary school	105 (78.1)	1	0.0016	1	0.0206
Beyond secondary school	197 (91.4)	2.94 (1.51, 5.82)		2.43 (1.15, 5.26)
Block 2
Disclosed relationship to healthcare professional (healthcare providers e.g., doctors/nurses, therapists, counselors, or other support services)
No	52 (65.4)	1	<0.0001		0.0002
Yes	238 (91.2)	5.42 (2.64, 11.16)		4.71 (2.13, 10.51)
Other sexual partners during current relationship
No	167 (83.8)	1
Yes	131 (90.1)	1.72 (0.87, 3.54)	0.1199
Relationship power dynamic (“More or equal say” about 100% of statements answered)
No	171 (83)	1
Yes	130 (92.3)	2.38 (1.16, 5.22)	0.0172
Relationship turmoil because of HIV
No	240 (89.6)	1		1	0.0146
Yes	61 (75.4)	0.35 (0.18, 0.73)	0.0055	0.35 (0.15, 0.81)
Block 3
Income
< $20,000	127 (80.3)	1	0.0072
≥ $20,000	170 (91.2)	2.5 (1.28, 5.01)
Region
Ontario	176 (89.8)	1	0.0062
British Columbia	34 (94.1)	1.52 (0.45, 7.86)
Prairies	39 (69.2)	0.26 (0.11, 0.59)
Quebec	34 (91.2)	1.05 (0.35, 4.16)
Atlantic Provinces	19 (73.7)	0.31 (0.11, 0.99)

p-values are log likelihood ratio tests; CI = confidence interval.

^‡ Nine participants with incomplete data were excluded; 12 participants who responded “do not know” to current viral load were coded as unsuppressed.

^a Factor modeled as a linear effect; p(tr) = p(trend).

Table 5. Multivariable model estimating association of factors with self-reported viral suppression by HIV-negative partners in current HIV-serodiscordant relationships in Canada (n = 226^‡).

Factor	Viral suppression N (%)	Crude odds ratio (95% CI)	p-value	Adjusted odds ratio (95% CI)^b	p-value
Block 1
Gender / sexual orientation
MSM	115 (91.3)	1	0.0049	1	0.0139
Heterosexual female	39 (79.5)	0.37 (0.14, 1.01)		0.42 (0.15, 1.19)
Heterosexual male	49 (71.4)	0.24 (0.1, 0.58)		0.24 (0.09, 0.61)
Other	23 (95.7)	1.49 (0.32, 14.32)		1.2 (0.23, 12.19)
Block 2
Disclosed relationship to healthcare professional (healthcare providers e.g., doctors/nurses, therapists, counselors, or other support services)		3.08 (1.41, 6.69)	0.0054
No	51 (72.5)
Yes	165 (89.1)
HIV diagnosis relative to relationship
Diagnosis before relationship start	34 (88.2)	1	0.0238
Diagnosis at same time of relationship start	123 (90.2)	1.32 (0.37, 3.93)
Diagnosis after relationship start	69 (75.4)	0.44 (0.13, 1.28)
Relationship turmoil because of HIV
No	180 (87.8)	1	0.0429
Yes	45 (75.6)	0.43 (0.19, 0.97)
Block 3
Income
< $20,000	61 (78.7)	1	0.0983
≥ $20,000	161 (87.6)	1.92 (0.88, 4.08)
Reside in city >500 K inhabitants
No	79 (77.2)	1	0.0123
Yes	147 (89.8)	2.57 (1.23, 5.45)
Region
Ontario	130 (87.7)	1	0.0005	1	0.0013
British Columbia	30 (96.7)	2.83 (0.66, 26.43)		2.69 (0.61, 25.55)
Prairies	31 (67.7)	0.3 (0.12, 0.74)		0.29 (0.11, 0.75)
Quebec	20 (100)	5.91 (0.74, 764)		4.61 (0.56, 600.62)
Atlantic Provinces	15 (60)	0.21 (0.07, 0.67)		0.21 (0.06, 0.69)

p-values are log likelihood ratio tests; CI = confidence interval.

^‡ Two participants with incomplete data were excluded; 14 participants responded “do not know” to current viral load of their HIV-positive partner and the response was categorized as unsuppressed.

^a Factor modeled as a linear effect; p(tr) = p(trend).