Sustained maternal human immunodeficiency virus viral load suppression and cascade of human immunodeficiency virus testing among exposed infants in Rwanda

Abstract

BACKGROUND

To prevent mother to child transmission (MTCT) of human immunodeficiency virus (HIV), sustained maternal viral load suppression (VLS) and early HIV testing among HIV exposed infants (HEI) is critical.

AIM

To investigate maternal viral load results and infant HIV testing uptake at 6–weeks, and 9–months and 18–months in Rwanda.

METHODS

Between 2015 and 2022, VLS (< 200 copies/mL) was measured among pregnant women living with HIV (WLHIV) from 38–healthcare facilities. Viral loads (VL) were measured at 6–months, 12–months and 24–months, respectively. For maternal VL, the unit of analysis was visit–pair, and the pairs were created to define those with VL < 200 copies/mL at two consecutive visits as having sustained VLS, persistent viremia (VL ≥ 200 copies/mL at two consecutive visits), viral rebound (VL < 200 copies/mL at prior visit only) and newly suppressed (VL < 200 copies/mL at subsequent visit only). HEI were considered to have persistent HIV testing if they had all three HIV tests. Poisson regression models with generalized estimating equations were used to estimate the adjusted incidence rate ratio (aIRR) and 95%CI for factors associated with sustained VLS and persistent HIV testing.

RESULTS

A total of 1145 mother-infant pairs were analyzed. Infant HIV testing uptake at 6– weeks, 9–months and 18–months was 1145 (100.0%), 1089 (95.1%), 1006 (87.9%) respectively. Nine hundred ninety–nine HEI (87.3%) tested for HIV persistently. At 18–months, the incidence of HIV among HEI was 8 (0.7%). Of 1145 mothers, 1076 (94.0%) had ≥ 2 VL results making a total of 2010 visit–pairs (142–single; 934–double visit–pairs). The incidence rate of sustained VLS, persistent viremia, viral rebound and new suppression were 91.0%, 1.3%, 3.6% and 4.0% respectively. Maternal disclosure of HIV status (aIRR = 1.08, 95%CI: 1.02–1.14) was associated with increased likelihood of sustained VLS. Having peer support (aIRR = 1.05 95%CI: 1.01–1.10) was associated with persistent HIV testing among HEI.

CONCLUSION

Sustained VLS is high among pregnant WLHIV in Rwanda. The low incidence of HIV among HEI may be attributed to high VLS levels. Targeted interventions, including enhanced HIV disclosure and peer support, are crucial for improving sustained VLS and increasing infant HIV testing uptake to reduce MTCT.

Key Words: Sustained viral load suppression; Mother to child human immunodeficiency virus transmission; Women living with human immunodeficiency virus; Human immunodeficiency virus exposed infants; Human immunodeficiency virus testing; Rwanda

Core Tip: Sustained maternal viral load suppression (VLS) is a key component for prevention of mother to child transmission of human immunodeficiency virus (HIV). We explored maternal viral load results, infant HIV testing uptake at 6–weeks, 9-months and 18–months and HIV infant HIV incidence in Rwanda. Of 1145 mothers, the incidence of sustained VLS was 91%. Infant HIV testing uptake at 6–weeks, 9–months and 18–months was 100.0%, 95.1% and 87.9%, respectively. Only 0.7% of infant were diagnosed with HIV at 18 months. Maternal disclosure of HIV status and having peer support were associated with increased likelihood of sustained VLS and infant HIV testing, respectively. High maternal VLS was critical to minimize HIV infection among HIV-exposed infants in this cohort.

INTRODUCTION

Global estimates indicate that 1.3 million women and girls living with human immunodeficiency virus (HIV) become pregnant annually[1]. Mother to child transmission (MTCT) of HIV contributes a significant proportion of pediatric HIV infections. Without any interventions, the percentage of pediatric HIV attributed by MTCT is between 15% and 45%[2]. By the year 2023, the rate of MTCT among children born to women living with HIV (WLHIV) at 24-months of follow up in Rwanda was 1%[3]. This rate is 65% lower than what was reported five years prior to 2023[4]. Although there is significant improvement of reduced MTCT, these transmissions still occur in Rwanda. Due to high morbidity and mortality associated with pediatric HIV infections[5,6], continued efforts to eliminate MTCT are critical.

Different interventions are known to prevent MTCT. These includes maternal serial HIV testing during pregnancy and breastfeeding period leading to early detection of HIV and linkage to care[7]. As of 2021, data from the Joint United Nations Programme on HIV/acquired immunodeficiency syndrome (UNAIDS) showed that HIV testing and treatment among women during pregnancy and breastfeeding period have averted over two million pediatric HIV infections since 2001[8]. Between June 2022 and July 2023, 383973 pregnant women with unknown HIV status in Rwanda attended their first antenatal clinic (ANC) and of these, 365759 (95.3%) tested for HIV and 1421 (0.4%) were living with HIV[3]. Over 99% of those living with HIV were kept on treatment. These data indicate great commitments towards elimination of MTCT of HIV in the country. Although nearly all pregnant mothers diagnosed with HIV were kept on treatment during that fiscal year, ensuring retention in care and optimal adherence of anti-retroviral therapy (ART) to achieve sustained viral load suppression (VLS) is paramount.

Rwanda has made significant progresses in achieving the UNAIDS 95-95-95 goals for HIV epidemic control[9]. Despite achieving these UNAIDS goals, since the adoption of the treat all policy, there is limited data on the trends of VLS among pregnant and breastfeeding WLHIV. However, reported VLS (HIV-RNA < 1000 copies/mL) among people living with HIV in the country is > 90% in multiple studies[10,11]. When VLS was redefined using a lower threshold (HIV-RNA < 200 copies/mL), reported VLS in the country was still high[12]. Although these levels of VLS were high, they only signify suppression at specific time points in the HIV care continuum. These cross-sectional reports of VLS are important, however, to meet the desired goals of HIV treatment, understanding trends of sustained VLS are more informative than VLS at specific time points. The scarcity of these trends data on sustained VLS among pregnant WLHIV and impact they may have on MTCT in the country necessitates additional studies.

Early detection and timely linkage to care are crucial for HIV exposed infants (HEI). Previous data from Rwanda indicated that the two-year mortality risk among HEI was three times higher than that of non-HEI[13]. One study in Rwanda that reported data during option B+ showed that an overall 24-month free-survival of 93.2% among HEI[14]. Since 2012, Rwanda implemented HIV testing among HEI at 6–weeks, 9–months, 18–months, and 24–months, however, these longitudinal HIV testing uptake data remain undocumented.

In this analysis of mother-infant pairs, we aim to explore key aspects of HIV treatment and prevention among WLHIV and their infants in Rwanda. Specifically, we examine the incidence of sustained VLS, viremia, viral rebound, and new suppression among women receiving antiretroviral therapy. Additionally, we assess the factors associated with sustained VLS to better understand the determinants of effective viral suppression. Furthermore, we investigate the prevalence and factors influencing persistent HIV testing among HEI and evaluate the incidence of HIV among HEI. These findings will provide critical insights into optimizing HIV care and prevention strategies in Rwanda.

MATERIALS AND METHODS

Study design and population

We conducted a retrospective cohort study using routinely collected data of 18 years or older pregnant (any gestation age) WLHIV and their HIV-exposed infants from thirty-eight healthcare facilities in Rwanda. Data were collected between July and November 2022. The cohort consisted of 1145 mother infant pairs. Although 1145 women-initiated treatment from these healthcare facilities, only 1076 (94.0%) had documented two or more viral load results. Maternal viral load outcome data for this analysis included those with two or more viral load results whereas infant HIV testing outcome included all 1145 infants.

Definition of variables

Viral loads (VL) were measured at 6–months, 12–months and 24–months. The minimum and maximum number of maternal viral load results were two and three. We created visit pairs using these viral load results to define the outcome of interest and create a unit of analysis. The main outcome of interest was sustained VLS. Sustained VLS was achieved if a person had suppressed viral load (< 200 copies/mL) at two consecutive visits. We chose a threshold of 200 copies/mL per the current definition of VLS from the Rwanda National HIV treatment guidelines[15]. Other secondary outcomes included: (1) Persistent viremia, if the viral load results was (≥ 200 copies/mL ) at two consecutive visits; (2) Viral load rebound, if the viral load was suppressed (< 200 copies/mL) at prior visit but unsuppressed (≥ 200 copies/mL) at subsequent visit; and (3) Newly suppressed, if the viral load was unsuppressed (≥ 200 copies/mL) at prior visit but suppressed at subsequent visit (< 200 copies/mL). In addition, we reported infant HIV testing uptake at 6–weeks, 9–months and 18–months and persistent HIV testing. HEI were considered to have tested persistently if they had all three HIV tests documented. Individual and healthcare facility characteristics were explored in this analysis. Individual characteristics evaluated included age (18–24 years, 25–34 years, 35–50 years), assignment to peer educator (assigned vs not assigned), disclosure of HIV status to others (disclosed, did not disclose), prior exposure to ART, pregnancy status at the start of ART, number of pregnancies since starting ART and place of delivery (home vs hospital). Facility characteristics included, type of healthcare facility (clinic vs health center), number of healthcare workers in a healthcare facility.

Trend in viral load outcomes during pregnancy and postpartum period

To compare trends in viral load outcome (suppression, detection, rebound) between our current analysis and data from other parts of the world, we searched literature for studies that reported these outcomes among WLHIV.

Sample size estimation

Sample size was estimated based on the following assumption: (1) The prevalence of first ANC visit within 12 weeks would be 50%; and (2) With the power of 80% and precision of 5%, the estimated sample size was 1296 participants. However, data for only 1145 women were retrieved.

Statistical analysis

Frequencies and proportions of categorical variables and median and interquartile range (IQR) for continuous variables were presented. Categorical participant characteristics were stratified by different virological outcomes. Since the unit of analysis for maternal viral load outcomes were based on visit pairs, to account for repeated measured data Poisson regression models with generalized estimating equations were used to compute incidence rate ratios (IRRs) and 95%CI for factors associated with sustained VLS. First, bivariate Poisson regression models were fit. Then, all variables with a P value < 0.2 on bivariate analysis were included in multivariable models. Since study participants were clustered within healthcare facilities; the latter were considered as clusters and random effects models were used to account for clustering in both bivariate and multivariable analysis. To assess for factors associated persistent infant HIV testing, log binomial regression models were used to estimate adjusted risk ratio (aRR) and 95%CI. Complete case analysis was conducted. Statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, United States).

RESULTS

Overall

Between 2015 and 2022, a total of 1145 mother-infant pairs from 38 health care facilities attended clinics and were identified. Maternal median age (IQR ) was 35 (30–40 years) years.

Incidence of viral load outcome

Of 1145 mothers, 1076 (94.0%) had ≥ 2 VL results making a total of 2010 visit–pairs (142–single; 934–double visit–pairs). The incidence of sustained VLS, persistent viremia, viral rebound and new suppression were 91.0%, 1.3%, 3.6% and 4.0% respectively (Table 1). Those who were married/living together, disclosed HIV status, or were pregnant when they initiated ART exhibited having more sustained VLS compared to those who were single, did not disclose HIV status and not pregnant at ART initiation respectively. Generally, there is no much variability in other viral load outcomes across different characteristics; however, divorced/widowed compared to married/living together (3.0% vs 1.0 %) had higher likelihood of persistent viremia. Those who were single compared to divorced/widowed (7.0% vs 2.6%) had higher likelihood of achieving new VLS. Those who delivered at home had higher proportion of new VLS compared to those who delivered at the healthcare facility (11.4% vs 3.8%).

Table 1 Incidence of sustained viral load suppression, persistent viremia, viral rebound and new suppression among pregnant women living with human immunodeficiency virus in Rwanda.

	Sustained viral load suppression		Persistent viremia		Viral rebound		Newly suppressed
	n/N	%	n/N	%	n/N	%	n/N	%
Overall	1830/2010	91.0	27/2010	1.3	73/2010	3.6	80/2010	4.0
Age (years)
13-24	185/199	93.0	4/199	2.0	6/199	3.0	4/199	2.0
25-34	599/661	90.6	8/661	1.2	22/661	3.3	32/661	4.8
35-50	1046/1150	91.0	15/1150	1.3	45/1150	3.9	44/1150	3.8
Marital status
Single	358/412	86.9	8/412	1.9	17/412	4.1	29/412	7.0
Married or living together	1311/1423	92.1	14/1423	1.0	52/1423	3.6	46/1423	3.2
Divorced/widowed	140/53	91.5	5/53	3.3	4/53	2.6	4/53	2.6
Missing	21/22	95.5	0/22	0.0	0/22	0.0	1/22	4.5
Assigned to peer educator
No	1184/1303	90.9	22/1303	1.7	46/1304	3.5	51/1304	3.9
Yes	646/707	91.4	5/707	0.7	27/707	3.8	29/707	4.1
Human immunodeficiency virus status disclosed
No	386/448	86.2	10/448	2.2	20/448	4.5	32/448	7.1
Yes	1444/1562	92.4	17/1562	1.1	53/1562	3.4	48/1562	3.1
Prior exposure to ART
No	627/706	88.8	14/706	2.0	26/706	3.7	39/706	5.5
Yes	1173/1269	92.4	13/1269	1.0	45/1269	3.6	38/1269	3.0
Missing	30/35	85.7	0/35	0.0	2/35	5.7	3/35	8.6
Pregnant at the start of ART
No	514/578	88.9	13/578	2.2	24/578	4.2	27/578	4.7
Yes	1218/1318	92.4	8/1318	0.6	43/1318	3.3	49/1318	3.5
Missing	98/114	86.0	6/114	5.3	6/114	5.3	4/114	2.4
Number of pregnancies since start
0	313/336	93.2	2/336	0.6	10/336	3.0	11/336	3.3
1-2	1312/1452	90.4	22/1452	1.5	56/1452	3.9	62/1452	4.3
≥ 3	205/222	92.3	3/222	1.4	7/222	3.2	7/222	3.2
Place of delivery
Hospital	1704/1869	91.2	25/1869	1.3	68/1829	3.6	72/1829	3.8
Home	38/44	86.4	0/44	0.0	1/44	2.3	5/44	11.4
Missing	88/97	90.7	2/97	2.1	4/97	4.1	3/97	3.1
Type of health care facility
Clinic	24/28	85.7	0/28	0.0	2/28	7.1	2/28	7.1
Health center	1806/1982	91.1	27/1982	1.4	71/1982	3.6	78/1982	3.9
Number of health care workers
1-5	833/925	90.0	13/925	1.4	37/925	4.0	42/925	4.5
6-10	675/745	91.1	11/745	1.5	30/745	4.0	25/745	3.4
≥ 3	318/340	93.5	3/340	0.9	6/340	1.8	13/340	3.8

ART: Anti-retroviral therapy.

Factors associated with sustained VLS: Modelling analysis

Maternal and facility characteristics associated with sustained VLS are presented in Table 2. On bivariate analysis, marital status, married/living together vs single (IRR = 1.06; 95%CI: 1.01–1.12); HIV status disclosure, disclosed vs did not disclose (IRR = 1.07; 95%CI: 1.02–1.13) and maternal pregnancy status at ART initiation, pregnant vs not pregnant (IRR = 1.04; 95%CI: 1.01–1.08) were associated with increased likelihood of achieving sustained VLS. On multivariable analysis, compared to those who did not disclose HIV status, those who disclosed were associated with 6% increased likelihood of achieving sustained VLS [adjusted IRR (aIRR) = 1.06; 95%CI: 1.00–1.11] . Women who attended at healthcare facilities with higher number of healthcare workers ≥ 11 vs 1-5, tended to have higher likelihood of achieving sustained VLS (aIRR = 1.04; 95%CI: 1.00–1.09), although this association was not statistically significant.

Table 2 Factors associated with sustained viral load suppression among pregnant women living with human immunodeficiency virus in Rwanda.

Characteristics	Bivariate		Multivariable
	IRR (95%CI)	P value	Adjusted IRR (95%CI)	P value
Age (years)
18-24	Reference
25-34	0.97 (0.92–1.03)	0.354
35-50	0.98 (0.93–1.03)	0.399
Marital status
Single	Reference		Reference
Married or living together	1.06 (1.01–1.12)	0.0201	1.04 (0.99–1.10)	0.106
Divorced/widowed	1.06 (0.98–1.14)	0.140	1.05 (0.97–1.13)	0.206
Assigned to peer educator
No	Reference
Yes	1.01 (0.97–1.04)	0.741
Human immunodeficiency virus status disclosed
No	Reference		Reference
Yes	1.07 (1.02–1.13)	0.0041	1.06 (1.00–1.11)	0.031
Prior exposure to ART
No	Reference		Reference
Yes	1.04 (1.01–1.08)	0.0261	1.01 (0.97–1.06)	0.609
Pregnant at the start of ART
No	Reference		Reference
Yes	1.04 (1.00–1.08)	0.0401	1.03 (0.98–1.08)	0.233
Number of pregnancies since start
0	Reference		Reference
1-2	0.97 (0.93–1.01)	0.176	0.99 (0.95–1.03)	0.630
≥ 3	0.99 (0.94–1.05)	0.854	0.99 (0.94–1.05)	0.881
Type of health care facility
Clinic	Reference
Health center	1.05 (0.86–1.30)	0.605
Number of health care workers
1-5	Reference		Reference
6-10	1.01 (0.97–1.05)	0.596	1.01 (0.97–1.04)	0.761
≥ 11	1.04 (1.00–1.08)	0.080	1.04 (1.00–1.09)	0.062

¹Indicates significant P value.

ART: Anti-retroviral therapy; IRR: incidence rate ratio; Ref: Reference.

HIV testing uptake

Infant HIV testing uptake at 6–weeks, 9–months and 18–months was 1145 (100.0%), 1089 (95.1%) 1006 (87.9%) respectively (Table 3). Nine hundred ninety–nine HEI (87.3%) tested for HIV persistently.

Table 3 Human immunodeficiency virus testing cascade among human immunodeficiency virus exposed infants in Rwanda.

	n	%
PCR done 6 weeks
Yes	1145	100.0
No	0	0.0
Serology done at 9 months
Yes	1089	95.1
No	56	4.9
Serology done at 18 months
Yes	1006	87.9
No	139	12.1
Tested persistently
Yes	999	87.3
No	146	12.7
Human immunodeficiency virus testing results
Positive	8	0.7
Negative	1137	99.3

Factors associated with persistent infant HIV testing: Modelling analysis

Bivariate and multivariable factors associated with infant HIV testing were presented in Table 4. On bivariate analysis, women who were assigned to peer educators during pregnancy had higher likelihood of having their infants tested for HIV persistently compared to those who were not assigned to peer educators (risk ratio = 1.04; 95%CI: 1.00–1.09). On multivariable analysis assignment to peer educators continued to be associated with persistent infant HIV testing with those women who were assigned having 6% increased likelihood of their infants being tested for HIV persistently compared to those who were not assigned to peer educators (aRR = 1.06; 95%CI: 1.01–1.11). Furthermore, women who attended at healthcare facilities with higher number of healthcare workers ≥ 11 vs 1-5, tended to have higher likelihood of having their infants tested for HIV persistently (aRR = 1.06; 95%CI: 0.99–1.13), although this association was not statistically significant.

Table 4 Maternal and facility characteristics associated with persistent human immunodeficiency virus testing among human immunodeficiency virus exposed infants in Rwanda.

Characteristics	Bivariate		Multivariable
	RR (95%CI)	P value	Adjusted RR (95%CI)	P value
Age
18-24	Reference
25-34	1.03 (0.95–1.13)	0.376
35-50	1.04 (0.95–1.13)	0.406
Marital status
Single	Reference
Married or living together	1.02 (0.96–1.08)	0.470
Divorced/widowed	0.97 (0.87–1.08)	0.568
Assigned to peer educator
No	Reference		Reference
Yes	1.05 (1.01–1.10)	0.0231	1.06 (1.01–1.11)	0.0211
Human immunodeficiency virus status disclosed
No	Reference		Reference
Yes	0.96 (0.92–1.01)	0.120	0.97 (0.92–1.02)	0.184
Prior exposure to ART
No	Reference
Yes	0.99 (0.95–1.04)	0.651
Pregnant at the start of ART
No	Reference		Reference
Yes	0.96 (0.92–1.01)	0.103	0.97 (0.93–1.02)	0.225
Number of pregnancies since start
0	Reference
1-2	1.01 (0.95–1.08)	0.747
≥ 3	0.97 (0.88–1.06)	0.507
Place of delivery
Hospital	Reference
Home	0.92 (0.77–1.12)	0.413
Type of health care facility
Clinic	Reference
Health center	1.11 (0.88–1.40)	0.393
Number of health care workers
1-5	Reference		Reference
6-10	0.98 (0.93–1.03)	0.463	0.99 (0.94–1.03)	0.571
≥ 11	1.04 (0.98–1.10)	0.154	1.06 (0.99–1.13)	0.091

¹Indicates significant P value.

ART: Anti-retroviral therapy; RR: Risk ratio; Ref: Reference.

Trends in viral load outcome during pregnancy and postpartum periods among WLHIV from different parts of the world

There are disparities in the definition of VLS for different studies reviewed, creating challenges in summarizing trends in comparing sustained VLS (Table 5)[16-31]. Moreover, the timing of viral load measurements throughout pregnancy and postpartum periods were also different. Despite of these challenges, while some studies showed increased proportions of women with good viral load outcomes (i.e. undetectable, suppression) from pregnancy to postpartum periods[16-19], others showed increased proportions of women with poor viral load outcomes[20-24] (i.e. detectable, non-suppression) over time[25-28], and others showed mixed results[29,30,32]. Several factors have been identified to be associated with different viral load outcomes. For example, younger age, and initiation of ART during third trimester was associated with increased risk of poor virologic outcome[16,24]. Depression and interpersonal abuse were also associated with increased odds of poor virologic outcomes[23,26]. Engagement in HIV care within 90 days of delivery[19], disclosure of HIV status to partners[27] and having a VL test performed during antenatal visits[17] were associated with better virologic outcomes.

Table 5 Trend in viral load outcomes during pregnancy and postpartum period among women living with human immunodeficiency virus from different parts of the world.

Ref.	Country	Study year	Outcome and definition	Percentages of outcome
				Time 1	Time 2	Time 3
Patel et al[28], 2018	United states	1996-2015	VLS (HIV RNA < 500 copies/mL) measured at start of pregnancy, end or pregnancy and 6 months postpartum	34.8	60.1	42.7
Lyatuu et al[16], 2021	Tanzania	2014-2016	VLS (HIV RNA < 400 copies/mL) measured between 0-11 months and at ≥ 36 months since enrolled at prevention of mother to child transmission	85.1	90.6
Landes et al[29], 2021	Malawi	2014-2016	Detectable viral load (HIV RNA > 40 copies/mL measured at enrollment, 12-months and 24-months post-partum	15.5	9.9	12.0
Boucoiran et al[20], 2017	Canada	1997-2015	Viral rebound HIV RNA > 50 copies/mL or > 400 copies/mL for 1997-1998 enrollees and measured within 1 month before delivery	0.0	6.0
Moyo et al[17], 2021	South Africa	2016-2017	Viraemia (HIV RNA ≥ 50 copies/mL) measured during pregnancy, at delivery and 24-months post-partum	53.6	36.9	33.5
Ntlantsana et al[18], 2019	South Africa	2016-2017	Viraemia (HIV RNA ≥ 50 copies/mL) measured at 4 weeks post enrollment among pregnant women those who were viremic at enrollment	100	53.3
Meade et al[21], 2018	United States	2011-2016	VLS (HIV RNA < 200 copies/mL) measured at delivery, 12-months and 24-months postpartum	59.0	15.0	0.0
Zanré et al[22], 2024	Canada	2012-2020	Undetectable VL (< 50 copies/mL), measured at delivery and 2-18 weeks postpartum	94.0	87.0
Jiang et al[31], 2024	Kenya	2015-2017	Viral load suppressed (HIV RNA < 1000 copies/mL) measure within the past 6-months and 24-months postpartum	100	83.8
Myer et al[23], 2017	South Africa	2013-2014	VLS (HIV RNA < 50 copies/mL). measured at enrollment and 12-months postpartum	100	70.0
Adams et al[19], 2015	United states	2005-2011	VLS (HIV RNA < 200 copies/mL). measured at 12-months and 24-months postpartum	31.0	34.0
Moyo et al[30], 2021	South Africa	2018-2020	Viral load un-suppression (HIV RNA > 1000 copies/mL). measured during pregnancy, delivery and up to 9-months postpartum	18.0	20.6	14.8
Flynn et al[24], 2021	Multiple countries	2011-2014	VLS (HIV RNA < 1000 copies/mL) measured at 6-weeks, 14-weeks, 26-weeks, and 50-weeks postpartum. Here we present 6-weeks, 14-weeks and 26-weeks data	91.0	89.0	88.0
Hatcher et al[25], 2022	South Africa	2013-2017	VLS (HIV RNA < 50 copies/mL) measured at 6-months, and 12-months postpartum	84.0	67.0
Brittain et al[26], 2020	South Africa	2013-2014	Viral load ≥ 50 copies/mL at delivery and 12-months postpartum among women who were diagnosed with HIV during pregnancy	21.0	32.0
Ngarina et al[27], 2015	Tanzania	2004-2006	Viral load ≥ 400 copies/mL measured at enrollment, 3-months, 6-months, 12-months and 24-months postpartum. Here we present proportion of VL ≥ 400 at enrollment, 6-months and 12-months postpartum	97.0	22.0	61.0

Multiple countries included India, Malawi, South Africa, Tanzania, Uganda, Zambia and Zimbabwe. HIV: Human immunodeficiency virus; VL: Viral loads; VLS: Viral load suppression.

DISCUSSION

From this cohort of mother-infant pairs in Rwanda, we evaluated trends of maternal viral load outcomes, infant HIV testing uptake and incidence. Overall, the incidence of sustained VLS, persistent viremia, viral rebound and new suppression were 91.0%, 1.3%, 3.6% and 4.0% respectively. Infant HIV testing uptake at 6–weeks, 9–months and 18–months was 100.0%, 95.1% and 87.9% respectively. Collectively, 87.3% of infants had documented HIV tests for all three testing visits with an overall HIV incidence of 0.7%. We have noted individual and facility characteristics which were associated with likelihood of achieving sustained VLS and infant HIV testing. For example, disclosure of HIV status and higher number of healthcare professionals in the facility were both associated with increased chance of achieving sustained VLS. Furthermore, assignment to peer educators and higher number of healthcare professionals in the facility were both associated with increased chance of infant HIV testing.

Our results showed WLHIV in Rwanda exhibited sustained VLS in higher percentages over time. These data indicate great progress towards elimination of MTCT. These high rates of sustained VLS explains low incidence of HIV among HEI in this setting. Globally, stigma against HIV has been a common challenging problem among most people living with HIV[33,34]. Stigma impairs retention in care and adherence to ART, critical components in the management of people living with HIV. In this cohort, we have noted that women who disclosed HIV status to others had higher likelihood of achieving sustained VLS. These results are similar to previously presented data[27,35]. Implementation of stigma prevention strategies and encouraging disclosure of HIV status is beneficial to achieving and maintaining sustained VLS and prevent MTCT of HIV. We also observed facilities with higher number of healthcare professionals facilitated sustained VLS. It is likely that, in these facilities, patients had enough time to have their treatment concerns being addressed to satisfaction leading to optimal adherence and achieving sustained VLS. On the contrary, shortage of healthcare professionals pose challenges to both patients and the healthcare staff. These may include minimal time to address patients concerns and staff burnout. These challenges may have negative impact on achieving sustained VLS. Balancing workload by improving healthcare staff to patient ratios would benefit achieving sustained VLS.

One hundred percent of HEI tested for HIV at 6–weeks and nearly 90% of them were followed up 18-months to ascertain their HIV status. These data are higher than what have been reported elsewhere. For example, only 76% of HEI tested for HIV at 6–weeks in Tanzania[36], 48% in Mozambique[37] and 73% in Zambia[38]. In South Africa, 9–months and 18–months HIV testing uptake was 39% and 24% respectively, which improved to 45% and 34% respectively with intense tracing efforts[39]. Early detection and linkage to care is critical to reduce morbidity and mortality associated with HIV in children. We observed women who were assigned to peer educators during pregnancy had higher chance of having their infants tested for HIV regularly compared to women who were not assigned to peer educators. Social support provided by peer educators is critical for successful navigations of prevention of MTCT services. In our previous publication on mother-infant pairs in Nigeria, we demonstrated that support provided by mentor mothers was associated with timely presentation for early infant diagnosis of HIV[40]. In addition, others have shown that, the median infant HIV testing uptake at 2–months and 9–months prior to the implementation of peer support via a mentor mothers was 26%–67% and 48%-100% respectively, compared to 67%–99% and 87%–100% respectively when mentor mother support was implemented[41]. Therefore, in addition to other factors, to address barriers of infant HIV testing, programs should consider assigning pregnant WLHIV to peer support as these supports have proven benefits in many settings[42]. Although lacking statistical significance, facilities with higher numbers of healthcare workers tended to have a higher likelihood of infants being tested. These results are consistent with others who noted that for every one increase in clinic staff, there is a four-fold increase in new HIV cases being identified[43].

A few limitations could be identified from this study. First, to assess sustained VLS, study participants needed to have multiple VL over time. In this analysis, about 13% of participants had only 2 viral load measurements leading to the contribution of a single visit pair. This limited exploration of long-term viral load outcomes for these patients. Second, the study was not primarily intended to address factors associated with sustained VLS and infant HIV testing. Because of that, other potential characteristics associated with VLS such as medication adherence were not collected. The lack of adherence data impaired our ability to recommend actionable plans to address medication nonadherence. We did not find significant association between number of healthcare workers and sustained VLS. Ideal metrics would be to compute a ratio of the number of healthcare workers per volume of patients attended. The absence of the average number of patients been attended limited our ability to use this ratio and assess its association with sustained VLS. The reliance on facility-based, routinely collected data may have excluded women lost to follow-up or those who delivered at home and never showed up at the clinics, there by introducing selection bias, skewing results toward better outcomes and reducing generalizability. Despite these limitations, the main strength of this study is the use of routinely collected data in an ideal clinic environment enabling our results to be generalizable. Furthermore, the use of multiple viral load results enabled us to assess trends of viral load outcomes over time which is more illuminating than reporting viral load outcome at a single time point. Moreover, we reported long term HIV testing uptake among HEI compared to other studies that focuses on the first six weeks of life.

CONCLUSION

Sustained VLS is high among pregnant WLHIV in Rwanda. The low incidence of HIV among HEI could be attributed to high levels of sustained VLS among these women. Our results corroborate results of previous research which also documented that disclosure of HIV status during pregnancy was associated with increased odds of long-term VLS[27]. We recommend encouraging HIV couple counselling and increased HIV disclosure to reduce MTCT of HIV. Furthermore, we also recommend targeted interventions focusing on peer support placement and increased healthcare staff to patient ratio is critical to achieve sustained VLS and infant HIV testing uptake in the efforts of reducing MTCT of HIV.