Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in the multi-country US-PEPFAR PROMOTE study: A cross-sectional analysis.

BACKGROUND: Despite recent efforts to scale-up lifelong combination antiretroviral therapy (cART) in sub-Saharan Africa, high rates of unsuppressed viremia persist among cART users, and many countries in the region fall short of the UNAIDS 2020 target to have 90% virally suppressed. We sought to determine the factors associated with unsuppressed viremia (defined for the purpose of this study as >200 copies/ml) among sub-Saharan African women on lifelong cART.
METHODS: This cross-sectional analysis was based on baseline data of the PROMOTE longitudinal cohort study at 8 sites in Uganda, Malawi, Zimbabwe and South Africa. The study enrolled 1987 women living with HIV who initiated lifelong cART at least 1-5 years ago. Socio-demographic, clinical, and cART adherence data were collected. We used multivariable Poisson regression with robust variance to identify factors associated with unsuppressed viremia.
RESULTS: At enrolment, 1947/1987 (98%) women reported taking cART. Of these, HIV-1 remained detectable in 293/1934 (15%), while 216/1934 (11.2%) were considered unsuppressed (>200 copies/ml). The following factors were associated with an increased risk of unsuppressed viremia: not having household electricity (adjusted prevalence risk ratio (aPRR) 1.74, 95% confidence interval (CI) 1.28-2.36, p<0.001); not being married (aPRR 1.32, 95% CI 0.99-1.78, p = 0.061), self-reported missed cART doses (aPRR 1.63, 95% CI 1.24-2.13, p<0.001); recent hospitalization (aPRR 2.48, 95% CI 1.28-4.80, p = 0.007) and experiencing abnormal vaginal discharge in the last three months (aPRR 1.88; 95% CI 1.16-3.04, p = 0.010). Longer time on cART (aPRR 0.75, 95% CI 0.64-0.88, p<0.001) and being older (aPRR 0.77, 95% CI 0.76-0.88, p<0.001) were associated with reduced risk of unsuppressed viremia.
CONCLUSION: Socioeconomic barriers such as poverty, and individual barriers like not being married, young age, and self-reported missed doses are key predictors of unsuppressed viremia. Targeted interventions are needed to improve cART adherence among women living with HIV with this risk factor profile.

Introduction

Since 2012, the rapid scale up of the World Health Organization (WHO) option B+ strategy among pregnant or breastfeeding women living with Human Immunodeficiency Virus (HIV) has resulted in a substantial reduction in maternal morbidity and mortality, as well as incident pediatric HIV infections[1]. Subsequently with the introduction of the Universal “Test and Treat” strategy, approximately 23.3 million people (including women) had access to combination Antiretroviral Therapy (cART) globally in 2018 [2]. Ensuring sustained adherence to and virologic suppression on cART is paramount in achieving the Joint United Nations Programme on HIV and AIDS (UNAIDS) 90-90-90 2020 strategy in ending the epidemic by 2030 [3-5].

Barriers to achieving the UNAIDS 2020 strategy regarding the ‘third 90’ persist in sub-Saharan Africa in part due to suboptimal cART adherence [4, 5]. In the absence of viral resistance, HIV viral load assessment is the proxy for adherence, therefore the contributing factors to both adherence and viremia may overlap. The most common factors identified as being associated with decreased adherence include individual factors like younger age below 24 years, forgetting the dosing time, depression, and substance use[6, 7]. The predominant contextual issue remains stigmatization and disclosure [4-7]. Other factors such as length of time on ART, level of education, personal motivation to start ART and satisfaction with health worker information availed [6-8]. Additionally, studies have shown that adherence to cART is better during pregnancy compared to the post-partum period [4, 8]. This may be attributed to less motivation to take ART since the risk of HIV transmission has ceased after cessation of breast feeding, as well as a possible break in transition from postnatal to general HIV care [4, 5].

Whereas the scale up of virologic monitoring in sub-Saharan Africa since 2013 has led to the availability of data regarding factors associated with virologic detectability, there is a paucity of literature as to which factors are most strongly associated with unsuppressed viremia among African pregnant women and mothers living with HIV[9]. What is known to date is that virological detectabilty in resource-limited settings has been associated with the presence of comorbidities like tuberculosis or psychiatric disease, higher pretreatment HIV RNA levels, repeat testers after suspected virologic failure and initiation of cART late in pregnancy[6, 7, 10–13]. The vast majority of existing studies have assessed factors associated with viremia >1000 or >400 copies/ml [6, 7, 10, 11]. Emerging antriretroviral drug resistance has been known to occur from levels of 200 copies/ml or above, and use of this threshold eliminates most cases of apparent viremia caused by viral load blips or assay variability[14, 15].Therefore this analyses sought to determine clinical and demographic risk factors associated with unsuppressed viremia above 200 copies/ml among a well characterized cohort of women living with HIV originally in the IMPAACT 1077BF/1077FF PROMISE (Promoting Maternal and Infant Survival Everywhere) clinical trial at the time of their entry into the PROMOTE Study. This cohort presents a unique opportunity to assess longer term treatment outcomes among women previously randomized to different antiretroviral (ARV) regimens (WHO option A and option B) during pregnancy and postpartum for the purpose of preventing perinatal HIV transmission[16]. Subsequently these women transitioned to lifelong cART in response to the START study which showed clear benefit of universal ART in June 2015(10). The PROMOTE study approach provides data from current public sector HIV care provision mixed with precise individualized clinical and laboratory data collected under trial settings.

Materials and methods

Design

The PEPFAR-PROMOTE study is a five-year observational cohort of sub-Saharan African women with HIV and their children previously enrolled in the PROMISE trial for the prevention of mother to child transmission of HIV [17]. The study is being conducted at eight research sites in four African countries: MUJHU/Kampala (Uganda), Blantyre and Lilongwe (Malawi), Harare Family Care, Seke North and St. Mary’s (Zimbabwe), PHRU/Johannesburg and CAPRISA Umlazi/Durban (South Africa). Commencing three months after the PROMISE trial closed-out, 1987 mothers and their children were recruited from September 2016 to August 2017 into the PROMOTE study. A cross-sectional analysis of unsuppressed viremia and related factors was conducted at enrolment.

Setting and study populations

At entry into the PROMISE trial, these women were pregnant, asymptomatic with pre-ART CD4+ T-cell counts ≥ 350 cells/mm3 and more details of eligibility criteria into the PROMISE trial have been previously published[16]. At enrollment into the PROMOTE study, women were either having a subsequent pregnancy or not pregnant (up to 4 years post-delivery) or breastfeeding babies born between PROMISE study close up and PROMOTE study start (Fig 1). Uganda site had women on both LPV/r (69%) and EFV (29.8%) based regimens, while Zimbabwe and Malawi had most women (97.3% and 98.3% respectively) on EFV based regimens only and South Africa had 96.6% of the women on the fixed dose combination (TDF/FTC/EFV).

Fig 1

Flow chart showing numbers from PROMISE trial to PROMOTE study.

Inclusion criteria

Women and children enrolled in the PROMISE trial from the 8 African PROMISE sites described above, who were willing to provide informed consent to enroll and continue follow-up in the PROMOTE study.

Exclusion criteria

Women who were judged by the site team as having social or other reasons which would make it difficult for the mother/child pair to comply with study requirements, for example those women who felt that continued follow up may predispose them to social harm and women who planned on relocating home out of the study catchment area and could not keep up with study visit requirements afterwards were excluded from the study.

Enrolment study procedures for women

Mothers were enrolled after appropriate counseling and consenting in the PROMOTE study. At the enrolment visit, trained study workers administered socio-demographic and ART adherence questionnaires. A complete medical history and physical examination, including WHO clinical staging, was performed. Included in a holistic package of comprehensive counseling was the provision of study-specific antiretroviral adherence counseling. Enrollment laboratory evaluations included: viral load and CD4+ cell count. Assays were done at various Good Clinical Laboratory Practice (GCLP) and Clinical Laboratory Improvement Amendments (CLIA) compliant site laboratories adherent to standards for proper collection, processing, labeling, and transport of specimens. The COBAS TaqMan and Abbot viral load assays were used with a lower limit of detectability at 20 copies/ml and 40 copies/ml respectively. All questionnaire responses and laboratory data were completed on designated case report forms (CRFs) by trained research site personnel.

Ethical considerations

The PROMOTE study was approved by all relevant institutional review boards (IRBs) in the U.S. and participating African research sites/countries. These include MUJHU/Kampala, Uganda: The Joint Clinical Research Centre (JCRC) IRB in Uganda and The Johns Hopkins Medical Institutions (JHMI) IRB in the U.S.; Umlazi, Durban, South Africa: Biomedical Research Ethics Committee and Kwazulu-Natal Department of Health; Harare, Seke North and St. Mary’s sites, Zimbabwe: Medical Research. Council of Zimbabwe (MRCZ) National Ethics Committee; Blantyre, Malawi: College of Medicine Research and Ethics Committee (COMREC) in Malawi and Johns Hopkins Medical Institutions (JHMI) IRB in the U.S.; Lilongwe, Malawi: National Health Sciences Research Committee (NHSRC) in Malawi and University of North Carolina, Chapel Hill (UNC-CH) Office of Human Research Ethics IRB in the U.S.; and PHRU, Johannesburg, South Africa: University of Witwatersrand Human Research Ethics Committee (Medical). All women provided written informed consent to enroll and be followed up for the duration of the study with their children and agreed to provide study samples for protocol lab safety assays, as well as for storage of blood and hair samples.

Statistical analysis

We analyzed baseline data from a multi-country cohort study to estimate the proportion of women living with HIV who had unsuppressed viremia (defined as viral load above 200 copies/ml) and to identify predictors of unsuppressed viremia. Furthermore, in accordance with WHO threshold for treatment failure, we performed sensitivity analyses and defined unsuppressed viremia as viral load above 1000 copies/ml. Fisher’s exact, Chi-square test of independence or Wilcoxon rank sum tests were used to test for an association between baseline characteristics and unsuppressed viremia. We used univariable and multivariable Poisson regression models with robust variance to identify the predictors of unsuppressed viremia, and calculated prevalence risk ratios to measure the strength of an association between baseline characteristics and unsuppressed viremia. This was done in two ways (i) each predictor was fitted in the model, and (ii) multivariable model with multiple predictors included in the model. Variables included in the multivariable analyses were chosen based on prior research on risk factors, clinical associations and biological plausibility. Due to site country variability in ARV regimen type, we did not adjust for the ARV regimen in any of the models to avoid multi-collinearity between country and ARV regimen variables. Variables with increased missing data (>20% of observations missing per variable) and variables that were highly correlated were not included in the multivariable model. In an exploratory analyses, women with detectable viral loads were further stratified into the following thresholds (<50, 50–200, 201–1000 and >1000 copies per ml based on the varied thresholds by various HIV cART committees in resource rich and resource limited settings)[18, 19].

Results

Overall, 1987 mothers were enrolled into the PROMOTE study, of whom 1947 (98%) women reported taking ART at the enrolment visit and HIV-1 viral load results were available for 1934. HIV-1 VL was above the limit of quantification in 293/1934 (15%). A total of 216/1934 (11.2%) presented with an unsuppressed viremia above 200 copies/ml. Furthermore, among the 293 women with detectable viral load, 24 (8.2%) had VL below 50, 53 (18.1%) had VL between 50 and 200, 50 (17.1%) had VL between 201 and 1000, while 166 (56.7%) had VL above 1000 copies/ml.

The individual and contextual baseline characteristics of the PROMOTE study have been reported elsewhere [17]. Table 1 displays baseline characteristics stratified by viral load below and above 200 copies/ml. With the exception of employment status, HIV status disclosure, condom usage, all the baseline variables were associated with unsuppressed viremia >200 copies/ml (Table 1). Notably, the prevalence of unsuppressed viremia > 200 copies/ml was the highest (17.5%) in Malawi compared to other countries. Additionally, 24% of women with recent hospitalization had unsuppressed viremia.

Table 1

Individual and contextual baseline characteristics.

Variable	Viral load≤200copies/ml (N = 1718)	Viral load >200 copies/ml (N = 216)	Total	p-value
Socioeconomic and demographic factors
Country, n (%)				<0.001
Uganda	319 (90.6%)	33 (9.4%)	352
Malawi	522 (82.5%)	111 (17.5%)	633
Zimbabwe	406 (90.6%)	42 (9.4%)	448
South Africa	471 (94.0%)	30 (6.0%)	501
Age (years), median(IQR)	31 (28–35)	29 (25–33)	31(27–35)	<0.001
Baseline CD4 cell count (cells/μL),median (IQR)	852(674–1064)	615(472–810)	825(646–1040)	<0.001
Marital status, n(%)				0.003
Other (single, divorced, widowed, separated)	325 (84.4%)	60 (15.6%)	385
Married/regular partner	1393 (89.9%)	156 (10.1%)	1549
Employment, n(%)a				0.076
Formal employment	383 (90.5%)	40 (9.5%)	423
Self-employment (small business)	532 (86.5%)	83 (13.5%)	615
Not employed/housewife	802 (89.7%)	92 (10.3%)	894
Highest level of education, n(%)				0.011
Secondary school completed or tertiary education	1227 (90.0%)	136 (10.0%)	1363
Lower level of education	491 (86.0%)	80 (14.0%)	571
Electricity in the premises, n(%)
Available	1209 (91.8%)	108 (8.2%)	1317	<0.001
Not available	509 (82.5%)	108 (17.5%)	617
Tap water in the premises, n(%)
Available	1139 (90.4%)	121 (9.6%)	1260	0.004
Not available	579 (85.9%)	95 (14.1%)	674
Travel time from home to clinic, n(%)b				0.044
Less than 30 minutes	444 (92.3%)	37 (7.7%)	481
30–60 minutes	757 (87.7%)	106 (12.3%)	863
1–2 hours	389 (87.2%)	57 (12.8%)	446
Greater than 2 hours	127 (88.8%)	16 (11.2%)	143
Disclosed HIV status to partner, n(%)c				0.465
Disclosed to partner	1201 (90.2%)	131 (9.8%)	1332
No disclosure	192 (88.5%)	25 (11.5%)	217
Partner’s HIV statusd
Positive	724 (90.2%)	79 (9.8%)	803	0.908
Negative	257 (89.9%)	29 (10.1%)	286
Condom usage during sex in last 3 months, n(%)e				0.811
Always	502 (89.8%)	57 (10.2%)	559
Sometimes	506 (88.6%)	65 (11.4%)	571
Never	265 (88.9%)	33 (11.1%)	298
Years on ART, median(IQR)	2 (1–2)	1 (1–2)	2 (1–2)	<0.001
Clinical factors
Admitted to hospital in the past 3 months, n(%)
Yes	22 (75.9%)	7 (24.1%)	29	0.036
No	1696 (89.0%)	209 (11.0%)	1905
Received TB treatment in the last 3 months, n(%)
Yes	8 (88.9%)	1 (11.1%)	9	1.00
No	1710 (88.8%)	215 (11.2%)	1925
Presence of abnormal vaginal discharge in the last 3 months, n (%)
Yes	80 (82.5%)	17 (17.5%)	97	0.047
No	1638 (89.2%)	199 (10.8%)	1837
Currently breastfeeding, n(%)b
Yes	163 (84.9%)	29 (15.1%)	192	0.071
No	1554 (89.3%)	187 (10.7%)	1741
ART related factors
ART regimen
EFV or NVP basedf	1007 (86.4%)	158 (13.6%)	1165	<0.001
LPV/r basedg	252 (90.0%)	28 (10.0%)	280
Fixed dose combination (TDF/FTC/EFV)	459 (93.9%)	30 (6.1%)	489
Since last scheduled visit, when was the last time you missed any of your ARV doses? n(%)h
Never missed any doses	1257 (91.8%)	113 (8.2%)	1370	<0.001
Within the last week	72 (79.1%)	19 (20.9%)	91
1–2 weeks ago	40 (85.1%)	7 (14.9%)	47
2–4 weeks ago	64 (83.1%)	13 (16.9%)	77
1–3 months ago	183 (85.9%)	30 (14.1%)	213
More than 3 months ago	61 (77.2%)	18 (22.8%)	79
Don't know or cannot remember	33 (91.7%)	3 (8.3%)	36
Number of days ARV doses missed in last four days, n(%)h
None	1638 (90.6%)	169 (9.4%)	1807	<0.001
One day	52 (86.7%)	8 (13.3%)	60
Two days	8 (66.7%)	4 (33.3%)	12
Three days	1 (50.0%)	1 (50.0%)	2
Four days	9 (34.6%)	17 (65.4%)	26
Don't know or cannot remember	2 (33.3%)	4 (66.7%)	6
Awareness of dosing instructions, n(%)h
Yes	826 (87.9%)	114 (12.1%)	940	0.017
No	842 (90.3%)	90 (9.7%)	932
Don't know or cannot remember	42 (100.0%)	0	42

a2 participants had missing data

b1 participant had missing data

camongst 1549 women with partners

damongst women whom their partner’s got tested and women knew their HIV status

eamongst women who report sexual activities, EFV/NVP basedf (ABC/3TC or AZT/3TC or TDF/3TC), LPV/r basedg (ABC/3TC or AZT/3TC or TDF/3TC or TDF/FTC)

hamongst those who were on ART at enrollment. Please note that we presented row percentages to display the proportion of patients with detectable viremia for each level of the variable.

Results from the multivariable model are shown in Table 2. Recent hospital admission and experiencing abnormal vaginal discharge in the last three months were associated with a 2.5-fold and almost 2-fold higher risk of detectable viremia respectively (adjusted prevalence risk ratio (aPRR) 2.48, 95% confidence interval (CI) 1.28–4.80, p = 0.007; aPRR 1.88; 95% CI 1.16–3.04, p = 0.010). In addition, the absence of socioeconomic factors such as electricity in the household premises was associated with a 74% higher risk of detectable viremia (aPRR 1.74, 95% CI 1.28–2.36, p<0.001). Married or women with regular partner had a 32% increased risk of presenting with detectable viremia (aPRR 1.32, 95% CI 0.99–1.78, p = 0.061) compared to women who were neither married nor have partners. Women who either missed some of their ART doses or cannot recall missing any doses since the last scheduled visit were more likely to present with detectable viremia (aPRR 1.63, 95% CI 1.24–2.13, p<0.001) compared to women who never missed any doses. Longer exposure to ART (aPRR: 0.75, 95% CI 0.64–0.88), p<0.001) and being older (aPRR 0.77, 95% CI 0.76–0.88, p<0.001) were associated with lower risk of detectable viremia. Other variables that were tested did not show strong association with detectable viremia such as: secondary school level completion or tertiary education attainment (aPRR 1.17, 95% CI 0.85–1.61, p = 0.326; compared to women with lower level of education); travel time to the cART clinic of 1 hour or more reduces the risk of detectable viremia by 12% when compared to travelling for less than an hour (aPRR 0.88, 95% CI 0.66–1.91, p = 0.417) and being aware of antiretroviral (ARV) medication dosing instructions (aPRR 1.08, 0.81–1.43, p = 0.612).

Table 2

Factors associated with detectable viral load >200 copies/ml.

	Multivariable1		Multivariable2
Variable	RR (95% CI)	p-value	aRR (95% CI)	p-value
Age (5-year increase)	0.77 (0.67–0.88)	<0.001	0.77 (0.67–0.88)	<0.001
Marital status (ref: married/regular partner)
Other (single, divorced, widowed, separated)	1.55 (1.19–2.04)	0.001	1.32 (0.99–1.78)	0.061
Employment (ref: formal employment)
Not employed	0.92 (0.65–1.32)	0.655	0.84 (0.58–1.21)	0.349
Self employed	1.09 (0.75–1.58)	0.646	1.11 (0.76–1.63)	0.577
Education (ref: secondary school not complete)
Secondary school complete	1.04 (0.78–1.38)	0.806	1.17 (0.85–1.61)	0.326
Electricity in the premises (ref: Yes)
No	1.62 (1.22–2.14)	<0.001	1.74 (1.28–2.36)	<0.001
Tap water in the premises (ref: Yes)
No	1.24 (0.96–1.62)	0.103	-	-
Travel time to clinic from home (ref: less than 1 hour)
1 hour or more	1.01 (0.77–1.33)	0.936	0.88 (0.66–1.19)	0.417
Condom use during sex in last three months (ref: Never)
Always	1.12 (0.74–1.72)	0.585	-	-
Sometimes	1.05 (0.71–1.56)	0.811	-	-
HIV status disclosure to partner (ref = Yes)
No	1.45 (0.94–2.25)	0.096	-	-
Abnormal vaginal discharge in last three months (ref: No)
Yes	2.12 (1.32–3.39)	0.002	1.88 (1.16–3.04)	0.010
Hospital admission in last three months (ref: No)
Yes	2.29 (1.20–4.36)	0.012	2.48 (1.28–4.80)	0.007
Aware of ARV medication dosing instructions (ref: No)
Yes	1.05 (0.79–1.41)	0.723	1.08 (0.81–1.43)	0.612
Missed ART doses since last visit (ref: None)
Missed some doses	2.01 (1.55–2.60)	<0.001	1.63 (1.24–2.13)	<0.001
Time since ART initiation(per 1-year increase)	0.68 (0.57–0.81)	<0.001	0.75 (0.64–0.88)	<0.001

1Each predictor fitted separately while adjusted for the country

2Multivariable model with multiple predictors

Notably, the most common reason for missing ART dosing was travelling without sufficient ARV supply and simply forgetting (Fig 2). Despite 11% of women not disclosing their HIV status to their male partners, this variable was not significantly associated with detectable viremia.

Fig 2

Reasons for missing ART dose.

Results of the sensitivity analyses using the WHO threshold of 1000 copies/ml did not show any major differences in the predictors of unsuppressed viremia (S3 Table).

Discussion

We found that 89% of the 1934 women initiated on antiretroviral treatment had suppressed viremia ≤200 copies/ml. This is slightly below the UNAIDS 90% target bearing in mind that the 200 copies/ml threshold is lower than the 1000 copies threshold used by UNAIDS. Among the women with detectable viremia, we identified that socio demographic, self-reported non-adherence, and clinical factors were associated with detectable viremia > 200copies/ml.

The higher proportion of women with detectable viremia than the UNAIDS target suggests that challenges to achieving the 3rd ‘90’ still persist even among women who are in an ideal setting (research) compared to the programs in resource limited settings[5, 9]. The association between socio demographic factors namely the absence of household electricity, a proxy for lower economic status, and detectable viremia above 200 copies/ml, were consistent with other literature that has highlighted that socio demographic factors are key predictors to poor adherence despite one being on cART [7, 20]. The association between younger age and unsuppressed viremia is consistent with current literature [5, 6]. Additionally, longer duration of cART use as protective of unsuppressed viremia is consistent with other literature as well [21]. In contrast to other literature, we found that employment, education and travel time to the clinic were not predictive of unsuppressed viremia in the PROMOTE cohort at baseline in the multivariable analyses [4, 20].

Based on self-reported adherence reports using the study questionnaire, we found that missed ART doses was significantly associated with unsuppressed viremia. Even though there is uncertainty regarding the reliability of self-reported adherence, this measure of adherence still remains as a cheap and easily determined mode of adherence monitoring in resource limited settings using appropriate tools[7]. We noted however that a small proportion of women (9.7%) were not aware of the dosing instructions. This factor was not a significant predictor of viral detectability. This finding may suggest that patient education is still lacking. Effective interventions like motivational adherence counseling ensure two-way input and steers away from the traditional methods of adherence counselling[5].

The most common reason for missed doses was travelling with insufficient ARV supply and forgetting, which are in line with other literature[20]. Re-emphasis on simple measures for example, setting an alarm (e.g. mobile phone alarm) and linking dosing with daily activities, should be part and parcel of adherence counseling. Provision of a pillbox is a tool used by participating South African sites. Traveling without ARV supply as a reason could fall in the “forgetting” category, or could be a cover for non-disclosure while visiting family homes. The reason of no privacy, be it at work or in home, also implies non -disclosure and feared stigmatization. Other reasons included being too busy and running out of treatment prior to visit. This is commonly due to life’s every day demands including work commitments. Countries are now working towards improving the access to medicines by means of decentralized dispensing for convenient pill collection. South Africa has implemented a new model where the dispensing services are contracted to private pharmacies[22]. Other means of differentiated care strategies to ensure patient convenience include multi-month prescriptions, fast-track refills and community adherence groups who assist with collection and distribution of cART as done in Malawi, Uganda and Zimbabwe[5, 23].

Relatedly, 11% of all the women had not disclosed their HIV status to their primary partner but this was not predictive of detectable viremia. This PROMOTE result is contrary to results of other studies citing non-disclosure as being associated with poor adherence in different HIV infected populations [20, 21]. Contextual issues however remain a major underlying contributor to detectable viremia and patients may not be forthcoming with these reasons for a missed dose. Challenges in the efforts to combat stigma related issues are persistent in hyperendemic settings even with advanced HIV outreach programmes with effective adherence promotion programmes.

Whereas adherence barriers like fear of side effects, pill burden, perceptions that ART is harmful, feeling sick and depressed have been associated with virologic detectability in various AIDS Clinical Trial Participants in United States, current first-line cART comprises low pill burden and improved safety profile with low toxicity[24]. Even if about 89% of women were receiving Efavirenz or Neviparine based regimen, ART regimen type was not significantly associated with virologic detectability. The aim of the health system and its providers is to ensure adherence to first-line cART regimens to prevent the need for second- and third-line cART which are more toxic and have greater pill burden.

More so, recent hospitalization was significantly associated with viral detectability. Recent hospitalization may suggest clinical failure in this subgroup of non-suppressed women. Hospital readmissions are frequent among HIV positive adults when compared to their HIV negative counterparts [25]. Persistent viremia above 200 copies/ml has been found to predispose one to higher morbidity, virologic failure and mortality especially among cases of delayed ART switch to 2nd line therapy [15, 26].

Recent abnormal vaginal discharge was also significantly associated with unsuppressed viremia. Vaginal discharge may imply Sexual Transmitted Infections (STIs) in these women. Studies have shown that presence of an STI increases risk of transmission of other STIs and increased morbidity [27]. Notably, literature on the effects of STIs on HIV viremia is limited. Condom use was not associated with viral detectability above 200 copies/ml in this analysis.

This study contributes much needed data regarding the factors associated with unsuppressed viremia among African pregnant and breastfeeding women receiving ART treatment for life; more so because viral load testing is only a fairly recent intervention in HIV treatment monitoring. These data demonstrate that socioeconomic barriers and age remain key predictors of viral detectability, as well as recent hospitalization and recent abnormal vaginal discharge.

The strengths in these analyses include that PROMOTE is one of the largest cohort studies of HIV infected women of child bearing age in Africa with a large sample size; and is being conducted in multiple sites in East and Southern Africa, which increases the generalizability of the findings. In addition, there is ongoing quality assurance facilitated by the data management center, and annual research site monitoring as part of the study. Limitations of this study include lack of cART resistance, depression, Intimate Partner Violence (IPV) and stigma data. Furthermore, this is a baseline cross sectional analysis hence virologic failure has not been confirmed. Self-report in response to questionnaires introduces the potential for recall bias. Additionally, this PROMISE-PROMOTE cohort may not be representative of women in the general population.

Future plans for the study

Follow up trends in viral load and adherence data over the 5 year follow up in PROMOTE will be presented when available; as will the relation of hair drug levels and drug resistance testing correlates. Point- of- care VL testing coupled with motivation adherence counseling and adherence risk assessment tool development are in the process of being implemented at the sites in Zimbabwe and Uganda respectively.

Conclusion

This baseline analysis of the PROMOTE study set out to evaluate what clinical and socioeconomic factors were associated with a detectable viremia of >200 copies/mL in sub-Saharan African mothers on lifelong cART. Baseline data demonstrate that socioeconomic barriers such as lack of electricity in household premises (a proxy for poverty), not being married, young age, and prior history of missing pill doses remain key predictors of viral detectability. This study suggests that the use of self-reported adherence to cART may still play a role in adherence determination in the absence of superior measures. The most common reasons given by mothers for missing cART doses emphasize the need for effective motivational adherence counseling coupled with economic empowerment, enabling women to improve adherence by using simple reminders and a differentiated care model tailored to mother’s needs. The PROMOTE study insights provide opportunities for possible development and improvement of targeted /cost effective implementation strategies to help support lifetime maternal adherence to both cART and HIV care.

Manuscript data.

(XLSX)

Click here for additional data file.

Variable name and label.

(XLSX)

Click here for additional data file.

Sensitivity analyses using the 1000 copies/ml WHO threshold.

(DOCX)

Click here for additional data file.

28 Jul 2019

PONE-D-19-17608

Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in a multi-country cohort study: US-PEPFAR PROMOTE study.

PLOS ONE

Dear Dr Atuhaire,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Sep 11 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Marcel Yotebieng

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Additional Editor Comments (if provided):

Reviews have made many suggestions as to how to improve the readability of the manuscript and validity of the results. A you revised the manuscript, make to not overstate the application/generilizability of the results.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

Reviewer #4: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

Reviewer #4: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

Reviewer #4: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

Reviewer #4: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors conducted an analysis to explore factors associated with viremia above 200 among women from multiple sites in Africa. I found the paper to be written well with a clear logical flow and tight focus. However, I have identified minor concerns that the authors should work on.

- Page 11, line 87 …are most strongly assoicated with 87 unsuppressed…..

Correct the typo

- Font size is not consistent, see beginning of results section, and probably other places

- Footnote in table2

a2 missing data, b1 missing data, camongst 1549

What do these numbers 2 and 1 supposed to represent?

- Reference 2, 2018 [ Is there something missing, it appears as if you started unclosed bracket?

Reviewer #2: July 21, 2019

Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in a mult-country cohort study: US-PEPFAR PROMOTE Study.

PLOS ONE

Summary: This manuscript describes factors related to viremia in WHLIV in sub-Saharan Africa within the PROMOTE multi-site study. The topic is of interest to clinicians and policy makers focused on caring for both WLHIV and pregnant/postpartum WLHIV. The manuscript is generally well written though does not present much novel data and has several gaps outlined below.

Suggested major revisions:

1. Introduction: Throughout the introduction the authors make assertions that they do not support with sufficient references. This includes Lines 77, 81, 82 and 109.

2. Introduction: description of the PROMISE study “…women randomized… to initiate varied ARV regimens during pregnancy….” Lines 99-103 is not well worded. Presumably women were not randomized in response to the START study though study protocols may have changed after START results were released?

3. Methods: inclusion criteria from PROMISE should be included since that defines the cohort currently being followed.

4. A bias could be introduced by allowing site teams to “judge social or other reasons which would make it difficult for mother/child pair to comply with study requirements” Is there any actual criteria to define this? How many mother/child pairs were excluded due to site judgement decisions? If this is a substantial number/% (>5%) authors should add to limitations.

5. Results: a flow diagram of PROMISE participants who then rolled over to PROMOTE and reasons for exclusion, missing baseline VL etc should be included

6. The pregnancy and postpartum status (e.g. months postpartum) should be reported either at time of VL reported or at least enrollment to PROMOTE to give context to findings.

7. Figure 1 does not add anything to the text presented.

8. If available, ART regimen should be included in Table 1. Since PROMISE randomized to different regimens and regimen is a known factor affecting viremia. Also would be interesting to note if women had changed from their PROMISE regimen.

9. Unclear what “awareness of dosing instructions” means and how this question was asked, given it is a yes/no response.

10. Discussion: Authors confuse the UNAIDS 90-90-90 targets. The third 90 actually represents 90% of women on ART or 73% of all WHLIV (diagnose 90%, put 90% of diagnosed women on ART had achieve viral suppression in 90% of those on ART).

11. Condom use result is randomly added to paragraph on reasons for ARV non-adherence

12. Separate paragraph on differentiated care models could be combined with section on travel and forgetting pills.

13. The statement “one would expect less disclosure issues and united communities taking treatment together” seems judgmental and not supported by literature demonstrating ongoing issues with stigma in high prevalence areas.

14. Authors do not effectively argue that pill burden and low toxicity should reduce barriers to adherence since they do not describe what ART regimens or formulations anywhere in results. Most countries still recommend efavirenz-based regimens for WHLIV which may have substantial side effects for a significant proportion of women.

15. Lines 276-284 seem to jump from result to result without an overlying theme. Why recent vaginal discharge is associated with viremia is not explored at all.

16. Lack of resistance data is a major limitation as the authors argue that the VL presented are all due to adherence. Additionally, lack of data on depression, IPV, stigma are also limitations.

17. Authors should also note bias in sample as all participants are from those retained in a long term study (PROMISE) which is not usually representative of the general population. This may have resulted in lower viremia than WLHIV generally.

18. While the authors found an association between self-reported adherence in a study setting and VL, this is often not the case in routine programmatic settings where patients have a disincentive to truthfully report adherence. This conclusion should be modified.

19. The author suggest motivational interviewing and differentiated care models will be effective but it is unclear how this will impact SES factors, young age, and marital status.

Minor revisions

20. Abstract: the conclusion suggests not being married and young age are socioeconomic barriers- suggest re-wording.

21. Methods: reword “one of the longest ongoing follow up epidemiologic multinational cohorts”

22. Lines 184- 185 repetitive phrasing (Notably, of note)

23. Discussion suggests SES is a known predictor of adherence while the data actually seem mixed on this and the references do not support this assertion.

24. Rephrase “steers away” line 240

Reviewer #3: This is an important manuscript addressing predictors of viremia in women of childbearing age living with HIV across sub-saharan Africa in an observational cohort study. This is generally very well presented. Below I present minor comments. In addition, a close review for typos/grammatical errors would be worthwhile.

Abstract

“Not being married” is mentioned in conclusion but not results. Consider adding to results (or dropping from conclusion) in abstract so the conclusions are drawn follow the results presented.

Intro

line 79

…”satisfaction with health worker information availed were conducive”… —> reword for clarity

Line 82

…”This may be attributed to less motivation to protect the child post delivery” —> perhaps clarify that the motivation may decline when the risk is no longer present (currently sounds as though the motivation declines but not clear that the risk has declined)

Methods - Eligibility

How long was the minimum time on ART prior to the enrollment VL - did you apply any restrictions for inclusion, e.g., at least 3 or 6 months of ART?

RESULTS

Starting Line 198: “The predictors of detectable viremia ….” - clarify you are referring to the multivariable anaysis

Line 196: This sentence:

The most common reason for 197 missing ART dosing was travelling without sufficient ARV supply and simply forgetting (Fig 2).

This is in the middle of a paragraph about Table 2 results and is jarring to jump here and jump back - perhaps reasons deserves its own paragraph

Table 1

= Clarify in the title or footer that these are row percents — that they are not baseline characteristics stratified by VL status

- since you are showing row percents, it would be helpful to include all levels of categorical variables., e.g, for highest level of education we only see completed secondary but would be nice to see the % non suppressed with less edu

Typo, for example -

Line 91-92

Thus the purpose of this analyses —> should be these or analysis

Reviewer #4: Please address the following major issues that need addressed before the paper could be published.

Title

1) Reading the title can be misleading in that it would make a reader think that you are performing a cohort study. Please reformulate to make the study design of the present analysis clear (which I think is cross-sectional), rather than focusing on the parent study.

Abstract

2) The designation “Africa women” may be misleading. It may be too general given that your sample is comprised of women from Eastern and Southern Africa. Countries such as Egypt and Libya are parts of Africa and if you believe that your findings apply to these countries, please elaborate so. However, please reformulate all instances of the word “Africa” to be more specific to your population of interest.

3) Please include the design of your analysis in the abstract.

4) In line 42, did you mean “to assess”

5) Please include the measure of association that you aimed to estimate in the method section.

6) The conclusion reads “Socioeconomic barriers such as poverty, not being married, young age, and self-reported missed doses remain key predictors of unsuppressed viremia.” I’d suggest restructuring the sentence because it seems that “not being married, young age…” are part of the socioeconomic barriers.

7) The word “remain” in the conclusion can be misleading because the present analysis is not longitudinal in nature. If you want to refer to a previous publication of the same group, please elaborate so.

Introduction

8) Line 66, please update the statistics about the number of people receiving cART, to a more recent date. Your statistics are from 2017.

9) Line 70, “ … 90-90-90 2020 Strategy in ending the epidemic by 2030 (2).” Does not read well. There is a capital letter in the middle of sentence.

10) Line 87. Please correct the word “assoicated”

11) Line 90. Suggest not starting “Tuberculosis” with capital letter.

12) Please provide a relevant reference to the following sentence “Emerging antriretroviral drug resistance has been known to occur from levels of 200 copies/ml or above, and use of this threshold eliminates most cases of apparent viremia caused by viral load blips or assay variability”

The WHO considered VL>1000 as threshold to define viral failure, to reduce pitfalls from viral blips. If you believe that VL<200 eliminates most viral blips, please provide the reference supporting this.

Method

13) Please focus on the description of the design of this present study. You were focusing on describing the parent study PROMOTE.

14) Please indicate what was defined as “high enrolling PROMISE sites”

15) Please briefly describe the intervention in PROMISE study, to relieve any concern that the PROMISE intervention could have an impact on viremia among women in your study sample.

16) By definition, exclusion criteria are characteristics of subjects who have met the inclusion criteria. Therefore, if “willing to provide informed consent to enroll and continue follow-up” is an inclusion criterion, it is would be redundant to write “unwilling to provide informed consent to continue follow-up” as exclusion criterion, because both represent the same thing. If you meant to say that participants who refused to provide consent after being enrolled are excluded, I would suggest making it clearer.

17) Please be specific about criteria used to make the judgment in line 132

18) Please add in the method section a description of covariates: how they were defined and coded. In addition, please indicate the justification for each coding scheme. If your group has used the coding in a previous manuscript, please provide relevant citation.

19) Could you add information about the testing procedure: blood specimen manipulation, storage, type of laboratory, existing quality control.

20) I do not think that the following sentence is relevant to your present objectives “Samples were stored for future HIV drug resistance testing (blood) and cumulative drug levels testing (hair).”

21) Please clearly indicate why did you use the Poisson regression and why you did not use log binomial regression.

22) You mentioned that you used prevalence risk ratio as measure of association. However, in the abstract you mentioned that you used prevalence rate ratio. I would suggest being consistent.

23) You wrote that each predictor was fitted in the model as first step of your modeling(line 159). However, in line 155 it reads “Fisher’s exact, Chi-square test of independence or Wilcoxon rank sum tests were used to test for an association between baseline characteristics and unsuppressed viremia”

Which one was used in the univariable analyses ?

24) “and (ii) multivariable model with all the predictors included in the model” This sentence has no verb

25) It seems as if you are contradicting yourselves by including the following phrases in the method section “…. multivariable model with all the predictors included in the model” AND “Variables included in the multivariable analyses were chosen based on prior research on risk factors, biological plausibility, and previously identified clinical associations”. The ALL in the first sentence may be understood as you included all covariates from the univariable analyses in the multivariable model.

26) The following sentence “… prior research on risk factors, biological plausibility, and previously identified clinical associations”

Using “Prior research” and “previously identified clinical associations” in the same sentence is redundant. In addition, please provide citation for the previously identified clinical associations.

Results

27) Please provide in the supplemental materials the results using VL>1000 as secondary outcome.

28) Please include a flowchart or a relevant table that summarizes the following information “Overall, 1987 mothers were enrolled into the PROMOTE study, of whom 1947 (98%) women reported taking ART at the enrolment visit and HIV-1 viral load results were available for 1934. HIV-1 VL was above the limit of quantification in 293/1934 (15 %). ”

29) Table 1 does not provide with an overall description of the sample, because it is stratified by viral suppression status. Please add a third column containing an overall frequency.

30) The notations p-values are not consistent, sometimes they are written as “<.001 ” . Sometimes, they are written as “<0.001”

31) For each association you describe in the result, please include the comparison group.

32) Please be consistent in the notation of copies/ml. Sometimes you use cp/ml, sometimes you use copies/ml

33) “In addition, the absence of socioeconomic factors such as electricity in the premises” The sentence does not read well. Which premises ?

34) Did you mean “travel time from the cART clinic “ or “travel time to the cART clinic “

35) I do not think that the following sentence is relevant with regards to your study objective. “Additionally, about 10% (n=29) of women with unsuppressed viremia had an HIV-uninfected partner.”

Discussion

36) The following sentence does not read well. “We identify sociodemographic, self-reported non-adherence, and clinical factors that were associated with detectable viremia > 200copies/ml. ”

37) LINE 230, you used “correlation” …. Correlation analyses and regression analyses are very different.

38) Line 292: The longitudinal design of the PROMOTE does not confer any strength to this present analysis, which is a cross-sectional analysis of the baseline data of a longitudinal analysis. However, the sample size is a strength. Please reformulate.

39) Please elaborate more on the ongoing quality assurance that would strengthen this present cross-sectional analysis.

40) There are additional limitations that ought to be discussed: potential recall bias due to self-report, the inability to confirm viral failure.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Reviewer #4: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Comments.docx

Click here for additional data file.

Submitted filename: revision plos one paper.docx

Click here for additional data file.

11 Sep 2019

Comments from Reviewer 1

The authors conducted an analysis to explore factors associated with viremia above 200 among women from multiple sites in Africa. I found the paper to be written well with a clear logical flow and tight focus. However, I have identified minor concerns that the authors should work on.

Comment: Page 11, line 87 …are most strongly associated with 87 unsuppressed…..

Response: Thank you for this comment. Correction made.

Comment: Font size is not consistent, see beginning of results section, and probably other places

Response: Font size has been revised

Comment: Footnote in table2 a2 missing data, b1 missing data, camongst 1549

What do these numbers 2 and 1 supposed to represent?

Response: These represent the number of participants with missing data for a specific variable. We have re-worded the footnote clearly to indicate that.

Comment: Reference 2, 2018 [ Is there something missing, it appears as if you started unclosed bracket?

Response: Bracket removed

Comments from Reviewer 2

Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in a multi-country cohort study: US-PEPFAR PROMOTE Study.

PLOS ONE

Summary: This manuscript describes factors related to viremia in WHLIV in sub-Saharan Africa within the PROMOTE multi-site study. The topic is of interest to clinicians and policy makers focused on caring for both WLHIV and pregnant/postpartum WLHIV. The manuscript is generally well written though does not present much novel data and has several gaps outlined below.

Suggested major revisions:

1. Introduction: Throughout the introduction the authors make assertions that they do not support with sufficient references. This includes Lines 77, 81, 82 and 109.

Response: Thank you for raising this. Additional references have been added.

2. Introduction: description of the PROMISE study “…women randomized… to initiate varied ARV regimens during pregnancy….” Lines 99-103 is not well worded. Presumably women were not randomized in response to the START study though study protocols may have changed after START results were released?

Response: Thank you for this comment. The sentence has been reworded to read better. The PEPFAR-funded PROMOTE study presents a unique opportunity to assess longer term treatment outcomes among women previously randomized to different antiretroviral (ARV) regimens during pregnancy and postpartum for the purpose of preventing perinatal HIV transmission in the PROMISE trial. Subsequently, all women transitioned to lifelong cART in response to the START study which showed clear benefit of universal ART in June 2015

3. Methods: inclusion criteria from PROMISE should be included since that defines the cohort currently being followed.

Response: Thank you for this comment. Inclusion criteria for PROMISE has been added.

4. A bias could be introduced by allowing site teams to “judge social or other reasons which would make it difficult for mother/child pair to comply with study requirements” Is there any actual criteria to define this? How many mother/child pairs were excluded due to site judgement decisions? If this is a substantial number/% (>5%) authors should add to limitations.

Response: Thank you for noting this. The criteria included those women who felt that continued follow up may predispose them to social harm and women who planned on relocating home out of the study catchment area and could not keep up with study visit requirements afterwards. We have added a table to indicate numbers from PROMISE study and the lost to follow up.

5. Results: a flow diagram of PROMISE participants who then rolled over to PROMOTE and reasons for exclusion, missing baseline VL etc should be included

Response: Thank you for this comment. We have included a table to show the numbers from PROMISE to PROMOTE including those that were lost to follow up at PROMOTE study entry. The reasons for non-enrolment were declined participation, some women felt that continued follow up may predispose them to social harm and were excluded, women who had relocated out of the study catchment area and could not keep up with study visit requirements afterwards were also excluded.

6. The pregnancy and postpartum status (e.g. months postpartum) should be reported either at time of VL reported or at least enrollment to PROMOTE to give context to findings.

Response: thank you for raising this. Additional text to indicate that enrolled women were either having a subsequent pregnancy or not pregnant (up to 4 years post-delivery in PROMISE) or breastfeeding babies born between PROMISE study close up and PROMOTE study start has been added. This information is also in Table 2.

7. Figure 1 does not add anything to the text presented.

Response: Figure 1 has been deleted.

8. If available, ART regimen should be included in Table 1. Since PROMISE randomized to different regimens and regimen is a known factor affecting viremia. Also would be interesting to note if women had changed from their PROMISE regimen.

Response: The summary of the regimens is shown in Table 1. Most women were switched from the PROMISE regimen (TDF/FTC/LPV/r) to TDF/3TC/EFV (standard of care regimen).

9. Unclear what “awareness of dosing instructions” means and how this question was asked, given it is a yes/no response.

Response: The question re-awareness of dosing instructions was phrased as “ Do you know if the ARV medications you are taking need to be taken on a schedule, such as “once a day” or “2 times a day” or “every 8 hours”? Yes/No/Don’t know.

10. Discussion: Authors confuse the UNAIDS 90-90-90 targets. The third 90 actually represents 90% of women on ART or 73% of all WHLIV (diagnose 90%, put 90% of diagnosed women on ART had achieve viral suppression in 90% of those on ART).

Response: Thank you for raising this. We have rephrased to state results in line with the UNAIDS 90% target.

11. Condom use result is randomly added to paragraph on reasons for ARV non-adherence

Response: This has been modified and condom use result has been added under the paragraph on vaginal discharge.

12. Separate paragraph on differentiated care models could be combined with section on travel and forgetting pills.

Response: Thank you for this suggestion. We have combined the section on travel with that of forgetting pills.

13. The statement “one would expect less disclosure issues and united communities taking treatment together” seems judgmental and not supported by literature demonstrating ongoing issues with stigma in high prevalence areas.

Response: Thank you for raising this important issue. We have deleted this statement.

14. Authors do not effectively argue that pill burden and low toxicity should reduce barriers to adherence since they do not describe what ART regimens or formulations anywhere in results. Most countries still recommend efavirenz-based regimens for WHLIV which may have substantial side effects for a significant proportion of women.

Response: Thank you for this query. We have provided data on ART regimens. About 88% of the women were on Efavirenz or Nevirapine based regimens.

15. Lines 276-284 seem to jump from result to result without an overlying theme. Why recent vaginal discharge is associated with viremia is not explored at all.

Response: This entire paragraph has been revised.

16. Lack of resistance data is a major limitation as the authors argue that the VL presented are all due to adherence. Additionally, lack of data on depression, IPV, stigma are also limitations.

Response: These limitations have been added to the study limitation section.

17. Authors should also note bias in sample as all participants are from those retained in a long term study (PROMISE) which is not usually representative of the general population. This may have resulted in lower viremia than WLHIV generally.

Response: Thank you for this comment. We have added this information as part of the limitations.

18. While the authors found an association between self-reported adherence in a study setting and VL, this is often not the case in routine programmatic settings where patients have a disincentive to truthfully report adherence. This conclusion should be modified.

Response: Thank you for this comment. The conclusion has been reworded to read “…. the use of self-reported adherence to cART may still play a role in adherence determination in the absence of superior measures.”

19. The author suggest motivational interviewing and differentiated care models will be effective but it is unclear how this will impact SES factors, young age, and marital status.

Response: Thank you for this comment. We have reworded to include the need for economic empowerment in addition to the motivational interviewing under the conclusion.

Minor revisions

20. Abstract: the conclusion suggests not being married and young age are socioeconomic barriers- suggest re-wording.

Response: Conclusion has been reworded.

21. Methods: reword “one of the longest ongoing follow up epidemiologic multinational cohorts”

Response: The methods section has been reworded.

22. Lines 184- 185 repetitive phrasing (Notably, of note)

Response: Thank you for noting this. This has been modified accordingly.

23. Discussion suggests SES is a known predictor of adherence while the data actually seem mixed on this and the references do not support this assertion.

Response: Additional references have been added.

24. Rephrase “steers away” line 240

Response: This has been rephrased.

Comments from Reviewer 3

Comment: This is an important manuscript addressing predictors of viremia in women of childbearing age living with HIV across sub-saharan Africa in an observational cohort study. This is generally very well presented. Below I present minor comments. In addition, a close review for typos/grammatical errors would be worthwhile.

Response: Thank you for this. A close review has been done

Abstract

Comment: “Not being married” is mentioned in conclusion but not results. Consider adding to results (or dropping from conclusion) in abstract so the conclusions are drawn follow the results presented.

Response: Thank you for picking this, it was an oversight. This has been rectified and marital status has been included in the results section. We would like to make a note that even though the p-value for marital status is greater than 0.05 but we felt that the 32% increased risk of detectable viremia for those not married is an important finding.

Comment : Intro line 79…”satisfaction with health worker information availed were conducive”… —> reword for clarity

Response: This sentence has been reworded for more clarity.

Comment: Line 82

…”This may be attributed to less motivation to protect the child post-delivery” —> perhaps clarify that the motivation may decline when the risk is no longer present (currently sounds as though the motivation declines but not clear that the risk has declined)

Response: Thank you for this comment. We have reworded for more clarity.

Methods – Eligibility

Comment: How long was the minimum time on ART prior to the enrollment VL - did you apply any restrictions for inclusion, e.g., at least 3 or 6 months of ART?

Response: The range time on ART to enrolment was 1 to 5 years. We did not apply any restrictions

RESULTS

Comment: Starting Line 198: “The predictors of detectable viremia ….” - clarify you are referring to the multivariable analysis

Response: The sentence has been revised to “Results from the multivariable model are shown in Table 2”

Comment: Line 196: This sentence:

The most common reason for 197 missing ART dosing was travelling without sufficient ARV supply and simply forgetting (Fig 2).

This is in the middle of a paragraph about Table 2 results and is jarring to jump here and jump back - perhaps reasons deserves its own paragraph

Response: A new paragraph has been created as recommended.

Table 1

Comment: = Clarify in the title or footer that these are row percents — that they are not baseline characteristics stratified by VL status

Response: The following sentence has been added in the footer of Table 1. “Please note that we presented row percentages to display the proportion of patients with detectable viremia for each level of the variable.”

Comment: since you are showing row percents, it would be helpful to include all levels of categorical variables., e.g, for highest level of education we only see completed secondary but would be nice to see the % non-suppressed with less education

Response: This has been inserted in Table 1 for other variables as well.

Comment: Typo, for example -Line 91-92

Thus the purpose of this analyses —> should be these or analysis

Response: This has been changed as suggested

Comment from Reviewer 4

Please address the following major issues that need addressed before the paper could be published.

Title

1) Reading the title can be misleading in that it would make a reader think that you are performing a cohort study. Please reformulate to make the study design of the present analysis clear (which I think is cross-sectional), rather than focusing on the parent study.

Response: Thank you for bringing this to our attention. The title has been amended accordingly to “Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in the multi-country US-PEPFAR PROMOTE study: A cross-sectional analysis.”

Abstract

2) The designation “Africa women” may be misleading. It may be too general given that your sample is comprised of women from Eastern and Southern Africa. Countries such as Egypt and Libya are parts of Africa and if you believe that your findings apply to these countries, please elaborate so. However, please reformulate all instances of the word “Africa” to be more specific to your population of interest.

Response: The authors have further clarified Africa to sub-Saharan Africa.

3) Please include the design of your analysis in the abstract.

Response: The study design has been included in the abstract as follows. “This cross-sectional analysis was based on baseline data of the PROMOTE longitudinal cohort study at 8 sites in Uganda, Malawi, Zimbabwe and South Africa”

4) In line 42, did you mean “to assess”.

Response: Yes, however this sentence has been deleted.

5) Please include the measure of association that you aimed to estimate in the method section.

Response: This was already described under the statistical analyses section. We used prevalence risk ratios to measure the strength of an association between baseline characteristics and unsuppressed viremia.

6) The conclusion reads “Socioeconomic barriers such as poverty, not being married, young age, and self-reported missed doses remain key predictors of unsuppressed viremia.” I’d suggest restructuring the sentence because it seems that “not being married, young age…” are part of the socioeconomic barriers.

Response: The sentence has been restructured in lines 60-61 “Socioeconomic barriers such as poverty, and individual barriers like not being married, young age, and self-reported missed doses, are key predictors of unsuppressed viremia.”

7) The word “remain” in the conclusion can be misleading because the present analysis is not longitudinal in nature. If you want to refer to a previous publication of the same group, please elaborate so.

Response: The sentence has been restructured in lines 60-61 as indicated above.

Introduction

8) Line 66, please update the statistics about the number of people receiving cART, to a more recent date. Your statistics are from 2017.

Response: Thank you for highlighting this. Statistics have been updated in line 73-74. “Subsequently with the introduction of the Universal “Test and Treat” strategy, approximately 23.3 million people (including women) had access to combination Antiretroviral Therapy (cART) globally in 2018.”

9) Line 70, “ … 90-90-90 2020 Strategy in ending the epidemic by 2030 (2).” Does not read well. There is a capital letter in the middle of sentence.

Response: This has been corrected in line 76.

10) Line 87. Please correct the word “assoicated”.

Response: Correction has been made in line 92.

11) Line 90. Suggest not starting “Tuberculosis” with capital letter.

Response: Correction has been done in line 95.

12) Please provide a relevant reference to the following sentence “Emerging antriretroviral drug resistance has been known to occur from levels of 200 copies/ml or above, and use of this threshold eliminates most cases of apparent viremia caused by viral load blips or assay variability”

The WHO considered VL>1000 as threshold to define viral failure, to reduce pitfalls from viral blips. If you believe that VL<200 eliminates most viral blips, please provide the reference supporting this.

Response: Thank you for noting this. 2 references (14, 15) have been included.

Method

13) Please focus on the description of the design of this present study. You were focusing on describing the parent study PROMOTE.

Response: The present study design has been clarified in line 120.

14) Please indicate what was defined as “high enrolling PROMISE sites”.

Response: This sentence has been removed since it is not relevant to this analysis. However, high enrolling sites were those sites that had more numbers of enrollments by the time accrual into the PROMISE trial had closed.

15) Please briefly describe the intervention in PROMISE study, to relieve any concern that the PROMISE intervention could have an impact on viremia among women in your study sample.

Response: Briefly described in lines 104-108 and 128-131. It reads ‘this cohort presents a unique opportunity to assess longer term treatment outcomes among women previously randomized to different antiretroviral (ARV) regimens (WHO option A and option B) during pregnancy and postpartum for the purpose of preventing perinatal HIV transmission. Lines 128-131 describes the regimens most women were receiving in each country.

16) By definition, exclusion criteria are characteristics of subjects who have met the inclusion criteria. Therefore, if “willing to provide informed consent to enroll and continue follow-up” is an inclusion criterion, it is would be redundant to write “unwilling to provide informed consent to continue follow-up” as exclusion criterion, because both represent the same thing. If you meant to say that participants who refused to provide consent after being enrolled are excluded, I would suggest making it clearer.

Response: Exclusion criteria have been restructured in lines 138-142.

17) Please be specific about criteria used to make the judgment in line 132.

Response: An example has been provided in lines 138-142.

18) Please add in the method section a description of covariates: how they were defined and coded. In addition, please indicate the justification for each coding scheme. If your group has used the coding in a previous manuscript, please provide relevant citation.

Response: There was no coding scheme used. Guided by the study’s schedule of evaluations, covariates were collected using baseline questionnaires and laboratory procedures at study entry, as indicated in lines 149-156. In Table 2, most variables were taken directly from the questionnaire and laboratory results. The levels of the variables that are combined (e.g. marital status, travel time to the cART clinic, missed ART doses) are provided in Table 1.

19) Could you add information about the testing procedure: blood specimen manipulation, storage, type of laboratory, existing quality control.

Response: In lines 150-153“Assays were done at various Good Clinical Laboratory Practice (GCLP) and Clinical Laboratory Improvement Amendments (CLIA) compliant site laboratories adherent to standards for proper collection, processing, labeling, and transport of specimens” has been added.

20) I do not think that the following sentence is relevant to your present objectives “Samples were stored for future HIV drug resistance testing (blood) and cumulative drug levels testing (hair).”

Response: This sentence has been removed.

21) Please clearly indicate why did you use the Poisson regression and why you did not use log binomial regression.

Response: The log binomial model failed to converge as more variables were added into the multivariable model. Therefore, we opted for Poisson regression with robust variance.

22) You mentioned that you used prevalence risk ratio as measure of association. However, in the abstract you mentioned that you used prevalence rate ratio. I would suggest being consistent.

Response: We have changed rate ratio to risk ratio

23) You wrote that each predictor was fitted in the model as first step of your modeling(line 159). However, in line 155 it reads “Fisher’s exact, Chi-square test of independence or Wilcoxon rank sum tests were used to test for an association between baseline characteristics and unsuppressed viremia”

Which one was used in the univariable analyses?

Response: We used to Fisher’s exact, Chi-square test of independence or Wilcoxon rank sum tests to test for an association between baseline characteristics and unsuppressed viremia. However, to determine the strength of the association and also to identify variables that are predictive of unsuppressed viremia, we used both univariable and multivariable Poisson models.

24) “and (ii) multivariable model with all the predictors included in the model” This sentence has no verb

Response: The sentence has been revised and it now reads: “and (ii) multivariable model with multiple predictors included in the model”

25) It seems as if you are contradicting yourselves by including the following phrases in the method section “…. multivariable model with all the predictors included in the model” AND “Variables included in the multivariable analyses were chosen based on prior research on risk factors, biological plausibility, and previously identified clinical associations”. The ALL in the first sentence may be understood as you included all covariates from the univariable analyses in the multivariable model.

Response: Thank you. The sentences have been fixed.

26) The following sentence “… prior research on risk factors, biological plausibility, and previously identified clinical associations” Using “Prior research” and “previously identified clinical associations” in the same sentence is redundant. In addition, please provide citation for the previously identified clinical associations.

Response: The sentence has been reformulated and citations have been included in the introduction (line 93, 96).

Results

27) Please provide in the supplemental materials the results using VL>1000 as secondary outcome.

Response: There were 166/1934 participants with VL>1000 and results are presented in Supplementary Tables 1 and 2. The following sentence was also inserted under the statistical analyses section. ‘Furthermore, in accordance with WHO threshold for treatment failure, we performed sensitivity analyses and defined unsuppressed viremia as viral load above 1000 copies/ml. “

28) Please include a flowchart or a relevant table that summarizes the following information “Overall, 1987 mothers were enrolled into the PROMOTE study, of whom 1947 (98%) women reported taking ART at the enrolment visit and HIV-1 viral load results were available for 1934. HIV-1 VL was above the limit of quantification in 293/1934 (15 %). ”

Response: A flow chart (figure 1) has been created to summarize the above information.

29) Table 1 does not provide with an overall description of the sample, because it is stratified by viral suppression status. Please add a third column containing an overall frequency.

Response: Table 1 has been amended to show total numbers

30) The notations p-values are not consistent, sometimes they are written as “<.001”. Sometimes, they are written as “<0.001”

Response: Thank you, this has been fixed

31) For each association you describe in the result, please include the comparison group.

Response: The results section has been revised accordingly.

32) Please be consistent in the notation of copies/ml. Sometimes you use cp/ml, sometimes you use copies/ml

Response: Thank you. Corrections have been made.

33) “In addition, the absence of socioeconomic factors such as electricity in the premises” The sentence does not read well. Which premises?

Response: This has been further clarified to household premises.

34) Did you mean “travel time from the cART clinic “ or “travel time to the cART clinic “

Response: It is travel time to the cART clinic. Corrections made.

35) I do not think that the following sentence is relevant with regards to your study objective. “Additionally, about 10% (n=29) of women with unsuppressed viremia had an HIV-uninfected partner.”

Response: This sentence has been removed.

Discussion

36) The following sentence does not read well. “We identify sociodemographic, self-reported non-adherence, and clinical factors that were associated with detectable viremia > 200copies/ml.”

Response: Thank you for noting this. The sentence has been revised to reads “Among the women with detectable viremia, we identified that socio demographic, self-reported non-adherence, and clinical factors were associated with detectable viremia > 200copies/ml.”

37) LINE 230, you used “correlation” …. Correlation analyses and regression analyses are very different.

Response: The sentence has been revised.

38) Line 292: The longitudinal design of the PROMOTE does not confer any strength to this present analysis, which is a cross-sectional analysis of the baseline data of a longitudinal analysis. However, the sample size is a strength. Please reformulate.

Response: Line 317-318 –Longitudinal design removed.

39) Please elaborate more on the ongoing quality assurance that would strengthen this present cross-sectional analysis.

Response: Quality assurance has been elaborated in lines 150-153 and 320-321.

40) There are additional limitations that ought to be discussed: potential recall bias due to self-report, the inability to confirm viral failure.

Response: Additional limitations have been included.

11 Oct 2019

Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in the multi-country US-PEPFAR PROMOTE study: A cross-sectional analysis

PONE-D-19-17608R1

Dear Dr. Atuhaire,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Marcel Yotebieng, M.D., MPH, Ph.D

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: All comments have been addressed

Reviewer #4: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: Yes

Reviewer #4: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

Reviewer #4: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

Reviewer #4: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

Reviewer #4: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: This is an excellent and important manuscript. It is much improved from the original submission and I have no further concerns.

Reviewer #4: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

Reviewer #4: No

17 Oct 2019

PONE-D-19-17608R1

Factors associated with unsuppressed viremia in women living with HIV on lifelong ART in the multi-country US-PEPFAR PROMOTE study: A cross-sectional analysis

Dear Dr. Atuhaire:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Marcel Yotebieng

Academic Editor

PLOS ONE