Skip to main content
SearchLoginLogin or Signup

Review 4: "Antiretroviral Therapy Retention, Adherence, and Clinical Outcomes among Postpartum Women with HIV in Nigeria"

Reviewers point out the relevance of the study, and only make minor comments mainly about missing data.

Published onMay 29, 2024
Review 4: "Antiretroviral Therapy Retention, Adherence, and Clinical Outcomes among Postpartum Women with HIV in Nigeria"
1 of 2
key-enterThis Pub is a Review of
Antiretroviral therapy retention, adherence, and clinical outcomes among postpartum women with HIV in Nigeria
Antiretroviral therapy retention, adherence, and clinical outcomes among postpartum women with HIV in Nigeria

Abstract While research involving pregnant women with HIV has largely focused on the antepartum and intrapartum periods, few studies in Nigeria have examined the clinical outcomes of these women postpartum. This study aimed to evaluate antiretroviral therapy retention, adherence, and viral suppression among postpartum women in Nigeria. This retrospective clinical data analysis included women with a delivery record at the antenatal HIV clinic at Jos University Teaching Hospital between 2013 and 2017. Descriptive statistics quantified proportions retained, adherent (≥95% medication possession ratio), and virally suppressed up to 24 months postpartum. Among 1535 included women, 1497 met the triple antiretroviral therapy eligibility criteria. At 24 months, 1342 (89.6%) women remained in care, 51 (3.4%) reported transferring, and 104 (7.0%) were lost to follow-up. The proportion of patients with ≥95% medication possession ratio decreased from 79.0% to 69.1% over the 24 months. Viral suppression among those with results was 88.7% at 24 months, but <62% of those retained had viral load results at each time point. In multiple logistic regression, predictors of loss to follow-up included having a more recent HIV diagnosis, higher gravidity, fewer antenatal care visits, and a non-hospital delivery. Predictors of viral non-suppression included poorer adherence, unsuppressed/missing baseline viral load, lower baseline CD4+ T-cell count, and higher gravidity. Loss to follow-up rates were lower and antiretroviral therapy adherence rates similar among postpartum women at our study hospital compared with other sub-Saharan countries. Longer follow-up time and inclusion of multiple facilities for a nationally representative sample would be beneficial in future studies.

RR:C19 Evidence Scale rating by reviewer:

  • Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.


Review: The concept behind this paper is important: “to quantify both ART adherence and viral suppression up to 24 months postpartum and identify risk factors for LTFU and unsuppressed viral load in postpartum women with HIV in Nigeria.” I found myself questioning the statistics and overwhelmed by the quantity of statistics. I suggest you have a statistician carefully review the statistics. See the specifics of my concerns in my comments on Table 1 below.

Although very important to identify predictive factors associated with loss to follow-up, the 7% LTFU seems quite low. An 87% viral suppression rate seems high but does not include the many with missed data and is misleading.

Comments on the paper are as follows:

  • Line 32 MTCT In the U.S. perinatal transmission or vertical transmission is preferred over mother to child transmission although MTCT may be clearer to readers.

  • Line 37 While MTCT research has historically centered on ART uptake and viral suppression among pregnant women with HIV through delivery and diagnostic outcomes among their neonates after delivery, less coverage has been given to ART continuation and clinical outcomes among these women postpartum. Good point

  • Line 59 HIV positive: Individual with HIV is preferred first person language

  • Line 64 What does APIN stand for?

  • Line 6-7 This retrospective clinical data analysis included women with a delivery record at the antenatal HIV clinic at Jos University Teaching Hospital between 2013 and 2017. You might want to comment on how regimens during this time frame may be different from after 2017, when ART for all pregnant women for life was implemented (B+). See next comment.

  • Lines 78-87 on ART eligibility: ART for all pregnant women B+ started in Jan 2017

  • Lines 98 MPR—I do not understand this statistic. Please give example of calculation of MPR.

  • Lines 102-103 Patients were excluded from the adherence analysis if they were receiving ART prophylactically. Was this PrEP?

  • Lines 103-105 Patients who became LTFU during the study period were included in the analysis up to the time interval during which their last clinic visit was recorded. Not sure this is statistically correct.

  • Lines 131-141—please ask someone other than me to review the statistical analysis

  • Lines 147-149-- The majority of women were married (66.8%), achieved primary or secondary education (61.7%), held non-income occupations (46.2%), and lacked previous ART experience (88.7%) at the time of ART enrollment in the PEPFAR HIV program. “Majority” is correct for 3 of the 4 demographic items you list. Suggest you make a separate sentence for non-income occupations.

  • Line 156, Table 1—please check with statistician. I don’t think it is correct to omit subjects with missing data from your denominator (concept of intent to treat). For example:

    • Education Statusa

    • Missing Data 42 .

    • No Formal 130 8.71

    • Primary/Secondary 921 61.69

    • Tertiary 442 29.6

    • This way of calculating may skew the interpretation of your results. If those missing data all had no formal schooling, that would change your percentages. I think you need to include the 42 with missing data in the total (100%) and reduce the percentages in the other 3 categories since no schooling may be a larger factor than you assessed. Same for all of the statistics in the table.

    • It looks like you did include all subjects, including missing data subjects, in the following:

    • Viral Load at Delivery

    • Missing Data 592 38.57

    • Suppressed (<1000 copies/mL) 797 51.92

    • Unsuppressed (≥1000 copies/mL) 146 9.51

  • Lines 187-189  Mean viral loads were 26,993 viral copies/mL (95% CI 12,551–188 41,435) at delivery, 9206 viral copies/mL (95% CI 2107-16,305) at 12 months, and 5796 viral189 copies/mL (95% CI 3256-8336) at 24 months.

    • I don’t think means are the appropriate statistic to use for this.

  • Lines 274-276 Studies in Malawi using prescription pick-up data and South Africa and Zambia using self-reported adherence found 67%, 63.9%, and 70.5% of postpartum women with optimal adherence, respectively: over what period of time?

  • Lines 331-317 Our study found 85.8% viral suppression at month 12 and 88.7% viral suppression at month 24 postpartum among those with viral load results. These numbers fall short of the 95% UNAIDS target for viral suppression. In comparison, South African studies found 14.7%–14.8% postpartum viral non-suppression [34,35]. Importantly, in our study, of the 1497 ART-eligible, only 50.0% had viral load results at month 12, and 57.3% had viral load results at month 24.

  • The low loss to follow-up (7%) needs more comment. What was special about this program? Saying that viral suppression was 88.7% at 24 months is misleading when only 57.3% had viral loads at 24 months.

tony montana:

The random number generator, a cornerstone in computational systems, generates unpredictable numbers pivotal for a myriad of applications, ensuring fairness and accuracy in simulations and processes. sayı seç