Skip to main content

Patterns of engagement in HIV care during pregnancy and breastfeeding: findings from a cohort study in North-Eastern South Africa

Abstract

Background

Eliminating mother-to-child transmission of HIV (MTCT) in sub-Saharan Africa is hindered by limited understanding of HIV-testing and HIV-care engagement among pregnant and breastfeeding women.

Methods

We investigated HIV-testing and HIV-care engagement during pregnancy and breastfeeding from 2014 to 2018 in the Agincourt Health and Demographic Surveillance System (HDSS). We linked HIV patient clinic records to HDSS pregnancy data. We modelled time to a first recorded HIV-diagnosis following conception, and time to antiretroviral therapy (ART) initiation following diagnosis using Kaplan-Meier methods.

We performed sequence and cluster analyses for all pregnancies linked to HIV-related clinic data to categorise MTCT risk period engagement patterns and identified factors associated with different engagement patterns using logistic regression.

We determined factors associated with ART resumption for women who were lost to follow-up (LTFU) using Cox regression.

Results

Since 2014, 15% of 10,735 pregnancies were recorded as occurring to previously (51%) or newly (49%) HIV-diagnosed women. New diagnoses increased until 2016 and then declined. We identified four MTCT risk period engagement patterns (i) early ART/stable care (51.9%), (ii) early ART/unstable care (34.1%), (iii) late ART initiators (7.6%), and (iv) postnatal seroconversion/early, stable ART (6.4%). Year of delivery, mother’s age, marital status, and baseline CD4 were associated with these patterns. A new pregnancy increased the likelihood of treatment resumption following LTFU.

Conclusion

Almost half of all pregnant women did not have optimal ART coverage during the MTCT risk period. Programmes need to focus on improving retention, and leveraging new pregnancies to re-engage HIV-positive women on ART.

Peer Review reports

Background

In sub-Saharan Africa (SSA) mother-to-child transmission (MTCT) rates ranged from 1.9% in Botswana to 31.9% in Somalia in 2019 [1]. The risk of MTCT starts at conception and continues until breastfeeding cessation, which ranges from 18 to 24 months in SSA countries [2,3,4]. Without intervention, MTCT risk is estimated at 5–10% during pregnancy, 10–15% during delivery, and 5–20% during breastfeeding, with the cumulative risk being 30–45% for infants who are breastfed for 18–24 months [5]. The adoption of Option B+ by most treatment programmes in SSA, including South Africa in 2015, has led to large reductions in MTCT rates [6, 7]. Option B+ has dramatically increased the number of pregnant women initiating ART, however retention is challenging, particularly in the postnatal period [8]. Reasons for disengagement include still feeling healthy, transition from PMTCT to adult ART services, stigma, and denial of HIV status [9,10,11,12]. In 2019, the United Nations Children’s Emergency Fund (UNICEF) released their roadmap for the elimination of mother-to-child transmission of HIV (EMTCT) [13], which identified missed opportunities for EMTCT including failure to prevent seroconversion during pregnancy or breastfeeding, mothers living with HIV (MLHIV) who do not receive antiretroviral therapy (ART), MLHIV who previously started ART and stopped, and MLHIV who started ART late in their pregnancy.

An increase in the proportion of people initiating ART shortly after being diagnosed with HIV, as per World Health Organisation (WHO) guidance [14], has in turn led to an increased spectrum of engagement in care patterns, with treatment interruption, reengagement in care and clinic transfers becoming increasingly prevalent [15,16,17,18]. Previous studies have used cohort and cross-sectional data to focus on the occurrence of specific events along the care cascade including linkage to care, ART initiation, and ART discontinuation [7, 19, 20]. However, these snapshots do not always capture the complexity of individual care trajectories. Monitoring these trajectories and understanding the factors that underlie patterns of disengagement and reengagement in care is becoming increasingly important to inform interventions to accelerate progress towards the second and third targets of the UNAIDS 95–95-95 goals and to reduce MTCT [21].

To inform interventions to address missed opportunities for EMTCT, we investigated HIV testing and engagement in care for pregnant and breastfeeding women in north-eastern South Africa. Using linked demographic surveillance and health facility data, we report on HIV testing events, identify periods of ART engagement and discontinuation during the pregnancy and postpartum periods for women living with HIV (WLHIV), and investigate factors associated with ART resumption during the MTCT risk period.

Methods

Setting and data sources

The Agincourt Health and Demographic Surveillance System (HDSS) is approximately 525 km2 with an estimated population of 120,000 residents living in 20,000 households within 31 villages [22]. The HDSS is located in Mpumalanga province, South Africa with HIV prevalence estimated at 14.1% across all ages [22,23,24]. In 2013, the total fertility rate was estimated as 2.4 children per woman [25]. Data sources and data used are summarised in Additional file 1.

Demographic surveillance

Fertility, mortality and migration data are collected annually from residents, based on a comprehensive household registration system, in operation since 1992 [22, 26]. Fieldworkers visit each household and interview the most knowledgeable adult available to obtain information on demographic events occurring since the last census [22, 27, 28]. The HDSS also collects verbal autopsy data to ascertain probable causes of death [29, 30]. Wealth quintiles were calculated using a list of household asset indicators collected as part of demographic surveillance. These calculations have been described in more detail elsewhere [30].

Point-of-contact interactive record linkage (PIRL)

Since 2014, HIV patient visits to ART health facilities in the area have been logged by fieldworkers and linked to HDSS resident records using Point-of-Contact Interactive Record Linkage (PIRL), described in detail elsewhere [31, 32]. In brief, a fieldworker conducts a short uptake interview with patients in the health facility waiting area. Consenting patients are asked for personal identifiers used to search the HDSS database using a probabilistic algorithm. Matches are confirmed in interaction with the patients, and names of household members are used as a key attribute to adjudicate between possible matches. Linkage data extraction for these analyses occurred on July 11, 2018.

Record review and patient tracing study

Lost to follow-up (LTFU) data are drawn from a study described in detail elsewhere [33, 34]. Briefly, on August 15, 2017, all patients within the PIRL database who were considered LTFU (more than 90 days late for a scheduled health facility visit) were recruited into a cohort and followed up to ascertain whether they were still alive and on treatment.

Data preparation and statistical analysis

All pregnancies occurring from 2014 onwards, recorded in the demographic surveillance database and ending in a live birth were eligible for this analysis. Women could contribute more than one pregnancy. Pregnancy data extraction for these analyses occurred on January 31, 2020, with the last recorded delivery occurring on November 11, 2018.

The Agincourt HDSS does not routinely conduct HIV serological surveys among its population. However, the availability of PIRL data presented a unique opportunity to update prevalence estimates [35], with linkage to an HIV-treatment health facility record presumed to indicate an HIV diagnosis. We identified HIV-treatment health facility records linked to each pregnancy through the PIRL database to determine dates of HIV diagnosis, ART initiation, and all HIV-associated health facility visits. Where no ART initiation date was reported, the first ART-related health facility visit date was used as the ART initiation date. Where the HIV diagnosis date was missing, it was assumed to have occurred the day before ART initiation (the most common occurrence for individuals with both HIV diagnosis and ART initiation dates available in our data). Conception for each pregnancy was estimated to have occurred 280 days prior to the delivery date. Demographic and health surveys data suggests that 10% of babies were still breastfed at 23 months in South Africa in 2016 [36], therefore, breastfeeding cessation was set at 24 months (730 days) following the delivery date to give the longest possible period of transmission risk.

Seroconversions and ART initiation

Any pregnancy for which an HIV diagnosis date occurred before the conception date was excluded from seroconversion analyses. Time to a first recorded HIV-positive result was calculated with the estimated conception date as the origin. We also calculated time to ART initiation with the date of the first recorded HIV-positive result as the origin. These time-to-event data were modelled using Kaplan-Meier methods.

Sequence states and outcomes

All pregnancies for which a first HIV-positive result was recorded before breastfeeding cessation were considered for this analysis. Each pregnancy was assigned a 1010-day sequence corresponding to 280 days of pregnancy and 730 days of breastfeeding. There were nine possible sequence states (Table 1).

Table 1 Definitions of all the possible sequence states

Using a state sequence-analysis approach, and utilising Optimal Matching distances, we identified distinct clusters of sequences with similar patterns of engagement in care during the MTCT risk period. Using logistic regression, we determined factors associated with belonging to a given cluster, after multiple imputation to account for missingness in the explanatory variables.

Factors associated with ART resumption

Using the record review and tracing study data, we restricted these analyses to women of reproductive age who were considered LTFU. Resumption of ART was defined as either reengagement or transfer, following loss to follow-up. Follow-up time began on the date of each patient’s last recorded clinic visit. Follow-up time was split into monthly intervals to account for pregnancy which was included as a time-varying covariate. We linked pregnancy data from the HDSS to identify new pregnancies, defined as a pregnancy occurring after LTFU that was not associated with previous HIV-treatment episodes. Women were considered pregnant if a monthly interval fell within the period of a new pregnancy identified through HDSS data. All other covariates remained constant in each time interval. A Cox regression model was used to determine the factors associated with ART resumption, with all other outcomes considered to be right-censored. Bi-variate analyses were conducted with a priori selected variables for which there was a plausible association to ART resumption. All variables with p < 0.1 were included in the multivariable Cox regression model. A parsimonious model was achieved using Wald tests. All models accounted for clustering at the clinic level and utilised robust standard errors.

Sequence analyses were conducted using the TraMineR R-package [37]. All other analyses were conducted using Stata 16 [38].

Results

Population characteristics

Between January 2014 and November 2018, 10,102 women had 10,831 pregnancies recorded in the HDSS, among which 96 (0.9%) ended in a still birth or abortion. The remaining 10,735 live births were recorded for 10,035 women (Table 2). Median age at delivery was 26 years (IQR: 21, 32).

Table 2 Sociodemographic characteristics of all pregnancies recorded in the Agincourt HDSS since 2014

Seroconversions and ART initiation

Of the 10,735 pregnancies ending in live births, 1628 (15.2%) were linked to an HIV-related health facility record. Of these, 823 (50.6%) had a first HIV-positive test recorded before the estimated conception date and were excluded from Kaplan-Meier (KM) models, 585 (35.9%) had a first test recorded during the pregnancy, 127 (7.8%) had a first test recorded during the breastfeeding period, and 93 (5.7%) had a first test occurring after breastfeeding cessation.

For women contributing more than one pregnancy to the analysis, previous pregnancies contributed analysis time up to the estimated conception date of the next pregnancy. The cumulative probability of having a positive HIV test by the end of the most recent pregnancy was 5.95, and 7.47% by the end of breastfeeding. The probability of having an HIV positive test varied by year of delivery (with a 1.5% reduction between 2016 and 2017 (p = 0.0201)) (Fig. 1), was higher for older women peaking in the 30–34 age group, for women in lower wealth quintiles, for divorced or widowed women, and for less educated women (all p < 0.001).

Fig. 1
figure 1

Kaplan-Meier curves showing the probability of an HIV-positive test during the vertical transmission risk period stratified by the year of delivery. This figure shows the cumulatative probability of having an HIV positive test linked to a pregnant or postpartum woman given there was no previous record or evidence of HIV diagnosis or clinic attendance for HIV care. This cumulative probability is stratified by delivery year showing a reduction in the cumulative probability in 2017 and 2018

The cumulative probability of ART initiation following a recorded HIV-positive test during pregnancy was 50.59% after 43 days, 75.04% after 202 days, 90.15% after 547 days, and 100% after 1687 days. This probability increased by year of delivery with a significant reduction in initiation time from 2016 (< 0.001) (Additional file 2).

Engagement sequences and clusters

We identified distinct patterns of clinic attendance connected to the pregnancy and breastfeeding period (Additional file 3), with 1477 women contributing 1628 pregnancies which were treated as independent for the purpose of this analysis. Of 1628 pregnant women, 93 were excluded from these analyses because the first recorded HIV positive test occurred after the breastfeeding period.

Of the remaining 1535 pregnant women, 710 (46.3%) had an unknown HIV status, 258 (16.8%) were HIV-positive but not on ART, 344 (22.4%) were on ART, 58 (3.8%) were late for their most recent appointment, and 165 (10.7%) were LTFU from HIV care at the beginning of their pregnancy. There was a yearly increase in the proportion of women already on ART at the beginning of their pregnancy starting at 0% in 2014 and increasing to 39.25% by 2018 (p < 0.001). Of 258 WLHIV but not on ART when they became pregnant, 133 (51.5%) had not initiated ART at delivery. Of 165 WLHIV who were LTFU when they became pregnant, 109 (66.1%) were still LTFU at delivery, however, 148 (89.7%) resumed ART by the end of the breastfeeding period (Fig. 2).

Fig. 2
figure 2

Chronograms of engagement in care during the vertical risk transmission period by status at the beginning of the pregnancy

The highest resumption happened in the 6 months leading up to delivery (33 (20.0%) second trimester, 28 (17%) third trimester), was its lowest in the 3 months after delivery (17 (10.3%) pregnant women) and increased during the breastfeeding period before levelling off (Additional file 4).

Cluster analysis identified four distinct patterns of engagement in care in the MTCT risk period: (i) women who initiated ART early and were stable in care (n = 796, 51.9%), (ii) women who initiated ART early but were not stable in care (characterised by frequent ART stoppages or late appointment attendance) (n = 524, 34.1%), (iii) women who were late ART initiators (n = 117, 7.6%), and (iv) postnatal seroconversion with early, stable ART (n = 98, 6.4%). In the logistic regression model, older women and women who delivered more recently were more likely to be in the early stable ART group. Younger women were more likely to seroconvert postnatally. Women who delivered in 2015 and 2016 were least likely to be in the late ART group or to seroconvert postnatally. Married women and women with a baseline CD4 > 500 cells/ μL were more likely to initiate ART late. (Additional file 5).

Factors associated with ART resumption

Of 767 (75.4%) women recorded as LTFU in the PIRL database, 62 (8.1%) were older than reproductive age (> 51 years) and were excluded from this analysis. Of the remaining 705, 279 (39.6%) initiated ART for PMTCT, 354 (50.2%) had been LTFU for less than a year, 468 (66.4%) had a baseline CD4 > 200 cells/ μL, and 84 (11.9%) had a pregnancy occurring during the follow-up time. The median age was 31 years (IQR: 27, 36) (Table 3).

Table 3 Factors associated with resumption of ART following loss to follow-up

Of these 705 women, 247 (35.0%) were found to have resumed ART (147 (59.5%) reengaged and 100 (40.5%) transferred). Women were out of care a median of 312 days (IQR: 182, 541) before resuming ART (Reengaged: 333 (170, 525), Transferred: 297.5 (185, 563), p = 0.6482). In the multivariable Cox model, a new pregnancy was associated with ART resumption (aHR: 2.79, 95% C.I: 1.31, 5.95) when compared to women who did not have a pregnancy occurring during the follow-up time. Longer time on ART before LTFU, higher baseline CD4, initiating ART after 2015, and being on a longer refill schedule (a proxy for stable patients) were associated with a higher hazard of ART resumption following LTFU. The hazard of ART resumption was lower the longer a patient had been LTFU (Table 3).

Discussion

Using linked demographic surveillance and health facility data, we investigated patterns of HIV testing and engagement in care for pregnant and breastfeeding women. Our findings showed a reduction in the number of women testing positive each year during the MTCT risk period. We also identified four patterns of engagement in care during the MTCT risk period representing varying times of seroconversion and different levels of care stability following ART initiation, with women’s age, marital status, baseline CD4 and the year of delivery predicting these patterns. We also found that following discontinuation of HIV care, ART resumption was more likely for women who had a new pregnancy.

An estimated 15% of pregnancies resulting in a live birth recorded in the Agincourt HDSS were to women who had utilised HIV-related healthcare. Of the pregnancies linked to HIV-related healthcare utilisation, 51% were to previously diagnosed women and 49% had no previous HIV-related healthcare data and were presumed to be new diagnoses. Although some of the women who were previously diagnosed had disengaged from care, encouragingly most resumed treatment over the study period.

Reengagement in care is becoming a more common phenomenon as more healthy patients initiate ART [15]. Our finding that a new pregnancy was associated with an increased probability of resumption was probably due to increased interaction with the health system, as well as a desire to protect their baby. Whilst about half of these women resumed ART before delivery, research shows that ART initiation early in the pregnancy is most effective at reducing MTCT risk [14, 39]. Furthermore, whilst ART resumption is a positive outcome, it represents a programmatic failure of Option B+ which aims to keep women on ART for life and extend protection throughout subsequent pregnancies. Treatment interruptions may also contribute to drug resistance [40, 41]. We found that 66.1% of WLHIV who were LTFU when they became pregnant and 51.5% of WLHIV but not on ART when they became pregnant were still not on treatment by delivery, indicating that ART programmes still have much work to do to initiate and keep less compliant patients on ART.

The cumulative probability of a new HIV diagnosis occurring during the MTCT risk period was estimated as 5.9% during pregnancy, increasing to 7.5% by the breastfeeding cessation. The proportion of women testing positive for HIV during pregnancy increased until 2016, reaching 7%, and then declined in 2017 to 5%. These are likely to be underestimates as some women might have attended private facilities or facilities outside the HDSS. Additionally, while UNAIDS estimated that 90.1% of pregnant women in South Africa were tested for HIV in 2019 [1], without a comprehensive HIV testing database where both negative and positive results are recorded, it is likely that not all seroconversions were captured. Nonetheless, the reduction in the number of women testing HIV positive in 2017 is encouraging. Option B+ and treat all were implemented as part of the paradigm shift towards using treatment as prevention. South Africa national guidelines moved to Option B+ in 2015 and to universal treatment in 2016 [6, 42], all of which have been widely adopted in the area covered by the Agincourt HDSS. Other countries have reported a surge in the volume of pregnant women testing positive and utilising PMTCT services as more women became eligible for ART following the implementation of Option B+ [43]. This was followed by a steady decline as more WLHIV became aware of their HIV status [44]. Our data suggests, there was a further reduction in 2018, however, further research and monitoring of this trend will be needed.

Over half of HIV-exposed pregnancies occurred to women who already knew their HIV status. There is a general recognition that the number of HIV-positive women having children who already know their HIV-status is increasing [44, 45], illustrating the importance of keeping women engaged in care and virally suppressed. Our finding that older age was associated with sustained retention on ART aligns with other studies [46,47,48,49]. Later year of delivery was also found to be associated with early stable ART with the odds of late ART decreasing over time potentially showing the benefits of simplified treatment guidelines under Option B+. We found that married women were more likely to initiate ART late which other studies have reported citing the need for their partner’s support [50,51,52]. As such, interventions that improve partner support such as couple testing or joint counselling are important to encourage early ART initiation for married women [10, 53, 54].

Teenage mothers were more likely to have a first positive test recorded postnatally which might suggest late seroconversion, late testing, or poorer health seeking behaviour, all of which have been reported as important factors among this age group in similar settings [55,56,57]. Older women were more likely to have already seroconverted or to be on treatment at the beginning of their pregnancy which possibly reflects their longer time at risk for HIV. There were lower odds of postnatal seroconversion in later years, possibly indicating reduced incidence of HIV in the wider population.

This study has some limitations. Firstly, the health facilities used may not capture every sentinel event as a small proportion of HDSS residents receive ART through private facilities or public facilities outside the HDSS. Furthermore, we did not have treatment adherence data. The assumption that a patient was in care until their next scheduled visit might therefore misclassify some patients. During the linkage process a very small proportion of residents declined to be linked and we did not manage to link everyone that claimed to be an HDSS resident. These linkage errors might bias these results downwards. Finally, there are biases arising from using routinely collected data. If women testing positive did not self-report a previous diagnosis, we would have erroneously categorised them as newly diagnosed. Additionally, if a woman tested positive but did not report to the PMTCT clinic to start treatment then she would have been missed as this data would only be available in the testing registers which were not linked to HDSS data. Finally, data from 2018 does not include clinic data past July 2018 or pregnancy data past November 2018, as such HIV incidence reduction seen in our analysis in 2018 might be an effect of the data.

This study also has some strengths. To our knowledge, this is the first study to use linkage of population and clinical records to estimate HIV incidence in pregnant women in SSA. The method of linking routinely collected data to answer this question shows promise and should be considered in future research. The use of multiple data sources also means we were able to crosscheck matching data and had access to new pregnancies data and socioeconomic variables which are not always available in clinical data.

Conclusions

In conclusion, our study shows the importance of new pregnancies in encouraging treatment resumption for women previously LTFU. This coupled with the reduction in the number and proportion of pregnant and postpartum women testing HIV positive should be considered as successes of the Option B+ programme. Our study also shows the growing scale of reengagement in care. Treatment programmes will need interventions to keep pregnant women on ART after delivery and ensure that pregnant women previously in care are reengaged promptly for their own health and to protect their unborn infants.

Availability of data and materials

The data that support the findings of this study are available from MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt), but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available by request here: https://www.agincourt.co.za/?page_id=1883 and with permission of the MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt).

References

  1. AIDSinfo | UNAIDS [Internet]. [cited 2018 Oct 11]. Available from: http://aidsinfo.unaids.org/

  2. Infant and young child feeding [Internet]. UNICEF DATA. [cited 2020 Dec 7]. Available from: https://data.unicef.org/topic/nutrition/infant-and-young-child-feeding/

  3. Dop M-C. Breastfeeding in Africa: will positive trends be challenged by the AIDS epidemic? Sante Montrouge Fr. 2002;12(1):64–72.

    Google Scholar 

  4. World Health Organization, UNICEF. Guideline Updates on HIV and infant feeding: The duration of breastfeeding, and support from health services to improve feeding practices among mothers living with HIV. [Internet]. 2016 [cited 2020 Jan 20]. Available from: http://www.ncbi.nlm.nih.gov/books/NBK379872/

    Google Scholar 

  5. De Cock KM, Fowler MG, Mercier E, de Vincenzi I, Saba J, Hoff E, et al. Prevention of mother-to-child HIV transmission in resource-poor countries: translating research into policy and practice. JAMA. 2000;283(9):1175–82. https://doi.org/10.1001/jama.283.9.1175.

    Article  PubMed  Google Scholar 

  6. South Africa: National Department of Health. National Consolidated Guidelines for the prevention of mother-to-child transmission of HIV (PMTCT) and the Management of HIV in Children, Adolescents and Adults [Internet]. 2015 [cited 2017 Mar 30]. Available from: https://www.health-e.org.za/2015/07/02/guidelines-national-consolidated-guidelines-for-pmtct-and-the-management-of-hiv-in-children-adolescents-and-adults/

    Google Scholar 

  7. Etoori D, Kerschberger B, Staderini N, Ndlangamandla M, Nhlabatsi B, Jobanputra K, et al. Challenges and successes in the implementation of option B+ to prevent mother-to-child transmission of HIV in southern Swaziland. BMC Public Health. 2018;18(1):374. https://doi.org/10.1186/s12889-018-5258-3.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Knettel BA, Cichowitz C, Ngocho JS, Knippler ET, Chumba LN, Mmbaga BT, et al. Retention in HIV Care During Pregnancy and the Postpartum Period in the Option B+ Era: A Systematic Review and Meta-Analysis of Studies in Africa. JAIDS J Acquir Immune Defic Syndr [Internet]. 2017; [cited 2018 Jan 7];Publish Ahead of Print. Available from: http://journals.lww.com/jaids/Abstract/publishahead/Retention_in_HIV_Care_During_Pregnancy_and_the.96776.aspx.

  9. McLean E, Renju J, Wamoyi J, Bukenya D, Ddaaki W, Church K, et al. “I wanted to safeguard the baby”: a qualitative study to understand the experiences of Option B+ for pregnant women and the potential implications for “test-and-treat” in four sub-Saharan African settings. Sex Transm Infect. 2017;93(Suppl 3).

  10. Cataldo F, Chiwaula L, Nkhata M, van Lettow M, Kasende F, Rosenberg NE, et al. Exploring the Experiences of Women and Health Care Workers in the Context of PMTCT Option B Plus in Malawi. J Acquir Immune Defic Syndr 1999. 2017;74(5):517–22.

    Article  Google Scholar 

  11. Katirayi L, Namadingo H, Phiri M, Bobrow EA, Ahimbisibwe A, Berhan AY, et al. HIV-positive pregnant and postpartum women’s perspectives about Option B+ in Malawi: a qualitative study. J Int AIDS Soc [Internet]. 2016;19(1) [cited 2017 Sep 16]. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4911420/.

  12. Clouse K, Schwartz S, Van Rie A, Bassett J, Yende N, Pettifor A. “What they wanted was to give birth; nothing else”: barriers to retention in option B+ HIV care among postpartum women in South Africa. J Acquir Immune Defic Syndr 1999. 2014;67(1):e12–8.

    Article  Google Scholar 

  13. UNICEF. Going the “last mile” to EMTCT: A road map for ending the HIV epidemic in children [Internet]. 2019 [cited 2020 Dec 7]. Available from: http://www.childrenandaids.org/sites/default/files/2020-06/1-EMTCT%20Whitepaper%20EN%20WEB.pdf

    Google Scholar 

  14. WHO. Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection Recommendations for a public health approach [Internet]. 2016. Available from: http://www.who.int/hiv/pub/arv/arv-2016/en/

    Google Scholar 

  15. Baldé A, Lièvre L, Maiga AI, Diallo F, Maiga IA, Costagliola D, et al. Re-engagement in care of people living with HIV lost to follow-up after initiation of antiretroviral therapy in Mali: who returns to care? PLoS One. 2020;15(9):e0238687. https://doi.org/10.1371/journal.pone.0238687.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Rotheram-Borus MJ, Tomlinson M, Scheffler A, Le Roux IM. Re-engagement in HIV care among mothers living with HIV in South Africa over 36 months post-birth. AIDS Lond Engl. 2015;29(17):2361–2. https://doi.org/10.1097/QAD.0000000000000837.

    Article  Google Scholar 

  17. Clouse K, Vermund SH, Maskew M, Lurie MN, MacLeod W, Malete G, et al. Mobility and Clinic Switching Among Postpartum Women Considered Lost to HIV Care in South Africa. J Acquir Immune Defic Syndr 1999. 2017;74(4):383–9.

    Article  Google Scholar 

  18. Phillips TK, Clouse K, Zerbe A, Orrell C, Abrams EJ, Myer L. Linkage to care, mobility and retention of HIV-positive postpartum women in antiretroviral therapy services in South Africa. J Int AIDS Soc [Internet]. 2018;21(Suppl Suppl 4) [cited 2020 Jan 24]. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6053482/.

  19. Gourlay A, Birdthistle I, Mburu G, Iorpenda K, Wringe A. Barriers and facilitating factors to the uptake of antiretroviral drugs for prevention of mother-to-child transmission of HIV in sub-Saharan Africa: a systematic review. J Int AIDS Soc. 2013;16(1):18588. https://doi.org/10.7448/IAS.16.1.18588.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Shamu S, Slabbert J, Guloba G, Blom D, Khupakonke S, Masihleho N, et al. Linkage to care of HIV positive clients in a community based HIV counselling and testing programme: a success story of non-governmental organisations in a south African district. PLoS One. 2019;14(1):e0210826. https://doi.org/10.1371/journal.pone.0210826.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Joint United Nations Programme on HIV/AIDS. Fast-Track: ending the AIDS epidemic by 2030. [Internet]. 2014 [cited 2019 Mar 14]. Available from: http://www.unaids.org/sites/default/files/media_asset/JC2686_WAD2014report_en.pdf

    Google Scholar 

  22. Kahn K, Collinson MA, Gómez-Olivé FX, Mokoena O, Twine R, Mee P, et al. Profile: Agincourt health and socio-demographic surveillance system. Int J Epidemiol. 2012;41(4):988–1001. https://doi.org/10.1093/ije/dys115.

    Article  PubMed  PubMed Central  Google Scholar 

  23. HIV statistics for South Africa | Prevalence, incidence, ARVs, deaths [Internet]. TB Facts | TB, tests, drugs, statistics. [cited 2019 Feb 23]. Available from: https://www.tbfacts.org/hiv-statistics-south-africa/

  24. Kabudula CW, Houle B, Collinson MA, Kahn K, Tollman S, Clark S. Assessing changes in household socioeconomic status in rural South Africa, 2001–2013: a distributional analysis using household asset indicators. Soc Indic Res. 2016;28(3):1–27. https://doi.org/10.1007/s11205-016-1397-z.

    Article  Google Scholar 

  25. Houle B, Pantazis A, Kabudula C, Tollman S, Clark SJ. Social patterns and differentials in the fertility transition in the context of HIV/AIDS: evidence from population surveillance, rural South Africa, 1993–2013. Popul Health Metr [Internet]. 2016;14 [cited 2020 Oct 28]. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4807549/.

  26. Garenne M, Kahn K, Collinson M, Gómez-Olivé X, Tollman S. Protective effect of pregnancy in rural South Africa: questioning the concept of “indirect cause” of maternal death. PLoS One. 2013;8(5):e64414. https://doi.org/10.1371/journal.pone.0064414.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Kahn K, Tollman S, Collinson MA, Clark S, Twine R, Clark B, et al. Research into health, population and social transitions in rural South Africa: data and methods of the Agincourt health and demographic surveillance system. Scand J Public Health Suppl. 2007;69:8–20. https://doi.org/10.1080/14034950701505031.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Kabudula CW, Houle B, Collinson MA, Kahn K, Gómez-Olivé FX, Clark SJ, et al. Progression of the epidemiological transition in a rural south African setting: findings from population surveillance in Agincourt, 1993–2013. BMC Public Health. 2017;17(1):424. https://doi.org/10.1186/s12889-017-4312-x.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Byass P, Hussain-Alkhateeb L, D’Ambruoso L, Clark S, Davies J, Fottrell E, et al. An integrated approach to processing WHO-2016 verbal autopsy data: the InterVA-5 model. BMC Med. 2019;17(1):102.

    Article  Google Scholar 

  30. Kabudula CW, Houle B, Collinson MA, Kahn K, Gómez-Olivé FX, Tollman S, et al. Socioeconomic differences in mortality in the antiretroviral therapy era in Agincourt, rural South Africa, 2001-13: a population surveillance analysis. Lancet Glob Health. 2017;5(9):e924–35. https://doi.org/10.1016/S2214-109X(17)30297-8.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Rentsch CT, Kabudula CW, Catlett J, et al. Point-of-contact Interactive Record Linkage (PIRL): A software tool to prospectively link demographic surveillance and health facility data. Gates Open Res [Internet]. 2017;1(8) Available from: https://gatesopenresearch.org/articles/1-8/v1.

  32. Kabudula CW, Clark BD, Gómez-Olivé FX, Tollman S, Menken J, Reniers G. The promise of record linkage for assessing the uptake of health services in resource constrained settings: a pilot study from South Africa. BMC Med Res Methodol. 2014;14(1):71. https://doi.org/10.1186/1471-2288-14-71.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Etoori D, Gomez-Olive FX, Reniers G, Rice B, Renju J, Kabudula CW, et al. Outcomes after being lost to follow-up differ for pregnant and postpartum women when compared to the general HIV treatment population in rural South Africa. JAIDS J Acquir Immune Defic Syndr [Internet]. 2020; [cited 2020 Jul 31];Publish Ahead of Print. Available from: https://journals.lww.com/jaids/Abstract/9000/Outcomes_after_being_lost_to_follow_up_differ_for.96160.aspx.

  34. Etoori D, Wringe A, Kabudula CW, Renju J, Rice B, Gomez-Olive FX, et al. Misreporting of patient outcomes in the South African national HIV treatment database: consequences for programme planning, monitoring, and evaluation. Front Public Health [Internet]. 2020;8 [cited 2020 Mar 24]. Available from: https://www.frontiersin.org/articles/10.3389/fpubh.2020.00100/full#supplementary-material.

  35. Gómez-Olivé FX, Angotti N, Houle B, Klipstein-Grobusch K, Kabudula C, Menken J, et al. Prevalence of HIV among those 15 and older in rural South Africa. AIDS Care. 2013;25(9):1122–8. https://doi.org/10.1080/09540121.2012.750710.

    Article  PubMed  PubMed Central  Google Scholar 

  36. South Africa: National Department of Health SSA. South Africa Demographic and Health Survey 2016 [Internet]. 2016 [cited 2021 Mar 14]. Available from: https://dhsprogram.com/pubs/pdf/FR337/FR337.pdf#page=215

    Google Scholar 

  37. Gabadinho A, Ritschard G, Müller NS, Studer M. Analyzing and visualizing state sequences in R with TraMineR. J Stat Softw. 2011;40(1):1–37. https://doi.org/10.18637/jss.v040.i04.

    Article  Google Scholar 

  38. StataCorp. Stata Statistical Software. College Station, TX: StataCorp LLC; 2017.

    Google Scholar 

  39. Chimbwesha CJ, Giganti MJ, Putta N, Chintu N, Mulindwa J, Dorton BJ, et al. Optimal Time on HAART for Prevention of Mother-to-Child Transmission of HIV. J Acquir Immune Defic Syndr 1999. 2011;58(2):224–8.

    Article  Google Scholar 

  40. Paredes R, Marconi VC, Lockman S, Abrams EJ, Kuhn L. Impact of antiretroviral drugs in pregnant women and their children in Africa: HIV resistance and treatment outcomes. J Infect Dis. 2013;207(Suppl 2):S93–100. https://doi.org/10.1093/infdis/jit110.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Rupérez M, Pou C, Maculuve S, Cedeño S, Luis L, Rodríguez J, et al. Determinants of virological failure and antiretroviral drug resistance in Mozambique. J Antimicrob Chemother. 2015;70(9):2639–47. https://doi.org/10.1093/jac/dkv143.

    Article  CAS  PubMed  Google Scholar 

  42. South Africa: National Department of Health. Implementation of the universal test and treat strategy for HIV positive patients and differentiated care for stable patients [Internet]. 2016 [cited 2019 Jul 24]. Available from: https://sahivsoc.org/Files/22%208%2016%20Circular%20UTT%20%20%20Decongestion%20CCMT%20Directorate.pdf

    Google Scholar 

  43. Ministry of Health, Government of Malawi. Integrated HIV Program Report July—September 2012. 2012.

    Google Scholar 

  44. Ministry of Health, Government of Malawi. Integrated HIV Program Report April-June 2017. 2017. Available from: https://hivstar.lshtm.ac.uk/files/2017/11/Integrated-HIV-Program-Report-April-June-2017-MALAWI.pdf.

  45. Jahn A. Option B+ in Malawi: have 4 years of “treat all” shown that 90–90-90 is achievable? In Boston; 2016. Available from: https://www.croiconference.org/abstract/option-b-malawi-have-4-years-treat-all-shown-90-90-90-achievable/

  46. Hoffman RM, Phiri K, Parent J, Grotts J, Elashoff D, Kawale P, et al. Factors associated with retention in Option B+ in Malawi: a case control study. J Int AIDS Soc. 2017;20(1):21464.

    Article  Google Scholar 

  47. Ford D, Muzambi M, Nkhata MJ, Abongomera G, Joseph S, Ndlovu M, et al. Implementation of Antiretroviral Therapy for Life in Pregnant/Breastfeeding HIV+ Women (Option B+) Alongside Rollout and Changing Guidelines for ART Initiation in Rural Zimbabwe: The Lablite Project Experience. J Acquir Immune Defic Syndr 1999. 2017;74(5):508–16.

    Article  Google Scholar 

  48. Musomba R, Mubiru F, Nakalema S, Mackline H, Kalule I, Kiragga AN, et al. Describing point of entry into care and being lost to program in a cohort of HIV positive pregnant women in a large Urban Centre in Uganda. AIDS Res Treat. 2017;2017:3527563–6. https://doi.org/10.1155/2017/3527563.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Mwapasa V, Joseph J, Tchereni T, Jousset A, Gunda A. Impact of Mother-Infant Pair Clinics and Short-Text Messaging Service (SMS) Reminders on Retention of HIV-Infected Women and HIV-Exposed Infants in eMTCT Care in Malawi: A Cluster Randomized Trial. J Acquir Immune Defic Syndr 1999. 2017;75(Suppl 2):S123–31.

    Article  Google Scholar 

  50. Dlamini-Simelane TTT, Moyer E. “Lost to follow up”: rethinking delayed and interrupted HIV treatment among married Swazi women. Health Policy Plan. 2017;32(2):248–56.

    PubMed  Google Scholar 

  51. Iroezi ND, Mindry D, Kawale P, Chikowi G, Jansen PA, Hoffman RM. A qualitative analysis of the barriers and facilitators to receiving care in a prevention of mother-to-child program in Nkhoma, Malawi. Afr J Reprod Health. 2013;17(4 Spec No):118–129.

  52. Elwell K. Facilitators and barriers to treatment adherence within PMTCT programs in Malawi. AIDS Care. 2016;28(8):971–5. https://doi.org/10.1080/09540121.2016.1153586.

    Article  PubMed  Google Scholar 

  53. Thomson KA, Telfer B, Opondo Awiti P, Munge J, Ngunga M, Reid A. Navigating the risks of prevention of mother to child transmission (PMTCT) of HIV services in Kibera, Kenya: barriers to engaging and remaining in care. PLoS One. 2018;13(1):e0191463. https://doi.org/10.1371/journal.pone.0191463.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Falnes EF, Moland KM, Tylleskär T, de Paoli MM, Msuya SE, Engebretsen IM. “It is her responsibility”: partner involvement in prevention of mother to child transmission of HIV programmes, northern Tanzania. J Int AIDS Soc. 2011;14(1):21. https://doi.org/10.1186/1758-2652-14-21.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Erlwanger AS, Joseph J, Gotora T, Muzunze B, Orne-Gliemann J, Mukungunugwa S, et al. Patterns of HIV Care Clinic Attendance and Adherence to Antiretroviral Therapy Among Pregnant and Breastfeeding Women Living With HIV in the Context of Option B+ in Zimbabwe. J Acquir Immune Defic Syndr 1999. 2017;75(Suppl 2):S198–206.

    Article  Google Scholar 

  56. Atanga PN, Ndetan HT, Achidi EA, Meriki HD, Hoelscher M, Kroidl A. Retention in care and reasons for discontinuation of lifelong antiretroviral therapy in a cohort of Cameroonian pregnant and breastfeeding HIV-positive women initiating “option B+” in the south west region. Trop Med Int Health TM IH. 2017;22(2):161–70. https://doi.org/10.1111/tmi.12816.

    Article  PubMed  Google Scholar 

  57. Dzangare J, Takarinda KC, Harries AD, Tayler-Smith K, Mhangara M, Apollo TM, et al. HIV testing uptake and retention in care of HIV-infected pregnant and breastfeeding women initiated on “option B+” in rural Zimbabwe. Trop Med Int Health TM IH. 2016;21(2):202–9. https://doi.org/10.1111/tmi.12637.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

The authors would like to thank all the participants in the study.

Funding

This study was made possible with support from the Economic and Social Research Council (ES/JS00021/1), the Bill and Melinda Gates Foundation for the MeSH Consortium (OPP1120138), the Bill and Melinda Gates Foundation ALPHA grant (OPP1164897), and the MRC SHAPE UTT grant (MR/P014313/1).

Author information

Authors and Affiliations

Authors

Contributions

The study was conceived by DE, AW and GR. Fieldwork was planned and executed by DE, FXGO and CWK. Data collection was supervised by DE. Analyses were conducted by DE with input from all authors. All authors contributed to the interpretation of the findings. The manuscript was drafted by DE with input from JR, BR and all the authors. All authors read and approved the final manuscript.

Corresponding author

Correspondence to David Etoori.

Ethics declarations

Ethics approval and consent to participate

Ethical approval was obtained from the London School of Hygiene and Tropical Medicine Research Ethics Committee, the University of Witwatersrand Human Research Ethics Committee, and the Mpumalanga Department of Health Public Health Research Committee. All methods were carried out in accordance with relevant guidelines and regulations and informed consent was obtained from all participants and from a parent and/or legal guardian of minors (subjects under 18) in the study.

Consent for publication

Not Applicable.

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1.

Flowchart showing data sources and data used for all analyses. A flowchart that illustrates the different databases and data that were used to conduct the different analyses for this manuscript.

Additional file 2.

Kaplan-Meier curves showing the probability of ART initiation following an HIV-positive test during pregnancy stratified by timing of the positive test and year of delivery. A panel of graphs showing ART initiation stratified by trimester at receipt of a positive HIV result and year of delivery.

Additional file 3.

Chronograms of engagement in care during the vertical risk transmission period by each engagement cluster. A panel of graphs showing engagement cluster chronograms.

Additional file 4.

Histogram of number of resumptions following a new pregnancy stratified by type of resumption (reengagement vs transfer) and year of delivery. A panel of graphs showing the frequency of clinic transfer or reengagement in care stratified by the year of delivery.

Additional file 5.

Factors associated with membership in each identified engagement cluster. A table showing factors associated with membership in each engagement cluster.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Etoori, D., Rice, B., Reniers, G. et al. Patterns of engagement in HIV care during pregnancy and breastfeeding: findings from a cohort study in North-Eastern South Africa. BMC Public Health 21, 1710 (2021). https://doi.org/10.1186/s12889-021-11742-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12889-021-11742-4

Keywords