Of 250 women initiated, 99 were LTFU of whom 46 (18.4%) within 28 days of delivery and 94 (37.6%) within a year of HAART initiation (Table 4). LTFU numbers per initiation year were 26 (31.7%), 39 (54.9%), and 34 (35.1%), with 46/99 (46.5%) of all LTFU occurring within the first 8 weeks after HAART initiation. LTFU rates were higher in 2009 compared with those in 2008 but improved again in 2010. LTFU women who were initiated on 36 weeks of gestation and later were significantly more likely to become LTFU within 4 weeks of initiation [7/27 (25.9%), vs. 23/223 (10.3%), P = 0.028].
We describe the characteristics of a cohort of pregnant women who initiated lifelong HAART during pregnancy in an urban public ART clinic in the Western Cape, South Africa, over 3 consecutive years.
When comparing the 2010 cohort with the 2009 cohort, we found that the median CD4 count at the first antenatal visit increased significantly. We also found that the median gestation at first antenatal visit and median gestation at HAART initiation had decreased. Not surprisingly, there was also a significant increase in the duration of HAART until delivery. This is in line with the revision of the national PMTCT protocol in April 2010,4 ensuring that more pregnant women are eligible for HAART referral by increasing the cutoff CD4 count from 250 to 350 cells per microliter and recommending fast-tracked HAART initiation within 2 weeks of diagnosis, from 14-week gestation. Compared with a study from Cape Town describing a cohort of HIV-infected pregnant women from 2005 with a median gestation at HAART initiation of 32 weeks,9 our cohort showed an earlier gestation at initiation from 31 weeks in 2008 to 25 weeks in 2010. As a consequence of this, the percentage of women receiving HAART for >8 weeks before delivery increased from 48.1% in 2008 (comparable with the 2005 rate in that study) to 73.7% in 2010. Thus, the new guideline has impacted the earlier initiation of therapy at higher CD4 count, an important factor in reducing MTCT.
Gestation at the first antenatal visit, HAART initiation, and delivery has been verified with antenatal records where possible. However, only 63% of the study participants booked before 24 weeks of gestation, and no accurate estimation of gestation by ultrasound can be established thereafter. Although booking gestation improved significantly in 2010 with a median of 17 weeks, still 9.2% of women booked at or after 30 weeks of gestation. Overall booking gestation is still late in pregnancy, limiting the window of opportunity for successful HAART initiation before delivery. Women should be encouraged to book early in pregnancy. Further research is needed to identify factors causing delays in booking.
The changes in the guideline do not seem to have expedited the referral process for HAART initiation. Although improvement was found in the median time between booking and actual HAART initiation in 2009, no further improvement was found after the introduction of the guideline in 2010, with the median time still >5 weeks. It is therefore clear that fast-tracking HAART initiation itself was not responsible for the reduction in MTCT, but rather that earlier HAART initiation at a higher CD4 count was the main contributor to this outcome.
Greater improvements might still be achievable by eliminating delays in the referral process for HAART initiation. In 2010, only 4.1% of pregnant women were initiated within the stipulated time period of 2 weeks after HIV diagnosis as per PMTCT guideline. This clearly identifies the current bottleneck of the referral system and poor integration of services. The median time between the first visit and HAART initiation at our facility was significantly shorter in 2010 at a median of 1.1 weeks, indicating that most treatment delay occurs during referral to our facility. Generally, delay is caused by awaiting laboratory results. Whether on-site rapid CD4 count measurement could shorten referral delay for pregnant women should be investigated, as this would minimize the time to establish HAART eligibility. Baseline blood could be drawn in the ANC facility upon referral, enabling ART clinicians with access to regional laboratory results to initiate HAART at the first visit if desired. Referral delay from the ANC facility to the ART facility might be caused by limited booking capacity for new referrals at the ART facility and requires investigation. Expansion of ART sites in the community with fast-track appointment slots for pregnant women is recommended. Integration of services with ANC health workers initiating and monitoring HAART is proposed, as this would eliminate the referral process entirely. Health policy changes toward universal HAART initiation for all HIV-positive pregnant women regardless of CD4 count would both reduce MTCT and simplify the initiation process.
The HIV transmission rate has decreased significantly with no transmission detected in 2010, compared with 7.0% in 2009. In 2010, the vertical HIV transmission rate at Tygerberg ID Clinic was comparable with the rates achieved in developed countries, which is an important milestone. Further studies are needed to show the impact on maternal health of starting therapy at higher CD4 counts. Our study confirmed that a longer duration of HAART in pregnancy was associated with decreased HIV transmission. No significant association was found between maternal age, maternal parity, booking CD4 count, booking gestation or delivery gestation, and HIV transmission. These findings are in agreement with previous reported data from a similar cohort in Cape Town.8
The availability of neonatal PCR results had significantly improved in 2010, with only 9.1% of PCR results unknown compared to 24% in 2009. Several factors contribute to the unavailability of neonatal PCR results, such as early neonatal death, migration, LTFU, errors in personal detail registration at the National Health Laboratory Service or Community Health Clinic, and lack of feedback systems between ART site and ANC and baby Community Health Clinics. Due to the retrospective nature of the study, information and selection bias may have occurred. As a result of the high number of LTFU after delivery, a number of baby birth dates and names could not be verified. This contributed to 17% of PCR results being unknown, possibly leading to the underestimation of transmission rates.
LTFU rates were unacceptably high at 37.6% within 1 year of initiation, with 18.4% of all LTFU occurring within 28 days of delivery date, which is generally within 8 weeks of initiation. Women initiating HAART in pregnancy are significantly more likely to become LTFU than are nonpregnant women initiating HAART.10,11 Interestingly, those women who were more established on therapy had a lower risk of LTFU. Several reasons could contribute to the poor retention in care around delivery, such as poor maternal health, fast-tracked initiation with too little time to come to terms with recent HIV diagnosis, inadequate support with no regular contact with health care services related to the pregnancy, poor socioeconomic circumstances undermining transport and childcare availability, relocation to family homesteads and multiple appointments at postnatal and baby clinics at different locations. Further research is needed to better understand how retention in care after delivery can be improved to optimize the healthcare status of these vulnerable women and neonates. A possible intervention could be the provision of a >1 month prescription at the last predelivery visit to bridge the stressful time around delivery and the early postnatal period.
The LTFU date was recorded as the date of the last clinic or pharmacy attendance without recorded transfer out procedures. We may have underestimated the time to actual nonadherence, as patients might have adhered until the very end of their provided medication, which is generally 28 days. It is also possible that women have self-referred to alternative ART sites to obtain further medication, without requesting clinical notes. Likewise, it is uncertain if women who have been transferred to other clinics did in fact end up attending these facilities, as no feedback system is in place between ART sites. It is however reasonable to assume that most women who have not returned for planned follow-up appointments have at least experienced some period of treatment interruption, as only the exact amount of needed medication till the next follow-up date is dispensed and ART is not readily available in local primary clinics or other ART sites without referral letters. Therefore, the magnitude of reported LTFU remains of great concern, especially as treatment interruption during breastfeeding might increase the transmission risk. Emphasis on retention in care should be promoted by developing efficient referral feedback systems between community ANC and baby clinics and strengthening patient tracing capacity.
The Tygerberg ID clinic functions predominantly as a community-based primary level ART care facility for those women in need of HAART who live in the local catchment area. However, as central referral hospital, the clinic also manages more complicated patients who are referred from other antenatal, ART or specialist clinics with specific comorbidity. The impact of these referred patients, together with the greater infrastructure at the Tygerberg ID clinic, interact in a complex way and limit the extrapolation of the data to other community-based ART clinics in South Africa.
Since the institution of the 2010 National PMTCT guideline, women initiated HAART earlier in pregnancy, and less MTCT of HIV transmission was evident. Longer duration of HAART before delivery was associated with a reduced risk of HIV transmission.
However, women are still initiating therapy relatively late in pregnancy and retention in care remains poor with unacceptable LTFU rates. With earlier access to PMTCT for pregnant women, the focus should now shift toward educating women of the benefits of early booking to effect immediate referral to ART facilities after HIV diagnosis and ensuring women remain in care in the postnatal period.
The authors gratefully acknowledge Sr Anneke Theron for collecting obstetric data, Justin Harvey for statistical support, Kathryn Stinson for her valuable comments on the manuscript, and the dedicated staff of the Tygerberg Infectious Diseases Clinic.
1. UNAIDS. Report on the Global AIDS Epidemic, 2010. Geneva, Switzerland: UNAIDS; 2010.
2. National Department of Health. The 2010 National Antenatal Sentinel HIV and Syphilis Prevalence Survey in South Africa. Pretoria, South Africa: Department of Health; 2011.
3. National Department of Health. Policy and Guidelines for the Implementation of the PMTCT Programme. Pretoria, South Africa: National Department of Health; 2008.
4. National Department of Health. Clinical Guidelines: PMTCT (Prevention of Mother-to-Child Transmission). Pretoria, South Africa: National Department of Health; 2010.
5. Cooper ER, Charurat M, Mofenson L, et al.. Combination antiretroviral strategies for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr. 2002;29:484–494.
6. Warszawski J, Tubiana R, Le Chenadec J, et al.. Mother-to-child HIV transmission despite antiretroviral therapy in the ANRS French Perinatal Cohort. AIDS. 2008;22:289–299.
7. Goga AE, Dinh TH, Jackson DJ; for the SAPMTCTE study group. Evaluation of the Effectiveness of the National Prevention of Mother-to-Child Transmission (PMTCT) Programme Measured at Six Weeks Postpartum in South Africa, 2010. South African Medical Research Council, National Department of Health of South Africa and PEPFAR/US Centers for Disease Control and Prevention. 2012.
8. Fitzgerald FC, Bekker LG, Kaplan R, et al.. Mother-to-child transmission of HIV in a community-based antiretroviral clinic in South Africa. S Afr Med J. 2010;100:827–831.
9. Stinson K, Boulle A, Coetzee D, et al.. Initiation of highly active antiretroviral therapy among pregnant women in Cape Town, South Africa. Trop Med Int Health. 2010;15:825–832.
10. Kaplan R, Orrell C, Zwane E, et al.. Loss to follow-up and mortality among pregnant women referred to a community clinic for antiretroviral treatment. AIDS. 2008;22:1679–1681.
11. Wang B, Losina E, Stark R, et al.. Loss to follow-up in a community clinic in South Africa—roles of gender, pregnancy and CD4 count. S Afr Med J. 2011;101:253–257.
Keywords:© 2013 by Lippincott Williams & Wilkins
HIV; antenatal; antiretroviral therapy; mother-to-child transmission; South Africa