WHO/HIV/2013.22
© World Health Organization 2013
GRADE tables: What ARV prophylaxis should be given to pregnant women living with HIV who do not receive lifelong ART? RANDOMIZED CONTROL TRIALS Author(s): Alexander W. Kay, Alicen B. Spaulding, Tara Horvath Date: 2012-11-27 Question: Should triple antiretroviral (ARV) prophylaxis versus zidovudine (AZT) + single-dose nevirapine (SD-NVP) be used for pregnant women versus maternal AZT and infant NVP for preventing the mother-to-child transmission of HIV? Settings: Burkina Faso, Kenya, Malawi, South Africa Bibliography: de Vincenzi 2011, Fao 2012, Jamieson 2012
Quality assessment
No. of patients
Effect Quality Importance
No. of studies
Design
Risk of bias
Inconsistency
Indirectness Imprecision
Other considerations
Triple ARV
ZDV + SDNVP
Relative (95% CI)
Absolute
Maternal mortality (48–52 weeks) 2 randomized trials no serious risk of bias no serious inconsistency serious1 very serious2 none 5/1261 (0.4%) 10/1080 (0.93%) RR 0.46 (0.16 5 fewer per 1000 (from ⊕OOO to 1.29) 8 fewer to 3 more) VERY LOW CRITICAL
Infant mortality (48–52 weeks) 2 randomized trials no serious risk of bias no serious inconsistency serious1 serious4 none 47/1250 (3.8%) 66/1072 (6.2%) RR 0.64 (0.45 to 0.92) 22 fewer per 1000 (from 5 fewer to 34 fewer) ⊕⊕OO LOW CRITICAL
Maternal morbidity (48–52 weeks) 3 2 randomized trials no serious risk of bias no serious inconsistency serious1 serious4 none 110/1261 (8.7%) 82/1080 (7.6%) RR 1.23 (0.94 17 more per 1000 (from to 1.62) 5 fewer to 47 more) ⊕⊕OO LOW CRITICAL
Infant morbidity (48–52 weeks) 3
1 This work was commissioned by the World Health Organization and carried out by The University of California, San Francisco (UCSF), Cochrane Review Group on HIV/AIDS
WHO/HIV/2013.22
© World Health Organization 2013
Quality assessment
No. of patients
Effect Quality Importance
No. of studies 2
Design
Risk of bias
Inconsistency
Indirectness Imprecision serious1
Other considerations none
Triple ARV 371/1250 (29.7%)
ZDV + SDNVP
Relative (95% CI)
Absolute
randomized trials
no serious risk of bias
no serious inconsistency
no serious imprecision
333/1072 RR 0.97 (0.86 9 fewer per 1000 (from ⊕⊕⊕O (31.1%) to 1.1) 43 fewer to 31 more) MODERAT E
CRITICAL
Infant HIV infections (48–52 weeks) 2 randomized trials no serious risk of bias no serious inconsistency serious1 none none 97/1250 (7.8%) 111/1072 RR 0.74 (0.57 (10.4%) to 0.95) 27 fewer per 1000 (from 5 fewer to 45 fewer) ⊕⊕⊕O MODERAT E CRITICAL
Retention in care (48–52 weeks) 5 2 randomized trials no serious risk of bias no serious inconsistency serious1 no serious imprecision none IMPORTANT 1050/1250 910/1072 RR 1 (0.96 to 0 fewer per 1000 (from ⊕⊕⊕O (84%) (84.9%) 1.03) 34 fewer to 25 more) MODERAT E
Maternal TB (48 weeks) 1 randomized trials no serious risk of bias no serious inconsistency serious1 very serious2 none 4/849 (0.47%) 4/668 (0.6%) RR 0.79 (0.2 1 fewer per 1000 (from to 3.13) 5 fewer to 13 more) ⊕⊕OO LOW IMPORTANT
1 2
Neither included intervention directly replicated option B versus A, and both interventions included women with CD4 cell counts <350 cells/mm3. Very few events. 3 Combined count of all clinical serious adverse effects. 4 Few events. 5 Defined in de Vincenzi (2011) as infant retention and Jamieson (2012) as retention of maternal-infant pairs.
2 This work was commissioned by the World Health Organization and carried out by The University of California, San Francisco (UCSF), Cochrane Review Group on HIV/AIDS
WHO/HIV/2013.22
© World Health Organization 2013
OBSERVATIONAL STUDIES Author(s): Alexander W. Kay, Alicen B. Spaulding Date: 2012-11-25 Question: Should triple antiretroviral (ARV) versus zidovudine (AZT) + single-dose nevirapine (SD-NVP) be used for pregnant women and their infants? Settings: Botswana Bibliography: Dryden-Peterson 2011
Quality assessment
No. of patients
Effect Quality Importance
No. of studies
Design
Risk of bias
Inconsistency
Indirectness Imprecision
Other considerations
Triple ART
AZT + SDNVP
Relative (95% CI)
Absolute
Infant mortality (follow-up 6 months) 1 observational studies serious1 no serious inconsistency serious2 very serious3 none 10/258 (3.9%) 10/170 (5.9%) RR 0.66 (0.28 to 1.55) 20 fewer per 1000 (from 42 fewer to 32 more) ⊕OOO VERY LOW CRITICAL
Infant HIV infections (follow-up 6 months) 1 observational studies serious1 no serious inconsistency serious2 very serious3 none 1/258 (0.39%) 9/170 (5.3%) RR 0.07 (0.01 to 0.57) 49 fewer per 1000 (from 23 fewer to 52 fewer) ⊕OOO VERY LOW CRITICAL
1 2
Reflecting Botswana treatment guidelines, mothers receiving triple ART had lower CD4 cell counts than mothers in the AZT group. Women were given triple ARV if they had CD4 cell counts ≤250 cells/mm3. 3 Very few events.
3 This work was commissioned by the World Health Organization and carried out by The University of California, San Francisco (UCSF), Cochrane Review Group on HIV/AIDS