Journal of the Pediatric Infectious Diseases Society 646 • JPIDS 2023:12 (December) • BRIEF REPORT Association of HIV Exposure and HIV Infection With In-hospital Mortality Among Hospitalized Infants <1 Year of Age, South Africa, 2016–2018 Nicole Wolter,1,2, Sibongile Walaza,1,3 Claire von Mollendorf,1 Anne von Gottberg,1,2 Stefano Tempia,1 Meredith L. McMorrow,4,5 Jocelyn Moyes,1,3 Florette Treurnicht,1,2 Orienka Hellferscee,1,2, Malefu Moleleki,1,2 Mvuyo Makhasi,1,3 Neydis Baute,6 and Cheryl Cohen1,3 1Centre for Respiratory Diseases and Meningitis, National Institute for Communicable Diseases of the National Health Laboratory Service, Johannesburg, South Africa, 2School of Pathology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, 3School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, 4Influenza Division, Centers for Disease Control and Prevention, Atlanta, GA, USA, 5Influenza Program, Centers for Disease Control and Prevention, Pretoria, South Africa, and 6Department of Paediatrics, Mapulaneng Hospital, Mpumalanga, South Africa We enrolled 1323 hospitalized infants aged <1 year in 2016– 2018, and examined the association between HIV status and in-hospital mortality. After controlling for confounders, HIV- exposed uninfected infants did not have an increased risk of mortality, whereas infants living with HIV had 4 times greater risk compared with HIV-uninfected infants. Key words. Africa; HIV exposure; HIV infection; infants; mortality. INTRODUCTION While infants living with HIV (HI) are known to be at increased risk of mortality compared with HIV unexposed uninfected (HUU) infants [1], the effect of HIV exposure on infant mor- tality is less well understood with much of the data from prior to widespread antiretroviral therapy (ART) use, and from non- African countries [2]. In a national antenatal survey in 2017, 31% of pregnant women in South Africa were living with HIV [3]. A national program for prevention of vertical transmission of HIV was im- plemented in 2002 and was expanded and improved over time so that from 2015 all women living with HIV were eligible for lifelong ART, irrespective of CD4 count, as well as the availa- bility of combination therapy [4]. This program has successfully reduced the risk of vertical HIV transmission; however, there is a growing population of HIV-exposed uninfected (HEU) in- fants. In a meta-analysis in 2016 of 22 studies conducted be- tween 1986 and 2013, the pooled estimate showed that HEU infants had more than a 70% increased risk of all-cause mor- tality compared with HUU infants within the first 2 years of life, although there was heterogeneity between the individual studies [5]. We aimed to determine whether HIV exposure and HIV infection were associated with an increased risk of in-hospital mortality among infants hospitalized with acute medical illness, in the context of an established ART program. METHODS The Infant Burden Study was conducted to assess the burden of disease associated with influenza and other respiratory pathogens among hospitalized infants at 3 sentinel surveil- lance hospitals located in 3 (KwaZulu-Natal, Mpumalanga, and North West) provinces from July 2016 through October 2018. Surveillance officers enrolled infants aged <1 year admitted to the medical ward with acute (symptom onset ≤10 days) respi- ratory and non-respiratory medical illness (excluding trauma/ surgical cases), including medical admissions to the ICU. Surveillance officers collected whole blood, and demographic and clinical information by structured interview with parent/ caregiver and hospital record review. Infants were followed up until discharge or in-hospital death. Infant HIV status was determined from standard of care testing during the admission, or through testing at enrollment by polymerase chain reaction (PCR). HUU infants were in- fants with a negative HIV result and a recently documented (<3 months) negative maternal HIV status. HEU infants were infants with a negative HIV result and a recently documented or verbally reported positive maternal HIV status, evidence that the mother was taking ART during pregnancy or post- partum. HI infants were infants with a recently documented positive HIV result, verbally reported by the parent/caregiver, or evidence that the infant was receiving ART. If the mother’s HIV status was unknown or a negative test result was from >3 months prior, the mother was offered voluntary counseling and testing. A cross-sectional analysis was conducted to assess the as- sociation between infant HIV status and in-hospital mortality. Factors associated with mortality were assessed using univar- iate random effects logistic regression, accounting for clustering Received 1 August 2023; editorial decision 6 November 2023; accepted 10 November 2023 Corresponding Author: Nicole Wolter, Centre for Respiratory Diseases and Meningitis, National Institute for Communicable Diseases, Private Bag X4, Sandringham, 2131, Gauteng, South Africa. E-mail: nicolew@nicd.ac.za. Journal of the Pediatric Infectious Diseases Society 2023;12(12):646–651 © The Author(s) 2023. Published by Oxford University Press on behalf of The Journal of the Pediatric Infectious Diseases Society. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. https://doi.org/10.1093/jpids/piad100 B R I E F R E P O R T D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 https://orcid.org/0000-0002-9526-0133 https://orcid.org/0000-0001-9246-8927 mailto:nicolew@nicd.ac.za BRIEF REPORT • JPIDS 2023:12 (December) • 647 by site. Potential confounders (P < .2) from this analysis were examined individually for their effect on the association be- tween HIV status and mortality, with evidence of confounding considered if the main effect adjusted odds ratio (aOR) differed by >10% from the unadjusted odds ratio (OR) for either HI or HEU compared with HUU infants. Length of hospital stay, ox- ygen therapy, and intensive care unit admission were considered to be on the causal pathway, and not assessed for confounding. Multivariable random effects logistic regression accounting for clustering by site was performed, adjusting for the identified confounders. A priori variables included age, vaccination status, and feeding type. Sensitivity analyses of the final model was performed by (1) comparing the risk of in-hospital mortality in HI com- pared with HIV-uninfected (grouping together both HUU and HEU) infants, and (2) restricting the analysis to infants aged <6 months. Statistical analyses were performed using Stata version 14.0. Ethical approval was obtained from the University of the Witwatersrand Human Research Ethics Committee (M140824) and University of KwaZulu-Natal Biomedical Research Ethics Committee (BE605/16). This surveillance was deemed non- research by the US Centers for Disease Control and Prevention. RESULTS During the study period, 1323 infants were enrolled, of which 1321 (99.8%) had known in-hospital outcome and 1296 (98.0%) had known HIV status. The most common admission diagnoses were pneumonia (538/1314, 40.9%) and diarrhea (191/1314, 14.5%), with 52.0% (683/1314) admitted with a respiratory di- agnosis (Supplementary Table 1). Two percent (27/1321) of in- fants died while in hospital. Overall, 55.8% (723/1296) of infants were HUU, 40.6% (526/1296) were HEU, and 3.6% (47/1296) were HI. Characteristics of enrolled infants by HIV status are shown in Table 1. The mortality ratio was 1.7% (12/721), 1.5% (8/526), and 12.8% (6/47) among HUU, HEU, and HI infants, respectively. Among the 38/47 (80.9%) HI infants with information avail- able, 92.1% (35/38) had mother’s that received ART during pregnancy. For 28 HI infants with available information, 21 (75.0%) were receiving ART at the time of admission. Of 523 HEU infants with available data, 513 (98.1%) had mother’s that were receiving ART during pregnancy. The OR for mortality for HEU infants was 0.99 (95% con- fidence interval (CI) 0.40–2.44) and for HI infants was 7.39 (95% CI 2.58–21.19), compared with HUU infants (Table 2). Infant mortality by demographic and clinical characteristics, stratified by HIV status, is shown in Supplementary Table 2. After adjusting for a priori confounders, as well as the iden- tified confounders (malnutrition and maternal education, Supplementary Table 3), HEU infants were not at increased risk of mortality compared with HUU (aOR 0.80, 95% CI .31–2.09), whereas HI infants had over 4 times increased risk of death compared with HUU (aOR 4.79, 95% CI 1.49–15.37) (Table 2). Malnutrition and not being fully vaccinated were also associ- ated with an increased risk of mortality. HI infants remained at increased risk of mortality com- pared with all HIV-uninfected (HUU and HEU) infants (Supplementary Table 4). When restricting to infants aged <6 months, HI but not HEU, were at increased risk of mortality compared with HUU infants (Supplementary Table 5). DISCUSSION In the context of a mature program for prevention of vertical HIV transmission this study found that HI infants, but not HEU infants, were at increased risk of in-hospital mortality compared with HUU infants. Malnutrition and not being fully vaccinated for age were also identified as risk factors for in-hospital mortality. The proportion of infants that were HEU and HI were higher than was expected based on the reported HIV prevalence (31%) of women who attended antenatal care in 2017 [3], as well as that reported in a cross-sectional survey conducted in 2012– 2013 (33.1% HEU and 2.6% HI) [6]. This may be due to these studies having been conducted among pregnant women and in- fants presenting for routine immunization, whereas our study was conducted among admitted infants, with HEU and HI in- fants more likely to be hospitalized [7]. Previous studies examining the association between HIV exposure and mortality have reported conflicting results. In a study conducted in 2010–2013 among infants aged <6 months hospitalized with lower respiratory tract infection in South Africa, HEU infants had 2.1-times increased risk of mortality compared with HUU infants [8], although feeding type was not adjusted for. In another study, among infants aged <6 months in South Africa in 2012–2013, the risk of hospitaliza- tion or death was 4-fold higher among HEU infants compared with HUU [9], although the CI included 1. A more recent study in a South African cohort of infants <1 year in 2017– 2019 showed that while infectious-cause hospitalization was higher for HEU than HUU infants, there was no difference in mortality [10]. Data from the Thembisa model showed that ART coverage among adult females in 2018 in South Africa was 63% [11]. Use of ART in the mothers of HEU infants, and thereby improved health and well-being of the mother during and after pregnancy, would have likely resulted in transfer of protective maternal antibodies as well as improved infant care and nutrition. Malnutrition was identified as a risk factor for mortality, with an increased mortality ratio in HUU, HEU, and HI infants com- pared with well-nourished infants with the same HIV status. Similarly, in the CHAMPS study in 7 countries in 2016–2020, D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 http://academic.oup.com/jpids/article-lookup/doi/10.1093/jpids/piad100#supplementary-data http://academic.oup.com/jpids/article-lookup/doi/10.1093/jpids/piad100#supplementary-data http://academic.oup.com/jpids/article-lookup/doi/10.1093/jpids/piad100#supplementary-data http://academic.oup.com/jpids/article-lookup/doi/10.1093/jpids/piad100#supplementary-data http://academic.oup.com/jpids/article-lookup/doi/10.1093/jpids/piad100#supplementary-data 648 • JPIDS 2023:12 (December) • BRIEF REPORT Table 1. Characteristics of Infants <1 Year of Age Enrolled in the Infant Burden Study, Overall and by HIV Status, South Africa, July 2016–October 2018 (N = 1323) Variable Categories No. (%) N = 1323 HIV Unex- posed Unin- fected (HUU) No. (%) N = 723 HIV-exposed Uninfected (HEU) No. (%) N = 526 HEU vs HUU P-valueg Living with HIV (HI) No. (%) N = 47 HI vs. HUU P-valueg Site Mpumalanga 214 (16.2) 114 (15.8) 73 (13.9) .075 16 (34.0) .001 KwaZulu-Natal 683 (51.6) 364 (50.4) 299 (56.8) 12 (25.5) North West 426 (32.2) 245 (33.9) 154 (29.3) 19 (40.4) Year 2016 177 (13.4) 105 (14.5) 62 (11.8) .369 8 (17.0) .493 2017 575 (43.5) 308 (42.6) 229 (43.5) 23 (49.0) 2018 571 (43.2) 310 (42.9) 235 (44.7) 16 (34.0) Sex Male 763 (57.7) 415 (57.4) 305 (58.0) .836 26 (55.3) .780 Female 560 (42.3) 308 (42.6) 221 (42.0) 21 (44.7) Race Black 1303 (98.5) 706 (97.7) 523 (99.4) .013 47 (100.0) .288 Non-black 20 (1.5) 17 (2.3) 3 (0.6) 0 (0.0) Age group (months) <1 108 (8.2) 56 (7.8) 51 (9.7) .344 1 (2.1) .177 1 to <3 364 (27.5) 207 (28.6) 137 (26.1) 10 (21.3) 3 to <6 335 (25.3) 188 (26.0) 120 (22.8) 15 (31.9) 6 to <9 277 (20.9) 148 (20.5) 118 (22.4) 8 (17.0) 9 to <12 239 (18.1) 124 (17.2) 100 (19.0) 13 (27.7) Underlying conditiona No 1284 (97.0) 698 (96.5) 513 (97.5) .316 46 (97.9) .625 Yes 39 (3.0) 25 (3.5) 13 (2.5) 1 (2.1) Malnutritionb No 975 (73.7) 556 (76.9) 382 (72.6) .145 21 (44.7) <.001 Yes 347 (26.2) 166 (23.0) 144 (27.4) 26 (55.3) Unknown 1 (0.1) 1 (0.1) 0 (0.0) 0 (0.0) Feeding type Exclusive breastfeeding 707 (53.4) 427 (59.1) 248 (47.2) <.001 21 (44.7) .066 Mixed feeding 240 (18.1) 142 (19.6) 82 (15.6) 12 (25.5) Formula feeding 333 (25.2) 130 (18.0) 182 (34.6) 14 (29.8) Unknown 43 (3.3) 24 (3.3) 14 (2.7) 0 (0.0) Prematurityc No 1082 (81.8) 585 (80.9) 428 (81.4) .839 42 (89.4) .149 Yes 241 (18.2) 138 (19.1) 98 (18.6) 5 (10.6) Birthweightd Normal 970 (73.3) 533 (73.7) 384 (73.0) .516 32 (68.1) .003 Low 282 (21.3) 153 (21.2) 121 (23.0) 7 (14.9) Unknown 71 (5.4) 37 (5.1) 21 (4.0) 8 (17.0) Vaccinatione Full coverage 964 (72.9) 530 (73.3) 378 (71.9) .646 37 (78.7) .013 No full coverage 307 (23.2) 167 (23.1) 132 (25.1) 5 (10.6) Unknown 52 (3.9) 26 (3.6) 16 (3.0) 5 (10.6) Admission diagnosisf Non-respiratory 631 (47.7) 320 (44.3) 283 (53.8) .003 17 (36.2) .470 Respiratory 683 (51.6) 399 (55.2) 239 (45.4) 30 (63.8) Unknown 9 (0.7) 4 (0.6) 4 (0.8) 0 (0.0) Maternal education level None/Primary 616 (46.6) 300 (41.5) 268 (51.0) .004 33 (70.2) .001 Secondary/Tertiary 694 (52.5) 414 (57.3) 254 (48.3) 14 (29.8) Unknown 13 (1.0) 9 (1.2) 4 (0.8) 0 (0.0) Outcome Survived 1294 (97.8) 709 (98.1) 518 (98.5) .473 41 (87.2) <.001 Died 27 (2.0) 12 (1.7) 8 (1.5) 6 (12.8) Unknown 2 (0.2) 2 (0.3) 0 (0.0) 0 (0.0) aUnderlying condition includes asthma, chronic lung, heart, liver or renal disease, stroke, sinusitis, organ transplant, anemia, immunosuppressive therapy, splenectomy, diabetes, burns, im- munoglobulin deficiency, autoimmune disease, nephrotic syndrome, cancer, spinal cord injury, seizure disorder, cerebral palsy, congenital heart disease, other congenital disorder, obesity, or chronic gastrointestinal problems. bMalnutrition defined as a weight-for-age <−2 standard deviations from the WHO mean Z-score. cPrematurity defined as gestational age at birth of <37 weeks. dLow infant birthweight defined as <2500 g. eVaccination defined as full vaccine coverage for age, using the Haemophilus influenzae type b vaccine given as part of the routine infant immunization schedule at 6,10, 14 weeks as a proxy. fAdmission diagnosis: respiratory diagnosis includes apnea, neonatal sepsis, bronchiolitis, pneumonia, tuberculosis, and bronchitis and non-respiratory diagnosis includes encephalitis, viral illness, diarrhea, febrile seizures, meningitis, sepsis (non-neonatal), and other diagnosis. gChi-squared test calculated for known data (unknown excluded). D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 BRIEF REPORT • JPIDS 2023:12 (December) • 649 Ta bl e 2. D em og ra ph ic a nd C lin ic al F ac to rs A ss oc ia te d W ith In -h os pi ta l M or ta lit y A m on g In fa nt s A ge d <1 Y ea r, So ut h A fr ic a, J ul y 20 16 –O ct ob er 2 01 8 Va ria bl e C at eg or ie s To ta l, N o. N = 1 32 1 N o. o f de at hs N = 2 7 M or ta lit y ra tio , % (9 5% C I) P- va lu e g O dd s ra tio h (9 5% C I) P- va lu e A dj us te d O R i ( 95 % C I) P- va lu e H IV s ta tu s H U U 72 1 12 1. 7 (0 .9 –2 .9 ) < .0 01 R ef er en ce R ef er en ce H E U 52 6 8 1. 5 (0 .7 –2 .3 ) 0. 99 (0 .4 0– 2. 44 ) 0. 97 4 0. 80 (0 .3 1– 2. 09 ) .6 52 H I 47 6 12 .8 (4 .7 –2 7. 8) 7. 39 (2 .5 8– 21 .1 9) < 0. 00 1 4. 79 (1 .4 9– 15 .3 7) .0 08 U nk no w n 27 1 3. 7 (0 .1 –1 9. 0) 2. 06 (0 .2 5– 16 .8 3) 0. 49 9 1. 94 (0 .2 1– 17 .9 1) .5 57 S ite M pu m al an ga 21 4 6 2. 8 (1 .0 –6 .1 ) < .0 01 N /A Kw aZ ul u- N at al 68 2 4 0. 6 (0 .2 –1 .5 ) N /A N or th W es t 42 5 17 4. 0 (2 .3 –6 .4 ) N /A Ye ar 20 16 17 7 5 2. 8 (0 .9 –6 .6 ) .6 67 R ef er en ce 20 17 57 4 10 1. 7 (0 .8 –3 .2 ) 0. 68 (0 .2 2– 2. 07 ) 0. 49 3 20 18 57 0 12 2. 1 (1 .1 –3 .7 ) 0. 81 (0 .2 7– 2. 10 ) 0. 70 2 S ex M al e 76 2 13 1. 7 (0 .9 –2 .9 ) .3 11 R ef er en ce Fe m al e 55 9 14 2. 5 (1 .3 –3 .9 ) 1. 47 (0 .6 8– 3. 18 ) 0. 32 4 R ac e B la ck A fr ic an 13 01 27 2. 1 (1 .4 –3 .0 ) .5 15 R ef er en ce N on -b la ck A fr ic an 20 0 0. 00 A ge g ro up (m on th s) < 1 10 8 3 2. 8 (0 .6 –8 .1 ) .1 38 1. 09 (0 .2 8– 4. 29 ) 0. 89 7 0. 49 (0 .0 9– 2. 76 ) .4 16 1 to < 3 36 3 4 1. 1 (0 .3 –2 .8 ) 0. 24 (0 .0 7– 0. 80 ) 0. 02 0 0. 15 (0 .0 4– 0. 63 ) .0 10 3 to < 6 33 5 8 2. 4 (1 .0 –4 .7 ) 0. 60 (0 .2 3– 1. 60 ) 0. 30 7 0. 51 (0 .1 8– 1. 44 ) .2 00 6 to < 9 27 7 3 1. 1 (0 .2 –3 .2 ) 0. 26 (0 .0 7– 0. 97 ) 0. 04 5 0. 20 (0 .0 5– 0. 82 ) .0 25 9 to < 12 23 8 9 3. 8 (1 .7 –7 .2 ) R ef er en ce R ef er en ce U nd er ly in g co nd iti on a N o 12 82 25 1. 95 (1 .3 –2 .9 ) .1 67 R ef er en ce Ye s 39 2 5. 1 (0 .6 –1 8. 5) 5. 48 (1 .1 2– 26 .7 9) 0. 03 6 M al nu tr iti on b N o 97 5 9 0. 92 (0 .4 –1 .8 ) < .0 01 R ef er en ce R ef er en ce Ye s 34 5 18 5. 2 (3 .1 –8 .3 ) 5. 19 (2 .2 9– 11 .7 9) < 0. 00 1 4. 80 (2 .0 0– 11 .5 1) < .0 01 U nk no w n 1 0 0. 0 Fe ed in g ty pe E xc lu si ve b re as tf ee di ng 70 5 13 1. 8 (1 .0 –3 .2 ) .9 46 R ef er en ce R ef er en ce M ix ed fe ed in g 24 0 5 2. 1 (0 .7 –4 .9 ) 1. 04 (0 .3 6– 2. 95 ) 0. 94 8 1. 02 (0 .3 2– 3. 24 ) .9 75 Fo rm ul a fe ed in g 33 3 8 2. 4 (1 .0 –4 .7 ) 1. 39 (0 .5 7– 3. 40 ) 0. 47 5 1. 27 (0 .4 8– 3. 35 ) .6 34 U nk no w n 43 1 2. 3 (0 .1 –1 2. 3) 1. 30 (0 .1 6– 10 .3 4) 0. 80 2 0. 57 (0 .0 5– 5. 96 ) .6 35 Pr em at ur ity c N o 10 81 23 2. 1 (1 .4 –3 .2 ) .6 48 R ef er en ce Ye s 24 0 4 1. 7 (0 .5 –4 .3 ) 0. 76 (0 .2 6– 2. 25 ) 0. 61 8 B irt hw ei gh td N or m al 97 0 21 2. 2 (1 .3 –3 .3 ) .4 57 R ef er en ce Lo w 28 0 6 2. 1 (0 .8 –4 .7 ) 1. 01 (0 .4 0– 2. 54 ) 0. 98 6 U nk no w n 71 0 0. 0 Va cc in at io ne Fu ll co ve ra ge 96 2 16 1. 7 (1 .0 –2 .7 ) .2 53 0. 49 (0 .2 1– 1. 13 ) 0. 09 5 0. 20 (0 .0 6– 0. 66 ) 00 8 N o fu ll co ve ra ge 30 7 9 2. 9 (1 .3 –5 .6 ) R ef er en ce R ef er en ce U nk no w n 52 2 3. 9 (0 .5 –1 3. 2) 1. 35 (0 .2 8– 6. 54 ) 0. 71 3 0. 62 (0 .1 0– 4. 02 ) .6 17 A dm is si on d ia gn os is f N on -r es pi ra to ry 63 0 15 2. 4 (1 .3 –3 .9 ) .6 63 R ef er en ce R es pi ra to ry 68 2 12 1. 76 (0 .9 –3 .1 ) 0. 64 (0 .3 0– 1. 40 ) 0. 26 4 U nk no w n 9 0 0. 0 D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 650 • JPIDS 2023:12 (December) • BRIEF REPORT malnutrition was identified as 1 of 2 most common underlying conditions among infant deaths [12]. Our study had a number of limitations. First, maternal HIV status was determined by numerous methods, including self-re- port for mothers living with HIV. It is possible that there was misclassification of infant’s exposure status. Second, we did not determine whether the infant died subsequent to hospital dis- charge. Third, data were not collected for maternal character- istics that have been associated with increased infant mortality risk such as maternal health, vital status, and age and there may be residual confounding. Fourth, we may have been under- powered to detect an association between HIV exposure and mortality. Strengthening efforts to further decrease vertical HIV trans- mission, such as early and repeat HIV testing of pregnant women and maintenance of maternal viral suppression, remain essential to prevent infant HIV infection. In addition, ensuring HI infants are diagnosed and started on treatment early, and that infants are well nourished and fully vaccinated will help to further reduce infant mortality. Supplementary Data Supplementary materials are available at the Journal of The Pediatric Infectious Diseases Society online (http://jpids.oxfordjournals.org). Notes Financial support. This work was supported by a research cooperative agreement with the United States of America Centers for Disease Control and Prevention (US CDC) [grant number 5U01IP001048] Potential conflicts of interest. C. C. has received grant support from Sanofi Pasteur, the Bill and Melinda Gates Foundation, US CDC, South African Medical Research Council and Wellcome Trust. A. v. G. and N. W. have received grant funding from the US CDC, the Bill and Melinda Gates Foundation and Sanofi. J. M. has received grant funding from Sanofi Pasteur. C. V. M. has received grant funding from Pfizer. All other authors report no potential conflicts. Author contribution: All authors approved the final manuscript as sub- mitted and agree to be accountable for all aspects of the work. Disclaimer: The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of their affiliated institu- tions or the agencies funding the study. Acknowledgements We acknowledge the following site investigators: Dr Omphile Mekgoe (Department of Paediatrics, Klerksdorp Hospital, Klerksdorp, South Africa), Dr Sumayya Haffejee (Department of Paediatrics, Matikwana Hospital, Hazyview, South Africa), Dr Fathima Naby (Department of Paediatrics, Pietermaritzburg Metropolitan Hospital, Pietermaritzburg, South Africa) and Dr Godfrey Siwele (Department of Medicine, Matikwana Hospital, Hazyview, South Africa), for the support they provided for the surveillance teams during participant enrollment. REFERENCES 1. Cohen C, Walaza S, Moyes J, et al. Epidemiology of viral-associated acute lower respiratory tract infection among children <5 years of age in a high HIV preva- lence setting, South Africa, 2009-2012. Pediatr Infect Dis J 2015; 34:66–72. 2. Evans C, Jones CE, Prendergast AJ. HIV-exposed, uninfected infants: new global challenges in the era of paediatric HIV elimination. Lancet Infect Dis 2016; 16:e92–e107.Va ria bl e C at eg or ie s To ta l, N o. N = 1 32 1 N o. o f de at hs N = 2 7 M or ta lit y ra tio , % (9 5% C I) P- va lu e g O dd s ra tio h (9 5% C I) P- va lu e A dj us te d O R i ( 95 % C I) P- va lu e M ot he rs /c ar eg iv er s ed uc at io n le ve l N on e/ Pr im ar y 61 6 18 2. 9 (1 .7 –4 .6 ) .1 03 R ef er en ce R ef er en ce S ec on da ry /T er tia ry 69 2 9 1. 3 (0 .6 –2 .5 ) 0. 61 (0 .2 6– 1. 41 ) 0. 24 4 0. 77 (0 .3 1– 1. 89 ) .5 70 U nk no w n 13 0 0. 0 a U nd er ly in g co nd iti on in cl ud es a ny o f t he fo llo w in g: a st hm a, c hr on ic lu ng , h ea rt , l iv er o r r en al d is ea se , s tr ok e, s in us iti s, o rg an tr an sp la nt , a ne m ia , i m m un os up pr es si ve th er ap y, s pl en ec to m y, d ia be te s, b ur ns im m un og lo bu lin d efi ci en cy , a ut oi m m un e di se as e, ne ph ro tic s yn dr om e, c an ce r, sp in al c or d in ju ry , s ei zu re d is or de r, ce re br al p al sy , c on ge ni ta l h ea rt d is ea se , o th er c on ge ni ta l d is or de r, ob es ity , o r ch ro ni c ga st ro in te st in al p ro bl em s. b M al nu tr iti on d efi ne d as a w ei gh t- fo r-a ge < − 2 st an da rd d ev ia tio ns f ro m t he W H O m ea n Z- sc or e. c P re m at ur ity d efi ne d as g es ta tio na l a ge a t bi rt h of < 37 w ee ks . d L ow in fa nt b irt hw ei gh t de fin ed a s < 25 00 g . e V ac ci na tio n de fin ed a s fu ll va cc in e co ve ra ge fo r ag e, u si ng t he H ae m op hi lu s in flu en za e ty pe b v ac ci ne g iv en a s pa rt o f th e ro ut in e in fa nt im m un iz at io n sc he du le a t 6, 10 , 1 4 w ee ks a s a pr ox y. f A dm is si on d ia gn os is : r es pi ra to ry d ia gn os is in cl ud es a pn ea , n eo na ta l s ep si s, b ro nc hi ol iti s, p ne um on ia , t ub er cu lo si s, a nd b ro nc hi tis a nd n on -r es pi ra to ry d ia gn os is in cl ud es e nc ep ha lit is , v ira l i lln es s, d ia rr he a, fe br ile s ei zu re s, m en in gi tis , s ep si s (n on -n eo na ta l), an d ot he r di ag no si s. g C hi -s qu ar ed t es t P- va lu e. h U ni va ria te m ix ed e ffe ct s re gr es si on m od el , a cc ou nt in g fo r cl us te rin g by s ite . i M ul tiv ar ia bl e m ix ed e ffe ct s re gr es si on m od el , a cc ou nt in g fo r cl us te rin g by s ite . Ta bl e 2. C on tin ue d D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 http://jpids.oxfordjournals.org BRIEF REPORT • JPIDS 2023:12 (December) • 651 3. Woldesenbet S, Kufa-Chakezha T, Lombard C, et al. Recent HIV infection among pregnant women in the 2017 antenatal sentinel cross-sectional survey, South Africa: assay-based incidence measurement. PLoS One 2021; 16:e0249953. 4. Burton R, Giddy J, Stinson K. Prevention of mother-to-child transmission in South Africa: an ever-changing landscape. Obstetr Med 2015; 8:5–12. 5. Brennan AT, Bonawitz R, Gill CJ, et al. A meta-analysis assessing all-cause mor- tality in HIV-exposed uninfected compared with HIV-unexposed uninfected in- fants and children. AIDS 2016; 30:2351–60. 6. Goga AE, Dinh T-H, Jackson DJ, et al; South Africa PMTCT Evaluation (SAPMCTE) Team. Population-level effectiveness of PMTCT Option A on early mother-to-child (MTCT) transmission of HIV in South Africa: implications for eliminating MTCT. J Glob Health 2016; 6:020405. 7. McMorrow ML, Tempia S, Walaza S, et al. The role of human immunodeficiency virus in influenza- and respiratory syncytial virus-associated hospitalizations in South African children, 2011-2016. Clin Infect Dis 2019; 68:773–80. 8. Cohen C, Moyes J, Tempia S, et al. Epidemiology of acute lower respiratory tract infection in HIV exposed uninfected infants. Pediatrics 2016; 137:e20153272. 9. Slogrove AL, Esser MM, Cotton MF, et al. A prospective cohort study of common childhood infections in South African HIV-exposed uninfected and HIV- unexposed infants. Pediatr Infect Dis J 2017; 36:e38–44. 10. Anderson K, Kalk E, Madlala HP, et al. Increased infectious-cause hospitalization among infants who are HIV-exposed uninfected compared with HIV-unexposed. AIDS 2021; 35:2327–39. 11. Johnson L, Dorrington R. Modelling the impact of HIV in South Africa’s prov- inces. 2022:1–134. https://www.thembisa.org/downloads. Accessed September 26, 2023. 12. Breiman RF, Blau DM, Mutevedzi P, et al. Postmortem investigations and identi- fication of multiple causes of child deaths: an analysis of findings from the Child Health and Mortality Prevention Surveillance (CHAMPS) network. PLoS Med 2021; 18:e1003814–19. D ow nloaded from https://academ ic.oup.com /jpids/article/12/12/646/7414050 by W itw atersrand H ealth Sciences Library user on 23 January 2024 https://www.thembisa.org/downloads