Table 2. Characteristics of included studies.

Study	Study type	Setting and time	Ages of adolescents Ages of comparison group	Inclusion and exclusion criteria	Outcomes and definitions
Ujah et al. 2005	Patient record review	Jos University Teaching Hospital, Nigeria Jan 1985-Dec 2001	≤15 20–24	Not reported	Maternal deaths; women who died during pregnancy or childbirth in the maternity ward of the hospital
Wort et al. 2006	Cross-sectional patient record review	Kilosa Hospital, Tanzania June 2001- October 2002	≤15, n = 34 ≥20, n = 281	Inclusion: vaginal births Exclusions: Twin births, stillbirths, cesarean section. Unclear whether triplets excluded.	Low birth weight, <2500 g
Van Dillen et al. 2008	Retrospective patient record review	Onandjokwe Lutheran Hospital, Namibia February 2002 –August 2002	14–17, n = 76 ≥20, n = 371	Inclusion: Primiparous women, singleton births Exclusion: Twin births, unknown age of mother	Low birth weight, <2500 g Perinatal mortality; all deaths occurring after 28 weeks gestation with weight greater or equal to 1000 g Maternal mortality Stillbirth (macerated)
Adam et al. 2009	Cross-sectional descriptive study	Khartoum teaching hospital, Sudan October 2007 –January 2008	≤16, n = 29 20–24, n = 203	Inclusion: Primiparous women, singleton births Exclusion: > 24 years old	Preterm birth, <37 weeks of gestation Low birth weight, <2500 g
Nkwabong et al. 2009	Retrospective patient record review	Yaoundé Teaching Hospital, Cameroon January 2004 –December 2004	≤15, n = 11 16, n = 13 17, n = 21 20–25, n = 403	Inclusion: Nulliparous teenagers or women aged 20–25 years old	Early neonatal death
Adeyinka et al. 2010	Retrospective case-control study, patient record review	The University College Hospital, Ibadan, Nigeria January 2007 –November 2008	<18, n = 45 20–35, n = 90	Inclusion: Adolescents aged <18 years old or adults aged between 20 and 35	Pre-eclampsia* Eclampsia** Low birth weight, <2500 g Stillbirths
Zeck et al. 2010	Retrospective patient record review	The Kilimanjaro Christian Medical Centre, Moshi, Tanzania.	<18, n = 209 22–27, n = 1341	Inclusion: Primiparas Within the set age categories	Intrauterine fetal death LBW, <2500 g at term Preeclampsia at the time of delivery
Tebeu et al. 2011	Case-control study, patient record review	Maroua Regional Hospital, Cameroon June 2005 –May 2007	13–16, n = 53 20–34, n = 330	Inclusion: Cases had hypertensive disorders in pregnancy Exclusion: Twin gestations, chronic hypertension	Pre-eclampsia/eclampsia***
Ganchimeg et al. 2013	Secondary analysis using facility-based cross-sectional patient record data from the WHO Global Survey on Maternal and Perinatal Health	Health institutions in 24 countries in Africa, Latin America and Asia. Africa: Algeria, Angola, Democratic republic of Congo, Niger, Nigeria, Kenya and Uganda. 2004–2005 in Africa	Africa: ≤15, n = 551 20–24, n = 10242	Inclusion: Health institutions with ≥1000 deliveries per year and capable of performing caesarean sections. Nulliparous women ≤24 years of age Singleton neonate of BW ≥500 g or gestational age ≥22 weeks if BW was missing	Low birthweight, <2500 g Preterm birth; births <37 completed weeks of gestation. Perinatal death**** Maternal deaths; intra-hospital deaths occurring on or before the 8th day postpartum
Lukonga et al. 2015	Cross-sectional population based data	National household survey Zambia Demographic and Health survey (ZDHS) 2007	12–17, n = 2649 (# of births) 25–29, n = 233 (# of births)	Inclusion: Women who gave birth to live infants within 5 years preceeding the survey Available and complete records of the babies for the first 28 days Exclusion: Incomplete or missing records for the baby	Neonatal death, up to 28 days postpartum
Banda et al. 2015	Cross-sectional data	Zambia. Data from national population census 2010,	10–14 20–24		Pregnancy-related deaths: Death of a woman while pregnant or within 42 days of termination of pregnancy
Ibrahim et al. 2015	Retrospective patient record review. Purposive sampling method from birth registry folders	Tamale Teaching Hospital, Northern Ghana, data collected from 2000 and 2010	<18, n = 12 25–34, n = 761		Low birth weight, <2500 g
De Wet et al. 2016	Secondary data analysis of Death Notification Forms (DNF)	South Africa. DNFs and household surveys from 2006–2012.	10–14 20–24	Inclusion: Completed forms where the deceased was female and pregnancy status was confirmed	Maternal death: Direct maternal causes of death while pregnant Pregnancy-related deaths: All deaths while pregnant
Njim et al. 2016	Retrospective patient record review	Bueau Regional Hospital, South-West region of Cameroon Jan. 2007 –Dec. 2012	≤16, n = 78 ≥20, n = 4450	Inclusion: Complete records of women who gave birth at the hospital Exclusion: Gestational age <28 completed weeks, multiple gestations, incomplete information	Preterm delivery Low birth weight
Mombo-Ngoma et al. 2016	Prospective multinational cohort study	Data from a RCT (MiPPAD) in Benin, Gabon, Mozambique and Tanzania Sept 2009 –Dec.2013	≤16, n = 248 20–30, n = 2376	Inclusion: HIV negative women, gestational age (GA) <28 weeks at first ANC visit, willing to participate and give birth at the study health facility Exclusion: Allergy to any of the study drugs. Any other ongoing serious condition. Stillbirths, multiple gestations. Tanzania data on PTD was excluded in the paper because of a systematic error in the assessment of GA	Low birth weight, <2500 g Preterm delivey, <37 completed weeks of gestation, estimated at ANC visit with bimanual palpation, and at delivery by the Ballard score
Moodley et al. 2016	Cross-sectional patient record review	A regional hospital in Durban, South Africa. July–Dec. 2011 and Jan.–June 2014	<18, n = 827 25–34, n = 3662	Inclusion: Women with viable pregnancies delivering neonate >_500 g with recorded birth outcomes Exclusion: Multiple births	Premature births, <37 completed weeks of gestation Low birth weight, <2500 g Stillbirth Small for gestational age (gestational age at delivery based on mothers last normal menstrual period, or ultrasound, or a combination of both)
Mombo-Ngoma et al. 2017	Analysis of two prospective cohort studies (MiPPAD and IDEA)	Two health institutions in Lambaréne and Fougamou, Gabon Sept 2009 –Nov. 2013	≤16, n = 66 20–30, n = 587	Inclusion: Participation in MiPPAD or IDEA trials Exclusion: Missing delivery data, miscarriage or stillbirth, multiple gestations	Low birth weight, <2500 g
Tessema et al. 2017	Secondary data from Global burden of Disease study from 2013.	Ethiopia, 1990–2013. Various data sources; sibling stories, censuses, maternal mortality surveillance and verbal autopsy analysis.	10–14 20–24 25–29		Maternal death: Death of a woman while pregnant or within 42 days of termination of pregnancy

* Hypertension (140/90 mmHg on two occasions 4h apart) and proteinuria (0,3g/dl) in the second half of pregnancy

** Associated with convulsions, oliguria (4400ml/24h), increased tendon reflex, pain in the right hypochondriac region

*** Women with a diastolic blood pressure of at least 90 mmHg or a systolic BP of at least 140 mmHg were considered to have hypertensive disorder in pregnancy

**** Perinatal deaths included fresh and macerated stillbirths and early neonatal deaths, defined as the inta-hospital death of a liveborn neonate during the first 7 days after delivery or earlier if the discharge occurred before 7 days