Paroxetine

Study	Type of data	Exposure measurement	Outcome assessment	Adjustment
Alwan, 2007	case control	Information on exposure to SSRIs and other potential risk factors during pregnancy were collected by standardized telephone interviews with mothers of case and control infants conducted in English or Spanish.	Case infants were ascertained through population-based birth-defects surveillance systems in eight U.S. states. Controls were selected randomly from the same geographic areas. Information on the case infants was reviewed by clinical geneticists unaware of the infants’ exposure status	Odds ratios are adjusted for maternal race or ethnic group, presence or absence of maternal obesity, presence or absence of maternal smoking, and family income. Infants whose mothers had prepregnancy type 1 or 2 diabetes mellitus are excluded.
Ames, 2021	case control	Maternal use of SSRIs during pregnancy were ascertained from all participants in three ways: self-report in a telephone interview shortly after study enrollment (SEED Caregiver Interview), self-report on the SEED maternal medical history form, and abstraction from prenatal medical records.	Children completed a multistage process. 1) Mother (mainly) completed the Social Communication Questionnaire. 2) Gold standard clinical assessments: Autism Diagnostic Observation Schedule, Autism Diagnostic Interview Revised, Mullen Scales of Early Learning and Vineland Adaptive Behavior Scales.	Adjusted ORs were adjusted by the following variables: maternal age (continuous), maternal race, maternal education, family income, and smoking history.
Andersen, 2014	population based cohort retrospective	Information on use of prescription medication was collected from the National Prescription Register (the Register of Medicinal Product Statistics), that contains individual-level data on all prescribed drugs dispensed at all pharmacies in Denmark.	The National Hospital Register was used to identified all registered cases of miscarriage (O021 and O03 according to the International Classification of Diseases, 10th Danish Revision) and induced abortion (ICD, 10th Danish Revision codes O04, O05 and O06).	Model adjusted for maternal age, number of previous miscarriages, income, year of outcome or censoring and educational length.
Anderson, 2020	case control	Information on exposure to SSRIs and other potential risk factors during pregnancy were collected by standardized telephone interviews with mothers of case and control infants, conducted 6 weeks to 24 months after the EDD.	Case infants were ascertained through population-based birth-defects surveillance systems in 10 U.S. states. Controls were selected randomly from the same geographic areas. Clinical data were abstracted from medical records and classified by clinician geneticists and other clinicians.	Adjusted for maternal race/ ethnicity, prepregnancy body mass index, education, and early pregnancy smoking and alcohol use
Bakker, 2010	case control	Information regarding medications dispensed before and during pregnancy is obtained from community pharmacies which keep complete records of dispersed medications. The use of the prescribed medications and the use of over-the-counter medication is verified in a telephone interview with the mother.	The registry is notified of infants and fetuses with a congenital malformation by physicians and midwifes on a voluntary basis. Reports are actively collected from obstetric, pediatric, pathology, cytogenetic departments. Data on malformations is obtained from the medical files and is coded (ICD).	Adjusted for year of birth.
Ban (Controls unexposed, disease free), 2014	retrospective cohort (claims database)	The Health Improvement Network (THIN), a nationally representative database that contains drug prescriptions.	All diagnoses of major congenital anomalies (MCAs) were identified in the children’s medical records using Read codes that we classified into 14 system-specific groups according to the European Surveillance of Congenital Anomalies (EUROCAT) subgroups.	Multivariable analyses were used to adjust for maternal age at the end of pregnancy, year of childbirth, Townsend deprivation quintile, maternal smoking history, body mass index before pregnancy, and maternal diabetes, hypertension, asthma, and epilepsy in the year before conception or during pregnancy.
Ban (Controls unexposed, sick), 2014	retrospective cohort (claims database)	The Health Improvement Network (THIN), a nationally representative database that contains drug prescriptions.	All diagnoses of major congenital anomalies (MCAs) were identified in the children’s medical records using Read codes that we classified into 14 system-specific groups according to the European Surveillance of Congenital Anomalies (EUROCAT) subgroups.	Multivariable analyses were used to adjust for maternal age at the end of pregnancy, year of childbirth, Townsend deprivation quintile, maternal smoking history, body mass index before pregnancy, and maternal diabetes, hypertension, asthma, and epilepsy in the year before conception or during pregnancy.
Bérard, 2017	retrospective cohort (claims database)	Prescription fillings dispensed to women identified in the cohort from the Quebec public prescription drug insurance database.	Major congenital malformations were identified in the Régie de l’assurance maladie du Québec (RAMQ) and the Quebec hospitalization archives (MedEcho) databases defined according to International Classification of Diseases (9th-10th).	Adjusted for maternal age, welfare status, diabetes, hypertension, asthma and other medication uses including benzodiazepines as well as healthcare usage in the year prior and during the first trimester.
Bérard, 2016	retrospective cohort (claims database)	The Public Prescription Drug Insurance database of Québec (drug name, start date, dose, and duration). Data on prescription filling for AD were validated against medical records and maternal reports.	The medical service databases: The Régie de l’assurance maladie du Québec (RAMQ): diagnoses, medical procedures, prescribers, and socioeconomic status of women and children), the Québec centralized hospitalization archives (MedEcho) and the Québec Statistics database.	Adjusted for use of ADs 1 year before the first day of gestation, use of ADs in the first trimester, infant characteristics (sex, year of birth), and maternal variables (maternal age at first day of gestation, high school completed, recipient of social assistance, living alone, chronic or gestational hypertension, chronic or gestational diabetes, and other psychiatric disorders).
Brown, 2017	retrospective cohort (claims database)	Ontario Drug Benefit database.	Canadian Institutes of Health Information Discharge Abstract Database and Ontario Health Insurance Plan database. Autism spectrum disorder was defined as 2 or more outpatient diagnoses (by pediatrician or psychiatrist), 1 or more diagnoses in hospital databases after the age of 2 years, or both.	Use of a high-dimensional propensity score (HDPS) including 500 covariates.
Calderon-Margalit, 2009	prospective cohort	Participants were interviewed during a prenatal visit prior to 20 weeks of gestation by trained research personnel using a structured questionnaire. Data on medications used during pregnancy were retrieved from both questionnaires and medical records.	Information on pregnancy outcome was ascertained by reviewing hospital labor and delivery medical records and clinic records after delivery.	Adjustments for maternal age, race, years of education, marital status, smoking duringpregnancy, preeclampsia, parity, and singleton/multiple pregnancy.
Chambers, 2006	nested case control	Trained study nurses who were unaware of the study hypothesis interviewed all the mothers. The telephone interview was detailed and structured, and included questions on the use of all medications (prescription and over-the-counter) from 2 months before conception to the end of the pregnancy.	Admission and discharge records from major referral hospitals and clinics were reviewed, logbooks from neonatal intensive care units were examined, and weekly telephone calls were made to collaborators at newborn nurseries in community hospitals.	Infants who were matched with patients according to the hospital in which they were born and their date of birth. No adjustment for this group pf exposure.
Chan (Controls exposed to TCA), 2024	retrospective cohort (claims database)	Study data were obtained from the Clinical Data Analysis and Reporting System, a database that contains prescribing/dispensing records.	Diagnoses of malformations were determined by pediatricians, and were ascertained using the ICD-9-CM codes from specialist outpatient and inpatient records.	Exclusion of abnormalities due to maternal infection or exposure to known teratogens. No adjustment for this group of comparison.
Chan (Controls unexposed, pop general), 2024	retrospective cohort (claims database)	Study data were obtained from the Clinical Data Analysis and Reporting System, a database that contains prescribing/dispensing records.	Diagnoses of malformations were determined by pediatricians, and were ascertained using the ICD-9-CM codes from specialist outpatient and inpatient records.	Exclusion of malfo due to maternal infection or exposure to known teratogens. Adjusted for age, parity, maternal pre-existing diseases (diabetes, hypertension, epilepsy...), gestational diabetes, hypertension, placental abnormalities, caesarean delivery, preterm delivery, maternal psychiatric disorders, substance/alcohol use disorders, medications (suspected teratogens, anticonvulsivant ...)...
Colvin, 2011	retrospective cohort (claims database)	The national Pharmaceutical Benefits Scheme (PBS), including dispensing in the community, private hospitals, and, since late 2004, public hospitals.	The Western Australia Data Linkage System (WADLS): Hospital Morbidity Data System, the Midwives’ Notification System, the Registry of Births and Deaths, and the Birth Defects Registry.	Preterm adjusted for previous preterm birth, smoke, Socio-Economic Indexes for Areas, parity, maternal age.
Colvin, 2012	retrospective cohort (claims database)	The national Pharmaceutical Benefits Scheme (PBS), a claims database that includes 80% of all prescriptions dispensed in Australia.	The Western Australia Data Linkage System (WADLS), which contains data from the Hospital Morbidity Data System, the Midwives’ Notification System, the WA Registry of Births and Deaths and the WA Register of Developmental Anomalies. The ICD-10-AM is used.	None for these outcomes.
Costei (Controls unexposed, NOS), 2002	prospective cohort	At the time of counseling (during pregnancy), detailed exposure data and information on all other drugs used concomitantly were collected by maternal interview.	After delivery, participants were contacted again to record the pregnancy outcome and neonatal complications, including withdrawal symptoms and breastfeeding.	Controls were matched for maternal age, gravity, parity, social drug use (alcohol and smoking), and nonteratogenic drug use (eg, acetaminophen, vitamins, and calcium supplements). Pregnant women who received other drugs known to cause withdrawal-type symptoms, such as opioids or benzodiazepines, were excluded.
Costei (Controls unexposed, sick), 2002	prospective cohort	At the time of counseling (during pregnancy), detailed exposure data and information on all other drugs used concomitantly were collected by maternal interview.	After delivery, participants were contacted again to record the pregnancy outcome and neonatal complications, including withdrawal symptoms and breastfeeding.	Controls were matched for maternal age, gravity, parity, social drug use (alcohol and smoking), and nonteratogenic drug use (eg, acetaminophen, vitamins, and calcium supplements). Pregnant women who received other drugs known to cause withdrawal-type symptoms, such as opioids or benzodiazepines, were excluded.
Dandjinou, 2019	nested case control	The Quebec Prescription Drug Insurance Database (drug name, start date, dosage and duration).	The medical service database (RAMQ: diagnoses and medical procedures), the Hospitalisation Archive Database (MedEcho: in-hospital diagnoses and procedures) and the Quebec Statistics Database (Institut de la statistique du Québec (ISQ):patient sociodemographic information).	Matched for gestational age and year of pregnancy. Adjusted for maternal age, area of residence, receipt of social assistance during pregnancy, physician-based diagnoses or filled prescriptions of medications for chronic comorbidities; physician-based diagnoses of maternal diseases; medication use other than antidepressants; history of antidepressant use and health service utilisation.
Davis, 2007	retrospective cohort (claims database)	Information on prescribed antidepressant medications was derived from the pharmacy database files available at each health system.	Information on patient age, health plan enrollment status, and inpatient and outpatient diagnoses and procedures was obtained from automated databases at each health system. Assessment of outcomes was limited to assigned ICD9 codes except for 3 malfo where medical charts were reviewed.	None.
De Jonge, 2013	case control	For controls: from the IADB, a population-based prescription database. For cases: the pharmacy records for 3 months before conception until delivery were obtained and verified in a telephone interview and only the actually used medication is registered.	Information on congenital malformations is obtained from the medical files, including pathology reports, and coded afterwards, according to the International Classification of Diseases (ICD) coding system by trained registry staff.	None
De Vera, 2012	nested case control	The Quebec’s Public Prescription Drug Insurance Plan.	Linkage of three administrative databases: (i) Régie de l’Assurance Maladie du Québec (RAMQ), (ii) MED-ECHO and (iii) Institut de la Statistique du Québec (ISQ).	Cases and controls matched on gestational age. Multivariable models were adjusted for sociodemographic variables, depression, anxiety, other comorbid medical conditions, medication use and health care utilization.
Diav-Citrin, 2008	prospective cohort	Details of exposure were collected at the initial contact with the TIS and before pregnancy outcome was known using a structured questionnaire. SSRIs and other exposures were also ascertained after delivery.	Follow-up was con- ducted by a telephone interview or mailed questionnaire to the woman or the child’s paediatrician to obtain details of the pregnancy outcome, gestational age at delivery, birth weight, congenital anomalies and neonatal complications.	Adjusted for gestational age at call, maternal age, smoking, TIS origine, concomitants psychiatric medication, multi fetal gestation, SSRI dose.
Dubnov-Raz, 2008	prospective cohort	Not specified.	For the neontates outcomes their medical charts and ECG tracings were extracted. For the purpose of this study, ECGs were interpreted by a single experienced pediatric cardiologist (Dr Fogelman), who was blinded to drug exposure.	Controls matched on gestational age. Exclusion of neonates born to women treated with any other chronic medication during pregnancy (whether known as QT prolonging or not), with gestational diabetes or hypothyroidism, with Apgar scores <7 (either 1 or 5 minutes) and those with cardiac structural abnormalities identified by echocardiography.
Einarson, 2008	prospective cohort	During the initial telephone contact, details of exposure and concurrent exposures are recorded on a standardized questionnaire form. Details about the exposure include duration, timing in pregnancy, dose, frequency, and medical indication for use of the drug.	At the follow-up interview, gestational findings and fetal outcomes are documented on a structured, standardized form by telephone interview. With the mother’s permission, this report is corroborated with the report of the physician caring for the baby.	None.
Einarson, 2009	prospective cohort	During the initial telephone contact, details of exposure and concurrent exposures are recorded on a standardized questionnaire.	At the follow-up interview, gestational findings, fetal outcomes, and neonatal health are documented on a structured form by telephone interview with each mother. The details are then corroborated with the report of the physician caring for the baby.	The 2 whole groups were matched for maternal age, smoking, and alcohol use. But no matching for individual substance.
Furu, 2015	population based cohort retrospective	The Nordic prescription registers include data on dispensed drugs, substance, brand name, and formulation, together with date of dispensing.	From the medical birth, patient, and malformation registers data on maternal characteristics, the pregnancy and delivery, and major birth defects were retrieved.	Adjusted for maternal age, year of birth, birth order, smoking, maternal diabetes, country, and use of other prescribed drugs (antiepileptics (atC code n03), anxiolytics and hypnotics (n05b and n05C), and angiotensin converting enzyme inhibitors (C09)).
Heuvelman, 2023	retrospective cohort (claims database)	The Clinical Practice Research Datalink (CPRD) contains an extensive code list to identify the name, formulation and dose of medications, which are mandatory fields in the prescription electronic record (according to protocol).	For child outcomes, the primary care clinical and referral records were examined for presence of disorders based on Read codes and for ADHD: prescription of ADHD medication (methylphenidate, dexamphetamine, atomoxetine, dextroamphetamine, amphetamine with dexamphetamine, or lisdexamphetamine).	Adjusted for maternal age, Charlson Comorbidity Index score, maternal disorders (alcohol-related, psychosis, anxiety, self-harm, bipolar disorder, eating disorders, personality disorders, sleep disorders and neuropathic pains), medications (for physical health problems, central nervous system agents, multiple antidepressants ...) smoking, any recorded severity of past depressive symptoms...
Huybrechts (Controls unexposed, NOS), 2014	cohort	The Medicaid Analytic eXtract data set that contains data on all filled outpatient medication prescriptions.	The Medicaid Analytic eXtract data set that contains data on all physician services and hospitalizations and the accompanying diagnoses and procedures. Congenital cardiac malformations were identified on the basis of International Classification of Diseases (ICD-9) codes.	None
Huybrechts (Controls unexposed, sick), 2014	cohort	The Medicaid Analytic eXtract data set that contains data on all filled outpatient medication prescriptions.	The Medicaid Analytic eXtract data set that contains data on all physician services and hospitalizations and the accompanying diagnoses and procedures. Congenital cardiac malformations were identified on the basis of International Classification of Diseases (ICD-9) codes.	Adjusted for the high-dimensional propensity score: including year of delivery, age, race, multiple gestation, antidepressant indications, proxies for depression severity, other chronic maternal illness, other psychotropic medication use, antidiabetic, antihypertensive and residual confounding for proxies of unmeasured confounders.
Jimenez-Solem (Controls unexposed, NOS), 2012	population based cohort retrospective	The drug redemptions were identified using the Register of Medicinal Product Statistics which has recorded drugs dispensed from Danish pharmacies.	Congenital malformations were identified through the Danish National Hospital Register.	Multivariable logistic regressions are adjusted for mother’s age, parity, income, education, smoking and year of conception.
Jimenez-Solem (Controls unexposed, sick), 2012	population based cohort retrospective	The drug redemptions were identified using the Register of Medicinal Product Statistics which has recorded drugs dispensed from Danish pharmacies.	Congenital malformations were identified through the Danish National Hospital Register.	No adjustment for this group of comparison.
Jordan, 2016	retrospective cohort (registry)	Anomalies registries were linked with prescription and healthcare databases covering their source populations (Danish national Prescription and Patient register; Norway National Prescription Database; and Wales’ health and social care linked electronic databank).	Three congenital anomalies registries that contribute to EUROCAT that routinely collected data on congenital anomalies. Major congenital anomalies were classified according to the EUROCAT standard subgroups (with correspondance with ICD10).	No adjustment for this exposure.
Kerr, 2018	case control	Within 6 months of delivery, trained Birth Defects Study (BDS) nurse- interviewers contacted mothers to complete a roughly 1-hr computer-assisted telephone interview, including medications during pregnancy.	Cases and controls were ascertained at participating hospitals or birth defect registries in the same areas.	Adjusted models included maternal age, race/ethnicity, education, study center, and study year (aORs were calculated when there were five or more exposed cases).
Kieler, 2012	population based cohort retrospective	The prescription registers.	from the medical birth registers and th cause of death registers, infants with persistent pulmonary hypertension of the newborn identified as an ICD-10 code P29.3 or I27.0.	Adjusted for maternal age, dispensed non-steroidal anti-inflammatory drugs and antidiabetes drugs, pre-eclampsia, chronic diseases during pregnancy, country of birth, birth year, level of delivery hospital, and birth order.
Kitchin, 2022	case control	Database for Pharmacoepidemiological Research in Primary Care, a computerized medical longitudinal population database of electronic medical records from 10.153 primary care practitioners and pediatricians distributed on nine Autonomous Regions (out of 17), which contains prescriptions.	Database for Pharmacoepidemiological Research in Primary Care, a computerized medical longitudinal population database of electronic medical records from 10.153 primary care practitioners and pediatricians, which contains medical diagnoses, medical visits, hospital admissions.	Controls individually matched to cases (maternal age, gestational age, and year of Last Menstrual Period date). Adjusted by number of GP visits, obesity, smoking, HTA, diabetes.
Kivistö, 2016	retrospective cohort	The data were gathered retrospectively from the hospital birth register.	The data were gathered retrospectively from the hospital birth register.	A regression model adjusted for age, overweight, nulliparity, prior termination, miscarriages, smoking, maternal alcohol consumption, chronic illness, gestational diabetes, and polyhydramnion.
Lee (Controls exposed to TCAs), 2025	retrospective cohort (claims database)	The Clinical Data Analysis and Reporting System (CDARS) that captures and links all clinical data, including prescribing and dispensing records.	The Clinical Data Analysis and Reporting System (CDARS) that captures and links all clinical data, including clinical information like diagnoses, clinic attendances, hospital admissions.	Singleton. No adjustment for this group of comparison.
Lee (Controls unexposed, general pop), 2025	retrospective cohort (claims database)	The Clinical Data Analysis and Reporting System (CDARS) that captures and links all clinical data, including prescribing and dispensing records.	The Clinical Data Analysis and Reporting System (CDARS) that captures and links all clinical data, including clinical information like diagnoses, clinic attendances, hospital admissions.	Singleton. Adjusted for age, parity, maternal diabetes, hypertension, dyslipidaemia, epilepsy, physical comorbidity burden, gestational diabetes and hypertensive disorders, maternal psychiatric disorders, substance/alcohol use disorders, drugs during pregnancy (antipsychotics, lithium, valproate, lamotrigine, carbamazepine, benzodiazepines, z-hypnotics, opioid), history of psychiatric admission...
Levinson-Castiel, 2006	retrospective cohort	The infants were identified from the delivery room records as they arrived at the nursery or from a medical history form completed by all mothers at admittance to the nursery. This form included notably type, dosage, and duration of treatment with SSRIs or other drugs.	The nurses and physicians in the center identified and recorded NAS symptoms. At discharge, the medical records were reviewed, and the Finnegan scores were assessed for possible errors.	No matching for this group of exposure.
Liu, 2017	population based cohort retrospective	Information on antidepressant use came from the Danish National Prescription Registry, that covers all prescriptions dispensed in Denmark since 1995.	Information on psychiatric diagnosis came from the Danish Psychiatric Central Research Register, that holds information on all inpatient and outpatient psychiatric treatment in Denmark. Primary outcome was first diagnosis of psychiatric disorders (ICD-10 codes F00-F99) in the offspring.	Adjusted for maternal age, primiparity, maternal psychiatric history, inpatient and outpatient psychiatric treatment from 2 years before pregnancy, dispensing of other psychotropic or antiepileptic during pregnancy, nb of non-psychiatric hospital visits during pregnancy, smoking, place of residence, marital status, highest education, income, year of delivery, and paternal psychiatric history.
Louik, 2007	case control	Detailed data are collected on all medications (prescription, over-the-counter, vitamins and minerals, and herbal products) used at any time from 2 months before conception through the end of the pregnancy, by mother completion of a 45-to-60-minute interview.	Research staff identify subjects by reviewing clinical and surgical logs, reviewing admission and discharge lists, and contacting newborn nurseries and labor and delivery rooms. Nonmalformed infants were identified in a population-based random sample of newborns in Massachusetts.	Adjusted for maternal age, maternal race or ethnic group, maternal education, year of last menstrual period/study center, parity, first-trimester smoking status, first-trimester alcohol consumption, any family history of a birth defect, history of a cardiac defect in a first-degree relative, prepregnancy body-mass index, seizures, diabetes mellitus, hypertension, infertility, use of folic acid.
Malm, 2011	population based cohort retrospective	The Drug Reimbursement Register that contains data on 98% of reimbursed prescription drug purchases.	The Medical Birth Register and the The Register of Congenital Malformations, which data on diagnoses during pregnancy and delivery and neonatal outcome data (including major malformations). Data are collected from all maternity hospitals and include all births and stillbirths.	Independent variables considered in the adjusted logistic model were maternal age at the end of pregnancy, parity, year of pregnancy ending, marital status, smoking during pregnancy, purchase of other reimbursed psychiatric drugs (including antiepileptics) during the first trimester, and maternal prepregnancy diabetes.
Marks (Controls exposed to Bupropion), 2021	retrospective cohort	Data were obtained from electronic medical records (EMRs) through the Regenstrief Institute in Indianapolis.	Diagnoses were extracted from the electronic database discharge summaries, delivery records, and/or International Classification of Diseases (ICD) codes. Clinical diagnosis were extracted from the delivery discharge summary written by the clinician caring for the infant.	All results reported as aOR (95% CI) controlling for maternal race, age, insurance, and gestational age at delivery.
Marks (Controls unexposed, sick), 2021	retrospective cohort	Data were obtained from electronic medical records (EMRs) through the Regenstrief Institute in Indianapolis.	Diagnoses were extracted from the electronic database discharge summaries, delivery records, and/or International Classification of Diseases (ICD) codes. Clinical diagnosis were extracted from the delivery discharge summary written by the clinician caring for the infant.	All results reported as aOR (95% CI) controlling for maternal race, age, insurance, and gestational age at delivery.
Martin, 2024	population based cohort retrospective	In the UK, prescription data were based on the prescriptions written by general practitioners (CPRD GOLD), whereas in Norway and Sweden, dispensation of prescription drugs from all ambulatory pharmacies was used (Norwegian Prescription Database, and Swedish Prescribed Drug Register).	The UK Clinical Practice Research Datalink that contains diagnoses made in primary care and secondary care data; Norway: Medical Birth Registry of Norway and the Norwegian Patient Registry; and Sweden: the Medical Birth Register of Sweden and the National Patient Register.	Singletons only. Adjusted for maternal age at delivery, early-pregnancy body mass index, parity, previous stillbirth, anti-seizure medication and antipsychotic use in the 12 months prior to pregnancy, smoking anytime during pregnancy, maternal depression or anxiety diagnosis prior to the start of pregnancy, proxy measures of socioeconomic position (SEP).
Maschi, 2008	prospective cohort	Maternal demographic data, indication for treatment and time of exposure were collected using a structured questionnaire.	A follow-up interview was performed to collect information concerning pregnancy outcome, details of labour and delivery and neonatal complications, including Neonatal Intensive Care Unit (NICU)/Special Care Nursery (SCN) admissions.	Controls matched for maternal age and gravidity.
Merlob, 2009	prospective cohort	A standardized pregnancy questionnaire is administered to all women on admittance to the maternity ward and reviewed by the attending neonatologist. The use of any drug during pregnancy is routinely recorded.	Every infant born at the center during that period was examined on the first day of life for cardiac murmur. Those with a persistent murmur on the second or third day of life were examined by a pediatric cardiologist and referred for electrocardiography and echocardiography.	None.
Nakhai-Pour, 2010	nested case control	The Régie de l’assurance maladie du Québec (RAMQ) database which provides prospectively collected data on filled prescriptions.	The Régie de l’assurance maladie du Québec (RAMQ) (physician-based diagnoses according to the ICD-9), the Med-Echo database (data on acute care hospital admissions) and the Institut de la statistique du Québec database (data on all births and deaths in Quebec).	Match. Adjusted for maternal age, social assistance status and place of residence; gestational age at index date; comorbidities (diabetes mellitus, cardiac disease, asthma, untreated thyroid disease, depression, anxiety, bipolar disorder); history of abortions; visits to physicians; duration of antidepressants; prenatal visits and other medication use in the year before and during pregnancy.
Nijenhuis (Controls exposed to TCA), 2012	retrospective cohort (claims database)	The pharmacy prescription database IADB.nl which contains pharmacy prescription data of an estimated population of 500 000 individuals from the Netherlands.	The use of laxatives (ATC code = A06) and antidiarrhoeal medication (ATC code = A07C; A07D; A07F; A07X) in the newborn, regarded as a proxy for constipation and diarrhoea, respectively.	None.
Nijenhuis (Controls unexposed, NOS), 2012	retrospective cohort (claims database)	The pharmacy prescription database IADB.nl which contains pharmacy prescription data of an estimated population of 500 000 individuals from the Netherlands.	The use of laxatives (ATC code = A06) and antidiarrhoeal medication (ATC code = A07C; A07D; A07F; A07X) in the newborn, regarded as a proxy for constipation and diarrhoea, respectively.	None.
Nordeng (Controls unexposed, NOS), 2012	cohort	The pregnant women completed 2 questionnaires during pregnancy at around gestational weeks 17 and 30, which included notably questions regarding medication use.	The Medical Birth Registry of Norway (MBRN) which contains detailed medical information and diagnostics regarding the infant originating from mandatory notification forms completed by midwives, obstetricians, and/or pediatricians at delivery and during the hospital stay.	Malfo: adjusted for maternal depression, maternal age at delivery, parity, and use of psychotropic drugs during pregnancy. Preterm, LBW: adjusted for level of depression, maternal age at delivery, education, parity, prepregnancy BMI, maternal asthma or cardiovascular disease, NSAID use, folic acid use, and smoking during pregnancy.
Nordeng (Controls unexposed, sick), 2012	cohort	The pregnant women completed 2 questionnaires during pregnancy at around gestational weeks 17 and 30, which included notably questions regarding medication use.	The Medical Birth Registry of Norway (MBRN) which contains detailed medical information and diagnostics regarding the infant originating from mandatory notification forms completed by midwives, obstetricians, and/or pediatricians at delivery and during the hospital stay.	No adjustment for this group of comparison.
Oberlander, 2008	retrospective cohort (claims database)	PharmaNet, a province-wide network recording all prescriptions dispensed by British Columbian pharmacists outside hospitals.	Five administrative sources housed in the BC Linked Health Database: British Columbia registry of births, hospital separation records, the PharmaCare registry of subsidized prescriptions; the Medical Services Plan physician billing records; and the registry of Medical Services Plan subscribers.	None (adjustment provided for risk difference but not for relative risk or odd ratio).
Oberlander, 2004	prospective cohort	Measure of plasma level of maternal SSRI medications.	Concern about respiratory or other symptoms on the part of the family physician, midwife or obstetrician led to assessment by the pediatrician. Physicians who had been asked to assess distressed newborns were partially blinded to the infants' prenatal exposure status. Physician reviewed chart.	None
Ozturk, 2016	prospective cohort	At the first contact, initiated via gynecologists, a detailed patient history form was used to notably record all drug exposures (dose, duration and timing in pregnancy).	Each newborn baby was checked at birth for signs of problems or complications.	None
Palmsten (Controls exposed to TCA), 2013	retrospective cohort (claims database)	Outpatient pharmacy-dispensing data.	Medicaid enrollment information was linked to inpatient and outpatient procedures and diagnoses. Outcome validity was assessed by reviewing delivery hospital medical records for a sample of 183 women. Preeclampsia defined with ICD-9 code for preeclampsia or eclampsia (642.4x–642.7x).	No adjustment for this group of comparison.
Palmsten (Controls unexposed, sick), 2013	retrospective cohort (claims database)	Outpatient pharmacy-dispensing data.	Medicaid enrollment information was linked to inpatient and outpatient procedures and diagnoses. Outcome validity was assessed by reviewing delivery hospital medical records for a sample of 183 women. Preeclampsia defined with ICD-9 code for preeclampsia or eclampsia (642.4x–642.7x).	Adjusted for delivery year, preeclampsia risk factors, depression severity proxies, other antidepressant indications, other medications, and healthcare utilization.
Palmsten b, 2013	retrospective cohort (claims database)	Data of prescription.	Women with an ICD-9 code for 666.x during the admission to hospital for delivery, or within three days after the delivery date, were classified as having postpartum hemorrhage. Atonic postpartum hemorrhage only (666.1x) and inpatient postpartum hemorrhage only, also considered.	Adjusted for delivery year, age, race, multiple pregnancy, diabetes, coagulopathy, number of outpatient and inpatient mood/anxiety disorder diagnoses, other mental health disorder, pain indication, sleep disorder, anticonvulsant, benzodiazepine, aspirin, heparin, low molecular weight heparin and warfarin dispensing, and number of outpatient visits and days in hospital during baseline.
Rai (Controls exposed to TCA), 2017	prospective cohort	Information on maternal use of antidepressants in pregnancy is derived from the medical birth register (as reported by pregnant women at their antenatal interview, at a median of 10 weeks’ gestation) and supplemented it with the prescribed drug register (available from July 2005).	Diagnostic information of autism were collected the national patient register, the Stockholm child and adolescent mental health register, and the habilitation registers (ICD-9 (299), ICD-10 (F84), or DSM-IV (299) codes).	No adjustment for this group of comparison.
Rai (Controls unexposed, disease free), 2017	prospective cohort	Information on maternal use of antidepressants in pregnancy is derived from the medical birth register (as reported by pregnant women at their antenatal interview, at a median of 10 weeks’ gestation) and supplemented it with the prescribed drug register (available from July 2005).	Diagnostic information of autism were collected the national patient register, the Stockholm child and adolescent mental health register, and the habilitation registers (ICD-9 (299), ICD-10 (F84), or DSM-IV (299) codes).	No adjustment for this group of comparison.
Rai (Controls unexposed, sick), 2017	prospective cohort	Information on maternal use of antidepressants in pregnancy is derived from the medical birth register (as reported by pregnant women at their antenatal interview, at a median of 10 weeks’ gestation) and supplemented it with the prescribed drug register (available from July 2005).	Diagnostic information of autism were collected the national patient register, the Stockholm child and adolescent mental health register, and the habilitation registers (ICD-9 (299), ICD-10 (F84), or DSM-IV (299) codes).	Models are adjusted for birth year, maternal depression, and antidepressant polypharmacy (binary variable for use of 2 or more antidepressants).
Reis (Controls exposed to TCA), 2010	population based cohort retrospective	Information on drug use is partly based on an interview conducted by the midwife at the first antenatal visit (in 90% of cases before the end of the first trimester) and partly on information from the antenatal care with respect to drugs prescribed later during the pregnancy by the attending doctor.	The Swedish Medical Birth Register, the Register of Birth Defects (previously known as the Register of Congenital Malformations) and the Patient Register (previous the Hospital Discharge Register).	Singletons only. No adjustment for this group of comparison.
Reis (Controls unexposed, NOS), 2010	population based cohort retrospective	Information on drug use is partly based on an interview conducted by the midwife at the first antenatal visit (in 90% of cases before the end of the first trimester) and partly on information from the antenatal care with respect to drugs prescribed later during the pregnancy by the attending doctor.	The Swedish Medical Birth Register, the Register of Birth Defects (previously known as the Register of Congenital Malformations) and the Patient Register (previous the Hospital Discharge Register).	Singleton only. Adjusted for pertinent variables, always including year of delivery, maternal age, parity, smoking, and body mass index (BMI).
Stephansson, 2013	population based cohort retrospective	The prescription registries in the Nordic countries include data on the dispensed item, substance, brand name, and formulation together with date of dispensing for more than 95% of the total outpatient population.	Information on stillbirth was obtained from the medical birth registries and neonatal and postneonatal deaths were obtained from the Nordic causes of death registries. All diagnoses and causes of death are classified according to the ICD-10 codes.	Adjusted for country and year of birth, maternal age, birth order, smoking in early pregnancy, and maternal diabetes and hypertension.
Van der Veere, 2020	prospective cohort	Not specified (but prospective cohort).	Children assessed at age 2.5 years using the Bayley Scales of Infant and Toddler Development (BSID, 3rd Edition). The examiner was blinded for maternal mental state and use of SSRI during pregnancy. Parents filled out the Child Behavior Checklist (CBCL)	None (adjusted for maternal depression and anxiety and maternal education, gender, gestational age, birth weight, severity of maternal depression and anxiety, but results not extractable).
Vial, 2006	prospective cohort	Maternal data and detailed history of drug exposures were collected during the first contact for individual risk counseling.	Follow-up of pregnancies were prospectively documented.	Women were matched for gestational age at the time of request.
Viktorin (Controls unexposed, NOS), 2017	population based cohort retrospective	Dispensations identified in the Swedish Prescribed Drug Register that holds information on all dispensed prescription drugs in Sweden.	A clinically ascertained diagnosis of offspring Autism spectrum disorder (ASD) was identified in the Swedish Patient Register, with an ICD-10 code according to: F84.0, F84.1, F84.2, F84.3, F84.4, F84.5, F84.8, or F84.9.	Adjusted for birthdate, maternal and paternal age, the father's psychotropic medication overlapping the pregnancy, the mother's one-time dispensations of antidepressant and/or other psychotropic medication that overlapped the pregnancy, for any diagnosis of depression in the mother's lifetime, for any diagnosis of psychiatric disorder subgroups in either the mother and/or father's life time.
Viktorin (Controls unexposed, sick), 2017	population based cohort retrospective	Dispensations identified in the Swedish Prescribed Drug Register that holds information on all dispensed prescription drugs in Sweden.	A clinically ascertained diagnosis of offspring Autism spectrum disorder (ASD) was identified in the Swedish Patient Register, with an ICD-10 code according to: F84.0, F84.1, F84.2, F84.3, F84.4, F84.5, F84.8, or F84.9.	Adjusted for birthdate, maternal and paternal age, the father's psychotropic medication overlapping the pregnancy, the mother's one-time dispensations of antidepressant and/or other psychotropic medication that overlapped the pregnancy, for any diagnosis of depression in the mother's lifetime, for any diagnosis of psychiatric disorder subgroups in either the mother and/or father's life time.
Wemakor, 2015	case control	Medication exposure information came from maternal medical/midwifery notes, created prospectively. Other additional data sources include paediatrician records (postnatal), medical geneticist records (postnatal), GP records of mother (prenatal), and maternal interviews (postnatal).	EUROCAT registries collect data using multiple sources of information: maternity, neonatal, and paediatric records; fetal medicine, cytogenetic, pathology, and medical genetics records; paediatric cardiology services; and hospital discharge and child health records. ICD 9 or 10 classification.	Analyses were adjusted for registry in order to adjust for con- founding that may arise if registries differ in both exposure and outcome prevalences.
Yaris, 2005	prospective cohort	Data surveyed by the interviews.	Examinations were made by obstetric ultrasound for the mothers and birth weight and height, gestational age, APGAR scores, congenital malformation and developmental problems for the babies. Periodic checks of all the babies in the neonatal period, were made the first year, and following years.	None for this group of exposure.
Yazdy, 2014	case control	The telephone interview were conducted by trained nurses within 1 year after delivery. It consisted in questions notably on illnesses and medications. If a mother reported using any medications, the timing and indication for use were noted.	Diagnosis of structural clubfoot was confirmed primarily by orthopedic records (77%); when medical records were not available, maternal report of 3 or more castings for the clubfoot was used to confirm a true structural clubfoot (23%).	Adjusted for maternal smoking, alcohol use, and body mass index.

master protocol