- Medication and Pregnancy KB

Study	Type of data	Exposure measurement	Outcome assessment	Adjustment
Cooper (Controls exposed to other treatment, sick), 2014	retrospective cohort (claims database)	The Tennessee Medicaid and Kaiser Permanente Northern California and Southern California claims databases.	Outcomes were identified from vital records and medical records: birth certificate check-boxes, infant claims data, fetal death or death certificates, records of maternal hospitalization. Medical records and autopsy reports were reviewed to confirm all adverse fetal outcomes.	No adjustment for this group of comparison.
Cooper (Controls unexposed, sick), 2014	retrospective cohort (claims database)	The Tennessee Medicaid and Kaiser Permanente Northern California and Southern California claims databases.	Outcomes were identified from vital records and medical records: birth certificate check-boxes, infant claims data, fetal death or death certificates, records of maternal hospitalization. Medical records and autopsy reports were reviewed to confirm all adverse fetal outcomes.	Propensity score including maternal age, race, ethnicity, years of education, parity, chronic health diagnoses (smoking, diabetes, epilepsy, asthma, renal disease, cancer, heart disease, hypertension, HIV, substance use, obesity, headaches), medications for chronic diseases, chronic immune mediated diseases... Singletons only. Exclusion of births with exposure to teratogenic medications.
Dawson, 2014	case control	Computer-assisted telephone interview which included detailed questions concerning exposures that occurred from 3 months prior to conception through the end of the pregnancy. Mothers were interviewed between 6 weeks and 24 months after their EDD.	A clinical geneticist and a pediatrician trained in cardiology review the abstracted clinical records for each case infant. Clinical records for control infants were not available for review.	None.
Howren, 2020	retrospective cohort	PharmaNet captures complete information on all drug prescriptions dispensed including drug identification number, dispensation date, dispensation quantity, dosage, and duration.	SGA was identified using ICD 9/10 codes in either the Medical Services Plan database (MSP), the Discharge Abstract Database (DAD), or in the BC Perinatal Database Registry (BCPDR) with data from obstetrical and neonatal medical records. Congenital anomalies were identified using the BCPDR.	Multivariate model. Potential confounders considered: maternal and pregnancy characteristics (BMI, age, prior. adverse pregnancy outcome...), maternal comorbidities (anxiety, depression, diabetes), other medication use before pregnancy and during pregnancy, and healthcare utilisation (considered markers of disease severity and health status).
Viktil (Controls exposed to other treatment, sick), 2012	population based cohort retrospective	The Norwegian Prescription Database (NorPD). All prescribed drugs dispensed from Norwegian pharmacies were recorded from 3 months prior to conception until labour and for men during 3 months prior to conception.	The outcomes of the births were recorded, and malformations classified according to the European Surveillance of Congenital Anomalies (EUROCAT) guidelines.	None for this group of exposure.
Viktil (Controls unexposed NOS), 2012	population based cohort retrospective	The Norwegian Prescription Database (NorPD). All prescribed drugs dispensed from Norwegian pharmacies were recorded from 3 months prior to conception until labour and for men during 3 months prior to conception.	The outcomes of the births were recorded, and malformations classified according to the European Surveillance of Congenital Anomalies (EUROCAT) guidelines.	None for this group of exposure.
Vinet, 2013	nested case control	Exposure was defined as having filled at least 1 prescription for methotrexate as recorded in the RAMQ prescription database.	Either one procedure code and/or a diagnostic code for induced abortion present in the MED_ECHO hospitalization and/ or RAMQ billing databases.	Age-matched controls. Adjusted for disease duration.
Weber-Schoendorfer (Controls exposed to other treatment, sick), 2014	prospective cohort	Data on exposure were obtained via structured telephone interviews and/or mailed questionnaires completed by the mother and/or her physicians.	Data on outcome were obtained via structured telephone interviews and/or mailed questionnaires completed by the mother and/or her physicians. Birth defects were classified as major or minor by 2 of the authors independently according to 2 standard classification systems.	Matched for disease and year of enrollment. Propensity score adjustment including maternal age, alcohol consumption, smoking status, and parity as covariates. For analyses involving the disease-matched cohort, taking disease-modifying drugs or systemic glucocorticoids was additionally included in the propensity score estimation.
Weber-Schoendorfer (Controls unexposed, disease free), 2014	prospective cohort	Data on exposure were obtained via structured telephone interviews and/or mailed questionnaires completed by the mother and/or her physicians.	Data on outcome were obtained via structured telephone interviews and/or mailed questionnaires completed by the mother and/or her physicians. Birth defects were classified as major or minor by 2 of the authors independently according to 2 standard classification systems.	Matched within the centers for year of ascertainment. Propensity score adjustment including maternal age, alcohol consumption, smoking status, and parity as covariates. For analyses involving the cohort of women without autoimmune diseases, taking disease-modifying antirheumatic drugs or systemic glucocorticoids was used as a covariate directly in addition to propensity score stratification.

master protocol