Reference

Study type

Characteristics

Intervention (I)

Controls (C)

Outcome measures and follow-up time

Results

Quality

assessment study

Cornel, 1996

Cohort study

N total = 4014

Setting: Northern Netherlands

Aim of the study:

To identify risk factors for malformations.

To investigate whether the occurrence of defects in any of the 32 diagnostic catergories were associated with any of 77 risk factors.

Inclusion criteria:

Exclusion criteria:

Case: Birth defect

E: risk factor

Controls: No birth defect

E: risk factor

Primary outcomes:

Birth defects (ICD 9)

Risk factor surveillance

Potential risk factors:

Maternal drug use, chronic illness, family history.

Secondary outcomes:

Follow up: 1981-1994

Primary outcomes:

An association between maternal use of psychotropic drugs (psycholeptics) and cleft lip with or without cleft palate (CLP) was present.

CPL and maternal use of oxazepam OR 8.17, 95% CI 1.26-42.2

Cleft lip with or without CLP and exosure to Anxiolytics:

Cases

E: 4

NE: 273

Controls:

E:11

NE: 3726

OR 4.96, 95% CI 1.14-16.88

Secondary outcomes: -

Clearly defined groups: Yes.

Selection bias: Possible, informed consent.

Method of assessing the outcome appropriately: Yes.

Selective loss to follow up:

No.

Identification confounders and correction in analysis:

No.

Funding: Not described.

Limitations:

- possible type 1 errors (only exposure defect combinations involving at least three infants)

- Lack of outcome specificity.

Diav-Citrin, 1999

Prospective cohort study

N total = 80

Setting: Canada

Aim of the study:

To compare pregnancy outcome following 1^st trimester exposure to zoplicone.

Inclusion criteria:

Not applicable.

Exclusion criteria:

Not applicable.

Exposure:

Zoplicone in 1^st trimester

N=40

Controls:

Matched controls with nonteratogenic exposure. N=40

Matched for age, cigarette smoking and alcohol consumption.

Telephone interview

Primary outcomes:

Malformations

Secondary outcomes:

Pregnancy outcome and offspring characteristics.

-Delivery method

-Preterm delivery

-Birth weight by GA

-Fetal distress

-NICU admission

Follow up: Motherrisk program 1993-1997

Primary outcomes:

No increase in rate of major malformations 0/31 (0%) for zoplicone group vs 1/37 (2.7%) for control group, p=1

Secondary outcomes:

No differences in outcome of pregnancy, delivery method, assisted deliveries, fetal distress, presence of meconium at birth, preterm deliveries, NICU admissions.

Newborns in the zoplicone group had significantly lower mean birth weight and lower gestatrional age compared to the control group.

Characteristics:

Clearly defined groups: yes.

Selection bias: Recall/report bias.

Method of assessing the outcome appropriately: Yes.

Selective loss to follow up: No.

Identification confounders and correction in analysis:

No. But they used matched control groups. No major differences in characteristics.

Funding: Not applicable.

Limitations:

- relatively smal sample size for number of variables -> underpowered.

- No information about co-medication

Hartz, 1975

Retrospective cohort (follow up study)

N= 740 (chlordiazepine)

Inclusion criteria:

-12 US hospitals

-GA>5months

-1958-1966

-Race white, black or puerto rican

Exclusion criteria:

-no consent

-Abortion <5 months

-Multiple pregnancies

-Missing data

-Rubella

Aim of the study:

To dentify the risk of antenatal chlordiazepoxide on malformations, metal development and childhood mortality

Primary outcomes:

-Malformations

Secundary outcomes:

-metal development

-mortality

Follow up: -

Primary outcomes:

Malformations;

-Early exposure 11/257 OR 0.8 (1.6-0.4)

-Late exposure 30/483 OR 1.2 (1.7-0.8)

Secundary outcomes:

-metal development

-mortality

4.8% (4.4% control), similar for early and late exposures

-Clearly defined groups:

Yes

-Selection bias:

Possible recall bias

-Method of assessing the outcome appropriately:

Yes

-Selective loss to follow up:

No, this was a follow up study

-Identification confounders and correction in analysis:

Yes, correction for race and socioeconomic status. Not for drugs, other medication, smoking, heridity etc.

-Funding:

Unknown

Kjaer, 2007

Case-time-control study

N total = 22 843 children with congenital abnormalities (cases) and 38 151 children with no abnormalities

(controls)

Aim of the study:

Congenital abnormalities following exposure to phenytoin, phenobarbital,

and diazepam in early pregnancy, i.e., second and third month, compared to mid-pregnancy, i.e., fifth and sixth month of pregnancy

Inclusion criteria:

-

Exclusion criteria:

congenital dislocation of the

hip based on Ortolani click, congenital

inguinal hernia, haemangioma,

umbilical hernia and hydrocele, specified

Mendelian and chromosomal congenital abnormality

Syndrome, Down’s syndrome

Children without Hungarian citizenship and

those with unknown mother’s

Controls with no

Hungarian citizenship and with unknown addresses.

Cases:

22.843 children with congenital abnormalities 

Controls:

38.151 children with no abnormalities

Controls were matched to cases on sex,

residence, and calendar birth week.

Primary outcomes:

Congenital abnormalities following exposure to phenytoin, phenobarbital,

and diazepam in early pregnancy, i.e., second and third month, compared to mid-pregnancy, i.e., fifth and sixth month of pregnancy.

Secondary outcomes:

Follow up: Hungarian Case-Control Surveillance of Congenital Abnormalities 1980-1996

Primary outcomes:

diazepam, OR for congenital abnormalities was 1.2 (95%CI: 1.0–1.4).

Statistically significant case-time-control odds ratios were seen for all drugs combined especially for neural tube defects, cleft lip—with or without cleft palate—, intestinal atresia/stenosis and limb deficiencies in offspring exposed in early compared to late pregnancy.

Diazepam:

Statistically significant case-time-control odds ratios were seen for diazepam, and among diazepam users, we found an increased risk of neural tube defects, cleft lip—with or without cleft palate—, and limb deficiencies in offspring exposed in early compared to late pregnancy.  The use of diazepam tended to be more frequent in late pregnancy than in the second and third months of gestation while this was not the case for phenytoin and phenobarbital (data not presented).

Secondary outcomes:

-

Clearly defined groups: Yes.

Selection bias: Possible, but study method pprotects

Method of assessing the outcome appropriately: yes.

Selective loss to follow up: Not described.

Identification confounders and correction in analysis: Yes. Logistic regression.

Funding: The Danish pharmaceutical Association

Limitations:

- Dependence on variability of drug use over time, with only those matched pairs that are discordant on exposure status contributing inforation about the effect.

- Maybe drug use later in pregnancy, no correcton for time -> effect bias.

Lin, 2004

Prospective cohort.

N total = 28,565 infants

N= 52 - Clonazepam

Aim of the study:

To obtain more information about

the potential teratogenicity of clonazepam

Inclusion criteria:

All births with a major  malformation, including live births prior to

hospital discharge, stillbirths older than 20 gestational

weeks, and elective terminations.

Exclusion criteria:

Conditions associated with prematurity (such as a patent ductus arteriosus or undescended testes

at <36 gestational weeks); “ultrasound only” diagnoses (such as unilateral renal agenesis that were not identifiable

by an examining pediatrician); positional deformities (e.g.,

flexible metatarsus adductus); minor anomalies (such as

glanular hypospadias); birth marks; and any genetic disorders

or chromosome abnormalities.

Intervention:

N= 43 monotherapy Clonazepam

Controls: None.

Primary outcomes:

Major malformations

(dose and timing medication).

Secondary outcomes:

-

Follow up: 32-month period from

May 1, 1998 to Dec. 31, 2000 at the Brigham and Women’s Hospital in Boston.

Primary outcomes:

During a 32-month period, 166 anticonvulsant-exposed infants were identified; 52 had been exposed to clonazepam, 43 as monotherapy, 9 as polytherapy.

Of the 43 mothers who used clonazepam alone, the treatment indication included seizures (1/43; 2.3%), migraine headaches (1/43; 2.3%), and psychiatric diagnoses including depression, bipolar disorder, panic attacks, anxiety, and obsessive compulsive disorder (41/43; 95.3%). In 19 of 41 (46.3%) mothers, there were two or more comorbid psychiatric disorders.

A total of 33 (76.7%) of the monotherapy infants were exposed during the first trimester. One (3.0%) infant had dysmorphic features, growth retardation, and a heart malformation (tetralogy of Fallot). Mother treated in the first trimester for depression with 1–2 mg per day of clonazepam.

No malformations in infants born to nine mothers who used clonazepam and other anticonvulsants.

Of these 33 mothers, 25 (75.7%) ingested clonazepam at the recommended frequency of one to two times per day. A total of 29 (87.9%) mothers reported using an appropriate dose of 0.5–2.0 mg per day. Three (9.1%) used more than 2.0 mg per day, and in one patient (3.0%) the dose was not reported.

Data about the timing of clonazepam exposure was available (or reasonably estimated) on all patients, having been provided through interview in 20 of 43 (46.5%) mothers.

Secondary outcomes:

Characteristics:

Brief questionnaire to determine medication use, dose, duration, and indication, e.g., seizure disorder, panic attacks.

Clearly defined groups: one group.

Selection bias: Possible.

Method of assessing the outcome appropriately:

Yes. But no logistic regression, no information about characteristics and possible confounding effect.

Selective loss to follow up:

Not described.

Identification confounders and correction in analysis:

No information or correction for confounders. No logistic regression analysis, no control groups.

Funding: No.

Limitations:

- small sample size – underpowered.

Strong:

- timing and dosis information.

Milkovich, 1974

Prospective cohort.

N total = 1904

Aim of the study:

To examine the teratogenic potential of the minor tranquilizers.

4 populations:

1. Meprobamate n=402

2. Chlordiazepoxide n=175

3. other drugs n=818

4. no drugs

n=509

Primary outcomes:

Incidence of severe congenital  anomalies

- Categorization severe, moderate, mild.

- Fetal deaths

According to gestational age at the time the diagnosis of anxiety was entered into the record.

Primary outcomes:

Rates of the anomalies after prescription of meprobamate or chlordiazepoxide in early pregnancy (the first 42 days) were higher 12.1/100 and 11.4/100 respectively than in children of the other-drug, 4.6/100, or no-drug cohorts 2.6/100.

When the reference date was later in pregnancy (43 days to termination), statistically insignificant differences in rates of the anomalies were observed among the 4 groups. These data suggest that meprobamate, and possibly chlordiazepoxide, may be teratogenic when taken during the first six weeks of gestation.

Type of anomaly:

Fetal deaths:

No significantly differences between the groups.

Secondary outcomes:

Characteristics:

There is agreement among the 4 study populations in mothers ages and racial backgrounds.

The three drug groups were not mutually exclusive. Fewer than 10 % of the mothers belonged in two cohorts. When this doubling occurred, the vast majority had received the second drug after the first trimester.

Clearly defined groups: yes.

Selection bias: Possible.

Method of assessing the outcome appropriately: Yes. But possibility of underreportion.

Selective loss to follow up:

Not mentioned.

Identification confounders and correction in analysis:

Not mentioned.

Funding: Yes. Grant from the national institute of child health and human development, national institutes of health.

Oberlander, 2008

Retrospective cohort

N= 968 (BDP onl;y)

Data used in this study came from five administrative

sources housed in the BC Linked Health Database

(Chamberlayne et al., 1998) (BC 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) linked to PharmaNet, a

province-wide network recording all prescriptions dispensed

by BC pharmacists outside hospitals

Aim of the study:

To determine a population-based incidence of congenital anomalies following prenatal exposure to

serotonin reuptake inhibitor (SRI) antidepressants used alone and in combination with a benzodiazepines (BZ).

Inclusion criteria:

-1979-2002

-live births

Exclusion criteria:

-Anticonvulsiva

Primary outcomes:

Major congenital anomalies

Secundary outcomes:

Cardiovascular congenital defects

Ventricular septal defects

Atrial septal defects

Primary outcomes:

Major congenital anomalies 31/968, OR 3.20 (2.27-4.51)

Secundary outcomes:

Cardiovascular congenital defects 5/968, OR 0.52 (0.22-1.20)

Ventricular septal defects 3/968, OR 0.31 (0.11-0.91)

Atrial septal defects 1/968, OR 0.10 (0.02-0.58)

-Clearly defined groups:

Yes

-Selection bias:

No

-Method of assessing the outcome appropriately:

Yes

-Selective loss to follow up:

No

-Identification confounders and correction in analysis:

Not always known which BDP were used, thereby sometimes comedication of more BDP

-Funding:

BC Ministry of Children and Family Development

through the Human Early Learning Partnership

Ornoy, 1998

Prospective cohort

N total = 1023

(599 benzo users and 424 controls)

Inclusion criteria:

- 1988 to 1996

- women called the Israeli TIS concerning exposures to BD during pregnancy

-pregant

Exclusion criteria:

-Not mentioned

Aim of the study:

To examine if BD causes an increase in spontaneous abortions or congenital anomalies

Intervention:

n= 460

Total percent use is greater than 100 due to use of more than one drug

by some women. There were 554 benzodiazepines used by 460 women.

Controls: 424 control pregnancies (nonteratogenic exposure)

The results are based on the information provided by the women (questionnaire) or their physicians

Primary outcomes:

-congenital anomalies,

-major medical problems

Secundary outcomes:

-sex, birth age and birth weight of the newborn

-spontaneous and induced abortions

-maternal age

Follow up: 460 pregnancies (76.6%)= 355 children in exposure group.

Primary outcomes:

The incidence of congenital anomalies (3.1%) was not different from that found in 424 control pregnancies (2.6%).

Secundary outcomes:

 There was a significantly higher incidence of induced abortions (14.1% vs. 4.7%, P = 0.00) and of spontaneous abortions (8.7% vs. 5.2%, P = 0.01).

-Clearly defined groups: Yes

-Selection bias: 1105 loss to follow up, not explained

-Method of assessing the outcome appropriately: No, they tested for significance with the

chi-square and Fisher exact test (two-tail) for differences in proportions between the BD exposed group and the control group. No multivariable analyses.

-Selective loss to follow up: (599-460=139 in BDP, in control 1390-424=966)

-Identification confounders and correction in analysis:

No identification or analyses. E.g. they mention the weakness in other studies regarding co-medication, but do not correct this in their own study.

-Funding:

Not mentioned

Reis, 2013

Retrospective cohort

N=1776

Swedish national health registers.

Aim of the study:

To investigate the proposed synergistic

teratogenic effect of use of selective serotonin receptor

inhibitors (SSRI) together with sedatives or hypnotics,

primarily benzodiazepines, during pregnancy.

Inclusion criteria:

Benzodiazepines or HBRA use

Data from the Swedish Medical Birth Register were

used.3 Women giving birth between 1 July 1995 and

31 December 2008 (n=1 290 672) were interviewed by

midwives at the first antenatal care visit, usually before

the end of the first trimester.

Exclusion criteria:

-

Primary outcomes:

Congenital malformations and cardiac defects

Primary outcomes:

Benzodiazepines:

Servere malformation 37/1000, OR 1.10 (0.79-1.54)

Cardiovasculair defect 13/1000, OR 1.30 (0.75-2.24)

HBRA:

Servere malformation 22/776, OR 0.86 (0.57-1.72)

Cardiovasculair defect  2/776, OR 0.26 (0.63-0.94)

-Clearly defined groups:

Yes

-Selection bias:

No

-Method of assessing the outcome appropriately:

Yes

-Selective loss to follow up:

Not apllicable

-Identification confounders and correction in analysis:

Yes, adjustment fora ge, parity, smoking and BMI, but not for inherited risk, concominant medication or drug use

-Funding:  The study was supported by grants from Evy and Gunnar

Sandberg’s Foundation, Lund (BK) and the Swedish Medical Research Council

(No. 2009-4740) (MR). The authors’ work was independent of the funders.

Rosenberg, 1983

Retrospective cohort.

N total = 3109

Setting: Boston, Philadelphia and Toronto

Aim of the study:

The present report provides further data on the risks of cleft lip and cleft palate from in utero exposure to diazepam.

Inclusion criteria:

Exclusion criteria:

Cases:

1. all children with any cleft lip with or without cleft palate (n=445)

2. all children with any cleft palate alone. (n=166)

Controls:

All other malformed children (n=2498)

N=1260 gastrointestinal defects

N=449 cardiovascular defects

N=383 central-nervous-system defects

N=308 musculoskeletal defects

N=263 various syndromes

N=259 genitourinary defects

N=254 neoplasms

N=116 respiratory defects

70 defects of eyes and ears

Primary outcomes:

Risk of oral cleft anomalies

Secondary outcomes:

-

Follow up: March 1976 – April 1982

Primary outcomes:

After control for identified confounders the corresponding relative risk of Diazepam exposure in the first 4 lunar months of pregnancy was 0.8 (95% CI 0.4-1.7) for cleft lip with or without cleft palate and 0.8 (95% CI 0.2-2.5) for cleft palate alone.

Secondary outcomes:

-

Characteristics:

-

Clearly defined groups: Yes.

Selection bias: Possible.

Method of assessing the outcome appropriately:

Selective loss to follow up:

No.

Identification confounders and correction in analysis:

Yes. Correction for confounders and logistic regression. (Year of interview, maternal age, geographic area, maternal cigarette smoking, maternal history of influenza or other infection during the 1^st trimester, maternal history of convulsive disorder or use of anticonvulsants during the pregnancy, maternal history of diabetes mellitus, sex of the child, use of benzodiazepines other than diazepam during pregnancy, history of cleft lip or palate in the mother, father, sibling, or half sibling, and suspicion by the mother that diazepam is a teratogen).

Funding: Not mentioned.

Limitations:

St. Clair, 1992

Aim of the study:

Inclusion criteria:

Exclusion criteria:

Primary outcomes:

Secundary outcomes:

Follow up:

Primary outcomes:

Secundary outcomes:

-Clearly defined groups:

-Selection bias:

-Method of assessing the outcome appropriately:

-Selective loss to follow up:

-Identification confounders and correction in analysis:

-Funding:

Wang, 2010

Retrospective cohort.

N= 14.982

2,497 mothers who received zolpidem treatment during pregnancy and those from

12.485 randomly selected mothers who did not receive the drug

Setting: Taiwan

Aim of the study:

The aim of comparing the risk of adverse

pregnancy outcomes in women who received zolpidem treatment for insomnia during pregnancy with that in women

who did not.

Inclusion criteria:

Live singleton births and

used prenatal care services

Exclusion criteria:

History of mental disorder.

Diagnosed with hypertension, diabetes, or coronary heart disease prior

to conceiving.

Exclusion for control group: prescribed

any type of medication for >30 days during pregnancy.

Exposure:

2497 mothers who received zolpidem

Defined women who had been prescribed zolpidem for

>30 days during pregnancy as those who had received zolpidem treatment.

Controls:

12.485 randomly picked mothers who received no zolpidem.

Matched with the study group in terms of age

Primary outcomes:

Adverse outcomes identified and assessed were delivery of low-birth-weight (LBW) infants, preterm

deliveries, delivery of small-for-gestational-age (SGA) infants, delivery of infants with congenital anomalies, and cesarean

delivery.

(Incidence)

Major malformations included in this study:

Hydrocephaly,

anencephaly, microcephaly, meningomyelocele, encephalocele, and spina bifida

Follow up: 1 January 2005 and 31 December

2005.

Primary outcomes:

The results show that the adjusted odds ratios (ORs) for adverse pregnancy outcomes—LBW infants, preterm deliveries, SGA infants, and caesarean delivery—were all higher in mothers who received zolpidem treatment during pregnancy, relative to the randomly selected controls (1.39 (95% confidence interval (CI) = 1.17–1.64), 1.49 (95% CI = 1.28–1.74), 1.34 (95% CI = 1.20–1.49), and 1.74 (95% CI = 1.59–1.90), respectively).

Major malformations:

There was no significant difference between the two groups in the

rates of congenital anomalies (0.48 vs. 0.65%, P = 0.329).

Characteristics:

In the total sample of 14,982 mothers, the mean age was 29.7 years with a SD of 4.9 years.

Mothers who received zolpidem were more likely to have lower educational levels (P < 0.001), gestational hypertension (P < 0.001), and anemia (P = 0.001) than mothers who did not take zolpidem. Also, in mothers who received the drug, the parity of the infants was higher than in the control group (P < 0.001).

Clearly defined groups: Yes.

Selection bias: Possible.

Method of assessing the outcome appropriately: Yes.

Selective loss to follow up:

No.

Identification confounders and correction in analysis: Yes.

Adjusting for other characteristics of the

mothers and infants (the characteristics of the mother

(level of education, gestational diabetes, gestational hypertension, preeclampsia/eclampsia, and anemia) and those of the infant (gender and parity). Maternal educational levels were categorized as elementary school or below, junior high school, senior high school, and college or above. Parity was recorded as 1, 2, or ≥3.)

Funding:

Conflict of Interest: The authors declared no conflict of interest.

Limitations:

- study was based on administrative data

- Data did not include complete information

on the social history of the patients (including smoking and alcohol use), on non prescription medications taken during pregnancy, and on the nutritional state of the mothers.

- No information on co-medication

Strong:

Duration and dosis included.