Laura Spencer, Tanya Bubner, Emily Bain, Philippa Middleton

Thyroid dysfunction pre-pregnancy and during pregnancy (both hyper- and hypothyroidism) is associated with increased risk of adverse outcomes for mothers and infants in the short- and long-term. Managing the thyroid dysfunction (e.g. thyroxine for hypothyroidism, or antithyroid medication for hyperthyroidism) may improve outcomes. The best method of screening to identify and subsequently manage thyroid dysfunction pre-pregnancy and during pregnancy is unknown. To assess the effects of different screening methods (and subsequent management) for thyroid dysfunction pre-pregnancy and during pregnancy on maternal and infant outcomes. We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (14 July 2015) and reference lists of retrieved studies. Randomised or quasi-randomised controlled trials, comparing any screening method (e.g. tool, program, guideline/protocol) for detecting thyroid dysfunction (including hypothyroidism, hyperthyroidism, and/or thyroid autoimmunity) pre-pregnancy or during pregnancy with no screening, or alternative screening methods. Two review authors independently assessed eligibility of studies, extracted and checked data accuracy, and assessed the risk of bias of included studies. We included two randomised controlled trials (involving 26,408 women) - these trials were considered to be at low risk of bias. Universal screening (screening all women) versus case finding (screening only those at perceived increased risk) in pregnancy for thyroid dysfunctionOne trial (4562 women) compared universal screening with case finding for thyroid dysfunction. Before 11 weeks' gestation, women in the universal screening group, and 'high-risk' women in the case finding group had their sera tested for TSH (thyroid stimulating hormone), fT4 (free thyroxine) and TPO-Ab (thyroid peroxidase antibody); women with hypothyroidism (TSH > 2.5 mIU/litre) received levothyroxine; women with hyperthyroidism (undetectable TSH and elevated fT4) received antithyroid medication.In regards to this review's primary outcomes, compared with the case finding group, more women in the universal screening group were diagnosed with hypothyroidism (risk ratio (RR) 3.15, 95% confidence interval (CI) 1.91 to 5.20; 4562 women; GRADE: high quality evidence), with a trend towards more women being diagnosed with hyperthyroidism (RR 4.50, 95% CI 0.97 to 20.82; 4562 women; P = 0.05; GRADE: moderate quality evidence). No clear differences were seen in the risks of pre-eclampsia (RR 0.87, 95% CI 0.64 to 1.18; 4516 women; GRADE: moderate quality evidence), or preterm birth (RR 0.99, 95% CI 0.80 to 1.24; 4516 women; GRADE: high quality evidence) between groups. This trial did not report on neurosensory disability for the infant as a child.Considering this review's secondary outcomes, more women in the universal screening group received pharmacological treatment for thyroid dysfunction (RR 3.15, 95% CI 1.91 to 5.20; 4562 women). No clear differences between groups were observed for miscarriage (RR 0.90, 95% CI 0.68 to 1.19; 4516 women; GRADE: moderate quality evidence), fetal and neonatal death (RR 0.92, 95% CI 0.42 to 2.02; 4516 infants; GRADE: moderate quality evidence), or other secondary outcomes: pregnancy-induced hypertension, gestational diabetes, congestive heart failure, thyroid storm, mode of birth (caesarean section), preterm labour, placental abruption, respiratory distress syndrome, low birthweight, neonatal intensive care unit admission, or other congenital malformations. The trial did not report on a number of outcomes including adverse effects associated with the intervention. Universal screening versus no screening in pregnancy for hypothyroidismOne trial (21,846 women) compared universal screening with no screening for hypothyroidism. Before 15 + 6 weeks' gestation, women in the universal screening group had their sera tested; women who screened 'positive' (TSH > 97.5th percentile, fT4 < 2.5th percentile, or both) received levothyroxine.Considering primary review outcomes, compared with the no screening group, more women in the universal screening screened 'positive' for hypothyroidism (RR 998.18, 95% CI 62.36 to 15,978.48; 21,839 women; GRADE: high quality evidence). No data were provided for the outcome pre-eclampsia, and for preterm birth, the trial reported rates of 5.6% and 7.9% for the screening and no screening groups respectively (it was unclear if these percentages related to the entire cohort or women who screened positive). No clear difference was seen for neurosensory disability for the infant as a child (three-year follow-up IQ score < 85) (RR 0.85, 95% CI 0.60 to 1.22; 794 infants; GRADE: moderate quality evidence).More women in the universal screening group received pharmacological treatment for thyroid dysfunction (RR 1102.90, 95% CI 69.07 to 17,610.46; 1050 women); 10% had their dose lowered because of low TSH, high fT4 or minor side effects. No clear differences were observed for other secondary outcomes, including developmental delay/intellectual impairment at three years. Most of our secondary outcomes, including miscarriage, fetal or neonatal death were not reported. Based on the existing evidence, though universal screening for thyroid dysfunction in pregnancy increases the number of women diagnosed with hypothyroidism who can be subsequently treated, it does not clearly impact (benefit or harm) maternal and infant outcomes.While universal screening versus case finding for thyroid dysfunction increased diagnosis and subsequent treatment, we found no clear differences for the primary outcomes: pre-eclampsia or preterm birth. No clear differences were seen for secondary outcomes, including miscarriage and fetal or neonatal death; data were lacking for the primary outcome: neurosensory disability for the infant as a child, and for many secondary outcomes. Though universal screening versus no screening for hypothyroidism similarly increased diagnosis and subsequent treatment, no clear difference was seen for the primary outcome: neurosensory disability for the infant as a child (IQ < 85 at three years); data were lacking for the other primary outcomes: pre-eclampsia and preterm birth, and for the majority of secondary outcomes.For outcomes assessed using the GRADE approach the evidence was considered to be moderate or high quality, with any downgrading of the evidence based on the presence of wide confidence intervals crossing the line of no effect.More evidence is needed to assess the benefits or harms of different screening methods for thyroid dysfunction in pregnancy, on maternal, infant and child health outcomes. Future trials should assess impacts on use of health services and costs, and be adequately powered to evaluate the effects on short- and long-term outcomes.