A comparison of 12 machine learning models developed to predict ploidy, using a morphokinetic meta-dataset of 8147 embryos.

STUDY QUESTION: Are machine learning methods superior to traditional statistics in predicting blastocyst ploidy status using morphokinetic and clinical biodata? SUMMARY ANSWER: Mixed effects logistic regression performed better than all machine learning methods for ploidy prediction using our dataset of 8147 embryos. WHAT IS KNOWN ALREADY: Morphokinetic timings have been demonstrated to be delayed in aneuploid embryos. Machine learning and statistical models are increasingly being built, however, until now they have been limited by data insufficiency. STUDY DESIGN, SIZE, DURATION: This is a multicentre cohort study. Data were obtained from 8147 biopsied blastocysts from 1725 patients, treated from 2012 to 2020. PARTICIPANTS/MATERIALS, SETTING, METHODS: All embryos were cultured in a time-lapse system at nine IVF clinics in the UK. A total of 3004 euploid embryos and 5023 aneuploid embryos were included in the final verified dataset. We developed a total of 12 models using four different approaches: mixed effects multivariable logistic regression, random forest classifiers, extreme gradient boosting, and deep learning. For each of the four algorithms, two models were created, the first consisting of 22 covariates using 8027 embryos (Dataset 1) and the second, a dataset of 2373 embryos and 26 covariates (Dataset 2). Four final models were created by switching the target outcome from euploid to aneuploid for each algorithm (Dataset 1). Models were validated using internal-external cross-validation and external validation. MAIN RESULTS AND THE ROLE OF CHANCE: All morphokinetic variables were significantly delayed in aneuploid embryos. The likelihood of euploidy was significantly increased the more expanded the blastocyst (P < 0.001) and the better the trophectoderm grade (P < 0.01). Univariable analysis showed no association with ploidy status for morula or cleavage stage fragmentation, morula grade, fertilization method, sperm concentration, or progressive motility. Male age did not correlate with the percentage of euploid embryos when stratified for female age. Multinucleation at the two-cell or four-cell stage was not associated with ploidy status. The best-performing model was logistic regression built using the larger dataset with 22 predictors (F1 score 0.59 for predicting euploidy; F1 score 0.77 for predicting aneuploidy; AUC 0.71; 95% CI 0.67-0.73). The best-performing models using the algorithms from random forest, extreme gradient boosting, and deep learning achieved an AUC of 0.68, 0.63, and 0.63, respectively. When using only morphokinetic predictors the AUC was 0.61 for predicting ploidy status, whereas a model incorporating only embryo grading was unable to discriminate aneuploid embryos (AUC = 0.52). The ploidy prediction model's performance improved with increasing age of the egg provider. LIMITATIONS, REASONS FOR CAUTION: The models have not been validated in a prospective study design or yet been used to determine whether they improve clinical outcomes. WIDER IMPLICATIONS OF THE FINDINGS: This model may aid decision-making, particularly where pre-implantation genetic testing for aneuploidy is not permitted or for prioritizing embryos for biopsy. STUDY FUNDING/COMPETING INTEREST(S): No specific funding was sought for this study; university funds supported the first author. A.Ca. is a minor shareholder of participating centres. TRIAL REGISTRATION NUMBER: N/A.

A core outcome set for trials in miscarriage management and prevention: An international consensus development study.

OBJECTIVE: To develop core outcome sets (COS) for miscarriage management and prevention. DESIGN: Modified Delphi survey combined with a consensus development meeting. SETTING: International. POPULATION: Stakeholder groups included healthcare providers, international experts, researchers, charities and couples with lived experience of miscarriage from 15 countries: 129 stakeholders for miscarriage management and 437 for miscarriage prevention. METHODS: Modified Delphi method and modified nominal group technique. RESULTS: The final COS for miscarriage management comprises six outcomes: efficacy of treatment, heavy vaginal bleeding, pelvic infection, maternal death, treatment or procedure-related complications, and patient satisfaction. The final COS for miscarriage prevention comprises 12 outcomes: pregnancy loss <24 weeks' gestation, live birth, gestation at birth, pre-term birth, congenital abnormalities, fetal growth restriction, maternal (antenatal) complications, compliance with intervention, patient satisfaction, maternal hospitalisation, neonatal or infant hospitalisation, and neonatal or infant death. Other outcomes identified as important were mental health-related outcomes, future fertility and health economic outcomes. CONCLUSIONS: This study has developed two core outcome sets, through robust methodology, that should be implemented across future randomised trials and systematic reviews in miscarriage management and prevention. This work will help to standardise outcome selection, collection and reporting, and improve the quality and safety of future studies in miscarriage.

Sporadic miscarriage: evidence to provide effective care.

The physical and psychological effect of miscarriage is commonly underappreciated. The journey from diagnosis of miscarriage, through clinical management, to supportive aftercare can be challenging for women, their partners, and caregivers. Diagnostic challenges can lead to delayed or ineffective care and increased anxiety. Inaccurate diagnosis of a miscarriage can result in the unintended termination of a wanted pregnancy. Uncertainty about the therapeutic effects of interventions can lead to suboptimal care, with variations across facilities and countries. For this Series paper, we have developed recommendations for practice from a literature review, appraisal of guidelines, and expert group discussions. The recommendations are grouped into three categories: (1) diagnosis of miscarriage, (2) prevention of miscarriage in women with early pregnancy bleeding, and (3) management of miscarriage. We recommend that every country reports annual aggregate miscarriage data, similarly to the reporting of stillbirth. Early pregnancy services need to focus on providing an effective ultrasound service, as it is central to the diagnosis of miscarriage, and be able to provide expectant management of miscarriage, medical management with mifepristone and misoprostol, and surgical management with manual vacuum aspiration. Women with the dual risk factors of early pregnancy bleeding and a history of previous miscarriage can be recommended vaginal micronised progesterone to improve the prospects of livebirth. We urge health-care funders and providers to invest in early pregnancy care, with specific focus on training for clinical nurse specialists and doctors to provide comprehensive miscarriage care within the setting of dedicated early pregnancy units.

Miscarriage matters: the epidemiological, physical, psychological, and economic costs of early pregnancy loss.

Miscarriage is generally defined as the loss of a pregnancy before viability. An estimated 23 million miscarriages occur every year worldwide, translating to 44 pregnancy losses each minute. The pooled risk of miscarriage is 15·3% (95% CI 12·5-18·7%) of all recognised pregnancies. The population prevalence of women who have had one miscarriage is 10·8% (10·3-11·4%), two miscarriages is 1·9% (1·8-2·1%), and three or more miscarriages is 0·7% (0·5-0·8%). Risk factors for miscarriage include very young or older female age (younger than 20 years and older than 35 years), older male age (older than 40 years), very low or very high body-mass index, Black ethnicity, previous miscarriages, smoking, alcohol, stress, working night shifts, air pollution, and exposure to pesticides. The consequences of miscarriage are both physical, such as bleeding or infection, and psychological. Psychological consequences include increases in the risk of anxiety, depression, post-traumatic stress disorder, and suicide. Miscarriage, and especially recurrent miscarriage, is also a sentinel risk marker for obstetric complications, including preterm birth, fetal growth restriction, placental abruption, and stillbirth in future pregnancies, and a predictor of longer-term health problems, such as cardiovascular disease and venous thromboembolism. The costs of miscarriage affect individuals, health-care systems, and society. The short-term national economic cost of miscarriage is estimated to be £471 million per year in the UK. As recurrent miscarriage is a sentinel marker for various obstetric risks in future pregnancies, women should receive care in preconception and obstetric clinics specialising in patients at high risk. As psychological morbidity is common after pregnancy loss, effective screening instruments and treatment options for mental health consequences of miscarriage need to be available. We recommend that miscarriage data are gathered and reported to facilitate comparison of rates among countries, to accelerate research, and to improve patient care and policy development.

Recurrent miscarriage: evidence to accelerate action.

Women who have had repeated miscarriages often have uncertainties about the cause, the likelihood of recurrence, the investigations they need, and the treatments that might help. Health-care policy makers and providers have uncertainties about the optimal ways to organise and provide care. For this Series paper, we have developed recommendations for practice from literature reviews, appraisal of guidelines, and a UK-wide consensus conference that was held in December, 2019. Caregivers should individualise care according to the clinical needs and preferences of women and their partners. We define a minimum set of investigations and treatments to be offered to couples who have had recurrent miscarriages, and urge health-care policy makers and providers to make them universally available. The essential investigations include measurements of lupus anticoagulant, anticardiolipin antibodies, thyroid function, and a transvaginal pelvic ultrasound scan. The key treatments to consider are first trimester progesterone administration, levothyroxine in women with subclinical hypothyroidism, and the combination of aspirin and heparin in women with antiphospholipid antibodies. Appropriate screening and care for mental health issues and future obstetric risks, particularly preterm birth, fetal growth restriction, and stillbirth, will need to be incorporated into the care pathway for couples with a history of recurrent miscarriage. We suggest health-care services structure care using a graded model in which women are offered online health-care advice and support, care in a nurse or midwifery-led clinic, and care in a medical consultant-led clinic, according to clinical needs.

Levothyroxine in Women with Thyroid Peroxidase Antibodies before Conception.

BACKGROUND: Thyroid peroxidase antibodies are associated with an increased risk of miscarriage and preterm birth, even when thyroid function is normal. Small trials indicate that the use of levothyroxine could reduce the incidence of such adverse outcomes. METHODS: We conducted a double-blind, placebo-controlled trial to investigate whether levothyroxine treatment would increase live-birth rates among euthyroid women who had thyroid peroxidase antibodies and a history of miscarriage or infertility. A total of 19,585 women from 49 hospitals in the United Kingdom underwent testing for thyroid peroxidase antibodies and thyroid function. We randomly assigned 952 women to receive either 50 µg once daily of levothyroxine (476 women) or placebo (476 women) before conception through the end of pregnancy. The primary outcome was live birth after at least 34 weeks of gestation. RESULTS: The follow-up rate for the primary outcome was 98.7% (940 of 952 women). A total of 266 of 470 women in the levothyroxine group (56.6%) and 274 of 470 women in the placebo group (58.3%) became pregnant. The live-birth rate was 37.4% (176 of 470 women) in the levothyroxine group and 37.9% (178 of 470 women) in the placebo group (relative risk, 0.97; 95% confidence interval [CI], 0.83 to 1.14, P = 0.74; absolute difference, -0.4 percentage points; 95% CI, -6.6 to 5.8). There were no significant between-group differences in other pregnancy outcomes, including pregnancy loss or preterm birth, or in neonatal outcomes. Serious adverse events occurred in 5.9% of women in the levothyroxine group and 3.8% in the placebo group (P = 0.14). CONCLUSIONS: The use of levothyroxine in euthyroid women with thyroid peroxidase antibodies did not result in a higher rate of live births than placebo. (Funded by the United Kingdom National Institute for Health Research; TABLET Current Controlled Trials number, ISRCTN15948785.).

Subclinical hypothyroidism and antithyroid autoantibodies in women with subfertility or recurrent pregnancy loss: Scientific Impact Paper No. 70 June 2022: Scientific Impact Paper No. 70 June 2022.

The thyroid is a gland located in the neck and is important for many processes in the body. Problems with the thyroid gland are common in women of reproductive age. It is essential to have a normal working thyroid gland in order to achieve a successful pregnancy. One of the most common problems with the thyroid is underactivity (known as hypothyroidism). An early, mild form of an underactive thyroid is called subclinical hypothyroidism. Often people with this condition do not have any symptoms. Another common problem is thyroid autoimmunity. Here, the immune system attacks the thyroid gland, sometimes leading to the development of abnormal thyroid function. This can be diagnosed by the presence of proteins in the bloodstream called antibodies. Mild thyroid problems and the presence of high levels of thyroid antibodies have been linked to miscarriage and premature birth. There is debate in medicine about whether there should be routine testing of thyroid function both in the general population and in individuals who are trying for a baby. In addition, the strategies used to manage certain thyroid problems are questioned. Discussions around testing and subsequent management particularly relate to women with a history of subfertility or repeated miscarriages. This Scientific Impact Paper provides information on thyroid testing and the management of mild thyroid problems and thyroid antibodies in women with a history of subfertility or recurrent miscarriages, using the latest evidence and guidelines. It concludes that there may be a role for treating these women with thyroxine tablets (the hormone produced by the thyroid gland) when subclinical hypothyroidism is present, and gives guidance on the cut-off levels for treatment.

An evaluation of the quality of online information on emergency contraception.

OBJECTIVES: The increasing availability of health information online combined with reduced access to health care providers due to the coronavirus pandemic means that more people are using the internet for health information. However, with no standardised regulation of the internet, the population is vulnerable to misinformation regarding important health information. This review aimed to evaluate the quality and readability of the online information available on emergency contraception (EC) options. STUDY DESIGN: In this descriptive study, a Google search was performed using the term 'emergency contraception options' on 13 April 2020 yielding 232 results. Seventy-one results were excluded (34 inaccessible, 37 contained no medical information). The remaining 161 results were categorised by typology and assessed for credibility (JAMA criteria and HONcode), reliability (DISCERN tool) and readability (Flesch-Kincaid Grade Level and Simple Measure of Gobbledygook). RESULTS: Of all webpages evaluated, the most common typology was governmental. Credibility of web pages was poor (average JAMA score of 1.47 out of 4). Only 10.6% of webpages were HONcode certified. The most common DISCERN category was Fair (29.81%), closely followed by Poor (27.95%) reliability. On average, readability levels were above the recommended grade level for health information. The intrauterine device was discussed least frequently (86.96%) of all the EC options. CONCLUSION: Online information was of low credibility, reliability and written above the recommended reading level. Clinicians should be aware of the poor quality of online information on EC options, and actively educate patients on what makes a source credible.

Myoinositol or D-chiro-inositol for PCOS symptoms in adolescents: a narrative review.

BACKGROUND: Polycystic ovary syndrome (PCOS) treatment in adolescents currently focuses on lifestyle interventions, with pharmacological treatment options often limited to hormonal contraceptives. Several of these carry broad side-effect profiles and are not always accepted by young girls. There is growing interest in non-hormonal therapies for PCOS. We aimed to collate the evidence on the use of myoinositol or D-chiro-inositol in the improvement of PCOS symptoms in symptomatic adolescents. CONTENT: A systematic literature review identifying key articles from inception to March 2023. Participants: Female adolescents (aged 12-19 years) with PCOS or PCOS-like features. Intervention: Myoinositol or D-chiro-inositol with or without additional interventions. Comparison: Any other treatment, including lifestyle interventions, hormonal therapy, metformin or no treatment. The main outcome measure were improvement in symptoms, quality of life and adverse effects. SUMMARY: Eight studies were included: two randomised open-label trials, one quasi-randomised and three non-randomised interventional studies, one case-control study and one cohort study. All studies showed improvements in some biochemical markers, metabolic parameters or clinical symptoms, but these were not reproducible across all studies. OUTLOOK: The benefit of myoinositol in adolescents with PCOS remains unclear, with limited high-quality evidence. This review highlights the need for robustly conducted research to inform clinical practice.

Interventions to prevent miscarriage.

The physical and psychological impact of miscarriage can be devastating. There are many lifestyle and therapeutic interventions that may prevent a miscarriage. In this review, we have outlined the key areas for health optimization to prevent pregnancy loss, drawing on the most up-to-date evidence available. The 3 key areas identified are lifestyle optimization in women, lifestyle optimization in men, and therapeutic interventions. The evidence demonstrates that the treatments to consider are first-trimester progesterone administration, levothyroxine in women with subclinical hypothyroidism, and the combination of aspirin and heparin in women with antiphospholipid antibodies.

Preconception Counseling and Care for Pregnant Women with Thyroid Disease.

Thyroid disease is associated with adverse maternal and fetal outcomes. Appropriate reference ranges should be used for the interpretation of test results, although universal screening for thyroid dysfunction is not warranted. Overt thyroid dysfunction requires careful consideration of medication adjustments and close monitoring. Mild thyroid hypofunction has been linked to adverse pregnancy outcomes including preterm delivery, and poor neurocognition in the offspring. This review summarizes the most recent evidence on the counseling and management of women with thyroid disease before and during pregnancy and highlights the areas of controversy in need of further research.

Evaluating interventions and adjuncts to optimize pregnancy outcomes in subfertile women: an overview review.

BACKGROUND: There is a wealth of information regarding interventions for treating subfertility. The majority of studies exploring interventions for improving conception rates also report on pregnancy outcomes. However, there is no efficient way for clinicians, researchers, funding organizations, decision-making bodies or women themselves to easily access and review the evidence for the effect of adjuvant therapies on key pregnancy outcomes in subfertile women. OBJECTIVE AND RATIONALE: The aim was to summarize all published systematic reviews (SRs) of randomized controlled trials (RCTs) of interventions in the subfertile population, specifically reporting on the pregnancy outcomes of miscarriage and live birth. Furthermore, we aimed to highlight promising interventions and areas that need high-quality evidence. SEARCH METHODS: We searched the Cochrane Database of Systematic Reviews and PubMed clinical queries SR filter (inception until July 2021) with a list of key words to capture all SRs specifying or reporting any miscarriage outcome. Studies were included if they were SRs of RCTs. The population was subfertile women (pregnant or trying to conceive) and any intervention (versus placebo or no treatment) was included. We adopted Grading of Recommendations, Assessment, Development and Evaluation (GRADE) for determining the quality of the evidence. Exclusion criteria were overview reviews, reviews that exclusively reported on women conceiving via natural conception, reviews including non-randomized study designs or reviews where miscarriage or live birth outcomes were not specified or reported. OUTCOMES: The primary outcome was miscarriage, defined as pregnancy loss <24 weeks of gestation. Data were also extracted for live birth where available. We included 75 published SRs containing 121 251 participants. There were 14 classes of intervention identified: luteal phase, immunotherapy, anticoagulants, hCG, micronutrients, lifestyle, endocrine, surgical, pre-implantation genetic testing for aneuploidies (PGT-As), laboratory techniques, endometrial injury, ART protocols, other adjuncts/techniques in the ART process and complementary interventions. The interventions with at least moderate-quality evidence of benefit in reducing risk of miscarriage or improving the chance of a live birth are: intrauterine hCG at time of cleavage stage embryo transfer, but not blastocyst transfer, antioxidant therapy in males, dehydroepiandrosterone in women and embryo medium containing high hyaluronic acid. Interventions showing potential increased risk of miscarriage or reduced live birth rate are: embryo culture supernatant injection before embryo transfer in frozen cycles and PGT-A with the use of fluorescence in situ hybridization. WIDER IMPLICATIONS: This review provides an overview of key pregnancy outcomes from published SRs of RCTs in subfertile women. It provides access to concisely summarized information and will help clinicians and policy makers identify knowledge gaps in the field, whilst covering a broad range of topics, to help improve pregnancy outcomes for subfertile couples. Further research is required into the following promising interventions: the dose of progesterone for luteal phase support, peripheral blood mononuclear cells for women with recurrent implantation failure, glucocorticoids in women undergoing IVF, low-molecular-weight heparin for unexplained subfertility, intrauterine hCG at the time of cleavage stage embryo or blastocyst transfer and low oxygen concentrations in embryo culture. In addition, there is a need for high-quality, well-designed RCTs in the field of reproductive surgery. Finally, further research is needed to demonstrate the integrated effects of non-pharmacological lifestyle interventions.

Evaluating the Progression to Hypothyroidism in Preconception Euthyroid Thyroid Peroxidase Antibody-Positive Women.

CONTEXT: Thyroid peroxidase antibody (TPOAb) positivity is prevalent in women of reproductive age and predisposes to thyroid dysfunction, particularly hypothyroidism, which has adverse effects on pregnancy. OBJECTIVE: This study aimed to report the rate of development of abnormal thyroid function among initially euthyroid TPOAb-positive women recruited into the TABLET trial, to identify factors associated with the development of hypothyroidism, and to compare outcomes between euthyroid and treated hypothyroid individuals. METHODS: This observational cohort study, conducted at 49 UK hospitals between 2011 and 2016, included euthyroid TPOAb-positive women 16 to 40 years of age with a history of miscarriage or subfertility, planning pregnancy, randomized to levothyroxine 50 mcg daily or placebo. Abnormal thyroid function, conception rate, and live birth rate (LBR) ≥34 weeks were analyzed. RESULTS: Among the women, 70/940 (7.4%) developed subclinical (SCH) or overt (OH) hypothyroidism: 27/470 taking levothyroxine and 43/470 placebo (relative risk [RR] 0.63; 95% CI, 0.39-1.00; P = 0.05); 83% of cases emerged prepregnancy. Baseline median serum TSH concentrations and TPOAb titers were significantly higher in those who developed hypothyroidism vs those who did not (P < 0.001). Treated SCH/OH demonstrated a higher failure-to-conceive rate compared with euthyroid women (adjusted RR 2.02 [1.56-2.62]; P < 0.001). The LBR ≥ 34 weeks was similar in the treated SCH/OH and euthyroid groups (adjusted RR 1.09 [0.77-1.55]; P = 0.6). CONCLUSION: Approximately 7% of euthyroid TPOAb-positive women will develop hypothyroidism within 1 year preconception or in pregnancy. Conception rates are lower in women with treated SCH/OH compared with euthyroid women, but LBR are comparable. Thyroid function in TPOAb-positive women should be monitored regularly, when trying to conceive, to ensure prompt diagnosis and appropriate treatment initiation.

The Prevalence of Thyroid Dysfunction and Autoimmunity in Women With History of Miscarriage or Subfertility.

OBJECTIVE: To describe the prevalence of and factors associated with different thyroid dysfunction phenotypes in women who are asymptomatic preconception. DESIGN: Observational cohort study. SETTING: A total of 49 hospitals across the United Kingdom between 2011 and 2016. PARTICIPANTS: Women aged 16 to 41years with history of miscarriage or subfertility trying for a pregnancy. METHODS: Prevalences and 95% confidence intervals (CIs) were estimated using the binomial exact method. Multivariate logistic regression analyses were conducted to identify risk factors for thyroid disease. INTERVENTION: None. MAIN OUTCOME MEASURE: Rates of thyroid dysfunction. RESULTS: Thyroid function and thyroid peroxidase antibody (TPOAb) data were available for 19213 and 19237 women, respectively. The prevalence of abnormal thyroid function was 4.8% (95% CI, 4.5-5.1); euthyroidism was defined as levels of thyroid-stimulating hormone (TSH) of 0.44 to 4.50 mIU/L and free thyroxine (fT4) of 10 to 21 pmol/L. Overt hypothyroidism (TSH > 4.50 mIU/L, fT4 < 10 pmol/L) was present in 0.2% of women (95% CI, 0.1-0.3) and overt hyperthyroidism (TSH < 0.44 mIU/L, fT4 > 21 pmol/L) was present in 0.3% (95% CI, 0.2-0.3). The prevalence of subclinical hypothyroidism (SCH) using an upper TSH concentration of 4.50 mIU/L was 2.4% (95% CI, 2.1-2.6). Lowering the upper TSH to 2.50 mIU/L resulted in higher rates of SCH, 19.9% (95% CI, 19.3-20.5). Multiple regression analyses showed increased odds of SCH (TSH > 4.50 mIU/L) with body mass index (BMI) ≥ 35.0 kg/m2 (adjusted odds ratio [aOR] 1.71; 95% CI, 1.13-2.57; P = 0.01) and Asian ethnicity (aOR 1.76; 95% CI, 1.31-2.37; P < 0.001), and increased odds of SCH (TSH ≥ 2.50 mIU/L) with subfertility (aOR 1.16; 95% CI, 1.04-1.29; P = 0.008). TPOAb positivity was prevalent in 9.5% of women (95% CI, 9.1-9.9). CONCLUSIONS: The prevalence of undiagnosed overt thyroid disease is low. SCH and TPOAb are common, particularly in women with higher BMI or of Asian ethnicity. A TSH cutoff of 2.50 mIU/L to define SCH results in a significant proportion of women potentially requiring levothyroxine treatment.