Prevalence, causes, and impact of non-visualized pregnancy losses in a recurrent pregnancy loss population.

STUDY QUESTION: Does the occurrence of non-visualized pregnancy loss (NVPL) affect future reproductive outcomes in patients with recurrent pregnancy loss (RPL)? SUMMARY ANSWER: The number of previous NVPLs is a significant predictor of subsequent live birth in patients with RPL. WHAT IS KNOWN ALREADY: The number of preceding miscarriages is a strong indicator for future reproductive outcomes. However, NVPL particularly has been sparsely addressed in previous literature. STUDY DESIGN, SIZE, DURATION: We performed a retrospective cohort study of 1981 patients attending a specialized recurrent pregnancy loss clinic (RPL) from January 2012 to March 2021. A total of 1859 patients met the inclusion criteria of the study and were included in the analysis. PARTICIPANTS/MATERIALS, SETTING, METHODS: Patients with a history of RPL, defined as ≥2 pregnancy losses before 20 weeks gestation, who attended a specialized RPL clinic in a tertiary care center were included. Patients' evaluation included parental karyotyping, antiphospholipid antibodies screening, uterine cavity assessment with hysterosalpingography (HSG) or hysteroscopy, maternal thyroid stimulating hormone (TSH) testing, and serum hemoglobin A1C testing. Other investigations were performed only when indicated such as testing for inherited thrombophilias, serum prolactin, oral glucose tolerance test, and endometrial biopsy. Patients were divided into three groups; patients who experienced NVPLs only (pure NVPLs group), patients with only visualized pregnancy losses (pure VPLs group), and patients with history of both NVPLs and VPLs (mixed group). Statistical analysis was performed using Wilcoxon rank-sum tests for continuous variables and Fisher's exact tests for categorical variables. Significance was detected when P values <0.05. A logistic regression model was used to determine the impact of NVPLs and VPLs numbers on any live birth subsequent to the initial RPL clinic visit. MAIN RESULTS AND THE ROLE OF CHANCE: The prevalence of patients with pure NVPLs, pure VPLs, and mixed losses was 14.7% (274/1859), 31.8% (591/1859), and 53.5% (994/1859), respectively. The prevalence of acquired and congenital uterine anomalies diagnosed by HSG or hysteroscopy was significantly different between pure NVPLs, pure VPLs, and mixed groups (16.8% versus 23.7% versus. 20.7%, respectively P = 0.05). There were no significant differences in the results of other RPL investigations or baseline demographics between the three groups. A logistic regression model controlling for maternal age at the initial RPL clinic visit and the follow-up duration showed that the numbers of NVPLs (odds ratio (OR): 0.77, CI: 0.68-0.88) and VPLs (OR: 0.75, CI: 0.64-0.86) are strong predictors for subsequent live births after the initial RPL clinic visit (P < 0.001). The odds of having a live birth decreased by 23% and 25% with each additional NVPL and VPL, respectively. LIMITATIONS, REASONS FOR CAUTION: This study may be limited by its retrospective design. Some of our data, including home pregnancy tests and obstetric history, are based on patient self-reporting, which could have overstated the true prevalence of NVPLs. Another limitation is the lack of available live birth data for all patients at the time of the analysis. WIDER IMPLICATIONS OF THE FINDINGS: To our knowledge, this is the first study to examine and analyze the reproductive outcomes of patients with pure NVPLs in a substantial cohort of patients with RPL. NVPLs seem to affect future live births the same way as clinical miscarriages, which supports their inclusion in RPL definitions. STUDY FUNDING/COMPETING INTEREST(S): This study was supported in part by Canadian Institute Heath Grant (CIHR): Reference Number/W11-179912 and Women's Health Research Institute (WHRI), Vancouver, BC, Canada. M.A.B: Research grants from Canadian Institute for Health Research (CIHR) and Ferring Pharmaceutical. M.A.B. is on the advisory board for AbbVie and Baxter. TRIAL REGISTRATION NUMBER: N/A.

Challenges and Opportunities of Deep Learning for Cough-Based COVID-19 Diagnosis: A Scoping Review.

In the past two years, medical researchers and data scientists worldwide have focused their efforts on containing the pandemic of coronavirus disease 2019 (COVID-19). Deep learning models have been proven to be capable of efficient medical diagnosis and prognosis in cancer, common lung diseases, and COVID-19. On the other hand, artificial neural networks have demonstrated their potential in pattern recognition and classification in various domains, including healthcare. This literature review aims to report the state of research on developing neural network models to diagnose COVID-19 from cough sounds to create a cost-efficient and accessible testing tool in the fight against the pandemic. A total of 35 papers were included in this review following a screening of the 161 outputs of the literature search. We extracted information from articles on data resources, model structures, and evaluation metrics and then explored the scope of experimental studies and methodologies and analyzed their outcomes and limitations. We found that cough is a biomarker, and its associated information can determine an individual's health status. Convolutional neural networks were predominantly used, suggesting they are particularly suitable for feature extraction and classification. The reported accuracy values ranged from 73.1% to 98.5%. Moreover, the dataset sizes ranged from 16 to over 30,000 cough audio samples. Although deep learning is a promising prospect in identifying COVID-19, we identified a gap in the literature on research conducted over large and diversified data sets.

Machine Learning Revealed New Correlates of Chronic Pelvic Pain in Women.

Chronic pelvic pain affects one in seven women worldwide, and there is an urgent need to reduce its associated significant costs and to improve women's health. There are many correlated factors associated with chronic pelvic pain (CPP), and analyzing them simultaneously can be complex and involves many challenges. A newly developed interaction ensemble, referred to as INTENSE, was implemented to investigate this research gap. When applied, INTENSE aggregates three machine learning (ML) methods, which are unsupervised, as follows: interaction principal component analysis (IPCA), hierarchical cluster analysis (HCA), and centroid-based clustering (CBC). For our proposed research, we used INTENSE to uncover novel knowledge, which revealed new interactions in a sample of 656 patients among 25 factors: age, parity, ethnicity, body mass index, endometriosis, irritable bowel syndrome, painful bladder syndrome, pelvic floor tenderness, abdominal wall pain, depression score, anxiety score, Pain Catastrophizing Scale, family history of chronic pain, new or re-referral, age when first experienced pain, pain duration, surgery helpful for pain, infertility, smoking, alcohol use, trauma, dysmenorrhea, deep dyspareunia, CPP, and the Endometriosis Health Profile for functional quality of life. INTENSE indicates that CPP and the Endometriosis Health Profile are correlated with depression score, anxiety score, and the Pain Catastrophizing Scale. Other insights derived from these ML methods include the finding that higher body mass index was clustered with smoking and a history of life trauma. As well, sexual pain (deep dyspareunia) was found to be associated with musculoskeletal pain contributors (abdominal wall pain and pelvic floor tenderness). Therefore, INTENSE provided expert-like reasoning without training any model or prior knowledge of CPP. ML has the potential to identify novel relationships in the etiology of CPP, and thus can drive innovative future research.

Active right atrial emptying fraction predicts reduced survival and increased adverse events in childhood pulmonary arterial hypertension.

BACKGROUND: Right atrial (RA) function has been studied rarely in childhood pulmonary arterial hypertension (PAH). We sought to determine if RA and right ventricular (RV) area changes measured by echocardiography predicted outcomes. METHODS: We reviewed data from children with PAH undergoing cardiac catheterization and echocardiography. RA and RV areas were obtained from the apical 4-chamber view. Clinical worsening indicated initiation of parenteral prostanoid therapy, heart and/or lung transplantation, Potts shunt surgery or death. RESULTS: We studied 57 children (27 females), median age 3 years (range 0.30-17 years), body surface area 0.56 m2 (0.2-1.8), follow up 3 years (0.21-8.35), time to clinical worsening was 1.14 years (0.03-6.14) and mortality was 1.55 years (range 0.88-4.95). We determined from receiver operator curves that RA active emptying fraction (RA EaF) ≥60% predicted clinical worsening (sensitivity 78%, specificity 69%, AUC 0.7) and mortality (sensitivity 100%, specificity 65%, AUC 0.82). RV fractional area change (RVFAC) <25% predicted clinical worsening (sensitivity 72%, specificity 79%, AUC 0.85) and death (sensitivity 67%, specificity 69%, AUC 0.77). The combination of RA EaF ≥60% and RVFAC <33% were best predictors of clinical worsening (sensitivity 72%, specificity 82%, partial AUC 0.65) and mortality (sensitivity 100%, specificity 77%, partial AUC 0.75). CONCLUSION: In childhood PAH, RA EaF ≥ 60% and RVFAC <25% were associated with poor outcomes. RA EaF ≥60% and RVFAC <33% were best predictors of clinical worsening and may be useful markers in children with PAH who require closer observation and more intensive therapy.

Hyperoxia Reduces Oxygen Consumption in Children with Pulmonary Hypertension.

High inspired oxygen concentration (FiO2 > 0.85) is administered to test pulmonary vascular reactivity in children with pulmonary hypertension (PH). It is difficult to measure oxygen consumption (VO2) if the subject is breathing a hyperoxic gas mixture so the assumption is made that baseline VO2 does not change. We hypothesized that hyperoxia changes VO2. We sought to compare the VO2 measured by a thermodilution catheter in room air and hyperoxia. A retrospective review of the hemodynamic data obtained in children with PH who underwent cardiac catheterization was conducted between 2009 and 2014. Cardiac index (CI) was measured by a thermodilution catheter in room air and hyperoxia. VO2 was calculated using the equation CI = VO2/arterial-venous oxygen content difference. Data were available in 24 subjects (males = 10), with median age 8.3 years (0.8-17.6 years), weight 23.3 kg (7.5-95 kg), and body surface area 0.9 m2 (0.4-2.0 m2). In hyperoxia compared with room air, we measured decreased VO2 (154 ± 38 to 136 ± 34 ml/min/m2, p = 0.007), heart rate (91 [Formula: see text] 20 to 83 [Formula: see text] 21 beats/minute, p=0.005), mean pulmonary artery pressure (41 [Formula: see text] 16 to 35 [Formula: see text] 14 mmHg, p=0.024), CI (3.6 [Formula: see text] 0.8 to 3.3 [Formula: see text] 0.9 L/min/m2, p = 0.03), pulmonary vascular resistance (9 [Formula: see text] 6 to 7 [Formula: see text] 3 WU m2, p = 0.029), increased mean aortic (61 [Formula: see text] 11 to 67 [Formula: see text] 11 mmHg, p = 0.005), pulmonary artery wedge pressures (11 [Formula: see text] 8 to 13 [Formula: see text] 9 mmHg, p = 0.006), and systemic vascular resistance (12 [Formula: see text] 6 to 20 [Formula: see text] 7 WU m2, p=0.001). Hyperoxia decreased VO2 and CI and caused pulmonary vasodilation and systemic vasoconstriction in children with PH. The assumption that VO2 remains unchanged in hyperoxia may be incorrect and, if the Fick equation is used, may lead to an overestimation of pulmonary blood flow and underestimation of PVRI.

Spectral analysis of the heart sounds in children with and without pulmonary artery hypertension.

BACKGROUND: Pulmonary artery hypertension (PAH) is difficult to recognize clinically. Digital stethoscopes offer an opportunity to re-evaluate the diagnosis of PAH. We hypothesized that spectral analysis of heart sound frequencies using recordings from a digital stethoscope would differ between children with and without PAH. METHODS: We recorded heart sounds using a digital stethoscope from 27 subjects (12 males) with a median age of 7 years (3 months to 19 years) undergoing simultaneous cardiac catheterization. 13 subjects had a mean pulmonary artery pressure (mPAp)<25 mm Hg (8-24 mm Hg). 14 subjects had a mPAp≥25 mm Hg (25-97 mm Hg). We applied the fast Fourier transform, power spectral analysis, separability testing, and linear discriminant analysis with leave-one-out cross-validation to the heart sounds recorded from the cardiac apex and 2nd left intercostal space (LICS) to examine the frequency domain. The significance of the results was determined using a t-test and rank-sum test. RESULTS: The relative power of the frequencies 21-22 Hz of the heart sounds recorded at the 2nd LICS was decreased significantly in subjects mPAp≥25 mm Hg versus<25 mm Hg. CONCLUSIONS: Heart sound signals of patients with PAH contain significantly less relative power in the band 21-22 Hz compared to subjects with normal PAp. Information contained in the frequency domain may be useful in diagnosing PAH and aid the development of auscultation based techniques for diagnosing PAH. In the future, utilizing the diagnostic information contained in heart sounds recordings may require analysis of both the time and frequency domains.