Improving dengue fever predictions in Taiwan based on feature selection and random forests.

BACKGROUND: Dengue fever is a well-studied vector-borne disease in tropical and subtropical areas of the world. Several methods for predicting the occurrence of dengue fever in Taiwan have been proposed. However, to the best of our knowledge, no study has investigated the relationship between air quality indices (AQIs) and dengue fever in Taiwan. RESULTS: This study aimed to develop a dengue fever prediction model in which meteorological factors, a vector index, and AQIs were incorporated into different machine learning algorithms. A total of 805 meteorological records from 2013 to 2015 were collected from government open-source data after preprocessing. In addition to well-known dengue-related factors, we investigated the effects of novel variables, including particulate matter with an aerodynamic diameter < 10 µm (PM10), PM2.5, and an ultraviolet index, for predicting dengue fever occurrence. The collected dataset was randomly divided into an 80% training set and a 20% test set. The experimental results showed that the random forests achieved an area under the receiver operating characteristic curve of 0.9547 for the test set, which was the best compared with the other machine learning algorithms. In addition, the temperature was the most important factor in our variable importance analysis, and it showed a positive effect on dengue fever at < 30 °C but had less of an effect at > 30 °C. The AQIs were not as important as temperature, but one was selected in the process of filtering the variables and showed a certain influence on the final results. CONCLUSIONS: Our study is the first to demonstrate that AQI negatively affects dengue fever occurrence in Taiwan. The proposed prediction model can be used as an early warning system for public health to prevent dengue fever outbreaks.

The association between interest of nutritional supplements and COVID-19 pandemic - evidence from Google Trends.

BACKGROUND: Due to the spread of the coronavirus disease 2019 (COVID-19) pandemic in 2020, the interest of nutritional supplements has emerged. Limited data are available on how the COVID-19 pandemic affects the search interest in nutritional supplements in Taiwan and worldwide. The study aims to investigate changes in public search interest of nutritional supplements pre- and during the COVID-19 pandemic. METHODS: Our World in Data dataset was used to collect both global and local (Taiwan) number of COVID-19 newly confirmed cases and deaths. Google Trends search query was being used to obtain relative search volumes (RSVs) covering a timeframe between 2019 to 2022. Spearman's rank-order correlation coefficients were used to measure relationships between confirmed new cases and deaths and RSVs of nutritional supplements. Multivariate analysis was conducted to examine the effect of domestic and global new cases and deaths on the RSVs of nutritional supplements. RESULTS: The mean RSVs for nutritional supplements were higher during the COVID-19 pandemic period (between 2020 to 2022) compared to the pre-pandemic period (year of 2019) for both Taiwan and worldwide. In terms of seasonal variations, except for vitamin D, the mean RSVs of probiotics, vitamin B complex, and vitamin C in winter were significantly lower compared to other seasons in Taiwan. The RSVs of nutritional supplements were not only affected by domestic cases and deaths but also by global new cases and deaths. CONCLUSIONS: The interests in nutritional supplements had substantially increased in response to the COVID-19 pandemic. The RSVs of nutritional supplements in Taiwan were not only influenced by global and domestic pandemic severity but also by seasons.

Macroalgae exhibit diverse responses to human disturbances on coral reefs.

Scientists and managers rely on indicator taxa such as coral and macroalgal cover to evaluate the effects of human disturbance on coral reefs, often assuming a universally positive relationship between local human disturbance and macroalgae. Despite evidence that macroalgae respond to local stressors in diverse ways, there have been few efforts to evaluate relationships between specific macroalgae taxa and local human-driven disturbance. Using genus-level monitoring data from 1205 sites in the Indian and Pacific Oceans, we assess whether macroalgae percent cover correlates with local human disturbance while accounting for factors that could obscure or confound relationships. Assessing macroalgae at genus level revealed that no genera were positively correlated with all human disturbance metrics. Instead, we found relationships between the division or genera of algae and specific human disturbances that were not detectable when pooling taxa into a single functional category, which is common to many analyses. The convention to use percent cover of macroalgae as an indication of local human disturbance therefore likely obscures signatures of local anthropogenic threats to reefs. Our limited understanding of relationships between human disturbance, macroalgae taxa, and their responses to human disturbances impedes the ability to diagnose and respond appropriately to these threats.

Global warming and recurrent mass bleaching of corals.

During 2015-2016, record temperatures triggered a pan-tropical episode of coral bleaching, the third global-scale event since mass bleaching was first documented in the 1980s. Here we examine how and why the severity of recurrent major bleaching events has varied at multiple scales, using aerial and underwater surveys of Australian reefs combined with satellite-derived sea surface temperatures. The distinctive geographic footprints of recurrent bleaching on the Great Barrier Reef in 1998, 2002 and 2016 were determined by the spatial pattern of sea temperatures in each year. Water quality and fishing pressure had minimal effect on the unprecedented bleaching in 2016, suggesting that local protection of reefs affords little or no resistance to extreme heat. Similarly, past exposure to bleaching in 1998 and 2002 did not lessen the severity of bleaching in 2016. Consequently, immediate global action to curb future warming is essential to secure a future for coral reefs.

Biogeographical disparity in the functional diversity and redundancy of corals.

Corals are major contributors to a range of key ecosystem functions on tropical reefs, including calcification, photosynthesis, nutrient cycling, and the provision of habitat structure. The abundance of corals is declining at multiple scales, and the species composition of assemblages is responding to escalating human pressures, including anthropogenic global warming. An urgent challenge is to understand the functional consequences of these shifts in abundance and composition in different biogeographical contexts. While global patterns of coral species richness are well known, the biogeography of coral functions in provinces and domains with high and low redundancy is poorly understood. Here, we quantify the functional traits of all currently recognized zooxanthellate coral species (n = 821) in both the Indo-Pacific and Atlantic domains to examine the relationships between species richness and the diversity and redundancy of functional trait space. We find that trait diversity is remarkably conserved (>75% of the global total) along latitudinal and longitudinal gradients in species richness, falling away only in species-poor provinces (n < 200), such as the Persian Gulf (52% of the global total), Hawaii (37%), the Caribbean (26%), and the East-Pacific (20%), where redundancy is also diminished. In the more species-poor provinces, large and ecologically important areas of trait space are empty, or occupied by just a few, highly distinctive species. These striking biogeographical differences in redundancy could affect the resilience of critical reef functions and highlight the vulnerability of relatively depauperate, peripheral locations, which are often a low priority for targeted conservation efforts.

Prediction of robotic neurorehabilitation functional ambulatory outcome in patients with neurological disorders.

INTRODUCTION: Conflicting results persist regarding the effectiveness of robotic-assisted gait training (RAGT) for functional gait recovery in post-stroke survivors. We used several machine learning algorithms to construct prediction models for the functional outcomes of robotic neurorehabilitation in adult patients. METHODS AND MATERIALS: Data of 139 patients who underwent Lokomat training at Taipei Medical University Hospital were retrospectively collected. After screening for data completeness, records of 91 adult patients with acute or chronic neurological disorders were included in this study. Patient characteristics and quantitative data from Lokomat were incorporated as features to construct prediction models to explore early responses and factors associated with patient recovery. RESULTS: Experimental results using the random forest algorithm achieved the best area under the receiver operating characteristic curve of 0.9813 with data extracted from all sessions. Body weight (BW) support played a key role in influencing the progress of functional ambulation categories. The analysis identified negative correlations of BW support, guidance force, and days required to complete 12 Lokomat sessions with the occurrence of progress, while a positive correlation was observed with regard to speed. CONCLUSIONS: We developed a predictive model for ambulatory outcomes based on patient characteristics and quantitative data on impairment reduction with early-stage robotic neurorehabilitation. RAGT is a customized approach for patients with different conditions to regain walking ability. To obtain a more-precise and clearer predictive model, collecting more RAGT training parameters and analyzing them for each individual disorder is a possible approach to help clinicians achieve a better understanding of the most efficient RAGT parameters for different patients. TRIAL REGISTRATION: Retrospectively registered.

Pain prediction model based on machine learning and SHAP values for elders with dementia in Taiwan.

INTRODUCTION: Pain conditions are common in elderly individuals, including those with dementia. However, symptoms associated with dementia may lead to poor recognition, assessment and management of pain. In this study, we incorporated the variables based on questionnaires into a machine learning algorithm to build a prediction model for the pain index of elderly individuals with dementia. MATERIALS AND METHODS: In this study, 113 cases were collected through questionnaires and used to build prediction models for the patient's pain index. Three machine learning algorithms were incorporated for comparison in this study. To interpret the prediction model, SHapley additive explanations values were used to depict the ranking importance of variables and the relationship between features and pain index. RESULTS: In the comparison of models, random forests with feature selection outperformed in terms of root mean square error and mean absolute error. A total of 11 features were selected based on embedded method. The results showed that the Karnofsky scale played a key role in predicting pain index for elderly individuals with dementia and was positively associated with pain index. Arthritis is the most important disease to predicting the pain index. CONCLUSIONS: Our findings provided the key insights to predict the pain index of elderly patients with dementia. In the future, it can be used to develop an application system or webpage, which can reduce the use of labour and improve the efficiency.

Poleward migration of tropical corals inhibited by future trends of seawater temperature and calcium carbonate (CaCO3) saturation.

Poleward range expansion of marine organisms is commonly attributed to anthropogenic ocean warming. However, the extent to which a single species can migrate poleward remains unclear. In this study, we used molecular data to examine the current distribution of the Pocillopora damicornis species complex in Taiwan waters and applied niche modeling to predict its potential range through the end of the 21st Century. The P. damicornis species complex is widespread across shallow, tropical and subtropical waters of the Indo-Pacific regions. Our results revealed that populations from subtropical nonreefal coral communities are P. damicornis, whose native geographical ranges are approximately between 23°N and 35°N. In contrast, those from tropical reefs are P. acuta. Our analysis of 50 environmental data layers demonstrated that the concentrations of CaCO3 polymorphs had the greatest contributions to the distributions of the two species. Future projections under intermediate shared socioeconomic pathways (SSP) 2-4.5 and very high (SSP5-8.5) scenarios of greenhouse gas emissions showed that while sea surface temperature (SST) isotherms would shift northwards, saturation isolines of two CaCO3 polymorphs, calcite (Ωcal) and aragonite (Ωarag), would shift southwards by 2100. Subsequent predictions of future suitable habitats under those conditions indicated that distinct delimitation of geographical ranges for the two species would persist, and neither would extend beyond its native geographical zones, indicating that tropical Taiwan waters are the northern limit for P. acuta. In contrast, subtropical waters are the southern limit for P. damicornis. We concluded that the decline in CaCO3 saturation would make high latitudes less inhabitable, which could be one of the boundary elements that limit poleward range expansion driven by rising SSTs and preserve the latitudinal diversity gradient (LDG) on Earth. Consequently, poleward migration of tropical reef corals to cope with warming oceans should be reevaluated.

XGBoost-Based Simple Three-Item Model Accurately Predicts Outcomes of Acute Ischemic Stroke.

An all-inclusive and accurate prediction of outcomes for patients with acute ischemic stroke (AIS) is crucial for clinical decision-making. This study developed extreme gradient boosting (XGBoost)-based models using three simple factors-age, fasting glucose, and National Institutes of Health Stroke Scale (NIHSS) scores-to predict the three-month functional outcomes after AIS. We retrieved the medical records of 1848 patients diagnosed with AIS and managed at a single medical center between 2016 and 2020. We developed and validated the predictions and ranked the importance of each variable. The XGBoost model achieved notable performance, with an area under the curve of 0.8595. As predicted by the model, the patients with initial NIHSS score > 5, aged over 64 years, and fasting blood glucose > 86 mg/dL were associated with unfavorable prognoses. For patients receiving endovascular therapy, fasting glucose was the most important predictor. The NIHSS score at admission was the most significant predictor for those who received other treatments. Our proposed XGBoost model showed a reliable predictive power of AIS outcomes using readily available and simple predictors and also demonstrated the validity of the model for application in patients receiving different AIS treatments, providing clinical evidence for future optimization of AIS treatment strategies.

Developing an International Classification of Functioning, Disability and Health Core Set for Pediatric Brain Tumor Survivors in Chinese Clinical Settings.

BACKGROUND: The International Classification of Functioning, Disability and Health (ICF) core set (CS) facilitates the standardization of functioning and impairment assessment for integration of holistic care. OBJECTIVE: This study developed an ICF CS for interviewing pediatric brain tumor survivors in Taiwan to help healthcare professionals in implementing disability assessment and management measures. METHODS: A group of 29 experts in 10 relevant fields with at least 5 years of experience working with children with brain tumors participated in this study. The first questionnaire contained 247 second-level ICF categories. The experts rated the significance of each category by using a 5-point Likert scale. Correlations between individual and group scores were calculated to determine consensus. Categories with an average rating of higher than 4 and for which greater than or equal to 80% (23) of the participants provided a rating of 4 or higher were included in the final CS. RESULTS: The final CS contained a total of 57 ICF categories: 20 from the Body Functions and Structures component, 36 from the Activities and Participation component, and 1 from the Environmental Factors component. CONCLUSION: The ICF CS for pediatric brain tumor survivors provides a framework for relevant healthcare professionals to deliver patient-centered care, ensuring that services focus on all areas of development. IMPLICATIONS FOR PRACTICE: Patient ratings for this ICF CS may serve as a new practical and effective patient-reported information tool for acquiring patient input and for the systematic monitoring of pediatric brain tumor survivors in clinical practice. Further research should be conducted on this CS to verify our findings.

Constructing a Learning Curve to Discuss the Medical Treatments and the Effect of Vaccination of COVID-19.

Acknowledging the extreme risk COVID-19 poses to humans, this paper attempted to analyze and compare case fatality rates, identify the existence of learning curves for COVID-19 medical treatments, and examine the impact of vaccination on fatality rate reduction. Confirmed cases and deaths were extracted from the "Daily Situation Report" provided by the World Health Organization. The results showed that low registration and low viral test rates resulted in low fatality rates, and the learning curve was significant for all countries except China. Treatment for COVID-19 can be improved through repeated experience. Vaccinations in the U.K. and U.S.A. are highly effective in reducing fatality rates, but not in other countries. The positive impact of vaccines may be attributed to higher vaccination rates. In addition to China, this study identified the existence of learning curves for the medical treatment of COVID-19 that can explain the effect of vaccination rates on fatalities.

Six years of demography data for 11 reef coral species.

Scleractinian corals are colonial animals with a range of life-history strategies, making up diverse species assemblages that define coral reefs. We tagged and tracked ~30 colonies from each of 11 species during seven trips spanning 6 years (2009-2015) to measure their vital rates and competitive interactions on the reef crest at Trimodal Reef, Lizard Island, Australia. Pairs of species were chosen from five growth forms in which one species of the pair was locally rare (R) and the other common (C). The sampled growth forms were massive (Goniastrea pectinata [R] and G. retiformis [C]), digitate (Acropora humilis [R] and A. cf. digitifera [C]), corymbose (A. millepora [R] and A. nasuta [C]), tabular (A. cytherea [R] and A. hyacinthus [C]) and arborescent (A. robusta [R] and A. intermedia [C]). An extra corymbose species with intermediate abundance, A. spathulata was included when it became apparent that A. millepora was too rare on the reef crest, making the 11 species in total. The tagged colonies were visited each year in the weeks prior to spawning. During visits, two or more observers each took two or three photographs of each tagged colony from directly above and on the horizontal plane with a scale plate to track planar area. Dead or missing colonies were recorded and new colonies tagged to maintain ~30 colonies per species throughout the 6 years of the study. In addition to tracking tagged corals, 30 fragments were collected from neighboring untagged colonies of each species for counting numbers of eggs per polyp (fecundity); and fragments of untagged colonies were brought into the laboratory where spawned eggs were collected for biomass and energy measurements. We also conducted surveys at the study site to generate size structure data for each species in several of the years. Each tagged colony photograph was digitized by at least two people. Therefore, we could examine sources of error in planar area for both photographers and outliners. Competitive interactions were recorded for a subset of species by measuring the margins of tagged colony outlines interacting with neighboring corals. The study was abruptly ended by Tropical Cyclone Nathan (Category 4) that killed all but nine of the more than 300 tagged colonies in early 2015. Nonetheless, these data will be of use to other researchers interested in coral demography and coexistence, functional ecology, and parametrizing population, community, and ecosystem models. The data set is not copyright restricted, and users should cite this paper when using the data.

Symbiont communities and host genetic structure of the brain coral Platygyra verweyi, at the outlet of a nuclear power plant and adjacent areas.

In the context of rising seawater temperatures associated with climate change, the issue of whether coral holobionts deal with this challenge by shuffling their associations with stress- and/or heat-tolerant Symbiodinium, by generating heat-resistant host genotypes, or both is important for coral survival. In this study, the composition of communities of the endosymbiont Symbiodinium and the population genetics of the coral host Platygyra verweyi were examined in a reef impacted by hot-water discharged from the outlet of a nuclear power plant in operation in Kenting, Southern Taiwan since 1984. The water at this site is 2.0-3.0 °C warmer than adjacent reefs in summer, which have an average seawater temperature of 29.0 °C. The data were compared with those for the same species at other sites within 12 km of the outlet site. Platygyra verwyei was associated with one or both of Symbiodinium types C3 (heat sensitive) and D1a (heat tolerant) at all sites with the latter being the dominant at the nuclear power plant outlet. The proportion of C3 in populations increased gradually with increasing distance from the hot-water discharge. Genetic analysis of the Platygyra verweyi host using mitochondrial and nuclear markers showed no genetic differentiation among sites. Changes in the composition of Symbiodinium types associated with P. verweyi among closely located sites in Kenting suggested that this coral might have acclimatized to the constant thermal stress by selective association with heat-tolerant Symbiodinium types, whereas the role of the host in adaptation was inconclusive.

Predicting clinical pregnancy using clinical features and machine learning algorithms in in vitro fertilization.

INTRODUCTION: Assisted reproductive technology has been proposed for women with infertility. Moreover, in vitro fertilization (IVF) cycles are increasing. Factors contributing to successful pregnancy have been widely explored. In this study, we used machine learning algorithms to construct prediction models for clinical pregnancies in IVF. MATERIALS AND METHODS: A total of 24,730 patients entered IVF and intracytoplasmic sperm injection cycles with clinical pregnancy outcomes at Taipei Medical University Hospital. Data used included patient characteristics and treatment. We used machine learning methods to develop prediction models for clinical pregnancy and explored how each variable affects the outcome of interest using partial dependence plots. RESULTS: Experimental results showed that the random forest algorithm outperforms logistic regression in terms of areas under the receiver operating characteristics curve. The ovarian stimulation protocol is the most important factor affecting pregnancy outcomes. Long and ultra-long protocols have shown positive effects on clinical pregnancy among all protocols. Furthermore, total frozen and transferred embryos are positive for a clinical pregnancy, but female age and duration of infertility have negative effects on clinical pregnancy. CONCLUSION: Our findings show the importance of variables and propensity of each variable by random forest algorithm for clinical pregnancy in the assisted reproductive technology cycle. This study provides a ranking of variables affecting clinical pregnancy and explores the effects of each treatment on successful pregnancy. Our study has the potential to help clinicians evaluate the success of IVF in patients.

Correction: Visual light perceptions caused by medical linear accelerator: Findings of machine-learning algorithms in a prospective questionnaire-based case-control study.

[This corrects the article DOI: 10.1371/journal.pone.0247597.].

Visual light perceptions caused by medical linear accelerator: Findings of machine-learning algorithms in a prospective questionnaire-based case-control study.

This study aimed to investigate the possible incidence of visual light perceptions (VLPs) during radiation therapy (RT). We analyzed whether VLPs could be affected by differences in the radiation energy, prescription doses, age, sex, or RT locations, and whether all VLPs were caused by radiation. From November 2016 to August 2018, a total of 101 patients who underwent head-and-neck or brain RT were screened. After receiving RT, questionnaires were completed, and the subjects were interviewed. Random forests (RF), a tree-based machine learning algorithm, and logistic regression (LR) analyses were compared by the area under the curve (AUC), and the algorithm that achieved the highest AUC was selected. The dataset sample was based on treatment with non-human units, and a total of 293 treatment fields from 78 patients were analyzed. VLPs were detected only in 122 of the 293 exposure portals (40.16%). The dataset was randomly divided into 80% and 20% as the training set and test set, respectively. In the test set, RF achieved an AUC of 0.888, whereas LR achieved an AUC of 0.773. In this study, the retina fraction dose was the most important continuous variable and had a positive effect on VLP. Age was the most important categorical variable. In conclusion, the visual light perception phenomenon by the human body during RT is induced by radiation rather than being a self-suggested hallucination or induced by phosphenes.

An Indo-Pacific coral spawning database.

The discovery of multi-species synchronous spawning of scleractinian corals on the Great Barrier Reef in the 1980s stimulated an extraordinary effort to document spawning times in other parts of the globe. Unfortunately, most of these data remain unpublished which limits our understanding of regional and global reproductive patterns. The Coral Spawning Database (CSD) collates much of these disparate data into a single place. The CSD includes 6178 observations (3085 of which were unpublished) of the time or day of spawning for over 300 scleractinian species in 61 genera from 101 sites in the Indo-Pacific. The goal of the CSD is to provide open access to coral spawning data to accelerate our understanding of coral reproductive biology and to provide a baseline against which to evaluate any future changes in reproductive phenology.

Leptoria phrygia in Southern Taiwan shuffles and switches symbionts to resist thermal-induced bleaching.

Symbiodiniaceae communities in some corals often shuffle or switch after severe bleaching events, one of the major threats to coral survival in a world with climate change. In this study we reciprocally transplanted five Leptoria phrygia colonies between two sites with significantly different temperature regimes and monitored them for 12 months. Our ITS2 amplicon deep sequencing demonstrated that L. phrygia acclimatized to maintain a strong and stable association with Durusdinium D17, D. trenchii, and D. glynnii, but also remained flexible and formed a short-term association with different Cladocopium. Most interestingly, two colonies shuffled between Durusdinium and Cladocopium without the occurrence of bleaching; one colony even switched its dominant Cladocopium after generic shuffling. Both dominant Cladocopium were originally rare with relative abundances as low as 0.024%. This is the first record of adult corals switching dominant symbiont without bleaching.

Demographic census confirms a stable population of the critically-endangered caryophyllid coral Polycyathus chaishanensis (Scleractinia; Caryophyllidae) in the Datan Algal Reef, Taiwan.

Polycyathus chaishanensis is a symbiotic caryophyllid coral described from a single population in a tidal pool off Chaishan, Kaohsiung, Taiwan. Due to its rarity, P. chaishanensis was declared a critically-endangered species under the Taiwan Wildlife Protection Act. In May 2017, a P. chaishanensis colony was discovered in the intertidal area of the Datan Algal Reef, Taoyuan, Taiwan. To determine whether this is a stable population in the algal reef, a demographic census-including data on occurrence, distribution, and colony size-was carried out in the algal reef in southern Taoyuan. Intertidal censuses and sediment collections were conducted at five different sections-Baiyu, Datan G1, Datan G2, Yongxing, and Yongan algal reefs-during the monthly spring low tide from July 2018 to January 2019. In total, 84 colonies-23 in Datan G1 and 61 in Datan G2-were recorded from a tidal range of - 160 to - 250 cm, according to the Taiwan Vertical Datum 2001 compiled by the Central Weather Bureau. No P. chaishanensis was found in Baiyu, Yongxing, or Yongan. The P. chaishanensis colony sizes ranged from 2.55 to 81.5 cm in diameter, with the larger P. chaishanensis present in the lower intertidal zone. Sediment was extremely high, with monthly site averages ranging from 3,818.26 to 29,166.88 mg cm-2 day-1, and there was a significant difference between sites and months, both of which affected the distribution of P. chaishanensis in the algal reef. Our study confirms the existence of a second population of P. chaishanensis in Taiwan, highlighting the importance of the Datan Algal Reef for the survival and protection of this critically-endangered caryophyllid coral and why it is so urgent that the reef should be conserved.

Comparison of Multivariable Logistic Regression and Other Machine Learning Algorithms for Prognostic Prediction Studies in Pregnancy Care: Systematic Review and Meta-Analysis.

BACKGROUND: Predictions in pregnancy care are complex because of interactions among multiple factors. Hence, pregnancy outcomes are not easily predicted by a single predictor using only one algorithm or modeling method. OBJECTIVE: This study aims to review and compare the predictive performances between logistic regression (LR) and other machine learning algorithms for developing or validating a multivariable prognostic prediction model for pregnancy care to inform clinicians' decision making. METHODS: Research articles from MEDLINE, Scopus, Web of Science, and Google Scholar were reviewed following several guidelines for a prognostic prediction study, including a risk of bias (ROB) assessment. We report the results based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Studies were primarily framed as PICOTS (population, index, comparator, outcomes, timing, and setting): Population: men or women in procreative management, pregnant women, and fetuses or newborns; Index: multivariable prognostic prediction models using non-LR algorithms for risk classification to inform clinicians' decision making; Comparator: the models applying an LR; Outcomes: pregnancy-related outcomes of procreation or pregnancy outcomes for pregnant women and fetuses or newborns; Timing: pre-, inter-, and peripregnancy periods (predictors), at the pregnancy, delivery, and either puerperal or neonatal period (outcome), and either short- or long-term prognoses (time interval); and Setting: primary care or hospital. The results were synthesized by reporting study characteristics and ROBs and by random effects modeling of the difference of the logit area under the receiver operating characteristic curve of each non-LR model compared with the LR model for the same pregnancy outcomes. We also reported between-study heterogeneity by using τ2 and I2. RESULTS: Of the 2093 records, we included 142 studies for the systematic review and 62 studies for a meta-analysis. Most prediction models used LR (92/142, 64.8%) and artificial neural networks (20/142, 14.1%) among non-LR algorithms. Only 16.9% (24/142) of studies had a low ROB. A total of 2 non-LR algorithms from low ROB studies significantly outperformed LR. The first algorithm was a random forest for preterm delivery (logit AUROC 2.51, 95% CI 1.49-3.53; I2=86%; τ2=0.77) and pre-eclampsia (logit AUROC 1.2, 95% CI 0.72-1.67; I2=75%; τ2=0.09). The second algorithm was gradient boosting for cesarean section (logit AUROC 2.26, 95% CI 1.39-3.13; I2=75%; τ2=0.43) and gestational diabetes (logit AUROC 1.03, 95% CI 0.69-1.37; I2=83%; τ2=0.07). CONCLUSIONS: Prediction models with the best performances across studies were not necessarily those that used LR but also used random forest and gradient boosting that also performed well. We recommend a reanalysis of existing LR models for several pregnancy outcomes by comparing them with those algorithms that apply standard guidelines. TRIAL REGISTRATION: PROSPERO (International Prospective Register of Systematic Reviews) CRD42019136106; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=136106.