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

Abstract

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 <i>τ<sup>2</sup></i> and <i>I<sup>2</sup></i>. 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; <i>I<sup>2</sup></i>=86%; <i>τ<sup>2</sup></i>=0.77) and pre-eclampsia (logit AUROC 1.2, 95% CI 0.72-1.67; <i>I<sup>2</sup></i>=75%; <i>τ<sup>2</sup></i>=0.09). The second algorithm was gradient boosting for cesarean section (logit AUROC 2.26, 95% CI 1.39-3.13; <i>I<sup>2</sup></i>=75%; <i>τ<sup>2</sup></i>=0.43) and gestational diabetes (logit AUROC 1.03, 95% CI 0.69-1.37; <i>I<sup>2</sup></i>=83%; <i>τ<sup>2</sup></i>=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. CLINICALTRIAL PROSPERO (International Prospective Register of Systematic Reviews) CRD42019136106; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=136106

Human and machine learning pipelines for responsible clinical prediction using high-dimensional data

“…Several health outcomes have been predicted by machine learning algorithms with satisfying performances for both categorical and numerical outcomes. [1][2][3][4] Recently, the most of well-known successes are deep-learning models that surpass human-level performance for diagnostic tasks. [5][6][7] However, prognostication with causal reasoning is warranted for prevention purpose.…”

“…For prognostication, these algorithms were found to outperform others for similar outcomes. 1 We also developed a deep-insight visible neural network (DI-VNN) protocol, 12 according to recent studies. 14,15 But, unlike any of those studies, our protocol allows a human to deeply explore the 'subconscious mind' of the machine learning prediction model to gain insights and to identify bias exploited for predictions.…”

Deep-insight visible neural network (DI-VNN) for improving interpretability of a non-image deep learning model by data-driven ontology

“…Clinical prediction in medicine have been extended by more machine learning algorithms with promising predictive performances, either to classify categorical or to estimate numerical outcomes. [1][2][3][4] Furthermore, recent deep-learning models were also be the most of well-known examples of prediction by machine, which surpasses human-level performance; yet, these were mostly for diagnosis. [5][6][7] Prognosis with causal reasoning, however, is more compelling.…”

Human and machine learning pipelines for responsible clinical prediction using high-dimensional data

“…Several health outcomes have been predicted by machine learning algorithms with satisfying performances for both categorical and numerical outcomes. [1][2][3][4] Recently, the most of well-known successes are deep-learning models that surpass human-level performance for diagnostic tasks. [5][6][7] However, prognostication with causal reasoning is warranted for prevention purpose.…”

“…For prognostication, these algorithms were found to outperform others for similar outcomes. 1 We also developed a deep-insight visible neural network (DI-VNN) protocol, 12 according to recent studies. 14,15 But, unlike any of those studies, our protocol allows a human to deeply explore the 'subconscious mind' of the machine learning prediction model to gain insights and to identify bias exploited for predictions.…”