Risk of bias of prognostic models developed using machine learning: a systematic review in oncology.
Ontology highlight
ABSTRACT: Background
Prognostic models are used widely in the oncology domain to guide medical decision-making. Little is known about the risk of bias of prognostic models developed using machine learning and the barriers to their clinical uptake in the oncology domain.Methods
We conducted a systematic review and searched MEDLINE and EMBASE databases for oncology-related studies developing a prognostic model using machine learning methods published between 01/01/2019 and 05/09/2019. The primary outcome was risk of bias, judged using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). We described risk of bias overall and for each domain, by development and validation analyses separately.Results
We included 62 publications (48 development-only; 14 development with validation). 152 models were developed across all publications and 37 models were validated. 84% (95% CI: 77 to 89) of developed models and 51% (95% CI: 35 to 67) of validated models were at overall high risk of bias. Bias introduced in the analysis was the largest contributor to the overall risk of bias judgement for model development and validation. 123 (81%, 95% CI: 73.8 to 86.4) developed models and 19 (51%, 95% CI: 35.1 to 67.3) validated models were at high risk of bias due to their analysis, mostly due to shortcomings in the analysis including insufficient sample size and split-sample internal validation.Conclusions
The quality of machine learning based prognostic models in the oncology domain is poor and most models have a high risk of bias, contraindicating their use in clinical practice. Adherence to better standards is urgently needed, with a focus on sample size estimation and analysis methods, to improve the quality of these models.
SUBMITTER: Dhiman P
PROVIDER: S-EPMC9261114 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:ObjectiveTo assess the methodological quality of studies on prediction models developed using machine learning techniques across all medical specialties.DesignSystematic review.Data sourcesPubMed from 1 January 2018 to 31 December 2019.Eligibility criteriaArticles reporting on the development, with or without external validation, of a multivariable prediction model (diagnostic or prognostic) developed using supervised machine learning for individualised predictions. No restrictions applied for study design, data source, or predicted patient related health outcomes.Review methodsMethodological quality of the studies was determined and risk of bias evaluated using the prediction risk of bias assessment tool (PROBAST). This tool contains 21 signalling questions tailored to identify potential biases in four domains. Risk of bias was measured for each domain (participants, predictors, outcome, and analysis) and each study (overall).Results152 studies were included: 58 (38%) included a diagnostic prediction model and 94 (62%) a prognostic prediction model. PROBAST was applied to 152 developed models and 19 external validations. Of these 171 analyses, 148 (87%, 95% confidence interval 81% to 91%) were rated at high risk of bias. The analysis domain was most frequently rated at high risk of bias. Of the 152 models, 85 (56%, 48% to 64%) were developed with an inadequate number of events per candidate predictor, 62 handled missing data inadequately (41%, 33% to 49%), and 59 assessed overfitting improperly (39%, 31% to 47%). Most models used appropriate data sources to develop (73%, 66% to 79%) and externally validate the machine learning based prediction models (74%, 51% to 88%). Information about blinding of outcome and blinding of predictors was, however, absent in 60 (40%, 32% to 47%) and 79 (52%, 44% to 60%) of the developed models, respectively.ConclusionMost studies on machine learning based prediction models show poor methodological quality and are at high risk of bias. Factors contributing to risk of bias include small study size, poor handling of missing data, and failure to deal with overfitting. Efforts to improve the design, conduct, reporting, and validation of such studies are necessary to boost the application of machine learning based prediction models in clinical practice.Systematic review registrationPROSPERO CRD42019161764.
| S-EPMC8527348 | biostudies-literature
Project description:BackgroundWhile many studies have consistently found incomplete reporting of regression-based prediction model studies, evidence is lacking for machine learning-based prediction model studies. We aim to systematically review the adherence of Machine Learning (ML)-based prediction model studies to the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) Statement.MethodsWe included articles reporting on development or external validation of a multivariable prediction model (either diagnostic or prognostic) developed using supervised ML for individualized predictions across all medical fields. We searched PubMed from 1 January 2018 to 31 December 2019. Data extraction was performed using the 22-item checklist for reporting of prediction model studies ( www.TRIPOD-statement.org ). We measured the overall adherence per article and per TRIPOD item.ResultsOur search identified 24,814 articles, of which 152 articles were included: 94 (61.8%) prognostic and 58 (38.2%) diagnostic prediction model studies. Overall, articles adhered to a median of 38.7% (IQR 31.0-46.4%) of TRIPOD items. No article fully adhered to complete reporting of the abstract and very few reported the flow of participants (3.9%, 95% CI 1.8 to 8.3), appropriate title (4.6%, 95% CI 2.2 to 9.2), blinding of predictors (4.6%, 95% CI 2.2 to 9.2), model specification (5.2%, 95% CI 2.4 to 10.8), and model's predictive performance (5.9%, 95% CI 3.1 to 10.9). There was often complete reporting of source of data (98.0%, 95% CI 94.4 to 99.3) and interpretation of the results (94.7%, 95% CI 90.0 to 97.3).ConclusionSimilar to prediction model studies developed using conventional regression-based techniques, the completeness of reporting is poor. Essential information to decide to use the model (i.e. model specification and its performance) is rarely reported. However, some items and sub-items of TRIPOD might be less suitable for ML-based prediction model studies and thus, TRIPOD requires extensions. Overall, there is an urgent need to improve the reporting quality and usability of research to avoid research waste.Systematic review registrationPROSPERO, CRD42019161764.
| S-EPMC8759172 | biostudies-literature
Project description:BackgroundRisk-of-bias assessments are now a standard component of systematic reviews. At present, reviewers need to manually identify relevant parts of research articles for a set of methodological elements that affect the risk of bias, in order to make a risk-of-bias judgement for each of these elements. We investigate the use of text mining methods to automate risk-of-bias assessments in systematic reviews. We aim to identify relevant sentences within the text of included articles, to rank articles by risk of bias and to reduce the number of risk-of-bias assessments that the reviewers need to perform by hand.MethodsWe use supervised machine learning to train two types of models, for each of the three risk-of-bias properties of sequence generation, allocation concealment and blinding. The first model predicts whether a sentence in a research article contains relevant information. The second model predicts a risk-of-bias value for each research article. We use logistic regression, where each independent variable is the frequency of a word in a sentence or article, respectively.ResultsWe found that sentences can be successfully ranked by relevance with area under the receiver operating characteristic (ROC) curve (AUC)?> 0.98. Articles can be ranked by risk of bias with AUC > 0.72. We estimate that more than 33% of articles can be assessed by just one reviewer, where two reviewers are normally required.ConclusionsWe show that text mining can be used to assist risk-of-bias assessments.
| S-EPMC4795562 | biostudies-literature
Project description:BackgroundRacial bias is a key concern regarding the development, validation, and implementation of machine learning (ML) models in clinical settings. Despite the potential of bias to propagate health disparities, racial bias in clinical ML has yet to be thoroughly examined and best practices for bias mitigation remain unclear.ObjectiveOur objective was to perform a scoping review to characterize the methods by which the racial bias of ML has been assessed and describe strategies that may be used to enhance algorithmic fairness in clinical ML.MethodsA scoping review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) Extension for Scoping Reviews. A literature search using PubMed, Scopus, and Embase databases, as well as Google Scholar, identified 635 records, of which 12 studies were included.ResultsApplications of ML were varied and involved diagnosis, outcome prediction, and clinical score prediction performed on data sets including images, diagnostic studies, clinical text, and clinical variables. Of the 12 studies, 1 (8%) described a model in routine clinical use, 2 (17%) examined prospectively validated clinical models, and the remaining 9 (75%) described internally validated models. In addition, 8 (67%) studies concluded that racial bias was present, 2 (17%) concluded that it was not, and 2 (17%) assessed the implementation of bias mitigation strategies without comparison to a baseline model. Fairness metrics used to assess algorithmic racial bias were inconsistent. The most commonly observed metrics were equal opportunity difference (5/12, 42%), accuracy (4/12, 25%), and disparate impact (2/12, 17%). All 8 (67%) studies that implemented methods for mitigation of racial bias successfully increased fairness, as measured by the authors' chosen metrics. Preprocessing methods of bias mitigation were most commonly used across all studies that implemented them.ConclusionsThe broad scope of medical ML applications and potential patient harms demand an increased emphasis on evaluation and mitigation of racial bias in clinical ML. However, the adoption of algorithmic fairness principles in medicine remains inconsistent and is limited by poor data availability and ML model reporting. We recommend that researchers and journal editors emphasize standardized reporting and data availability in medical ML studies to improve transparency and facilitate evaluation for racial bias.
| S-EPMC9198828 | biostudies-literature
Project description:PurposeThe purpose of this systematic review was to assess risk of bias in existing prognostic models of hospital-induced delirium for medical-surgical units.MethodsAPA PsycInfo, CINAHL, MEDLINE, and Web of Science Core Collection were searched on July 8, 2022, to identify original studies which developed and validated prognostic models of hospital-induced delirium for adult patients who were hospitalized in medical-surgical units. The Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies was used for data extraction. The Prediction Model Risk of Bias Assessment Tool was used to assess risk of bias. Risk of bias was assessed across four domains: participants, predictors, outcome, and analysis.ResultsThirteen studies were included in the qualitative synthesis, including ten model development and validation studies and three model validation only studies. The methods in all of the studies were rated to be at high overall risk of bias. The methods of statistical analysis were the greatest source of bias. External validity of models in the included studies was tested at low levels of transportability.ConclusionsOur findings highlight the ongoing scientific challenge of developing a valid prognostic model of hospital-induced delirium for medical-surgical units to tailor preventive interventions to patients who are at high risk of this iatrogenic condition. With limited knowledge about generalizable prognosis of hospital-induced delirium in medical-surgical units, existing prognostic models should be used with caution when creating clinical practice policies. Future research protocols must include robust study designs which take into account the perspectives of clinicians to identify and validate risk factors of hospital-induced delirium for accurate and generalizable prognosis in medical-surgical units.
| S-EPMC10434879 | biostudies-literature
Project description:BackgroundWith the rising prevalence of diabetes, machine learning (ML) models have been increasingly used for prediction of diabetes and its complications, due to their ability to handle large complex data sets. This study aims to evaluate the quality and performance of ML models developed to predict microvascular and macrovascular diabetes complications in an adult Type 2 diabetes population.MethodsA systematic review was conducted in MEDLINE®, Embase®, the Cochrane® Library, Web of Science®, and DBLP Computer Science Bibliography databases according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist. Studies that developed or validated ML prediction models for microvascular or macrovascular complications in people with Type 2 diabetes were included. Prediction performance was evaluated using area under the receiver operating characteristic curve (AUC). An AUC >0.75 indicates clearly useful discrimination performance, while a positive mean relative AUC difference indicates better comparative model performance.ResultsOf 13 606 articles screened, 32 studies comprising 87 ML models were included. Neural networks (n = 15) were the most frequently utilized. Age, duration of diabetes, and body mass index were common predictors in ML models. Across predicted outcomes, 36% of the models demonstrated clearly useful discrimination. Most ML models reported positive mean relative AUC compared with non-ML methods, with random forest showing the best overall performance for microvascular and macrovascular outcomes. Majority (n = 31) of studies had high risk of bias.ConclusionsRandom forest was found to have the overall best prediction performance. Current ML prediction models remain largely exploratory, and external validation studies are required before their clinical implementation.Protocol registrationOpen Science Framework (registration number: 10.17605/OSF.IO/UP49X).
| S-EPMC10012374 | biostudies-literature
Project description:IntroductionApproximately 7,000 newborns die every day, accounting for almost half of child deaths under 5 years of age. Deciphering which neonates are at increased risk for mortality can have an important global impact. As such, integrating high computational technology (e.g., artificial intelligence [AI]) may help identify the early and potentially modifiable predictors of neonatal mortality. Therefore, the objective of this study was to collate, critically appraise, and analyze neonatal prediction studies that included AI.MethodsA literature search was performed in PubMed, Cochrane, OVID, and Google Scholar. We included studies that used AI (e.g., machine learning (ML) and deep learning) to formulate prediction models for neonatal death. We excluded small studies (n < 500 individuals) and studies using only antenatal factors to predict mortality. Two independent investigators screened all articles for inclusion. The data collection consisted of study design, number of models, features used per model, feature importance, internal and/or external validation, and calibration analysis. Our primary outcome was the average area under the receiving characteristic curve (AUC) or sensitivity and specificity for all models included in each study.ResultsOf 434 articles, 11 studies were included. The total number of participants was 1.26 M with gestational ages ranging from 22 weeks to term. Number of features ranged from 3 to 66 with timing of prediction as early as 5 min of life to a maximum of 7 days of age. The average number of models per study was 4, with neural network, random forest, and logistic regression comprising the most used models (58.3%). Five studies (45.5%) reported calibration plots and 2 (18.2%) conducted external validation. Eight studies reported results by AUC and 5 studies reported the sensitivity and specificity. The AUC varied from 58.3% to 97.0%. The mean sensitivities ranged from 63% to 80% and specificities from 78% to 99%. The best overall model was linear discriminant analysis, but it also had a high number of features (n = 17).Discussion/conclusionML models can accurately predict death in neonates. This analysis demonstrates the most commonly used predictors and metrics for AI prediction models for neonatal mortality. Future studies should focus on external validation, calibration, as well as deployment of applications that can be readily accessible to health-care providers.
| S-EPMC8887024 | biostudies-literature
Project description:BackgroundThe lack of nonparametric statistical tests for confounding bias significantly hampers the development of robust, valid, and generalizable predictive models in many fields of research. Here I propose the partial confounder test, which, for a given confounder variable, probes the null hypotheses of the model being unconfounded.ResultsThe test provides a strict control for type I errors and high statistical power, even for nonnormally and nonlinearly dependent predictions, often seen in machine learning. Applying the proposed test on models trained on large-scale functional brain connectivity data (N= 1,865) (i) reveals previously unreported confounders and (ii) shows that state-of-the-art confound mitigation approaches may fail preventing confounder bias in several cases.ConclusionsThe proposed test (implemented in the package mlconfound; https://mlconfound.readthedocs.io) can aid the assessment and improvement of the generalizability and validity of predictive models and, thereby, fosters the development of clinically useful machine learning biomarkers.
| S-EPMC9412867 | biostudies-literature
Project description:ObjectiveEvaluate the completeness of reporting of prognostic prediction models developed using machine learning methods in the field of oncology.Study design and settingWe conducted a systematic review, searching the MEDLINE and Embase databases between 01/01/2019 and 05/09/2019, for non-imaging studies developing a prognostic clinical prediction model using machine learning methods (as defined by primary study authors) in oncology. We used the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement to assess the reporting quality of included publications. We described overall reporting adherence of included publications and by each section of TRIPOD.ResultsSixty-two publications met the inclusion criteria. 48 were development studies and 14 were development with validation studies. 152 models were developed across all publications. Median adherence to TRIPOD reporting items was 41% [range: 10%-67%] and at least 50% adherence was found in 19% (n=12/62) of publications. Adherence was lower in development only studies (median: 38% [range: 10%-67%]); and higher in development with validation studies (median: 49% [range: 33%-59%]).ConclusionReporting of clinical prediction models using machine learning in oncology is poor and needs urgent improvement, so readers and stakeholders can appraise the study methods, understand study findings, and reduce research waste.
| S-EPMC8592577 | biostudies-literature
Project description:IntroductionPre-eclampsia is one of the most serious clinical problems of pregnancy that contribute significantly to maternal mortality worldwide. This systematic review aims to identify and summarise the predictive factors of pre-eclampsia using machine learning models and evaluate the diagnostic accuracy of machine learning models in predicting pre-eclampsia.Methods and analysisThis review adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. This search strategy includes the search for published studies from inception to January 2023. Databases include the Cochrane Central Register, PubMed, EMBASE, ProQuest, Scopus and Google Scholar. Search terms include 'preeclampsia' AND 'artificial intelligence' OR 'machine learning' OR 'deep learning'. All studies that used machine learning-based analysis for predicting pre-eclampsia in pregnant women will be considered. Non-English articles and those that are unrelated to the topic will be excluded. PROBAST (Prediction model Risk Of Bias ASsessment Tool) will be used to assess the risk of bias and the applicability of each included study.Ethics and disseminationEthical approval is not required, as our review will include published and publicly accessible data. Findings from this review will be disseminated via publication in a peer-review journal.Prospero registration numberThis review is registered with PROSPERO (ID: CRD42023432415).
| S-EPMC10496701 | biostudies-literature