Risk of bias of prognostic models developed using machine learning: a systematic review in oncology

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.

Performance of externally validated machine learning models based on histopathology images for the diagnosis, classification, prognosis, or treatment outcome prediction in female breast cancer: A systematic review

“…Although PROBAST 42 was originally created for statistical (regressionbased) prediction models, most of its items are applicable to ML-based prediction model studies. 75 As ML models need large sample sizes, insufficient sample sizes when developing and validating ML models can be considered a major design flaw. 75 Although obtaining good results in one or more EV datasets is important evidence of generalizability, 15 it is essential to note that this doesn't guarantee that the model will perform well in all other settings.…”

External Validation of Prognostic Models in Critical Care: A Cautionary Tale From COVID-19 Pneumonitis

“…The models included in this study represent a broad scope of models in the literature and for the majority (8/13 articles), there were insufficient details to allow external validation. This is not the first study to identify deficiencies in reporting quality and that the adoption of the TRIPOD guidelines by authors to clearly report their models is imperative (2, 40–42).…”

Updating M6 pregnancy of unknown location risk‐prediction model including evaluation of clinical factors

“…Machine learning is now playing an increasingly central role in analytics, given its ability to solve complicated problems 47,48 . However, it can be subject to bias evolving from validation procedures, and calibration is often insufficient 47,48 .…”

“…Machine learning is now playing an increasingly central role in analytics, given its ability to solve complicated problems 47,48 . However, it can be subject to bias evolving from validation procedures, and calibration is often insufficient 47,48 . In fact, publications comparing machine learning to statistical logistic regression have concluded equivocal performance [48][49][50] .…”