Dataset Information

Association of Clinician Diagnostic Performance With Machine Learning-Based Decision Support Systems: A Systematic Review.

ABSTRACT:

Importance

An increasing number of machine learning (ML)-based clinical decision support systems (CDSSs) are described in the medical literature, but this research focuses almost entirely on comparing CDSS directly with clinicians (human vs computer). Little is known about the outcomes of these systems when used as adjuncts to human decision-making (human vs human with computer).

Objectives

To conduct a systematic review to investigate the association between the interactive use of ML-based diagnostic CDSSs and clinician performance and to examine the extent of the CDSSs' human factors evaluation.

Evidence review

A search of MEDLINE, Embase, PsycINFO, and grey literature was conducted for the period between January 1, 2010, and May 31, 2019. Peer-reviewed studies published in English comparing human clinician performance with and without interactive use of an ML-based diagnostic CDSSs were included. All metrics used to assess human performance were considered as outcomes. The risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) and Risk of Bias in Non-Randomised Studies-Intervention (ROBINS-I). Narrative summaries were produced for the main outcomes. Given the heterogeneity of medical conditions, outcomes of interest, and evaluation metrics, no meta-analysis was performed.

Findings

A total of 8112 studies were initially retrieved and 5154 abstracts were screened; of these, 37 studies met the inclusion criteria. The median number of participating clinicians was 4 (interquartile range, 3-8). Of the 107 results that reported statistical significance, 54 (50%) were increased by the use of CDSSs, 4 (4%) were decreased, and 49 (46%) showed no change or an unclear change. In the subgroup of studies carried out in representative clinical settings, no association between the use of ML-based diagnostic CDSSs and improved clinician performance could be observed. Interobserver agreement was the commonly reported outcome whose change was the most strongly associated with CDSS use. Four studies (11%) reported on user feedback, and, in all but 1 case, clinicians decided to override at least some of the algorithms' recommendations. Twenty-eight studies (76%) were rated as having a high risk of bias in at least 1 of the 4 QUADAS-2 core domains, and 6 studies (16%) were considered to be at serious or critical risk of bias using ROBINS-I.

Conclusions and relevance

This systematic review found only sparse evidence that the use of ML-based CDSSs is associated with improved clinician diagnostic performance. Most studies had a low number of participants, were at high or unclear risk of bias, and showed little or no consideration for human factors. Caution should be exercised when estimating the current potential of ML to improve human diagnostic performance, and more comprehensive evaluation should be conducted before deploying ML-based CDSSs in clinical settings. The results highlight the importance of considering supported human decisions as end points rather than merely the stand-alone CDSSs outputs.

SUBMITTER: Vasey B

PROVIDER: S-EPMC7953308 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND:Computerized decision support systems (CDSSs) are software programs that support the decision making of practitioners and other staff. Other reviews have analyzed the relationship between CDSSs, practitioner performance, and patient outcomes. These reviews reported positive practitioner performance in over half the articles analyzed, but very little information was found for patient outcomes. OBJECTIVE:The purpose of this review was to analyze the relationship between CDSSs, practitioner performance, and patient medical outcomes. PubMed, CINAHL, Embase, Web of Science, and Cochrane databases were queried. METHODS:Articles were chosen based on year published (last 10 years), high quality, peer-reviewed sources, and discussion of the relationship between the use of CDSS as an intervention and links to practitioner performance or patient outcomes. Reviewers used an Excel spreadsheet (Microsoft Corporation) to collect information on the relationship between CDSSs and practitioner performance or patient outcomes. Reviewers also collected observations of participants, intervention, comparison with control group, outcomes, and study design (PICOS) along with those showing implicit bias. Articles were analyzed by multiple reviewers following the Kruse protocol for systematic reviews. Data were organized into multiple tables for analysis and reporting. RESULTS:Themes were identified for both practitioner performance (n=38) and medical outcomes (n=36). A total of 66% (25/38) of articles had occurrences of positive practitioner performance, 13% (5/38) found no difference in practitioner performance, and 21% (8/38) did not report or discuss practitioner performance. Zero articles reported negative practitioner performance. A total of 61% (22/36) of articles had occurrences of positive patient medical outcomes, 8% (3/36) found no statistically significant difference in medical outcomes between intervention and control groups, and 31% (11/36) did not report or discuss medical outcomes. Zero articles found negative patient medical outcomes attributed to using CDSSs. CONCLUSIONS:Results of this review are commensurate with previous reviews with similar objectives, but unlike these reviews we found a high level of reporting of positive effects on patient medical outcomes.

Project description:Underuse and overuse of diagnostic tests have important implications for health outcomes and costs. Decision support technology purports to optimize the use of diagnostic tests in clinical practice. The objective of this review was to assess whether computerized clinical decision support systems (CCDSSs) are effective at improving ordering of tests for diagnosis, monitoring of disease, or monitoring of treatment. The outcome of interest was effect on the diagnostic test-ordering behavior of practitioners.We conducted a decision-maker-researcher partnership systematic review. We searched MEDLINE, EMBASE, Ovid's EBM Reviews database, Inspec, and reference lists for eligible articles published up to January 2010. We included randomized controlled trials comparing the use of CCDSSs to usual practice or non-CCDSS controls in clinical care settings. Trials were eligible if at least one component of the CCDSS gave suggestions for ordering or performing a diagnostic procedure. We considered studies 'positive' if they showed a statistically significant improvement in at least 50% of test ordering outcomes.Thirty-five studies were identified, with significantly higher methodological quality in those published after the year 2000 (p = 0.002). Thirty-three trials reported evaluable data on diagnostic test ordering, and 55% (18/33) of CCDSSs improved testing behavior overall, including 83% (5/6) for diagnosis, 63% (5/8) for treatment monitoring, 35% (6/17) for disease monitoring, and 100% (3/3) for other purposes. Four of the systems explicitly attempted to reduce test ordering rates and all succeeded. Factors of particular interest to decision makers include costs, user satisfaction, and impact on workflow but were rarely investigated or reported.Some CCDSSs can modify practitioner test-ordering behavior. To better inform development and implementation efforts, studies should describe in more detail potentially important factors such as system design, user interface, local context, implementation strategy, and evaluate impact on user satisfaction and workflow, costs, and unintended consequences.

Project description:BackgroundArtificial intelligence (AI), such as machine learning (ML), shows great promise for improving clinical decision-making in cardiac diseases by outperforming statistical-based models. However, few AI-based tools have been implemented in cardiology clinics because of the sociotechnical challenges during transitioning from algorithm development to real-world implementation.ObjectiveThis study explored how an ML-based tool for predicting ventricular tachycardia and ventricular fibrillation (VT/VF) could support clinical decision-making in the remote monitoring of patients with an implantable cardioverter defibrillator (ICD).MethodsSeven experienced electrophysiologists participated in a near-live feasibility and qualitative study, which included walkthroughs of 5 blinded retrospective patient cases, use of the prediction tool, and questionnaires and interview questions. All sessions were video recorded, and sessions evaluating the prediction tool were transcribed verbatim. Data were analyzed through an inductive qualitative approach based on grounded theory.ResultsThe prediction tool was found to have potential for supporting decision-making in ICD remote monitoring by providing reassurance, increasing confidence, acting as a second opinion, reducing information search time, and enabling delegation of decisions to nurses and technicians. However, the prediction tool did not lead to changes in clinical action and was found less useful in cases where the quality of data was poor or when VT/VF predictions were found to be irrelevant for evaluating the patient.ConclusionsWhen transitioning from AI development to testing its feasibility for clinical implementation, we need to consider the following: expectations must be aligned with the intended use of AI; trust in the prediction tool is likely to emerge from real-world use; and AI accuracy is relational and dependent on available information and local workflows. Addressing the sociotechnical gap between the development and implementation of clinical decision-support tools based on ML in cardiac care is essential for succeeding with adoption. It is suggested to include clinical end-users, clinical contexts, and workflows throughout the overall iterative approach to design, development, and implementation.