Evaluation and Mitigation of Racial Bias in Clinical Machine Learning Models: Scoping Review.

BACKGROUND: Racial 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. OBJECTIVE: Our 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. METHODS: A 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. RESULTS: Applications 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. CONCLUSIONS: The 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.

Feasibility study of a force-sensing resistor device for landmark detection for neuraxial procedures and anesthesiologists' impact on device improvement.

The standard of care for neuraxial procedures is landmark palpation to determine the site of placement. This research study aimed to test the primary feasibility of VerTouch™, a force-sensing resistor device, to identify landmarks for the initiation of neuraxial procedures. Patients were recruited at the time of labour epidural, or when consenting for spinal anaesthesia for caesarean delivery at a single centre. The clinical team used the device to create a pressure map image of the bony spine. If they felt confident, they used the device guide to make a mark on the patient's back. If the mark was used, total insertions and redirections, combined as passes, of the needle during the procedure were counted for secondary outcome analysis. A total of 101 parturients were recruited, and the provider felt confident making a mark based on the imaging in 96.9% of cases. Device success (completion using the mark and ≤4 total passes) occurred in 91.4% of cases. This feasibility study showed that the primary outcome, the provider using the device to make a mark, was successful. In addition, based on comparing passes with historical data, the Vertouch™ device showed promise for future use to minimise needle manipulation in neuraxial procedures.

Effect of Quantification of Blood Loss on Activation of a Postpartum Hemorrhage Protocol and Use of Resources.

OBJECTIVE: To determine if quantification of blood loss (QBL) would result in fewer activations of postpartum hemorrhage (PPH) protocols than visual estimation of blood loss (EBL) after cesarean birth and to track the use of related resources. DESIGN: Prospective observational trial. SETTING: A tertiary academic medical center in the midwestern United States. PARTICIPANTS: A total of 42 cases of cesarean birth. METHODS: We visually estimated blood loss during cesarean birth and quantified blood loss with colorimetric testing after the surgery. We compared EBL to QBL in four categories, from no hemorrhage to severe PPH, and documented resources used for women placed on the institutional PPH protocol by EBL who did not meet criteria for PPH by QBL. RESULTS: The median EBL was 1,275 ml (interquartile range = 1,100-1,510 ml), and the median QBL was 948 ml (interquartile range = 700-1,267 ml, p < .001). Twenty-four (57%) instances of PPH based on visual EBL would not have been classified as such based on QBL. The most frequently used resources in these cases included laboratory testing and administration of uterotonics. CONCLUSION: Use of QBL during cesarean births would have reduced the number of identified PPHs by more than 50% over visual EBL and may have reduced the resources used as part of care.