Critical thinking skills can enhance diagnostic reasoning. This article discusses how deliberate reflection can improve diagnosis and how deliberate reflection can be integrated into digital decision support tools.

Enhancing clinical reasoning skills is an emerging strategy to reduce diagnostic errors. This meta-analysis of 26 studies estimated that cognitive reasoning tools can lead to small improvements in diagnostic accuracy.

This article describes the Yorkshire contributory factors framework, which identifies factors contributing to safety errors across four hierarchical levels (active errors, situational factors, local working conditions, and latent factors) and two cross-cutting factors (communication systems and safety culture). The authors apply this framework to a case of missed mass on breast imaging and discuss how its use can help health systems effectively learn from error and develop systematic, proactive programs to improve safety and manage safety issues.

In addition to being used for patient-specific clinical purposes, data within the electronic health record (EHR) may be used for other purposes including epidemiological research. Researchers in the UK developed and tested a clinical decision support system (CDSS) to evaluate changes in the types and number of observations that primary care physicians entered into the EHR during simulated patient encounters. Physicians documented more clinical observations using the CDSS compared to the standard electronic health record. The increase in documented clinical observations has the potential to improve validity of research developed from EHR data.

Using the events at the United Kingdom’s Mid Staffordshire Trust hospital as a case study, the authors discuss the impact of ‘collective denial’ on organizational processes and safety culture. The authors suggest that safeguards allowing for self-reflection and correction be implemented early in the safety reporting process, and that employees be granted power to speak up about safety concerns.

There is increasing interest in the use of artificial intelligence (AI) to improve breast cancer detection. This study developed and validated an AI algorithm using mammography readings from five institutions in South Korea, the United States, and the United Kingdom. The AI algorithm alone showed better diagnostic performance in breast cancer detection compared to radiologists without AI assistance (area under the curve [AUC] of 0.94 vs. 0.81, p<0.0001) or radiologists with AI assistance (0.88; p<0.0001). AI improved performance of radiologists and was better at detecting mass cancers, distortion, asymmetry, or node-negative cancers compared with radiologist alone.

Advanced computing holds promise for reducing missed diagnoses of cancer. This metanalysis found that computer-aided diagnosis effectively detects melanoma; however, studies were low in quality. The authors suggest that these systems may help assist dermatologists in overcoming the limitations of human cognition for performing repetitive tasks.

Machine learning may have the potential to improve clinical decision-making and diagnosis. In this study, machine-learning algorithms generally performed better than human experts in accurately diagnosing 7 types of pigmented skin lesions and the top 3 algorithms performed better than the 27 physicians.

Despite growing support for patient involvement in safety and quality improvement, little is known about engaging patients as partners in reducing diagnostic error. This commentary summarizes research on how sharing notes with patients can improve the timeliness of follow-up to confirm a diagnosis, identify documentation errors, and strengthen communication between the clinical team and the patient. The authors discuss challenges to the successful implementation of this strategy and areas of focus needed for future development. A PSNet interview discussed use of OpenNotes to engage patients in their care.

Researchers assessed the impact of two different virtual patient models containing error-based scenarios on medical students at six different institutions across three countries. They found that the use of branched decision-making logic did not change students' motivation as compared to a linear virtual patient model without such logic.

Medical errors can have emotional consequences for clinicians. This commentary emphasizes the importance of organizational support for second victims to ensure that these providers receive assistance from their colleagues to remain healthy and productive. The authors suggest that peer support programs are also required in organizations with blame-free cultures to support providers who feel guilt after an error.

Artificial intelligence (AI) may have the capacity to improve diagnosis. Researchers found that an AI system was able to detect breast cancer using mammography with accuracy similar to that of the average of the 101 radiologists whose interpretations were included in the study.

This simulation study randomized physicians to identify the correct diagnosis in a standardized case, either with the aid of a debiasing exercise or without any prompting. Even though the participants believed that the debiasing tool was effective, it did not improve diagnostic accuracy. These results underscore the challenge of enhancing diagnostic cognition.