Diagnostic uncertainty can lead to misdiagnosis and delayed treatment. This article provides an overview of the literature on diagnosis-related uncertainty, where uncertainty occurs in the diagnostic process and outlines recommendations for managing diagnostic uncertainty.

Cognition is a recognized human factor that can contribute to medical error. This systematic review explored whether mindful practice can improve diagnosis in healthcare. Of the 33 included studies, the majority were non-empirical; however, the authors tentatively conclude that mindful practice may be a promising method to improve diagnostic accuracy and reduce error.

Based on analysis of four years of data submitted to the Child Health Patient Safety Organization (CHILDPSO), researchers sought to identify types of serious safety events and contributing factors. Three main groups of serious safety events were identified: patient care management, procedural errors, and product or device errors. Contributing factors included lack of situational awareness, process failures, and failure to communicate effectively.

This systematic review provides an update to McDonald et al’s 2013 review of strategies to reduce diagnostic error.  Technique (e.g., changes in equipment) and technology-based (e.g. trigger tools) interventions were the most studied intervention types. Future research on educational and personnel changes would be useful to determine the value of these types of interventions.

Artificial intelligence (AI) and big data have potential to impact health care improvement in a variety of ways. This commentary examines the intersection of AI, diagnosis, and inequality. The authors highlight several innovations that could reduce health disparities and diagnostic errors in at-risk patient populations if adopted.

Delayed diagnosis and treatment of breast cancer can lead to poor outcomes. Based on multi-year data from one health system, the authors of this cohort study found that black women with screen-detected breast cancers were more likely than white women to experience diagnostic delays, including delays in diagnostic evaluation and biopsy. The delay in diagnosis was also associated with an increase in breast cancer mortality.

Overdiagnosis of breast cancer and the resulting overtreatment can cause physical, emotional, and financial harm to patients. Analysis of observational data and modelling indicates overdiagnosis accounts for less than 10% of invasive breast cancer in patients aged 50-69. Understanding rates of overdiagnosis can assist in ascertaining the net benefit of breast cancer screening.

Incorrect interpretation of radiologic images can result in delayed diagnosis or unneeded additional tests and treatment. This commentary describes the visual illusions radiologists use in detecting and categorizing abnormalities, and recommends further research into the ways visual illusions are used in order to improve diagnostic safety.

Misdiagnosis of cardiovascular conditions can lead to serious patient harm. This systematic review explored misdiagnosis of heart failure. Based on 10 included studies, the rate of heart failure misdiagnosis ranged from 16.1% (in an inpatient setting) to 68.5% (when general practitioners referred patients to specialists). Included studies found that heart failure is frequently misdiagnosed as chronic obstructive pulmonary disease (COPD).

The goal of the patient centered medical home (PCMH)  model is to reorganize primary care to provide team-based, coordinated, accessible health care. This study used a consensus process with input from a physician panel to examine ambulatory patient safety concerns (e.g., medication safety, diagnostic error, treatment delays, communication or coordination errors) in the context of the PCMH model and explore variability in the implementation of patient safety practices.

Missed or delayed diagnosis of sepsis can lead to serious patient harm. This study used a Symptom-Disease Pair Analysis of Diagnostic Error (SPADE) “look-forward” analysis to measure potential misdiagnosis of sepsis in patients discharged from the emergency department (ED) with treat-and-release fluid and electrolyte disorders (FED) or altered mental status (AMS). FED and AMS were associated with a spike in sepsis hospitalizations in the 7-day period following the ED visit. The authors suggest SPADE could be used to compare sepsis diagnostic performance across institutions and regions; develop interventions for targeted subgroups; and update early warning systems for sepsis diagnosis.

Prehospital triage tools are used to differentiate between patients who need emergency care at a major trauma center (MTC) and those that may receive adequate care at a non-MTC.  Accurate triage tools are necessary to ensure that patients are not over- or undertriaged. This review found high variability in sensitivity and specificity across geriatric triage tools indicating some patients may not be receiving the specialized trauma care they need. The authors highlight several future research targets including development of relevant reference standards and balancing the risk between over- and undertriage.

Diagnostic errors continue to be a patient safety concern, including in pediatric critical care. This systematic review explored the prevalence, impact, and contributing factors to diagnostic errors in the pediatric intensive care unit (PICU). The most common diagnostic errors occurred in cardiovascular, infectious, congenital, and neurologic conditions; cognitive and systems factors were associated with diagnostic errors. Future research should focus on disease- and systems-level determinants. 

This study explored the role of machine-learning based clinical decision support (CDS) algorithms to support (rather than replace) human decision-making and the impact on diagnostic performance. This systematic review of 37 studies found limited evidence that the use of machine learning-based CDS systems contributes to improved diagnostic performance among clinicians. Interobserver agreement, user feedback, and clinician override were the most commonly reported outcomes. The authors emphasize the importance of further evaluation of human-computer interaction.

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.

Challenges to identifying and measuring diagnostic errors, particularly in the era of COVID-19, persist. The authors of this perspective proposed a new framework of diagnostic process safety to measure the quality and safety of diagnostic processes. The framework focuses on three measurable components – do not miss diagnoses, red flags, and diagnostic pitfalls. This framework can provide a structured approach for designing and testing specific measures of diagnostic process safety.

Reducing harm related to diagnostic error remains a major focus within patient safety. Based on input from an international group of experts and stakeholders, the authors identified priority questions to advance diagnostic safety research. High-priority areas include strengthening teamwork factors (such as the role of nurses in diagnosis), addressing system factors, and strategies for engaging patients in the diagnostic process.

Delays in cancer diagnosis can lead to treatment delays and patient harm. This study linking primary care and cancer registry data found that more than one-quarter of bladder and kidney cancer patients presenting with fast-tract referral features did not achieve a timely diagnosis. These findings suggest inadequate adherence to guidelines intended to help identify patients with high risk of cancer based on the presence of alarm signs and symptoms.