Older adults are at increased risk of hospitalization due to COVID-19 infections. This study examined the potential severe drug-drug interactions (DDI) among hospitalized older adults taking two or more medications at admission and discharge. There was a significant increase in prescription of proton pump inhibitors and heparins from admission to discharge. Clinical decision support systems should be used to assess potential DDI with particular attention paid to the risk of bleeding complications linked to heparin-based DDIs.

The prescribing of potentially inappropriate medications, particularly among older adults, is an ongoing quality and safety concern. Among adults 65 years and older, this study found that clinical decision support integrated with a new electronic health record system significantly reduced potentially inappropriate medications.   

Older patients are vulnerable to adverse drug events due to comorbidities and polypharmacy. This cross-sectional study from Spain reviewed prescriptions for 593 older adults aged 65-75 years with multiple comorbidities and documented polypharmacy to estimate the prevalence of potentially inappropriate prescribing using the STOPP and Beers Criteria. Potentially inappropriate prescribing was detected in over half of patients. The most frequently detected inappropriate prescriptions were for prolonged use of benzodiazepines (36% of patients) and prolonged use of proton pump inhibitors (45% of patients). Multiple risk factors associated with potentially inappropriate prescribing were identified, including polypharmacy and use of central nervous system drugs.

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.

Use of computerized provider order entry (CPOE) is increasingly widespread. This systematic review found that while CPOE with clinical decision support reduced certain medication errors associated with prescribing, CPOE led to the introduction of new errors.

Artificial intelligence (AI) has the potential to improve health care. This narrative review summarizes short-, medium-, and long-term safety concerns associated with AI implementation in medical care. The authors provide quality control questions to help those involved in developing AI systems detect areas of concern.

Although clinical decision support is intended to improve safety, decision support alerts often result in alert fatigue and overrides. This prospective observational study examined overrides for exceeding the maximum dose of a medication in the intensive care unit. Researchers determined that insulin was the most frequent medication for which a maximum dosage alert was overridden. In almost 90% of cases, the overrides were deemed clinically appropriate. The authors conclude that more intelligent clinical decision support for medication dosing is needed to balance safety with alert fatigue in the intensive care unit. A past PSNet perspective discussed the challenges of implementing effective medication decision support systems.

This systematic review of clinical decision support for safe medication use found that such systems are incompletely implemented and lack standardization and integration of patient-specific factors. The authors suggest that reducing alert fatigue and employing human factors principles would enhance decision support effectiveness.

Clinical decision support systems in electronic health records (EHRs) aim to avert adverse events, especially medication errors. However, alerts are pervasive and often irrelevant, leading patient safety experts to question whether their modest improvement in safety outweighs the harms of alert fatigue. This study assessed provider overrides of a commercial EHR's medication alerts in intensive care units at one institution. Providers overrode most alerts, and the majority of those overrides were appropriate. Inappropriate overrides occasionally led to medication errors and did so more frequently than appropriate overrides. A recent WebM&M commentary recommends employing human factors engineering to make clinical decision support more effective.

Diagnostic error is a growing area of focus within patient safety. Artificial intelligence has the potential to improve the diagnostic process, both in terms of accuracy and efficiency. In this study, investigators compared the use of automated deep learning algorithms for detecting metastatic disease in stained tissue sections of lymph nodes of women with breast cancer to pathologists' diagnoses. The algorithms were developed by researchers as part of a competition and their performance was assessed on a test set of 129 slides, 49 with metastatic disease and 80 without. A panel of 11 pathologists evaluated the same slides with a 2-hour time limit and one pathologist evaluated the slides without any time constraints. The authors conclude that some of the algorithms demonstrated better diagnostic performance than the pathologists did, but they suggest that further testing in a clinical setting is warranted. An accompanying editorial discusses the potential of artificial intelligence in health care.

The recent National Academy of Medicine report on improving diagnosis cited the need for enhanced clinical decision support. This pre–post study used a simulation approach (standardized patients) to compare visits with and without use of a diagnostic clinical decision aid embedded in the electronic health record. The patients' visit satisfaction ratings did not differ in the visits with and without the decision support, although more patients in the decision support group noted that physicians focused more on the computer than the patient. The physicians reported high overall satisfaction with the decision tool, but they noted that it required inputting more clinical documentation during the visit, resulting in more time directed at the electronic health record. The authors conclude that the clinical decision support tool interface should be improved in order to facilitate adoption of real-time diagnostic support.

Despite increasing focus on diagnostic error, it remains a controversial patient safety issue. The Institute of Medicine recently suggested that further research is needed regarding electronic tools to improve diagnosis. Differential diagnosis generators provide a list of possible diagnoses for a problem. The investigators conducted a systematic review and found that differential diagnosis generators have been shown to improve diagnostic accuracy when a clinician has an opportunity to re-review the case using the software in pre-post studies. The degree of improvement varied between studies. The effect on actual clinician behaviors—such as test ordering, clinical outcomes, and cost—is unclear. Clinicians need prospective studies in order to determine whether such tools enhance diagnosis in actual practice. A recent PSNet perspective discussed future research avenues to ensure progress in diagnostic safety.

Clinician use of clinical guidelines is known to be less than optimal. According to this review, evidence indicates that nurse utilization of computerized clinical guidelines resulted in care process improvements, but further research is needed to determine if there is a correlation between increased provider access to guidance and patient safety.

This narrative review found that while computerized provider order entry combined with clinical decision support systems effectively prevented medication prescribing errors, there was no clear effect on clinical adverse drug event rates. This finding may be due to alert fatigue and other unintended consequences of the technology.

An electronic clinical decision support system for prescribing in the emergency department used tiered alerts with higher and lower urgency information in order to avoid alert fatigue.