In the quest for more efficient healthcare systems, the adoption of electronic prescribing has become a priority for many healthcare providers. However, as we delve into the potential risks associated with these systems, a critical question arises: are we inadvertently creating new avenues for medication errors? This article explores the intriguing and often overlooked issue of 'look-alike sound-alike' (LASA) medication errors in the context of electronic prescribing systems.
The Human Factor in Electronic Prescribing
One of the primary concerns with electronic prescribing systems is the potential for LASA errors. These errors, which can occur when drug names sound or look similar, are a significant threat to patient safety. The tragic case of Sidra Aliabase, who died due to a prescribing error, highlights the devastating consequences of such mistakes.
The introduction of electronic systems was supposed to reduce medication errors, with the UK government claiming a potential 30% reduction. Yet, as we examine the data, a different picture emerges. While electronic prescribing may have mitigated some errors, it has also introduced new challenges, particularly with LASA medicines.
Incident Data: A Complex Picture
Obtaining accurate data on LASA incidents is a challenge in itself. The transition from the National Reporting and Learning System (NRLS) to the Learn from Patient Safety Events (LFPSE) service has led to potential dual reporting, making it difficult to ascertain the true extent of LASA errors. Additionally, the way incidents are reported, often in free text, further complicates data extraction.
Despite these challenges, the available data suggests that LASA errors may have shifted rather than increased. Experts like Professor Bryony Dean Franklin and Julia Scott, a pharmacist and CIO, suggest that while paper-based prescribing had its errors, electronic systems have introduced a different set of challenges.
Mitigating Errors: A Balancing Act
One strategy to prevent LASA errors in paper-based systems is 'tall-man lettering', where certain letters in drug names are capitalized to distinguish them. However, implementing this in electronic systems is not straightforward. Julia Scott raises important questions: can we integrate tall-man lettering into e-prescribing systems? How do we design drop-down menus to minimize errors while maintaining usability?
The integration of AI and clinical decision support systems is seen as a potential solution. Scott suggests that AI could bring sophisticated prompts, especially when integrated with electronic patient records. However, she also warns of the 'flip side' - the potential for errors with ambient voice technology (AVT) or 'AI scribes'. AVT, which generates transcripts from patient-clinician conversations, could introduce a new category of sound-alike errors, akin to the days of verbal orders.
Other Methods and the Role of AI
Beyond tall-man lettering, systems like Touchdose, a clinical decision support system, offer a different approach. By prescribing based on indication, Touchdose reduces the likelihood of LASA errors. A study showed a significant reduction in prescribing errors when using Touchdose.
The potential of AI in analyzing LASA error reporting is also highlighted. With the new LFPSE system, there's hope that AI can provide a more in-depth analysis of patient safety events. However, as Franklin points out, under-reporting is a significant issue. Only a fraction of prescribing and administration errors are reported, making it challenging to grasp the true scale of the problem.
The Future: AI and Beyond
While AI offers promising solutions, it also brings new challenges. Scott acknowledges the need to tackle environmental and ethical concerns associated with AI. Despite these challenges, the potential benefits of AI in enhancing medication safety are significant. As we move forward, we must invest in the skills and knowledge required to utilize AI safely.
In conclusion, while electronic prescribing systems offer many advantages, they also present new risks. The issue of LASA errors is a complex one, requiring a nuanced understanding of human behavior and system design. As we continue to innovate, the challenge lies in finding the right balance between efficiency and patient safety. The integration of AI and a deeper understanding of cognitive mechanisms that lead to errors will be crucial in this journey. The future of medication safety lies in our ability to adapt and learn from these challenges.
