Gleneagles Hospital

Carol Chow

Automated pharmaceutical technology innovations for medication management are well documented and commonly used nowadays to reduce medication errors, improve efficiency and quality of care. However, these innovations have led to new and unforeseen challenges. Various computerized processes are implemented at different points of the medication-use pathway and these are divided into the following care points:-

• Computerized prescription order entry systems (CPOE)

• Hospital / Pharmacy information system (HIS)

• Automated dispensing cabinets / machine (ADC)

• Medication repackaging system (unit dose packaging)

• Barcode medication administration (BCMA)

• Computerized medication administration records (CMARs)

•‘Smart’ infusion pumps, with drug libraries information on common concentrations, dosing units and limits.

Despite how pharmacy departments have worked with technology vendors and the information technology sector to improve the usability of systems, it is still not error-proof. In this article, I shall focus on medication management from the pharmacist’s perspective and will discuss the first four points above.

Let’s begin with CPOE and HIS. When these two systems are directly interfaced, they should be capable of issuing medication safety alerts, supporting drug-drug and drug-disease interactions and providing a basis for clinical decision making. One of the first challenges lies in the prescribers’ adoption rate, especially for individuals who are not as computer literate, requiring extensive training to become familiar with HIS. Secondly, prompts that alert against drug interaction, allergy or overdose often lead to alert-fatigue and eventually become easy to ignore. One recent case involved one of the drugs known to induce Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) – Allopurinol. Proactive measures were taken since 2008 to perform mandatory Human leucocyte antigen-B (HLA-B)*5801 gene testing before prescribing some main causative drugs like allopurinol. The prescription was issued via CPOE, the system was designed to alert prescribers to order the HLA-B*5801 allele test and to educate patients on early signs of hypersensitivity reaction. Sadly, a 73 year old woman passed away in November 2022, since the involved HIS generated too many alerts. “Alert fatigue” in this case, caused the alerts to become a distraction which prescribers unconsciously ignored. This illustrates the importance in striking the correct balance of alert frequency in order to achieve patient safety.

Moving on to automated dispensing cabinets and medication repackaging systems. These have been shown to have a positive impact on dispensing efficiency and accuracy, medication security, as well as freeing up pharmacists’ time on clinical services.

Unfortunately, automated systems still require human intervention in providing the correct medication / information, to ensure that safety is not compromised. There is little standardization across automated systems, which can complicate matters when you employ staff who trained on a different system. Previous encounters highlight the necessity to scan medication before filling it into the medication cabinet, especially for look alike, sound[1]alike drugs, and drugs with multiple strengths. When various innovative measures are placed at different points of care, BCMA is able to pick up differences in medication strength before the medication is administered.

 It is important to remember that technologies such as barcode systems are meant to facilitate and support, rather than replace human judgement. This was demonstrated in a recent incident in a hospital pharmacy when an inventory executive scanned the bar-code on an invoice and the order number on the packaging, the two of which correctly matched. However, although the type of medication listed is correct, the barcode system incorrectly dispensed medication that was double the required strength. This highlights the need for human intervention, through additional checks and verification processes, to ensure the proper functioning of technological systems.

It is also important to not only train staff to use these technologies, but also to understand the concepts behind them. We should strive to integrate AI and automation in a way that enhances rather than replaces human involvement in the healthcare system. Personally, I welcome the introduction of new technologies and seeing these incorporated into not only the medication management process, but also the patient journey- sometimes even before a medication reaches the commercial stage. My recent involvement in coordinating medical research made me realize the frustration in seeing 90 percent of drug candidates fail during the early phase of clinical trials -making it far more challenging to develop translational treatments. The introduction of Artificial Intelligence (AI) in clinical research has the potential to be useful in many ways: from patient matching to handling data and drug discovery, benefiting patients especially those suffering from the rarest conditions for which limited treatments are available. I am excited to be part of a healthcare team, at the vanguard of a revolution in technology assisted patient care.