Phi Lambda Sigma is the international pharmacy leadership society. Our mission is to support pharmacy leadership commitment by recognizing leaders and fostering leadership development.

0
Contact Us
PLS Updates The PLS Pill

The PLS Pill: AI – Friend or Foe?

Updated on 5 min read

The PLS Pill: AI – Friend or Foe? 

Written by: International Affairs Committee Vice-Chair Emi Radetich, PharmD, PGY1 Resident at University of Utah Hospital and Clinics

 

It’s been all the rage… Whether you are a fan of Black Mirror or a humble viewer of recent news, you have probably heard of this buzzword: artificial intelligence, or AI. Some people were quick to ask ChatGPT some of life’s deepest questions, while others fear the capabilities of AI. Whatever group you fall into, it’s important to gain a baseline understanding of what AI is and how it affects the profession of pharmacy. AI will not go away any time soon, so each person should have agency in the opinions they form and the choices they make surrounding AI.

 

So first of all… what is AI? AI is the idea that computers are able to perform tasks typically associated with human intelligence. AI is carried out via the modality of an advanced computer programming method called machine learning. Thank goodness computer scientists are straightforward in nature, because machine learning is exactly what it sounds like – a machine that is learning. But who does the machine learn from? The answer is… us. So maybe artificial intelligence isn’t so artificial if it comes from us. People feed these machines data and the machine develops its own algorithm from that data. Which leads me to introduce…

 

The Risks of AI

If we feed data to a machine and it learns to retrieve conclusions and predictions based on that data, we must be extremely cautious about what type of information we give the machine. Humans are inherently biased creatures, and our biases, such as systemic discrimination, can be amplified by AI if we aren’t careful. Moreover, in the context of healthcare, the data that will inevitably be used will be patient data. How can we ensure that patient data will remain private, and what are the implications of informed consent? Who is liable if there are problems caused by the AI’s generative thinking, such as harmful output? And since AI can come up with its own ideas, does that threaten existing jobs? These concerns are just scratching the surface of potential issues we may face with AI in a healthcare setting. 

 

Benefits of AI

Now that we understand some of the risks, we can understand what needs to be done to reduce risk to reap maximum benefit. So what are some benefits of AI? In relation to the profession of pharmacy, AI can augment pharmacist and pharmacy technician workflow. For example, AI can predict personalized treatment for common ailments, such as strep throat management in an urgent care. AI can also optimize routine monitoring and communication with patients via utilization of virtual health assistants that use Natural Language Processing (NLP). Pharmacy operations can be improved via optimization of inventory management, automated dispensing, and product verification, allowing technicians more opportunity to practice at the top of their licenses. Industry is another area that can improve with AI-assisted drug discovery and development, reducing burden on research scientists. Overall, any area in medicine can benefit from workflow enhancements, while keeping in mind the risks that AI can bring.

 

The Current Stance on AI

The aforementioned information is a basic overview of risks and benefits to AI. What do official organizations have to say about it? The American Society of Health-System Pharmacists (ASHP) released a statement regarding AI in pharmacy, keeping in mind that the use of AI is an evolving landscape. To help contextualize the profession’s role in implementing AI, ASHP recommends that pharmacists ask themselves these key questions:

 

  1. Which medication- or therapy-related tasks are appropriate for AI to address?
  2. How should AI models be evaluated?
  3. For each type of use case, which AI learning approach(es) is (are) most appropriate?

 

While not all-inclusive of everything that can be considered, these three questions are a great starting point. The statement later explores the gains and deficits of AI in pharmacy, and how different areas of expertise can work together to shape the future of AI. 

 

Conclusion

With some background knowledge about AI, I hope you feel that you have a great launch point to develop a more informed opinion. I encourage you to learn more about it via podcasts, digital health newsletters and blogs, reliable tech news sources, and conversations with experts in the field. Perhaps you will find that AI is not a friend nor a foe, but rather a frenemy all along. 

 

If you have any questions regarding AI, pharmacy informatics, or any component of this post, please reach out via LinkedIn or email me at emi@philambdasigma.org 



References

MIT Technology Review https://www.technologyreview.com/

Recommended Articles

Please enable JavaScript in your browser.

Copyright 2022 Phi Lambda Sigma | 40 Bailey Lane, Uniontown, PA 15401 | Phi Lambda Sigma | Privacy Policy