Artificial Intelligence (AI) is rapidly transforming the healthcare industry and pharmacy practice. However, many clinicians lack a foundational understanding of AI technology and may have difficulty applying it in their own practice setting.
So, what exactly is “Artificial Intelligence'' or “AI”? It is the science of building machines to perform tasks that typically require human intelligence. AI is broken down into three branches: Machine Learning, Natural Language Processing (NLP), and Computer Vision.
Machine Learning is a way to train a model to make accurate predictions. An example of machine learning is how a GPS processes traffic patterns and predicts how long it will take at a certain time of day/day of the week to reach your destination. Based on historical and real-time data, the algorithm can predict that a commute could take 10 minutes longer on Friday evening due to rush hour traffic. This is a simple example, yet it highlights the countless possibilities used in machine learning. The basic steps to developing a machine learning algorithm are to:
- Formulate a question
- Collect data
- Select, train, and evaluate the model
- Make predictions
Natural Language Processing (NLP) is a way to decipher the complicated human language and produce a simple, understandable output. A great example of this is smart speakers activated in our homes. For instance, there are many ways to ask a smart speaker about the weather. You could utilize a voice command to activate the smart speaker and ask out loud “What is the high for today? Will it rain?” and so forth. NLP analyzes the language, generates a human response, and provides an outcome such as “The high today is 65 degrees Fahrenheit and bring your umbrella because there is a 100% chance of rain.”
Computer vision uses pattern recognition and deep learning to recognize what’s in a picture or video. Have you ever wondered how social media pages, such as Facebook, can tell you who exactly to tag in your picture? This is because computer vision technology can process and then “recognize” facial features of individuals in the photo.
Applying artificial intelligence to healthcare delivery
Now that we have covered the basics of AI, how can these advanced technologies be used in healthcare delivery? Several goals include:
- Increase efficiency
- Digital task shifting and managing staff shortages
- Population health applications to enable targeted and differentiated services
- Earlier detection of diseases
- Improve quality of clinical decision making
- Continuous patient monitoring
AI applications used in pharmacy practice can be broken down into prevention, diagnosis, and treatment. Examples include:
- Nguyen et al. utilized machine learning to predict minimum inhibitory concentrations (MICs) of Salmonella strains and predict their susceptibilities to 15 antibiotics.1
- Doctors Without Borders developed a microbiology tool for use in resource-limited settings that uses computer vision to read “zones of inhibition” to advance patient care.2
Key factors to consider when evaluating an AI solution for healthcare:
- Feature engineering – data elements that go into the model
- Clinical expertise – did the data science team collaborate with clinicians in building the model?
- Validation – How well does the model predict the outcome of interest using validation data?
- Application and clinical benefits - how is it being used? Having a decent model that makes accurate predictions yet has no applicable clinical benefits has no place in the real world.
- Interpretation - Is the application easy to interpret and understand?
- User education and training – clinicians should be educated on what the model output means and how it may impact patient care decisions.
- Workflow integration - how and when will the clinician receive the information to impact clinical decision making?
The impact of AI in C. diff detection
Wolters Kluwer leveraged this technology and developed a Clostridioides difficile infection (CDI) risk score. C. diff infections negatively impact patient outcomes and organizations’ financial and quality performance measures. According to CDC data released in 2019, over 223,9000 CDI cases and 12,8000 deaths were reported related to CDI. Quality measures for CDI exist in both the Hospital-Acquired Condition (HAC) Reduction Program3 and Value-Based Purchasing (VBP) Program4 measures.
This means poor performance in these programs results in reduced payments by Medicare. Patients with hospital-acquired CDI have longer length of stay, higher readmission rates, higher total cost, and higher mortality than those who avoid it. Studies have shown Antimicrobial Stewardship (AMS) programs that reduce antimicrobial use in targeted antibiotics (i.e., clindamycin, fluoroquinolones, etc.) result in reduced CDI incidence.5-7