By Thomas Davis, DNAP, MAE, CRNA

It is an understatement to say that Artificial Intelligence (AI) has made exponential progress in the past 3 years, or even the past 3 months.  The science fiction of the future is quickly becoming the reality of today.  A year ago, AI was creeping into our lives, today it has quickly become a way of life. 

AI is driving cars, programming personal devices, and writing papers for aspiring college students.  Likewise, AI is now guiding healthcare decisions and guiding the therapies that are prescribed by healthcare professionals.  AI quickly analyzes vast volumes of data and gives insights into development of best practice guidelines.  Moving forward, it will modify the everyday workflow of the healthcare system. 

Harvard Medical School compared X-rays that were read by AI versus those read by a radiologist and found that AI quickly and accurately does an initial screen and creates a lists of  diagnoses, however, AI lacks the medical context specific to the patient and may miss subtle clues on the radiograph.  The radiologist views both the image and the AI interpretation and then fine tunes the report by adding a “second opinion” to the AI analysis.

Likewise, The Pathologist journal reports that AI plays an important role in the initial screening of pathology slides however the expertise of the pathologist is still needed to put the findings into the context of the individual patient and make a final determination of the results.

When I look into my crystal ball, I see many opportunities for AI to influence or radically change our practice of anesthesia.  One application, pharmacogenetics driven by AI interpretation, has the potential to revolutionize our daily practice by aligning drug selection with the patient’s individual genetic makeup.

Pharmacogenetics seeks to understand an individual’s pharmacodynamic and pharmacokinetic tendencies based on their genetic makeup.  As anesthesia students, we learned about metabolism via the P450 system and the genetic link to malignant hyperthermia.  Modern technology, including genetic screening, has opened the door to understand many more genetic related responses to the drugs that we give.  Author Arash Behrooz notes “the cytochrome P450 (CYP) superfamily of enzymes is of particular interest when considering phase I reactions. There are 59 P450 proteins which are categorized into 18 families and 43 subfamilies, the majority of which are expressed in the smooth endoplasmic reticulum of hepatocytes, with CYP1, CY2 and CYP3 being the main families involved in human drug metabolism.”   Pharmacogenetics will tell the provider which patients are likely to be high versus low metabolizers of drugs based on their individual DNA profile.

An AI overview of pharmacogenetics in anesthesia found that an individual’s genetic makeup can affect many areas of our practice including:

• Muscle Relaxants: Variations in the BChE gene dictate butyrylcholinesterase (pseudocholinesterase) deficiency. Patients with this variant struggle to break down succinylcholine, leading to severely prolonged, dangerous muscle paralysis.
• Malignant Hyperthermia: Genetic mutations in the RYR1 gene predispose certain patients to malignant hyperthermia, a life-threatening, hypermetabolic crisis triggered by volatile inhalation anesthetics and succinylcholine.
• Opioid Sensitivity: Polymorphisms in the OPRM1 gene can significantly alter a patient’s pain threshold and their required dose of opioids (e.g., morphine, fentanyl, oxycodone) to achieve adequate pain control.
• Drug Metabolism: Genetic variations in the Cytochrome P450 (CYP450) enzyme systems specifically CYP2D6 and CYP2C19—heavily influence how rapidly a patient metabolizes common sedatives (like diazepam) and antiemetics (like ondansetron).
• Postoperative Nausea & Vomiting (PONV): Variations in ABCB1 (P-glycoprotein) and CYP450 genes play a massive role in why some patients suffer from intractable PONV, allowing providers to proactively choose alternative non-triggered antiemetics.

An article released by the NIH titled “Pharmacogenetics and anesthetic drugs: Implications for perioperative practice” listed the following as genetic mutations that affect the patient’s response to the drugs that we commonly use.

Drug	Genetic variation	Result
Propofol	UGT1A9	Higher induction dose required
	1818T/C	Longer time needed for loss of consciousness
Sevoflurane	RYR1	Malignant hyperthermia
Ketamine	CYP2B6	Reduced drug clearance
Lidocaine	SCN9A	Reduced efficacy
Fentanyl	OPRM1	Variations in effective dose
Succinylcholine	BChE	Reduced hydrolysis, prolonged duration of action
	RYR1	Malignant hyperthermia
Rocuronium	SLCO1B1	Reduced elimination, prolonged duration
Ondansetron	ABCB1	Increased bioavailability, increased antiemetic effect

As we move forward, I predict that a basic genetic profile will become a part of everyone’s medical record and that AI will be used to evaluate the DNA profile and highlight genetic patterns which predict patient responses to medications including anesthetic drugs.  Currently, genetic testing is not covered by insurance unless it is required to diagnose a specific disease.  In the future, genetic testing will be as common as fingerprinting and available to all.  For those who are on medications and would like to know more about their personal pharmacogenetic profile,  the company Clarity X offers genetic testing that shows interactions with 285 prescription medications.  The out-of-pocket cost is $375 for the comprehensive test and list of interactions with medications.

Pharmacogenetics is but one of the ways that our practice will change.  As we move forward, I foresee AI doing the preop chart review, linking to the genetic profile, and printing an anesthetic plan similar to the flight plan that is given to a pilot before takeoff. 

Now it’s your turn…what technology changes do you see in the future of anesthesia practice?

Tom is an experienced leader, educator, author, and requested speaker.  Click here for a video introduction to Tom’s talk topics.