hand hygiene
hand hygiene
By Robert P . Lee
Artificial Intelligence and Hand Hygiene
In simple terms , AI is a natural extension of computers , now automated and expedited to gather information but , of importance , analyze that data as well .
AI can assemble all the data and consider the relationship of this data to outcomes .”
In discussions with our nursing team , I was reminded that at the onset of their careers they operated off the “ buddy system ” and nursing was a team sport — each nurse with a partner who complimented their provision of care . This is now unusual , as nurses today typically manage increasing numbers of patients each shift without any buddy system . Additionally , numerous travel / temp nurses are required to maintain operational efficiency . These factors challenge quality and safety standards with removal of checks and balances in the system . How do we reinsert these checks and balances back into the system ? Is artificial intelligence ( AI ) a potential solution ? Let ’ s explore these issues .
In simple terms , AI is a natural extension of computers , now automated and expedited to gather information but , of importance , analyze that data as well . AI can assemble all the data and consider the relationship of this data to outcomes . With this new ability , AI can predict expected outcomes . AI can create algorithms with the data , both accelerating the analysis and actually providing solutions based on the data .
Let ’ s review some definitions of terms related to AI .
Artificial intelligence is comprised of computer systems that perform tasks that previously required human input , for example , visual perception , speech recognition , and decision-making . Usually involves pattern recognition then followed by an action or a decision [ I think it is more than pattern recognition , but rather analysis and decision-making capacity , i . e ., analyzing and acting on the data with proposed actions , unrelated to human analysis .]
Machine learning is a subdomain of AI . The computer system “ learns ” from datasets to make predictions and recommendations about new data , as opposed to executing a set of programmed rules . In classic machine learning , programmers design and tune these algorithms .
Deep learning is a subdomain of machine learning . The computer uses a mathematical structure inspired by neural networks to learn from very large datasets to make predictions about new data . The neural network builds the algorithms automatically by finding novel relationships between inputs and outputs . The algorithms cannot be analyzed by humans as they involve millions of small decisions about data . 1
With the advent of hand hygiene technology and real time location systems ( RTLS ), we now have real-time workflow data , specific to individuals , spaces , assets , and equipment , which is highly accurate and simultaneously provides this data 24 / 7 . As this data is linked to electronic medical records / electronic health records ( EMRs / EHRs ), we now have patient and facility health data at our fingertips . These linked databases allow AI to develop algorithms that can provide timely , accurate and precise information , allowing decisions and actions to be predictive rather than retrospective
I remember when computers were introduced to our medical research lab where I worked at Tufts New England Medical Center in the late 1960s . Receiving one of the first IBM boxes at the cancer research lab , the chief researcher said to me , “ Figure this thing out .” My fellow co-op colleagues and I spent days on this project using all the punch cards to determine the computer ’ s capabilities and how it could bring value to the laboratory .
Data is the new currency in the marketplace and AI is the next evolution in computer technology , an exponential change and potential advance . Hand hygiene technology is helping us to gather and harvest this data and AI should allow an increase in the speed and analysis of the data with much more informed and timely decisions . When a healthcare provider uses hand hygiene technologies , it is like the “ buddy system ” discussed above . The badge technology gathers the data and the internal algorithms help to interpret this data , providing feedback , reminders , reporting and analytics that is equivalent to a buddy system . Additionally , if you are still using secret shopper to audit , train and educate your staff , hand hygiene technology and AI will bring you into the 21st century , not only helping you discard the clipboards and notepads , but providing real-time analysis and direction of the appropriate actions based on the data .
We are all challenged at times by the increasing monitoring of our every action , whether via video monitors in our communities or the data collected during our professional endeavors in the healthcare arena . However , this will certainly continue as it provides information essential to improving patient care , whether diagnostic assessments or infection prevention endeavors . The key is using the information not in a punitive or negative feedback loop but rather information to improve our performance around preventing healthcare-associated infections ( HAIs ). I doubt anyone in the infection prevention arena would argue that we cannot do better , and the hope is that AI and informed , prospective decisions will provide a more effective armamentarium to take the next step .
A recent study at Stanford University assessed the application of automatic detection of hand hygiene using computer vision technology , a critical step in the development of novel ways to apply AI-based monitoring to improve patient care and safety . The lead author notes , “ The potential is really limitless when you are monitoring something continuously
20 • www . healthcarehygienemagazine . com • august 2024