3 Ways AI Can Help Nurses

By Giorgio Piatti & Allyson Paiewonsky, Advisory Board
Twitter: @AdvisoryBd

There’s a real, and untapped, opportunity for artificial intelligence (AI) and automation to alleviate the burden on nurses and help them focus on top-of-license care. The nursing workforce is incredibly burned out due to COVID-19 and the ongoing staffing crisis. As many facilities struggle to keep a sufficient number of nurses staffed, nurses continue to be bogged down by administrative tasks that keep them from practicing at top-of-license. Leaders must start using technology to optimize nurses’ time if they want to keep up with care demand under the current staffing resources.

Many early applications of AI and automation focused primarily on back-office functions like revenue cycle. And in instances where this technology is applied to clinician workflows, they are typically focused on physicians rather than nurses. But leaders can’t forget about nurses as they plan to invest in and implement new technologies.

1. Automation takes on rote administrative tasks to increase efficiency and reduce manual errors

Nurses increasingly spend more time on care-adjacent tasks like paperwork and the repeated collection of basic patient information, which is an inefficient use of time and leads to job dissatisfaction. Automation of administrative tasks takes some of the administrative burden off nurses and prevents the need for manually inputting information that can also be more prone to error.

For example, one of the hospitals within California-based Dignity Health with a 23-bed surgical unit, and Applied Science Inc., an automation vendor, partnered to develop a system known as ADEPT that automatically enters data from remote patient monitoring devices to patient records.

Prior to the implementation, the hospital had observed an error rate of about 20% for manually entered data as a result of typographical and rounding errors. Post implementation, errors were reduced to 0% and delays in data entry were reduced by anywhere from 5 minutes to 2 hours.

2. AI supports nurses to make faster, equally accurate clinical decisions

One of the more commonly performed tasks in the care process for a nurse is diagnosing abnormalities and identifying treatment paths, all while taking vitals and keeping the patient stable. Nurses are required to do this under time constraints, so high levels of stress and potential errors are sometimes more common. AI can optimize nursing diagnostic practices to be more consistently accurate, more time-efficient, and less stressful for nurses.

Back propagation neural networks (BPNs) are a method of machine learning that are used to support clinical decision-making. BPNs monitor outputs from data sets and give more weight to the inputs that correspond to accurate results.

A hospital in Taiwan used a BPN to help nurses receive instructions and address patient care with relevant data. The solution was a success, showing that time efficiency among nurses can improve without sacrificing quality.

Nurse satisfaction in patient diagnosing almost doubled, the time to make a diagnosis decreased from 35.5 minutes to 19.8 minutes, and the BPN-produced diagnoses matched nursing recommendations in 91% of cases.

3. Automated triaging helps nurses focus on high acuity cases

Nurses must care for a wide range of patients with varied needs. The overwhelming demand for care means that nurses end up spending some of their limited capacity on low-acuity patients that didn’t really need a nurse’s attention. Automation solutions can help take on lower-acuity cases so that nurses can focus more time and energy on the patients who require more urgent attention.

AI-powered chatbots or nurse assistants receive patient health data and proactively identify patients who need nurse intervention before the situation escalates. Additionally, these assistants address requests that don’t require a nurse’s assistance, like reminding patients to take medication or answering patient questions about treatments. PaSent is an Ai solution that learns from discharge letters to proactively recommend resources and answer patients’ questions in lieu of nurses.

This article was originally published on the Advisory Board blog and is republished here with permission.