AI, Machine Learning Use Data to Improve Patient Satisfaction

A new artificial intelligence (AI) approach has the potential to improve health outcomes by analyzing patient satisfaction surveys and anticipating patient needs, according to Penn State researchers.

The research team collaborated with Geisinger in applying artificial intelligence to machine learning algorithms to make sense of patient satisfaction data and produce helpful recommendations for hospitals and healthcare providers.

The study used anonymous patient satisfaction datasets collected between 2009 and 2016 to test the algorithm. The data, gleaned from EHRs, included services and clinical information. It also included patient satisfaction survey answers about hospital care and services as well as post-care satisfaction levels.

“Patient health care is like a journey. They need to interact with multiple health professionals across different service units throughout the entire length of stay,” Ning Liu, lead author of the study, a fall 2019 Penn State doctoral recipient in industrial engineering, and current data scientist at Microsoft, said in a statement.

“It’s important for providers to understand the needs of each patient group, like those receiving surgery, cancer treatments or emergency visits. We wanted to know what is most important for each group, and how do we interpret that from the data we receive?”

Patient satisfaction is a top priority for many hospitals; therefore it is important for providers to be aware of areas that need improvement. The machine learning algorithm, with the use of artificial intelligence, scanned the patient data, turning it into information that the researchers could use and interpret.

According to the study, patient experience was most influenced by the respect patients were shown by doctors and nurses, the promptness and helpfulness in addressing their concerns or complaints, and pain management quality.

Researchers and healthcare providers could use this information to improve patient satisfaction by anticipating what different groups of patients care about most and ensuring patients walk away with a positive experience.

“A key performance indicator for hospitals is patient satisfaction,” said Soundar Kumara, Allen E. Pearce and Allen M. Pearce professor of industrial engineering at Penn State. “So, the question becomes, ‘how do we analyze and explain why patients rate a hospital the way that they do?’”

“In the context of hospitals, interpretability of data becomes critical. The major impact of the work lies in the AI models we have developed for interpreting the machine learning methodologies results. This work is among the first in this space.”

The machine learning system provides an opportunity for hospitals to improve overall health outcomes, as patient satisfaction is strongly associated with greater compliance and increased treatment adherence, according to the researchers.

The use of AI in patient engagement has provided a more personalized and convenient healthcare experience. Whereas systems such as Penn State’s machine learning algorithm can benefit providers, tools like patient portals and symptom checkers can largely benefit the patient.

AI chatbots, when perceived as competent and human-like, can also boost patient satisfaction. The ability to receive healthcare information in a timelier manner using AI tools may lead to better health outcomes.

Back in 2019, in a Pegasystems survey of 2,000 healthcare consumers, almost half (42 percent) of patients said they are comfortable with their doctors using AI to make healthcare decisions.

Sixty-two percent of patients in the same survey said they wanted their doctor’s primary source of communication to be a phone call, rather than email, because it is more personalized.

Patients seem to prefer quick and personalized medical service and AI tools have the potential to address those desires. The use of AI also has potential to speed up the process of addressing and improving patient satisfaction, as demonstrated by the Penn State researchers’ machine learning model.