Hospital Acquired Infections (HAI) are more than just a thorn in the side of the healthcare industry – they’re a serious problem. According to CDC data, one in 20 patients — about 1.7 million people — will get an HAI, such as a urinary tract or surgical site infection, or pneumonia. They contribute at least in part to about 99,000 deaths a year.
A number of developers are mounting technologies they say might bring those numbers down. Using a plethora of technologies and approaching the problem from different angles, they’re aiming at altering the environmental factors that have made HAIs a threat to so many hospital patients.
Yuri Malina, for example, co-founded Evanston, Ill.-based SwipeSense, Inc., shortly after graduating from Northwestern University. When he was a student, he and a classmate, Mert Iseri, took part in a community project that tackled an important question: Why weren’t hospital personnel washing their hands more frequently?
He spent three months researching the topic and found two problems. “The first was convenience — doctors and nurses have to wash their hands about 100 times a day, and they have to stop what they’re doing and go to a sink or wall (sanitizer) dispenser. How it’s set up makes it difficult for it to become a habit or second nature.”
The second problem was a mismatch in perception. “People think that they are about 98 percent compliant with CDC guidelines for washing your hands. But the CDC says that, on average, hospital staff are about 40 percent compliant with those guidelines.”
So they founded SwipeSense, which is now marketing a sanitizer dispenser that lets personnel clean their hands on the go. It clips on to the scrubs of a doctor or nurse and carries enough gel to last over the course of a shift. It also sends signals to a central database that records when and how often personnel are sanitizing their hands.
The company did a short trial at Northwestern Memorial Hospital in Chicago with two hospital units using the device and two other units for comparison. Malina admitted to having a small sample size, but the study suggested that having the device resulted in a 64 percent increase in hand hygiene usage. Northwestern Memorial along with Rush Health Systems, also in Chicago, became SwipeSense’s first customers.
In case doctors and nurses object to wearing the sanitizing devices on their clothes, SwipeSense can retrofit existing wall sanitizer and soap dispensers.
“That enables us to get data from any dispenser in the hospital,” Malina added. “It can interface with a badge, so information can be garnered even from visiting physicians. It allows us to capture hand hygiene for the whole institution.”
Another developer, Novaerus, Inc., headquartered in both Dublin and Tampa, Fla., set its sights on sterilizing the air in a patient’s room, according to CEO Kevin Maughan. “We spend billions of dollars a year on infection control, and spend time on cleaning hands and wiping down surfaces, but nobody treats the air,” he said.
Novaerus air sterilization units, based on existing technologies that have long been used in aerospace and semiconductor industries, emit a plasma field generating particles that kill harmful microbes in a patient room or common area. The company reports that the device kills 99.99 percent of airborne microbes and reduces microbial surface counts by up to 68 percent.
Maughan attributed many HAIs to increased antibiotic usage, which often depresses a patient’s immune system.
“As soon as patients show any signs of infections, they’re given antibiotics,” he said. More than fifty percent of infections are viral, so antibiotics won’t do anything for them. They have to give them more and more, with stronger doses. Eventually they can contract an HAI — then as curtains in the room are opened and closed, and bed linens are changed, these things become airborne.”
“Our premise is we’ll reduce antibiotic usage by getting rid of airborne infections,” added Maughan. Novaerus devices are currently being used at just over 200 healthcare facilities in the U.S., primarily in Florida, Georgia, North Dakota and Minnesota.
Denys Proux, project manager at Xerox Research Centre Europe in Meylan, France, described yet another approach to HAI reduction, implemented using Xerox’s FactSpotter text mining technology.
Like the U.S., France has been hit hard by HAI’s, costing the nation about a billion euros (or $1.3 billion U.S.) annually. The French government had made some attempts at collecting HAI data, but encountered repeated difficulties with data that physicians were reporting, according to Proux. FactSpotter was utilized as part of the three-year Assistant de Lutte Automatisée et de Détection des Infections (ALADIN) project, which was launched in 2010. The goal was determining whether patient records could point to risks for infections, so providers could prevent them even before they started.
“Not all infections are the same,” Proux said. “Some are skin infections. Some are infections you might get from a surgery. Knowing the impact, whether in terms of deaths or just additional days spent in the hospital, is very important.”
“We asked, ‘Where can we to find the commonalities in the data, so as to avoid having HAI’s from occurring?’” Proux added.
The study collected data from ICU, brain surgery, stomach surgery and knee issue patients at the Hospices Civil de Lyon. Studying and redacting the data was a massive undertaking, as no sensitive information could leave the hospital.
“They were a lot of personal data in these reports. We had to move them outside of the hospital so we could work on them. We removed any personal information, such as doctor’s names, nurse’s names, addresses — any information that could go back to the patient,” Proux said.
FactSpotter compiled the information into a narrative order, making note of information such as drugs, bacteria, infections, temperatures or viruses. “It was very important to keep the chronology of the events, in order to make the distinction between hospital acquired infection and some other infection,” according to Proux, who said the system was able to detect 93 percent of problems, with an 87 percent accuracy rate.
“Text is still an important communication channel,” Proux said, adding that he and his researchers were constantly grappling with the question of whether their collected data would be sufficient in predicting infections. “Fortunately, the answer is yes.”
All the developers said a combination of techniques is needed to effectively combat HAI’s. “It’s very much, technology plus current best practices,” Maugham said.
But most important, suggested Malina, was a shift towards embracing change.
“The general culture of a hospital is a resistance to new ideas and they need to be very transparent and open to what is not working.”