According to the Centers for Disease Control and Prevention (CDC), at any given time, about one in every 20 patients acquires an infection related to their hospital care. Healthcare-associated infections (HAIs) not only affect patient lives, but also add to the nation’s growing healthcare costs.

This is one reason why Urbana, Ill.-based Carle Hospital and Physician Group has implemented ICNet, an automated infection surveillance system which monitors an array of clinical data streams and provides real-time alerts that allow infection prevention teams to identify affected patients more quickly and avert potential disease transmission.

ICNet, based in the United Kingdom serving more than 1,000 hospitals globally, has also provided Carle’s 345-bed hospital and multiple regional clinics with a pharmacy module that interfaces with the health system’s electronic medical record (EMR) system to ensure appropriate prescribing of antibiotics, detect and prevent adverse drug events, and provide more accurate and powerful reports on a wide range of quality-of-care metrics.

Daniel Bronson-Lowe, Ph.D., senior infection preventionist at Carle, recently spoke with Healthcare Informatics Assistant Editor Rajiv Leventhal to talk about more specifically about the implementation of ICNet at the hospital, which has been in place since October 2012.

Patients who get healthcare-related infections are likely to have their hospital stays extended. Is that a concern that led to Carle implementing ICNet?

Well, a couple of things really drove it for us. First, with all of the new reporting [as mandated by federal healthcare reform] coming up, there are so many pieces of information that we are required to keep track of and generate data sets to submit—to the National Healthcare Safety Network (NHSN), for example—and it was becoming very difficult to do some of those manually. We just don’t have enough people to do all those investigations—but with a system like ICNet, it does a great deal of the initial work for you. Instead of looking at a list that is 200 cases that needs to be investigated, it might be able to narrow it down to half a dozen depending on the situation, making it much more feasible. There is also a lot of information related to reporting surgical site infections, and we are expecting national organizations are going to want even more tracked in the future. It becomes a very laborious task for a human to pull all of the various pieces of information that have to be provided for all the procedures that are done; a computer system can encapsulate all of that information in a file that is ready to go. That was a big reason we needed this system.

Trying to identify all those points where attention is needed was another reason. If you have to spread your attention over all of the information coming from surgery and all of the information coming from the lab, it becomes very difficult to catch everything and drill down into the ones that are most important. But if you have a system that can do a lot of that initial work for you and bring together that combination of events that truly need attention, you are then able to focus on those important ones.

How is ICNet being specifically implemented at your hospital?

Implementation involves getting several data feeds into the system. We have our laboratory data feed, our radiology transcripts, a pharmacy data feed, and a surgery feed, in addition to a few others that we will be trying to add in the future. This system takes all of that and puts it together. At the core, it’s either running reports for the information it has or setting up alert systems so that if a certain combination of criteria is met across those various feeds for a particular patient or a cluster of events, it can fire off alerts which let you know you need to go take a look at something. Basically, by filtering data and reducing the need for manual identification of potential infections, the software will allow Carle’s infection preventionists to spend more time improving care processes and working with staff. ICNet will also be invaluable to us in reporting surgical data to state and federal agencies. It would be very difficult for an organization of our size to do that manually.

What kind of alerts does this include?

It varies. Right now, alerts that we have include notifications for one of the MDROs (multidrug-resistant organisms) that we are watching for. That alert would fire off and then we can take that information to set up flags within our EMR so people know that patient will need isolation moving forward.  We also have alerts built for a specific surgery. For example, if a CABG (coronary artery bypass graft) has been performed, the system will watch for a set amount of time for indicators of possible infections.  Maybe lab results come back indicating that a wound swab was done or an organism was indentified. Or it could fire off because the white blood cells rose during that time period. These are things you can set up. Then it is generating an alert saying these are individuals for whom those combinations of events have been met.  So I will go in and take a look at that person, look at the medical record, and investigate what is taking place and see if that is indeed a surgical site infection or just a random selection of events that came together.

What are some lessons learned or specific challenges you have faced since implementing ICNet?

The system might be very well developed and ready to take any data you give it, but sometimes peeling that data out of your systems and records on site can be tricky. We have had some difficulty, particularly with the pharmacy feed, getting all the variables that we need out of our current system, so that feed can go into ICNet. There were certain pieces of information and data that were not available in the initial data feed we thought we were going to use. In part, we were originally looking at having a pharmacy data feed that focused on orders, but what we really needed was a data feed that focused on administration of medication so we could have a much more accurate identification of when patients received that medication and for how long. We have been trying to find a good data feed that provides all of those pieces of information to ICNet.