Infection Control Hotline

By Marcia Patrick, R.N.

Q: Why is tracking infection rates so important?

Tracking infections is important to identify trends and clusters (or outbreaks) of infection. Calculating rates also is important because often, raw numbers don’t give the true picture. Say that three patients have nosocomial, central line-related bacteremias. Is that three of three? Three of 30? Three of 300? Selecting the correct denominator also is important. The more precise the denominator, the better the information produced. In this same example, one could use admissions or discharges to the unit the patients were on. It would give us a percentage of admitted or discharged patients who acquired bloodstream infections. But perhaps not all the admissions had a central line.

The denominator should include only patients at risk for central line infection, also known as the “population at risk.” A better denominator choice would be the number of central line days. This is obtained by capturing, at the same time each day, the number of patients who have one or more central lines. The daily count is added to obtain the number of central line days for the month. The rate is calculated by dividing the number of nosocomial central line infections that meet the definition, divided by the number of central line days, multiplied by 1,000.

This yields the number of central line-related infections per 1,000 central line days. Assuming that the CDC infection definitions and surveillance methods are used, that rate can be benchmarked against the National Healthcare Safety Network (NHSN) rates for central line infections. If other definitions are used, then the data cannot be benchmarked against CDC’s findings.

The other advantage to using standardized definitions and surveillance methods is that rates can be compared with similar units in organizations using the same processes. Comparing rates can help identify regional or even statewide outbreaks that can be defined, studied and hopefully, quickly remediated.

Q: What kind of technology is available to help hospitals both track and report infection rates?

Fortunately, the technology for tracking and reporting infection rates and identifying opportunities for improvement is now available and getting better all the time. Particularly in larger institutions, it’s hard to really see everything that is going on. A sudden cluster of infections or specific organisms is obvious and will be investigated. But there are a lot of what I call “percolators.” An organism comes and goes, or a particular infection shows up infrequently, so we don’t recognize the presence of a problem.

Patients move around the hospital frequently. If we’re looking at the location of a patient when the culture was obtained, we might not realize that the individual was in the ICU or other unit for several days prior, and the likelihood is that the infection originated there.

Data-mining programs can track patient locations, organisms, culture sites, etc., and identify trends and patterns early before they become outbreaks. These programs can identify opportunities for improvement, too.

We formerly conducted targeted surveillance in our ICUs for central line infections, ventilator associated pneumonias and catheter-related urinary tract infections. We did very little on the non-ICU units. With data mining, we get information—opportunities for improvement—on every unit.

We use a program (MedMined from Cardinal Health) that showed us that catheter-related UTIs were a much greater problem than we believed, and that they were contributing to an increased length of stay. We did a massive training program on aseptic insertion and maintenance of Foley catheters (our “bundle”) and had a tremendous decrease in UTIs.

We were able to reduce length-of-stay on one of our med-surg units from 4.1 to 3.7 days. We used a rapid-cycle improvement process (Plan-Do-Check-Act) to make changes at the unit level—they then owned the problem and the solutions.

The program was rolled out to other non-ICU units over several months, with similar, positive results. One thing we learned early on was that about 35 percent of our UTIs were probably present on admission, but the culture wasn’t performed until after 48 hours.

We discussed with the nurses the need to obtain the specimen and get an order when they encountered nasty looking urine on admission if the patient already had a Foley, or if they encountered it when inserting a catheter. We reviewed signs and symptoms of urinary tract infections, and what to look for. This was very successful.

The CDC is working with a number of data-mining vendors to interface their products with NHSN. This will allow direct reporting of data from the data-mining program to NHSN, and save the additional data entry. We must automate, as much as possible, our data collection and reporting.  The data still have to be identified and developed by the infection preventionist.  Several studies that show data from hospital chart coding is not accurate when it comes to infections.

It’s a good way to obtain “yes” or “no” data—e.g., did the patient receive aspirin on arrival? Was the antibiotic given within one hour of the incision? Those questions lend themselves to coded data. The presence or absence of a ventilator-associated pneumonia does not. Coders can only pull out what is written in the chart.

I’ve known physicians who call any respiratory issue “pneumonia” and others who won’t use the “p” word even when the patient has a florid case of it. This goes back to using standardized definitions and surveillance methods.

Q: What do rate comparisons among hospitals tell infection preventionists?

Rate comparisons can be valuable to see how organizational rates stack up against national data. Of course, we’re all aiming for zero. However, benchmarking can help prioritize areas that need work. The critical part is using the same definitions, applied the same way and the same surveillance methods as the organization being used as the benchmark. Rates can’t be compared if different infection definitions or surveillance methods are used.

Q: Do all hospitals have to report infection rates and do they report the same things?

Requirements for reporting infections vary by state. Over the past few years, a number of states have passed legislation that requires reporting of infections. Each state’s list of what must be reported, how it is to be reported, the definitions and surveillance methods used is different. In some states, reporting is encouraged but not mandated. The idea is to give consumers data with which to make decisions about their health care.

In many instances, definitions and surveillance methods were not specified, leading to data that were not comparable, although they were usually displayed in a table format that made it appear that the numbers could be compared. To me, bad data is worse than no data.

In Washington state, APIC members worked closely with the legislature and the Washington State Hospital Association for more than three years. We were able to convince lawmakers that we needed to use the standardized surveillance and definitions from NHSN as well as use NHSN for reporting our data rather than developing a whole new system.

The CDC has more than 30 years of experience with this; it’s a free program and probably offers the best hope of having comparable data. An interesting development is the impact of publicly reporting rates on surveillance activities. As infection preventionists, we have historically used very broad, inclusive definitions so we don’t miss any cases.

It lets us identify trends and patterns of infection and colonization. For public reporting, we want to use narrow definitions that only include true cases of infection. We will have to wait and see where this goes.

We need to do a better job educating the public on what reporting is based upon, what are good comparator data, and what the rates mean. Infection prevention and epidemiology is a distinct discipline with rules and guidelines that are not apparent to the lay public. We must be certain to ensure that data are gathered, reported and interpreted correctly and consistently and that limitations of the data are delineated.

Q: What do you foresee happening with infection rate reporting in the future?

I believe more states will require reporting. I encourage infection preventionists to work with their legislators or state health departments before laws are passed, so the resulting data is meaningful for consumers. Connect with colleagues in states that already have reporting to benefit from their experiences. Know what you want to accomplish before trying to convince legislators to adopt a particular bill or proposal.

APIC chapters in Washington did a lot of work by e-mail to come to a consensus on how we wanted to approach the issue of mandatory reporting. We identified our allies and potential foes. There was a lot of individual and group education of legislators and the state House and Senate health committees. We also came to a point where the legislature agreed to our proposals for reporting and using NHSN. APIC worked hard to keep our patients’ best interests at the forefront of our efforts, but also to be realistic about what infection preventionists would be able to do within existing resources.

We had to help the legislature temper their enthusiasm for data collection with the reality of taking valuable infection prevention time away from activities that actually reduce the risk of infection—it required a lot of education.

About this column

This column presents answers and practical guidance to some of the most commonly asked questions of suppliers and educators in the infection control and sterile processing communities. To submit a question to the column, e-mail Bob Kehoe, executive editor, at [email protected].

This article first appeared in the January 2009 issue of Materials Management in Health Care.

To respond to this article, please click here.