Small improvements in culture contamination rates yield large cost savings.
In 2014, the reduction of blood and urine culture contamination by 50 percent was a personal project and goal for the laboratory department in which I work, which serves a 100-bed acute care hospital with a robust outreach program. Like many others, our blood culture contamination rate bounced around at the three percent benchmark. A 50 percent reduction would require a 1.3 percent reduction in false-positive, or contaminated, blood cultures. How this one percent improvement saved $1.8 million in annual care costs was a journey with a few early failures, an unexpected twist, and a very happy ending.
Analyzing the problem
We began with a root-cause analysis, which led us to focus on improper collection techniques. In response, we initiated systemwide mandatory blood collection process education, followed by tracking the contamination rate by individual collector. While this did provide improvement, the percentage was 0.02 percent of our 1.3 percent goal. Clearly, we had not truly identified the root cause of contaminated blood cultures.
Returning to perform a second root-cause analysis, the team defined a more specific cause within the broad area of improper collection technique. It was determined that the required 30-second vigorous cleansing followed by a natural 30-second air dry before venipuncture was not occurring consistently. To assist staff in meeting the timing requirements, very inexpensive timers were purchased and placed everywhere blood culture supplies were housed. Spot checks by observation were conducted. Again, improvement occurred. We were disappointed, however, to find that the improvements were again minimal, bringing us only slightly closer to our goal. We had only achieved a 0.07 percent improvement. Without a doubt, we were missing something vital in our root-cause analysis.
Returning to the drawing board for the third time, the team took a two-pronged approach. First, a task force was trained to keenly observe the process in action. Second, a detailed data analysis was performed on the organisms by identification. After two weeks, the team reconvened with two key pieces of data. First, the collection techniques and timing were being followed to the letter, yet contamination rates remained constant. Second, in addition to coagulase-negative Staphlycoccus, Viridians Streptococcus, Corynebacterium and diphtheroids were most commonly isolated. That led us to ask a key question that we had missed in the original root-cause analyses: "Where were supplies obtained from?" The answer was, so to speak, our "smoking gun": "From drawers, pockets, cabinets, shelves, all over the place.... "
This lead the team to a key correlative event. Many times, an IV is started and blood cultures are drawn as the first blood from the Luer lock with a vacutainer holder. If there were the same level of lack of skin cleanliness, is the same level of bacteria being introduced into the blood stream as the blood culture bottle? To test the theory, intravenous (IV) catheter tips were cultured, and the results led the team to a significant discovery: The blood agar was almost perfectly pristine. The difference between the IV catheter and the blood cultures collection materials was the packaging. One was packaged for sterility; the other came from areas where it was subject to the accumulation of biomass from the environment. Tourniquets came from pockets, drawers, and multiple handlings. Vacutainer holders came from trays sitting open to the air and multiple handlings. Blood culture bottles came from shelves, drawers, and boxes open to the air and multiple handlings.
But how could the packaging be a central issue? The tops of the bottles were always cleaned according to the manufacturer's instructions and the industry's best practices. Had we really found the smoking gun?
To test the hypothesis, kits were produced in sterile processing using fresh, previously unopened boxes of all the key products except gloves, tourniquets, sterile 2x2 gauze pads, chlorohexidine scrubs, alcohol pads, vacutainer holders, butterfly needles, syringes, and packing bags. The kits were used exclusively for ninety days, and the contamination rate fell by 70 percent.
The joint work group of laboratory and nursing educated their peers and deployed the blood culture collection kits. For the requisite six months required to form a new habit, supply volumes fed the process, and the rate of contamination, or false-positive blood cultures, steadily declined. At the end of the study, our kit inventory depleted, staff returned to use of supplies from their normal hiding places. The blood culture contamination rate immediately rebounded to sub-optimal levels. It became clear that although our facilities are very clean and our HAI rate very low, there was a clear benefit to use collection kits.
With the volume of needed kits, limited production space, and the discomfort of bringing kit component boxes into the sterile processing field, we selected a vendor partner to produce the kit. The final kit was designed by a joint task force of nursing and laboratory staff. It contained the collection supplies used by both teams, but not packaged in a sterile packaging. Simple, direct instructions were written and added to each kit to ensure that every time, every collector had all the tools to perform the best collection possible.
With the return of collection kits and reminder education, false-positive blood culture result percentages immediately began falling. It took three root-cause work sessions to find the correct root cause and design a solution, but the results were dramatic. Figure 1 illustrates the power of small, targeted changes.
Looking back and forward
Two years have passed since we uncovered our root cause of contaminated blood cultures. While the effect of staff turnover remains a challenge, the scale of the fluctuation is approximately one percent. Overall, the goal to decrease blood and urine culture contamination by 50 percent has been reached, in fact beyond initial hopes. In blood culture contamination improvement alone, the organization has seen more than a 70 percent decrease in false-positive blood cultures or contaminated blood cultures. The result has been $1.8 million annual savings. The savings stem entirely from the difference in managing a patient with a false-positive blood culture and a true positive blood culture. The false-positive patients cost an average of $800 in additional laboratory and imaging testing and $1,900 in additional antibiotic use, and average an increased length of stay of 0.75 days. An assessment of the long-term impact on the reduction of antibiotic resistance within the community would be an interesting academic study, and would undoubtedly lead to an increase in savings.
When we began this process, we would not have imagined that a one-percent improvement would have such a significant financial and operational impact or that the root cause of false-positive blood cultures was how supplies are managed. Finding the resources for small improvements beyond the industry standard resulted in results beyond projections. In a time when staffing and financial resources are challenged, this demonstrates that a one-percent improvement can have a $1 million impact, even in a small process, in a small patient population.
(1.) Harding AD, Bollinger S. Reducing blood culture contamination rates in the emergency department. J Emerg Nurs. 2013;39(1|:e1-6.
(2.) Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev. 2006;19(4):788-802.
(3.) Slagle M, Richardson T. Evaluation of Cost Associated With False Positive Blood Cultures. University of Montana 2015.
Ginger Baker, LBBR MBA, MS, MT (AAB), practices in the healthcare areas of high reliability, operational excellence and laboratory operations. She focuses on the highest levels of quality, safety and fiscal operations in her current role of System Director of Operational Excellence with Appalachian Regional Healthcare based in Hazard, KY.
Caption: Figure 1: The journey of blood culture contamination reduction.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||SPECIAL FEATURE|
|Author:||Baker, Ginger A.|
|Publication:||Medical Laboratory Observer|
|Date:||Aug 1, 2017|
|Previous Article:||What is a table of critical limits?|
|Next Article:||Table of critical limits.|