College of American Pathologists
Printable Version

  Contaminated blood cultures: taking bold steps to lower rates


CAP Today




April 2012
Feature Story

Anne Paxton

Afrequent failure that is costly for a hospital can rightly be called a serious problem, but when the failure stems from avoidable error, it’s especially troub-ling. Blood culture contamination, a cause of false-positive results, diagnostic errors, and frustration, is one such failure, says Dennis Ernst, MT(ASCP), executive director of the Center for Phlebotomy Education, Corydon, Ind.

While high rates of blood culture contamination are common, they don’t have to be permanent, he says. Institutions can slash their contamination rates to two percent or less by mustering the will to tackle the causes. Ernst compares blood culture contamination to having an obnoxious brother-in-law who won’t move out: “As much as you’d rather deal with it tomorrow, it’s something that just has to be confronted today.”

Ernst, who recently served as a member of a Centers for Disease Control and Prevention evidence review panel establishing best practices for blood collection, has helped many institutions with the problem of contamination, starting with a hospital where he was employed several years ago. “We had a whopping 10 percent rate of contaminated blood cultures. They asked me to do in-service training, and we brought it down to one percent.” It was a 400-bed hospital, so the net savings probably amounted to hundreds of thousands of dollars, he says.

In a recent webinar, “Seizing Control of Blood Culture Contamination Rates,” Ernst outlined how management policies and techniques—including methodical site preparation, banning repalpation, and avoiding line draws—can help hospitals reduce blood culture contamination significantly, if personnel are properly trained and managed.

Not every hospital can see as dramatic an improvement as Ernst’s former employer. But he finds the American Society for Microbiology benchmark of three percent contaminated cultures to be a bit “timid” as an aspiration for hospitals. “I’m an advocate for better results,” he says. “I like people to shoot for contamination levels that are better than the benchmark.”

He is quick to point out that laboratory personnel are usually properly trained in the technique of drawing blood cultures and the threat of contamination; their average contamination rate, studies show, is one to three percent. “Blood specimen collection in general has always been the domain of the lab,” he said in an interview with CAP TODAY. “But in the last 20 years or so it has been diversified as a procedure and delegated to non-laboratory health care professionals. Historically, they’ve not been supported with the same educational component about the risks. That’s why we see such high rates of blood culture contamination coming from non-laboratory draws, especially the emergency room.” Ernst cites several studies showing that a decentralized staffing model with more non-laboratory personnel is associated with blood culture contamination rates of 4.2 to 8.4 percent. Similarly, emergency department staff have been found to have relatively high rates—five to 7.4 percent.

Unfortunately, the first preanalytical step in blood culture collection, deciding the number of sets of cultures ordered, is left to the physician and out of the lab’s control, Ernst says. “We draw multiple sets to rule out contamination, because the chances that multiple sets will be contaminated by skin flora is exceedingly remote. If all sets are positive, it’s highly diagnostic for bacteremia or septicemia. Doctors are taught about it in medical school, and for the most part they are up to date, but we still get orders occasionally for one blood culture, which is absolutely useless. It has very little predictive value.” The ASM recommends two to four blood culture sets, so that if only one is positive the physician can conclude it is a contaminated culture. Five sets of cultures, however, is considered overkill, he notes.

But if the cultures provide conflicting information, that puts the physician in a quandary. “Let’s say there is one set that is positive and it’s one species known to be a contaminant like Staphylococcus aureus. But the patient has a fever. Then we’re off to the races. The physician is likely to treat, because the physician isn’t going to take chances with the person’s life.” That’s just one example, Ernst says, of how the person drawing blood can change the way a patient is diagnosed and medicated. “If the culture really was contaminated during the collection, the patient is being treated for something he doesn’t have, the physician gets distracted from an accurate diagnosis, and thousands of dollars get allocated where they don’t belong.”

Those figures are backed up by multiple studies that have estimated the cost of an inpatient stay stemming from a contaminated culture at between $2,889 and $8,720 on average, with increased lengths of stay estimated at up to 3.3 days. “Even if we take the lower number,” Ernst says, “just think how much we can save the facility if we as phlebotomists, nurses, and so on, focus on technique.”

On an annualized basis, one study has shown that a reduction from five percent to one percent in the contamination rate recouped an estimated $400,000 at one hospital, while other studies have projected as much as $4.1 million savings in patient charges. “It can really add up to some serious change,” Ernst says. But it’s not only the facility costs that need to be factored into the equation. “If there is an increased length of stay, that subjects patients to treatments they don’t need, it extends the risk of hospital-acquired infections, and it keeps them from returning to their families and reclaiming their daily lives.”

Shooting for zero contaminated specimens is probably not realistic, Ernst points out. “Within every facility’s blood culture contamination rate is a contribution from transient bacteremias. Those occur when organisms enter the bloodstream from extra-vascular sources temporarily, then are promptly dispatched by the patient’s immune system. So if you’re drawing the patient at the same time as a transient bacteremia, it’s still going to look like you have a false-positive.”

For preventable contamination, site preparation is the first element of blood culture collection that should be addressed in a systematic fashion. “Historically, the protocol has been to perform a friction scrub 30 to 60 seconds to get dead skin cells off the surface and expose the underlying bacteria to the antiseptic that will follow, then apply the antiseptic in concentric circles of increasing diameter, for an antiseptic field that might be two or three inches in diameter,” Ernst explains. “The process of applying in concentric circles is always mentioned in textbooks.” But there are no data to back up the recommendation, he says. “The thinking is there are still bacteria on the skin; you don’t want to bring any bacteria toward the site where you’re going to put the needle. But even if you do bring something in, it’s going to be killed.” The Clinical and Laboratory Standards Institute has already dropped this traditional technique from its blood culture guideline because it is not an evidence-based practice in infection control today, so Ernst expects it to go by the wayside.

The antiseptic itself is not so much the issue, whether it’s multiple alcohol preps, tincture of iodine, chlorhexidine, povidone (Betadine), or less commonly now, iodophors. It’s easy to violate rules about drying time, Ernst says. “The term is something of a misnomer because it’s not so much drying time as contact time. Whether the antiseptic is dry after 30 seconds is not the point. It just takes that long for the chlorhexidine or povidone or tincture of iodine to kill bacteria. But 30 seconds can seem like an eternity in some situations such as the ER.”

Pediatric patients have special needs that sometimes lead personnel to rush through the drawing procedure, and not pay strict attention to the contact time rules for the antiseptic. Partly as a result, an especially disheartening statistic from one study is that 26 percent of pediatric outpatients are hospitalized unnecessarily because blood cultures drawn from them are contaminated, Ernst says. “I understand why the statistic is what it is, but it’s important that you take those steps to ensure these most delicate patients don’t get hospitalized if it’s not necessary.”

After site preparation, anyone drawing blood must take care not to repalpate if they want to avoid the possibility of contaminated cultures, Ernst continues. “You can’t repalpate the site after it’s been cleansed. If you do that, you have to re-cleanse the site. But some people just have to feel that vein one last time before putting in the needle, and it’s an ongoing struggle to keep them from doing it. We have to realize that gloves are not sterile and we’re putting bacteria right back where we worked so hard to remove it.”

In an attempt to compensate, some personnel cleanse their gloved finger with an antiseptic before repalpating, a practice that is accepted under ASM Cumitech guidelines. But Ernst warns against it. “People think ‘I’ll cleanse my finger just like I cleansed the skin,’ but there’s an essential difference. They’re probably not going to stand there and wait 30 seconds for the required contact time, so they’re just fooling themselves into thinking they’ve sterilized their fingertip.” Ernst doesn’t advise going against institutional policy, but if a hospital or other facility aims to keep contamination rates low, it should not take the risk of allowing repalpating.

A third policy that Ernst strongly advises for bringing contamination rates down: requiring venipunctures instead of doing line draws. In the best-practices panel he participated in recently at the CDC, it was decided that one such recommended practice was not to draw a blood culture during an IV start. “Drawing blood samples from an intravascular device happens all the time in the ED, and it’s the primary reason the contamination rates are so high there.”

There is strong evidence, Ernst says, that the facilities most successful in lowering contamination rates are those that limit all line draws. “If you can get administrative support for that, you would make pretty giant strides toward improving your rates.” He cites one hospital that established the policy that all ED blood culture specimens had to be collected by separate venipuncture. “They cut their contamination rate by more than six percentage points, which is huge. I know other facilities that ban this practice too, and they have some pretty fantastic outcomes.” While there is a perception that patients don’t want venipunctures, “if you told them ‘I’m going to prevent a contaminated culture,’ most people would want you to perform a separate puncture,” Ernst says.

Tracking performance in cutting contamination rates and training personnel regularly are essential practices once new policies are in place, Ernst says. He recommends training phlebotomists first. “They are not only the easiest to train but the most essential. They are going to be your gold standard. So putting them first is like training the trainer.”

Once the non-laboratory personnel have been trained as well, the performance of hospital departments and individual personnel should be tracked. “The hot spots of contamination are sometimes centered on specific departments like the ER, sometimes on individuals. But you have to find where contamination flourishes, because knowing where the highest rates are makes you more vigilant.” Using collector feedback introduces an element of competition that is highly effective. “Facilities that employ collector feedback can cut their rates in half,” he says. “Nobody wants to manage a department with the worst contamination rate, and you can’t hide in a group if your own rate is being tracked. So you’re motivated to get out of the spotlight as quickly as possible.” As a manager, he adds, “make sure your staff knows that that number is what your hopes rise and fall on each month.”

The benefits of cutting contamination rates are substantial, but Ernst believes hospitals should start by tackling the main contributors systematically: the need to address proper antiseptic application, no repalpation, and no blood cultures from line draws or IV starts. “With those three things, I think they would have tremendous success.” He would add a fourth item: “Once you get your rates down, continue to track, and continue to educate. You can’t just turn your back on this problem because it will rear its ugly head again and you’ll be right back to square one.” It’s not only advisable but essential to continually reinforce and monitor blood collection practices, he says, because “getting blood culture contamination rates under control is not a process with an endpoint.”

Anne Paxton is a writer in Seattle.