- Meet Consumers/Patients
|Newsletters: Septic Complications of TPN|
Septic Complications of TPN
Lyn Howard, MB, FRCP, FACP
The following article is based on a talk given by Lyn Howard, Head of the Division of Clinical Nutrition at Albany Medical Center and Medical & Research Director of the Oley Foundation. The talk was given October 15, 1996, at the Albany College of Pharmacy to the Upstate New York chapter of ASPEN (UNYSPEN). It addresses chiefly catheter sepsis in the hospitalized patient, but many of these issues are pertinent to home patients, and from time to time most home patients are hospitalized.
My presentation today is really an evolving discussion about the devices and methods for keeping infection at bay in people with central venous catheters. The big picture first; there are an estimated 20 million vascular catheters placed per year mostly, of course, in the hospital setting. Approximately 3 million of these are central lines. The infection rate for central lines in hospital patients is in the order of four percent, which translates into 120,000 catheter related blood stream infections (CRBSI) per year. (Fortunately, the infection rate for homePN patients is much lower; in adults it averages one CRBSI every two years.)
A study by Pittet which looked at CRBSIs in hospital patients who were critically ill, found that the average cost per survivor was $40,000. It required eight extra days in the intensive care unit and extended their total hospital stay by 14 days. Additionally, there was a 35 percent mortality attributable to catheter related sepsis. Obviously this is a very sick population, but CRBSI is an extraordinarily expensive complication in any patient.
These costs and the fact that CRBSI rates in hospitalized patients are increasing, are a huge concern and thus subject to annual review by the Center for Disease Control. In fact, there has been a two to three fold increase in catheter related sepsis over the past ten years, particularly in the larger medical centers which presumably are treating the more complicated patient.
We’ve always had a problem in critically ill patients with gram negative bacterial sepsis and that really hasn’t changed. But the likelihood that the sepsis is related to a staphylococcal species has increased six-fold. There’s also been a four fold increase in fungal species, such as candida, and a two to three fold increase in enterobacter species. These are all organisms that frequently infect patients on parenteral nutrition.
Now a word or two about what we know about the factors that make infection more likely. First of all, it depends on the severity of the patient’s illness and therefore the hospital location of the patient. If we look at central venous catheters in patients in the MICU, there’s a likelihood of four septic events per 1,000 catheter days. That number increases to 13 if the patient is in a burn ICU. Perhaps this is no big surprise, since we all know that a burn patient has a loss of skin barrier to deal with and keeping a central line clear of infection becomes a big challenge in burn patients. Line sepsis also depends on the type of catheter used. For example, peripheral venous catheters have a much lower incidence of infection. In fact, 90 percent of all catheter related blood stream infections are associated with central lines and only 10 percent with peripheral lines.
How does the catheter or the blood get infected? There’s data now which shows that, during the first week or two, the likely mechanism for most infections is the organism going down the outside of the catheter. However, after a week or two when the Teflon barrier has become embedded, the most common cause of infection relates to organisms becoming imbedded in the proteinatious material built up on the internal lining of the catheter. And so it’s from the hub and the internal surface of the catheter that infections become an issue in the long haul.
We characterize catheter related infections into four types. First, we have the local infection at the exit site. This is easily visible. Second, we have tunnel infections where there is inflammation that extends all along the subcutaneous part of the catheter. This type of infection is also easy to recognize. Third there are systemic infections which develop either from the catheter or from what we are putting through the catheter. There have only been a few well documented episodes of infusates being contaminated, so solution contamination is an uncommon cause of infection. Fourth and last, the catheter can be secondarily infected by the blood from another source within the body (so called hemetagenous spread), such as a lung infection, urosepsis or an intra-abdominal abscess.
Another challenging question is how to properly document an infection that relates to the catheter. If we take the catheter out and roll it on a plate in a specified manner, described by Dr. Dennis Maki, then we can count the colony forming units and make a quantitative assessment. Positive cultures are strong evidence that the catheter was indeed infected. A second method for assessment entails sonicating the catheter in a bacterial growth medium, a technique which dislodges bacteria from the catheter. This method is associated with much higher counts of colony forming units. Finally there are techniques for gram straining the catheter and showing that the organisms are imbedded in the catheter material or adhering precipitate. But it should be noted that all of these methods require pulling the catheter out, and as many of you know, we are often in clinical situations where that isn’t really a good option.
So what can we do to quantitate a catheter related infection while leaving the catheter in? Typically we culture the blood, both from the central line and from the periphery, and demonstrate a positive blood culture with the difference in the number of colony forming units being five times higher in the central as opposed to peripheral culture. Obviously culturing the exit site, the tunnel or even the hub, is not difficult. The common organisms as I mentioned earlier, are staphylococcus, especially coagulose negative species and candida species. It’s usually possible to clear 90 percent of staph. coag. negative leaving the line in. So some attempt to clear the infection, prior to pulling that line, is especially reasonable if access to this patient is limited. The chance of clearing a coag. positive staph is only 50 percent. Since that organism puts the patient at high risk for a serious endocarditis, most practitioners would pull the line if the cultures grow staph coag. positive. If a fungus is cultured, then the line has to be pulled, since you can’t clear a fungal infection without taking the line out.
If we have this propensity and we have these costs, what can be done to avoid catheter related blood stream infections? I was interested in reviewing this field because I’m responsible for the placement of a lot of catheters. I think those of us involved in nutrition support have to stay on top of this data. What are the good prospective, randomized control trials? How can we move our response beyond personal preference to where the hard information is?
One place to start is with the site where the central line is placed. The site with the highest risk of infection is a catheter inserted below the waist in the femoral vein; the second highest risk is in the internal jugular vein; and least not is in the subclavian vein. This is why most practitioners try to access through the subclavian vein, even though sometimes we do have to use other approaches.
A second factor to look at is the catheter itself. There is no strong data to show increased infection rates with catheters that are multilumenal versus a single lumen other than those used for TPN. In the TPN literature, there are at least three good studies showing that as soon as a catheter is shared for several purposes and is no longer a dedicated TPN line, you go from an acceptable infection rate of less than 4 percent to one as high as 50 percent. So although triple lumen catheters are a wonderful device in many respects, the closeness of the ports and the presence of the TPN, a great growth medium, significantly increase the risk of infection.
The material the catheter is made from is also an important factor. It’s been shown that silicon and polyurethane, which have a much smoother surface compared to PVC (polyvinyl chloride), are associated with fewer episodes of blood stream infection and fewer episodes of thrombosis. It’s actually harder for staphylococcal and fungal organisms to adhere to something as smooth as a silicon surface. Thus, adherence may be an important mechanism for organisms getting a foot hold.
Now we move to the issue of the insertion technique. A lot of practitioners prefer that catheters, which are going to be used for extended periods of time, be inserted in the operating room (OR). However, it’s been shown that you don’t have to be in the OR, however you do have to use maximal barrier technique to reduce the risk of an infection. There are two good studies in which the physician inserted the central line using maximal barrier technique, which means gloves, gown, mask, and a large drape; they had a 0.6 percent incidence of infection. When they used just a small fenestrated drape and gloves, but no gown or mask, they had a three percent incidence of infection. So it’s not so much where you put in the line that’s important, but rather the precautions that you take when you do it.
What about those of us who have grown up with long term catheters who are very attached to tunneling? I, for one, was a little sad to see that there are now two prospective, randomized control trials showing no benefit from tunneling, at least in terms of infection. One thing we can say, is that tunneling certainly stabilizes the catheter. Once the tunnel and the cuff of adhering Teflon, which is just as close to the exit site, has grown into the skin, it becomes an important insurance that this piece of spaghetti won’t just slide out.
What about these new coated catheters? Two trials, one by Maki, one by Flowers, have shown a significant reduction of blood stream infections when a silver chelated collagen cuff was used. These cuffs deal with early infections that go down the tunnel along the outside of the catheter. Similarly, antiseptic catheters bonded with chlorohexidine or silver sulfadiazine, have been associated with a significant decrease in infection. Cefazolin bonding has also been associated with a significant decrease in catheter related blood stream infections; and vancomycin and heparin flushes have been shown to significantly decrease the most common infection of staphyloccocal origin. Obviously we are all concerned about exposing patients long term to modest doses of antibiotics and provoking resistant forms of bacteria to develop. So far, where that’s been carefully studied, it hasn’t happened yet. But we are probably sitting on a hot seat in this area.
One issue that used to be heavily emphasized, especially in ICU settings, was that of frequent line changing. In fact, some of the earlier studies indicated that if you didn’t change the line every seven to 10 days, you were going to see a mounting rate of sepsis. Well longer term studies have shown that actually this is not true. What happens, is that during the first 10 days there is a dramatic increase in the likelihood of infection, but that rate gradually flattens and eventually stabilizes. In long term catheters, the actual rate of hospital line infection is a stable rate of one percent per day. So, in fact, you face a high rate of infection initially, and this probably relates to the practitioner having disrupted the barrier aspect of the skin when putting in the line, setting the patient up for tunnel migration of bacteria. However, once the patient has gone beyond that phase, the practitioner is really dealing with issues related to the internal surface of the catheter, and for such a problem frequently changing that line is not the best approach to preventing infection. Only 20 percent of unexplained fevers in a TPN patient will turn out to be catheter related sepsis. For this reason, we usually don’t change a central line unless there is a strong clinical indication that the line is contaminated. If you choose to change that line over a guidewire, then you will see fewer mechanical complications than when you reinsert the catheter in a new site. When the catheter is taken out by changing it over a guidewire, the proper quantitative bacterial techniques can be done. If you discover the catheter is indeed contaminated, then it should be taken out and a new catheter inserted in a clean site. This is probably the most reasonable thing and conservative; you’ve kept your access and you didn’t take the risks of new sticks until you knew for sure that the catheter was infected.
A Clean Catheter Site
The next question is how to clean the catheter site. In a study by Maki, a large number of patients were randomized to use 10 percent povidone-iodine (Betadine), 70 percent isopropyl alcohol or a solution of two percent chlorohexidine. These researchers found the likelihood of a local infection and a bacterial blood infection was significantly reduced with chlorohexidine. For this reason, most practitioners now use two percent chlorohexidine as a mechanism for cleaning the line. One drawback is that chlorohexidine isn’t as easy to use at Betadine. It’s not alcohol based, so it takes longer for the site to dry before a dressing can be applied.
Another practice that many physicians prefer, is the use of transparent dressings. We like them because we can see what’s happening. We can see whether there is a redness developing around the insertion site. We can keep an eye on the tunnel. Recently there have been some good studies done on transparent dressings. Over a peripheral catheter, it’s clear that there is no increased incidence of infection when using the transparent dressing. However, over central catheters, the relative risk of infection is higher for people who are using transparent dressings than for those using dry gauze and tape. This is probably due to the fact that central catheters are often in an area that’s warmer, where sweating leads to moisture and the build up of organisms underneath the bandage. Thus, for patients who get frequent exit site or tunnel infections, gauze and tape is probably preferable. In well established lines, the dressing is not likely to influence line sepsis since these relate to the internal surface of the catheter. There are no studies yet examining whether one can, in fact, discontinue dressings over an established catheter.
Now what about IV or nutrition support teams? Does it make any difference if hospitals have a protocol that a team follows in a routine fashion? The answer is it does, and very significantly. In one study, doctor infection rates were compared to nurses specially trained to change the lines in a pediatric ICU. When the dressing was done by the surgical resident who was just coming in and out, not following an immaculate protocol, the rate of catheter related infection was as high as 28 percent. When the nurse specialist was doing it, the infection rate was down to 3 percent. So an organized approach to the care of lines and to the changing of dressings, is critical.
New Access Devices
Moving now to some of the newer devices; where do we stand on them and what are we learning? There’s a lot of interest in devices that have automatic valves, whether we are talking about Groshong catheters, or some of the newer needleless systems. It’s certainly easier for patients to deal with such a situation because they do not require a hub cap and do not require the patient to aspirate any air, and so on. There is concern, however, because the valves have many surfaces on which bacteria can lodge, and where freeing them with a flush is difficult. In a preliminary study, Maki has shown that this concern is legitimate; that a device that has had a needless system attached to it has a much heavier contamination than one that hasn’t.
A new port called the Cath-link has been introduced as a way of eliminating the need for needles. We have had some experience with this catheter in our institution. These devices may turn out to be a benefit to the relatively short term or intermittent user, perhaps for someone getting chemotherapy once a month; however, for people who are using their access over and over again, such as a long term TPN patient who is hooking up every night, this is not a viable approach. This port limits the direction from which you can insert your intracath, and the skin will only take so many sticks from one direction before it starts breaking down.
In a study from the Anderson Cancer Center in Texas, researchers noticed that when a heparin lock is used, you often see blood coming back up the catheter. That’s not too surprising. The patient has been anticoagulated in the area of the catheter, so the blood is able to move up into the catheter. If long term our problem is the precipitation of fibrin and proteinatious material on the internal surface of the catheter, the blood in the line becomes a mechanism by which platelets and fibrin can be deposited on the internal surface of the catheter. Because of this, the use of heparin becomes worrisome. The study was in leukemic patients who were using their port intermittently. They were given either a urokinase or a heparin lock once a month. Those patients using heparin were three to four times more likely to develop thrombosis or sepsis. This is why many physicians have switched to using urokinase at least on a preventive basis once a month. This enzyme digests and cleans the proteinatious material lining the internal surface of the catheter.
Recently there is increasing concern that the internal surface of ports don’t stay as gleamingly clean as we had hoped. In fact, the traditional port collects a lot of sludge in the reservoir. Those of you who have dealt with ports know that from time to time they just seem to block; even though the needle is in, there’s something that prevents infusion. Recent studies have shown that it’s common to have a certain amount of thrombos sitting permanently in the port. Presumably, every now and then a chunk breaks off and blocks the exit valve. One can actually demonstrate such a clot radiologically without taking the port out. So what can we do about this problem? Well, using urokinase once a month may help. But there have also been thoughts about creating a round interior to the port with a centrifugal exit, rather than a beehive-like interior and a perpendicular exit. When a round port is infused, any fibrin material is consistently flushed out, preventing any build up. In addition, there are now ports that have a very large diaphragm. We’re interested in this port, both because of its circular interior and because this device can be accessed from a much wider number of directions, allowing the skin of patients who are sticking themselves every night to hold up for much longer. So far our patients using this new port have reported easy port access. These new ports are now under review to see whether they reduce the skin breakdown and the build up of sludge.
In summary I have reviewed studies which discuss the cause and prevention of catheter sepsis, studies that challenge us to change our hard held ideas. I also touched on some new devices that are now being introduced.
Copyright © 1995 The Oley Foundation